Full text of "Annals of the Cape provincial Museums"

See other formats

Wit '.»:’■ w.’- .■■Vj:^‘.:r.tivv ViV'"*''

w i

ilLi;

■j;

■■a '’■•i

;

■1

; r

»'V-

. ' J

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 1 31st August 1988

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

L':

A further contribution to the knowledge of the ethology of the genus
Ceramius Latreille (Hymenoptera: Masaridae) in the southern and western

Cape Province of South Africa

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

ABSTRACT

Accounts of the ethology of Ceramius cerceriformis Saussure, C. jacoti Richards, C. rex
Saussure (previously known only from three specimens) and C. socius Turner are given with in
addition ethological notes on C. peringueyi Brauns, C. clypeatus Richards, C. nigripennis
Saussure, C. braunsi Turner, C. beyeri Brauns, and C. lichtensteinii (Klug). The contribution
made to the understanding of the species grouping in the genus Ceramius Latreille is discussed
in particular with respect to the division of one of these species groups.

INTRODUCTION

The present paper is the sixth in a series of publications (Gess, 1965, 1968 and 1973; Gess
and Gess, 1980 and 1986) dealing with the systematics and ethology of southern African species
of the genus Ceramius Latreille. In the two most recent of the above publications accounts were
given of some aspects of the ethology of the three species commonly occurring in the eastern
Cape Province: C. capicola Brauns, C. linearis Klug and C. lichtensteinii (Klug), and of some of
the more westerly occurring species: C. bicolor (Thunberg), C. clypeatus Richards, C. nigri-
pennis Saussure and C. socius Turner.

The genus has been divided into eight species groups (Richards, 1962; Gess and Gess,
1986). Gess and Gess (1986) found that the nest data available indicated that nest characters
are sufficiently different between groups and similar within a group to make them useful
taxonomic characters for use in conjunction with morphological characters in defining these
groups.

The present paper contributes further ethological data derived from observations and
investigations made in the Oudtshoorn district (7-12. xii. 1986), the Clanwilliam district
(7-13. X. 1987), the Hester Malan Nature Reserve, Springbok (15-21. x. 1987) and at Tierberg, in
the Prince Albert district (26.xi-5.xii.1987). These data add to the knowledge of three of the
groups and give a first account for a fourth group. They also contribute further foraging data
indicating preferences within groups for plants of a common family.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988
DESCRIPTIONS OF THE STUDY AREAS

The Oudtshoorn district

Three sites along the Kammanassie River to the south of Oudtshoorn were sampled for
wasps. These sites were: Frischgewaagd Site 1 (33° 40' 05"S, 22° 14' 18"E); Frischgewaagd Site
2 (33° 39' 29"S, 22° 13' 18"E); and Onverwacht (33° 37' 35"S, 22° 14' 18"E).

The area which lies in a rain shadow being immediately inland of the Outeniqua Mountains
receives an annual rainfall of 240 mm. Rain may be received throughout the year, however,
spring and late summer are the wettest periods. The soils above the flood plain are relatively
coarse grained and are of the Cretaceous Enon Formation. Those of the flood plain are light
coloured, finer textured and are of diverse provenance having been carried down from further
east by the river. The area lies within Acocks’ Veld Type 26, False Karroid Broken Veld
(Acocks, 1953 and 1975). The area is characterized by dwarf scrub, with a noticeable succulent
element, and with taller shrubs mainly along water courses (Fig. 1).

Fig. 1. Onverwacht, Oudtshoorn district, 6.x. 1987.

Tierberg, Prince Albert district

The farm Tierberg lies 25 km to the east of Prince Albert just north of the Swartberg and
is consequently drier than the Oudtshoorn sites receiving an annual rainfall of only 170 mm.

GESS & GESS; KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Rain may be received throughout the year but March is often the wettest month. The Tierberg
River which crosses the farm rises to the north east in an area of sandstone and mudstone of the
Beaufort Series, and passes over shale and sandstone of the Ecca Series, shale with bands of
chert and containing phosphate nodules and tillite with bands of shale of the Dwyka Series, and
to the south west a band of shale and quartzite of the Witteberg Series. The soils of the area are
therefore varied and those of the flood plain of mixed origin.

The area which lies within Acocks’ Veld Type 26, False Karroid Broken Veld (Acocks, 1953
and 1975) is characterized by low-growing dwarf scrub, dominated by Compositae and Mesem-
bryanthemaceae, with taller shrubs mainly along the watercourses and on the koppies (Fig. 2).

Fig. 2. Tierberg, Prince Albert district, 4.xii.l987.

The site (33° 42'S, 22° 16' 24"E) favoured for nesting by Ceramius species is an area of
relatively level ground flanking the Tierberg River in the vicinity of perennial pools. The soil in
the area is coarse and gritty, semi-non-friable to non-friable with a high enough clay element to
make it malleable when mixed with water.

The Clanwilliam district

Clanwilliam lies in the Olifants River Valley with to the east the Cederberg Mountains and
to the west a hilly area with beyond it the coastal plain. Apart from the sandy coastal plain the

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

whole area is classified geologically as Table Mountain Series. The soils are derived from
quartzitic sandstone and shale and are therefore a mixture of sand and clay the proportions of
each varying from area to area resulting in their being of variable friability.

The Olifants River Valley to the north of, around and some way to the south of Clanwilliam
lies in Acocks’ Veld Type 31, Succulent Karoo, and further to the south extending almost to
Citrusdal in Type 26, Karroid Broken Veld (Acocks, 1953 and 1975). The high lying areas lie in
Veld Type 69, Macchia (Fynbos) and the coastal plain in Type 34, Strandveld. Moll et al. (1984)
re-described the major vegetation categories in and adjacent to the Fynbos Biome. They
categorize the area in the immediate vicinity of Clanwilliam a “Mosaic of Dry Mountain Fynbos
and Karroid Shrublands” and the high lying areas to the west and east “Mesic Mountain Fynbos”,
changing to “Dry Mountain Fynbos” further to the west on the fringes of the coastal plain.

The sites at which Ceramius species were collected were Klein Alexandershoek
(32° 20' 20"S, 18° 46'E) (see Gess and Gess, 1986: Fig. 4) situated in Mesic Fynbos, a site 5 km
west of Clanwilliam on the road to Graafwater situated in dry Mountain Fynbos, a sparsely
vegetated slope above the Clanwilliam Dam (32° 11' 30"S, 18° 53' 42"E) and Kransvlei
(32° 14' 3"S, 18° 50' 49'E) (Fig. 3) (see also Gess and Gess, 1986: Fig 20) which lies in a transition
area, the vegetation being a mosaic of Succulent Karoo, Karroid Broken Veld and Fynbos.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA; MASARIDAE)

The Hester Malan Nature Reserve, Springbok

The Hester Malan Nature Reserve lies approximately 12 km to the east of Springbok in the
Carolusberg, in the region of Namaqualand termed Namaqualand Klipkoppe characterized by
rocky hills and Eindoorn Granite domes and receiving an annual rainfall of 100-200 mm. The
soils are coarse and sandy, however, where they have not been disturbed they are non-friable.
When trampled the structure is readily broken down and the soils become friable. In the river
bed and in water run-off channels on the slopes the soil is friable sand.

The chosen study area was 6 km by vehicle track north from the offices of the Department of
Nature Conservation in a hilly area drained by the headwaters of the Droedap River (Fig. 4). At the
time of the study, 15-21. x. 1987, the river bed was dry except for a series of pools where water from
beneath the sand comes to the surface, trickles across rocks and sinks once more beneath the
surface.

Fig. 4. Hester Malan Nature Reserve to the east of Springbok in the Carolusberg. 21.x. 1987.

The ground on either side of the river rises at a relatively steep gradient, on the west-facing
slope continuing at a steep gradient to the rocky hillcrest and on the east-facing slope evening
out to a relatively level stretch before reaching the rocky steep slope of the hilltops. The soils of
the east-facing slope are more stable than those of the west-facing slope, support a greater
variety of plants and are favoured by ground nesting wasps.

The area lies within Acocks’ Veld Type 33, Namaqualand Broken Veld (Acocks, 1953 and
1975). The east-facing slope, that favoured for nesting by Ceramius spp. is characterized by dwarf
shrubs. The dominant plants are Mesembryanthemaceae. Also common are perennial Compositae.

ETHOLOGICAL ACCOUNTS

The ethological accounts are presented in the context of the eight species groups, based on
morphological characters, to which the species have been allocated (Richards, 1962; Gess and

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Gess, 1986). The species composition of each group and for which species ethological data are
available are indicated. Group 2 has been subdivided and the suggestion that it should form two
separate groups is discussed.

Group 1.

C. fonscolombei Latreille (Fonscolombe, 1835; Bequaert, 1940 in Richards 1962, foraging
record only), C. caiicasicus Ed. Andre (Fahringer, 1922 in Richards 1962, foraging record
only), C. bureschi Atanassov.

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret,
c-h. ?

i. Resedaceae and Plumbaginaceae.

Group 2.

Richards (1962) placed together in this group four species, (C. cerceriformis Saussure,
C. schulthessi Brauns, C. clypeatus Richards and C. peringueyi Brauns), which to him were both
poorly known and taxonomically confused. Gess (1965) removed from C. peringueyi the male
assigned to that species by Richards and included this male together with a previously
undescribed female in C. richardsi Gess. The previously unknown female of C. clypeatus was
described at the same time. In a subsequent paper Gess (1973) sank C. schulthessi into synonomy
with C. cerceriformis . In Gess and Gess (1986) the emended Group 2 was shown to consist of
C. cerceriformis Saussure, C. clypeatus Richards, C. peringueyi Brauns and C. richardsi Gess.
These four species may be separated from the other Ceramius species by the following two
morphological characters in combination: mid tibia with two spurs and propodeum with
spine-like processes or at least very blunt angular projections. The inconsistency of form of the
clypeus exhibited by members of the group, however, leads to the view that this group is an
unnatural one and consequently it is here split into two as follows: Group 2a consisting of
C. cerceriformis and C. peringueyi which have the clypeus of both sexes unmodified (assumed to
be so for the still unknown male of C. peringueyi); and Group 2b consisting of C. clypeatus and
C. richardsi which have the clypeus of both sexes much modified with upturned apical teeth.

Group 2a.

C. cerceriformis Saussure (Gess and Gess, present paper) and C. peringueyi Brauns (Gess
and Gess, present paper, foraging data only).

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret.

c. ?

d. Nest with a relatively long main shaft.

e. First cell terminating the main shaft.

f. Succeeding cells terminating relatively short secondary shafts.

g. Secondary shafts sub-horizontal, grouped to one side of main shaft, not all at same or
different depths.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

h. A constructed mud-cell within an excavated-cell.

i. Forage plants of the family Mesembryanthemaceae.

Ceramius cerceriformis Saussure

Geographic distribution

Ceramius cerceriformis has been recorded from various sites in the western and southern
Cape Province: Calvinia and the Touws River district in the south west; several sites from
Bitterfontein northwards to Springbok in Namaqualand; and from Oudtshoorn and Willowmore
in the south east (Richards, 1962; Gess 1965, 1968, and 1973). The present observations were
made in the Hester Malan Nature Reserve, Springbok where a sample of 16 females and 6 males
was taken and at Tierberg, in the Prince Albert district where 10 females were taken.

Plants visited

There are only three records of flower visiting by C. cerceriformis . One female was collected
on “purple flowers of Mesembryanthemum sp. (sensu lato) (Aizoaceae)” 10 km South of Garies,
7/8. X. 1967 (F. W. Gess and W. H. R. Gess) (Gess, 1968), and a second female was collected on
flowers of Psilocaulon acutisepalum (Berger) N. E. Br. at Mesklip (29° 48'S, 17° 52'E), 1.x. 1985
(F. W. Gess and S. K. Gess). This species is also recorded as visiting the white flowers of
Mesembryanthemum crystallinum (L.) N. E. Br. (Mesembryanthemaceae) at Willowmore,
31.x. 1967 (C. F. Jacot Guillarmod) (Gess, 1973).

Provision

Pollen was obtained from provisioned cells from both sites. The pollen grains from both
sites are spherical, thin, smooth walled, 25 p in diameter and of the “mesem” pollen type.

Water collection

Females of C. cerceriformis were observed collecting water from small pools in the river bed
and from a puddle in a rut in the road in relatively close proximity to the single nest located in
the Hester Malan Nature Reserve. Whilst filling their crops these wasps stood on the very wet
soil at a short distance from the water’s edge.

Description of the nesting areas

Only two nests of C. cerceriformis were located. Nest 1, located in the Hester Malan Nature
Reserve, was in a bare area between low growing plants in the gently sloping area between the
steeper slope above the river and the rocky hills on the western side of the river, that is facing
east. The other, Nest 2, located at Tierberg, Prince Albert district, was between low growing
plants in more or less level ground above the steep bank of the river.

Description of the nest

Both nests (Fig. 5) consisted of a multi-cellular subterranean burrow surmounted by a short
cylindrical mud turret. The turret walls, which were unusually thin, were constructed from mud
pellets cemented together and smoothed on the inside. The burrow consisted of a main shaft
which descended sub-vertically and then turned sharply to one side to end in a sub-horizontal
cell, one of a group of cells, six cells in Nest 1 and two in Nest 2, all/both lying to the same side
of the main shaft and all terminating sub-horizontal secondary shafts. The main shaft of Nest 2
maintained a diameter of 5,5 mm along its entire length, 135 mm, whereas that of Nest 1 though

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Fig. 5. Vertical plans of turrets and underground workings of nests of C. cerceriformis Saussure (x 1). Nest 1, Hester
Malan Nature Reserve, 20.x. 1987 and Nest 2, Tierberg, 27. xi. 1987. For key to lettering see Table 1.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

of the same diameter over the greater part of its length was enlarged to form a bulb 10 mm in
diameter and 15 mm long at a depth of 95 mm, its total length being 155 mm.

The sealed cells, five in Nest 1 and one in Nest 2 contained constructed mud-cells. The
lateral shafts leading to these sealed cells had been filled with compacted earth and their
entrances to the main shaft had been sealed off with mud.

The cells were all orientated sub-horizontally and were arranged in two ranks of three
ranging in depth from 151-185 mm and all to one side of the main shaft.

The constructed mud-cells were collected. Those from the Hester Malan Reserve are 22,
5 mm long and those from Tierberg are 19 mm long. The diameter at the widest point which is
towards the inner end of the cell is 8,5-9 mm in the former and 8 mm in the latter. The walls
of all are approximately 1 mm thick, rough on the outside and smoothed on the inside. The
completed cells were sealed with a concave mud plate constructed just within the mouth of the
cell.

The provision is a firm pollen loaf which is loose in the cell. It is approximately 13 mm long
and is 6 mm in diameter over the greater part of its length, however, the first quarter is
noticeably narrower being only 2,5 mm in diameter (sample of the five loaves obtained from
Nest 1).

Sheltering

The female nest builder at Tierberg was found to be sheltering in her nest at night.

Male behaviour

Males were not observed at flowers or in association with the nest. They were, however,
observed at water where they were seen to be pairing with females.

Ceramius peringueyi Brauns

Geographic distribution

Ceramius peringueyi has previously been recorded from Stellenbosch, Het Kruis and
Paleisheuvel to the south west of Clanwilliam (Gess, 1965) and is now recorded in addition from
a site 5 km east of Vredendal on the road to Vanrhynsdorp.

Plants visited

Fourteen females of C. peringueyi were collected foraging on pinkish white flowers of
Psilocaulon acutisepalum (Berger) N. E. Br. (Mesembryanthemaceae) at the site 5 km east of
Vredendal on the road to Vanrhynsdorp situated in Acocks’ Veld Type 31, Succulent Karoo.

Group 2b.

C. clypeatus Richards (Gess and Gess, 1986 and present paper, foraging data only) and
C. richardsi Gess.

a.-h. ?

i. Forage plants, yellow flowered species of Leguminosae.

Ceramius clypeatus Richards

Geographic distribution

Ceramius clypeatus has been recorded from various districts of the western Cape: the
Citrusdal district; the Clanwilliam district; and the Het Kruis district to the south west of

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Clanwilliam (Richards, 1962; Gess, 1965; Gess and Gess, 1986). The present observations were
made in the Clanwilliam district. Fourteen females were collected.

Plants visited

Gess and Gess (1986) recorded the capture of two males foraging on the yellow flowers of
Aspalathus desertorum Bol. (Leguminosae) at Klein Alexandershoek, Clanwilliam district. This
was a surprising record as previously all known forage plants of Ceramius spp. in southern Africa
were either of the family Mesembryanthemaceae or Compositae. During the period 7-14. x. 1987
flowering plants in the Clanwilliam district were sampled for wasp visitors. C. clypeatus was
found foraging only on Aspalathus desertorum from which specimens were taken at the
Clanwilliam Dam (14.x. 1987, 10 females) and at a site 5 km west of Clanwilliam on the road to
Graafwater (12.x. 1987, 4 females).

Group 3.

C. nigripennis Saussure (Gess and Gess, 1986 and present paper), C. toriger Schulthess
(Gess, 1968, foraging record only), C. braunsi Turner (Gess and Gess, present paper,
foraging record only), C. jacoti Richards (Gess and Gess, present paper), and possibly
C. micheneri Gess.

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret.

c. Nest perennial (confirmed for C. nigripennis only).

d. Nest with relatively short main shaft.

e. First cell terminating main shaft.

f. Succeeding cells terminating extremely short secondary shafts.

g. Secondary shafts sub-vertical and all of comparable depth.

h. A constructed mud-cell within an excavated-cell.

i. Forage plants of the family Compositae.

Ceramius nigripennis Saussure

Geographic distribution

Ceramius nigripennis seems to be a Namaqualand species, having been recorded from
various sites in that region: Swart Doringrivier, between Bitterfontein and Garies; Garies;
Kamieskroon; Bowesdorp; Mesklip; and a site 8 miles east of Springbok on the road to Pofadder
(Richards, 1962; Gess, 1965 and 1968; Gess and Gess, 1986). During the present study it was
found to be extremely abundant in the Flester Malan Nature Reserve, Springbok.

Plants visited

Gess and Gess (1986) recorded C. nigripennis provision to contain pollen which matched
that of the orange flowered Dimorphotheca sinuata DC. (Compositae). In the Hester Malan
Nature Reserve a female was observed on the yellow flowers of Pentzia suffruticosa (L.) Hutch,
ex Merxm. (Compositae) and a male was collected on the yellow flowers of a composite cf.
Berkheya.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA; MASARIDAE)

Water collection

In the Hester Malan Nature Reserve females were observed in large numbers collecting
water mainly from small pools but also from trickles over rocks and sand in the river bed. Whilst
filling their crops these wasps stood on the water (Fig. 6).

Fig. 6. Female C. nigripennis Saussure filling her crop with water whilst standing on the water surface, Hester Malan
Nature Reserve, 16.x. 1987. Actual length of female 15 mm.

Male behaviour

Male C. nigripennis were commonly observed at water in company with the females with
which they were seen to pair.

Description of nesting area

Thousands of nests of C. nigripennis were located in the Hester Malan Nature Reserve on
the west slope above the river, from the river bank to the upper limit of the steeper slope. As
at Mesklip (Gess and Gess, 1986) the nests were grouped in relatively small aggregations in close
proximity to the bases of bushes on the edge of bare patches. The north sides of bushes seem to
be particularly favoured.

Description of the nest, method of construction of the nest, oviposition and provisioning.

See Gess and Gess, 1986.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988
Ceramius jacoti Richards

Geographic distribution

Ceramius jacoti has been recorded from various sites in the south western Cape: Hex River;
Touws River district; Constable; Bloutoring Station; a site 50 km east of Touws River; the
Montagu district including Ouberg Pass; Ladismith; and Oudtshoorn in the east (Richards, 1962;
Gess, 1965). The present observations were made in the Oudtshoorn district. A sample of 23
females was collected. These additional records are: Frischgewaagd Site 1 , 7.xii.l986, 1 female
(H. W. Gess); Frischgewaagd Site 2, 7-8. xii. 1986, 5 females (F. W. Gess), 7 females (H. W.
Gess), 8 females (R. W. Gess), and 1 female (S. K. Gess); and Onverwacht, 9-12. xii. 1986,
1 female (S. K. Gess).

Plants visited

Ceramius jacoti was collected foraging on the yellow flowers of Pteronia incana DC.
(Compositae), Brandrivier road, 2 miles from the junction with the Ladismith-Riversdale road,
30. ix. 1972, by C. F. Jacot Guillarmod (Gess, 1973). All those specimens from Frischgewaagd
Site 2, listed above, were collected foraging on the yellow flowers of Senecio rosmarinifolius
L. f. (Compositae).

Provision

Pollen was obtained from a fully fed larva of C. jacoti and compared with that of Senecio
rosmarinifolius with which it was found to be identical.

Water collection

Only one female was observed at water, a puddle in a rut in the road.

Male behaviour

Males were not observed. However, their absence from the nesting area is of interest.
Description of the nesting area

A nesting area of C. jacoti was located at Onverwacht. The nesting site was a sloping bank
on the edge of a bare area on the vegetated flood plain.

Description of the nest

The nest (Figs 7, 8 and 9) consists of a subterranean burrow surmounted by a curved tubular
mud turret roughly smoothed on the outside and well smoothed on the inside with no interstices
left open. The turret and shaft opening are of equal diameters. The relatively short main shaft
descends sub-vertically. Initially it is of the same diameter as the shaft entrance but after a short
distance the diameter increases by approximately 50 per cent and then decreases forming a
“bulb” of approximately equal length to the upper part of the burrow. Beneath the bulb the
shaft continues for a short distance ending in a cell. The excavated-cell approximates the “bulb”
in diameter. Within it is a constructed mud-cell.

As the five nests investigated were new nests and none had reached the stage of cell closure,
the description of the nest must perforce for the present remain incomplete.

Method of construction of the nest, oviposition and provisioning
Water is required for nest construction.

GESS & GESS; KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Fig. 7. C. jacoti Richards, nest turret (x 1,8), Onverwacht, Oudtshoorn, 9.xii.l986.

The turret is constructed at an early stage in nest excavation. At the commencement of
turret construction pellets are laid down around the shaft entrance in such a way that the turret
will have the same inner diameter as that of the shaft, that is 5 mm (sample of five, no variation).
The walls of the turret are approximately 1,5 mm thick. The shaft diameter is maintained
constant to a depth of 20-25 mm (average 22 mm, sample of 4) after which the diameter is
increased to form a “bulb” 7-9 mm (average 7,7 mm, sample of 4,) in diameter and 20-24 mm
(average 21 mm, sample of 4) in length. Beneath the “bulb” the diameter of the shaft returns
to 5 mm for a short distance to form a neck to the cell. An excavated-cell of diameter 7-9 mm
is prepared and within this a mud-cell is constructed with its walls approximately 1 mm thick and
the inner surface of the walls well smoothed.

Only five nests were located. These were excavated and found to be new nests, four of
which contained one cell each. No egg or pollen loaf was obtained. One cell contained a large
fully fed larva from which pollen was extracted and identified (see Provision).

Sheltering

A female C. jacoti was found sheltering in one of the nests.

Associated insects

One of the nests was found to have been invaded by a megachilid bee which had constructed
a leaf cell within the nest.

Fourteen Allocoelia capensis (Smith) (Chrysididae) were collected from flowers of Senecio
rosmarinifolius where they were found in company with foraging C. jacoti. A. capensis has
previously been recorded as a parasite of C. lichtensteinii (Brauns, 1910) and in association with
this wasp in its nesting areas (Gess and Gess, 1980).

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Fig. 8. Excavation showing burrow of C. jacoti Richards cut through
vertically, constructed-cell intact, (x 1,8), Onverwacht, Oudtshoorn,
9.xii.l986.

GESS & GESS; KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Fig. 9. Vertical plans of turrets and underground workings of two nests of C. jacoti Richards (x 1). Onverwacht,

Oudtshoorn, 9.xii.l986.

Ceramius braunsi Turner

Geographic distribution

Ceramius braunsi has been recorded from: Olifants River, between Citrusdal and
Clanwilliam; Pakhuis Pass to the east of Clanwilliam; Vanrhynsdorp to the north; Worcester to
the south east; and Willowmore, the most easterly record (Richards, 1962; Gess, 1965, 1968 and
1973). The present observations concerning this species were made at Kransvlei in the
Clanwilliam district during the period 7-13. x. 1987. A sample of ten females was collected.

Plants visited

Two females of C. braunsi were collected foraging on the yellow flowers of Athanasia
trifurcata (L. ) L. (Compositae) growing at Kransvlei, Clanwilliam district.

Water collection

Six females of C. braunsi were collected on very wet sand, that is supersaturated sand
covered with a film of water, at the edge of the dam at Kransvlei, Clanwilliam district.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Group 4.

C. beyeri Brauns (Brauns, 1910; Gess, 1973, forage record only; Gess and Gess, present
paper) and probably C. damarinus Turner

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret,
c-h. ?

i. Forage plants of the family Mesembryanthemaceae.

Ceramius beyeri Brauns

Geographic distribution

Ceramius beyeri Brauns has previously been recorded from: Calvinia; Blaukrans near
Calvinia; Cape Town, Beaufort West {sic)\ Willowmore; Cradock; Somerset East; and
Grahamstown (Richards, 1962; Gess, 1973). The present observations were made at Tierberg,
Prince Albert district, where three females and one male were collected (F. W. , S. K. and R.
W. Gess, 26.xi. 1987-5. xii. 1987).

Plants visited

Gess (1973) recorded a female C. beyeri visiting the “whitish flowers” of “mesems”
(Mesembryanthemaceae) at the Bible Monument, Grahamstown. During the present
investigation one female was collected visiting the white, flushed with pink, flowers of
Sphalmanthus cf. bijliae (N. E. Br.) L. Bol. (Mesembryanthemaceae).

Description of nesting area

The nesting site of C. beyeri at Tierberg was at the base of a stone in a bare area of level
ground in dwarf scrub on the left bank of the Tierberg River in close proximity to water.

Description of the nest

A single nest of C. beyeri was located. It consisted of a burrow surmounted by a short
cylindrical mud turret constructed from pellets cemented together and smoothed on the inside
so that no interstices remained. The inner diameter of the turret was equal to that of the shaft,
that is 5 mm, and its height was 7 mm. The shaft descended vertically for 20 mm. Unfortunately
as no further nests were located the nest plan and the nature of the cells remain unknown.

Group 5.

C. lichtensteinii (Klug) (Brauns, 1910; Gess and Gess, 1980)

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret.

c. Nest perennial.

d. Nest with relatively long main shaft.

e. No cell terminating the main shaft.

f. Cells terminating extremely short secondary shafts.

g. Secondary shafts sub-horizontal and each successive cell deeper.

h. A constructed mud-cell within an excavated-cell.

i. Forage plants, shrublets of the family Mesembryanthemaceae.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Ceramius lichtensteinii (Klug)

Geographic distribution

Ceramius lichtensteinii is, for Ceramius, a widespread species having been recorded from
the Tankwa Karoo in the west to Grahamstown in the east and from the Eastern Cape Province
north to Kroonstad in the Orange Free State (Richards, 1962; Gess, 1965, 1968 and 1973).

Description of the nesting area

Gess and Gess (1980) recorded that C. lichtensteinii at Hilton and Clifton, both Eastern
Cape, show a preference for nesting on raised ground despite the abundance of apparently
suitable bare areas of horizontal ground which are favoured for nesting by C. capicola and
C. linearis. It seems of interest therefore to note that at Tierberg, Prince Albert district, where
pseudo-colonies of this wasp were extremely common at the time of the authors’ visit in
December 1987, horizontal ground and raised sloping ground were equally favoured.

Plants visited

In the Grahamstown district. Eastern Cape Province males and females of C. lichtensteinii
have been recorded visiting flowers of several species of Mesembryanthemaceae (Gess, 1973 and
Gess and Gess, 1980). At Tierberg both males and females of these wasps were commonly seen
on the white, flushed with pink, flowers of Sphalmanthus cf. bijliae (N. E. Br.) L. Bol.
(Mesembryanthemaceae) .

Description of the nest

The nesting of C. lichtensteinii, C. capicola and C. linearis at Hilton has been described in
detail (Gess and Gess, 1980). The diameter of the main shaft of the nests was stated to be related
to the size of the wasps. The average diameter of the main shaft of the nests of C. lichtensteinii
at Hilton was 6,2 mm. It is of interest that the average diameter of the main shaft of the nests
of this species at Tierberg was 8,0 mm (sample of 22). This significantly greater diameter can be
related to the significantly larger size of individuals of this species at the Tierberg Site. The width
of heads measured across the eyes is 5,03 mm (average of 30, range 5,0-5, 5 mm) for females
from Hilton as compared with 6,35 mm (average of 30, range 6, 0-7,0 mm) for females from
Tierberg. In other words the diameter of the shafts of the nests at Tierberg is 29 per cent greater
than that of nests at Hilton and the width of the heads of the wasps is 26 per cent greater.

Group 6.

C. cfl/jfer Saussure, C. metanotalis Richards, C. rex Saussure (Gess and Gess, present paper)

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret.

c. Nest perennial.

d. Nest with a relatively short main shaft.

e. First cell terminates the main shaft.

f. Succeeding cells terminate secondary shafts.

g. Secondary shafts sub-horizontal, in a group to one side, neither all at the same depth nor all
at different depths.

h. A constructed mud-cell within an excavated-cell.

i. Forage plant, the yellow flowered Berkheya spinosissima (Thunb.) Willd. (Compositae).

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Ceramius rex Saussure

Geographic distribution

Ceramius rex is very rare in collections, apparently being known from only three specimens
prior to the present study. The type specimen, a female, was described from the Cape Colony
by de Saussure in 1855. Richards (1962) without explanation and without having seen the type
specimen sank C. rex into synonymy with C. lichtensteinii (Klug). The second specimen, a
female, also not seen by Richards, was collected at Caries in Namaqualand in 1931 (S. A. M.
Staff). This specimen was stated by Turner (1935) and Cess (1965), who supported his view with
a detailed description of this specimen, to correspond to the description of C. rex Saussure. In
an unsuccessful attempt to obtain further specimens. Cess (F. W.) collected at Caries during
early October 1966 and again at the same time in 1967. The third specimen collected by chance
by Dr and Mrs H. Townes at Caries (25. ix. 1970) was the long unknown male of C. rex. A
description of this male was given by Cess (1973).

The search for the elusive C. rex was taken up again in September/October 1985 by Cess
and Cess (the present authors) and was finally brought to a successful conclusion in October
1987 when it was discovered nesting not uncommonly at a site in the Hester Malan Nature
Reserve, Springbok, 100 km due north of Caries. A sample of 39 specimens, 28 females and 11
males, was taken during the period 15-21. x. 1987, 12 nests were investigated, a forage plant was
identified and both males and females were taken at water.

From the scanty distribution data available it seems likely that C. rex is of rather restricted
occurrence and may probably be termed a Namaqualand species.

Plants visited

During the period 15-21. x. 1987 a wide range of plants in flower in the Hester Malan Nature
Reserve was sampled for wasp visitors. Only one instance of plant visiting by C. rex was
recorded, a female, foraging on the yellow flower capitulae of Berkheya spinosissima (Thunb.)
Willd. (Compositae).

Provision

Provision was obtained from the cells of C. rex from three nests. Pollen grains from the
provision and from flowers of Berkheya spinosissima were compared. It was found that the provision
contained pollen grains identical in size and structure to those from B. spinosissima flowers and also
grains of apparently the same structure but of a noticeably smaller size (ratio 9:13).

Water collection

Females were observed collecting water from small pools in the river bed. Whilst filling her
crop a wasp stands on the water surface.

Male behaviour

Males were not seen on flowers nor in the nesting area, however, they were collected at
water where they were observed pairing with the females. It seems that they wait for the females
at and around their water collecting source.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)
Description of the nesting area

A nesting site of C. rex was located in the Hester Malan Nature Reserve. The area favoured
for nesting by these wasps is gently sloping and lies between the steeper slope above the river and
the rocky hills on the western side of the river, that is facing east. The nests occurred in bare
areas either singly or more commonly grouped. The nests in the groups were old, old re-used and
new suggesting that a newly emerged female tends to nest either in the nest of her origin or else
in a new nest excavated by herself in close proximity to the one from which she emerged.

Description of the nest

The nest (Fig. 10) of C. rex consists of a multi-cellular burrow surmounted by a curved
tubular mud turret. The turret is constructed of mud pellets roughly smoothed on the outside
and well smoothed on the inside. Some interstices are left open. The turret and shaft opening are
of equal diameters. The main shaft descends sub-vertically and is initially of the same diameter
as the entrance but widens after some distance. The wider section of the main shaft varies in
length from nest to nest. In some nests it is no more than a turning “bulb” below which the shaft
continues downwards with a diameter equalling that of the upper section of the shaft. In others
the entire lower section of the sub-vertical shaft is wide. In some nests the diameter of this lower
wide section of the shaft fluctuates so that its sides are very uneven. The shaft at the base of its
sub-vertical section curves outwards to terminate in a cell which lies sub-horizontally.
Sub-horizontal secondary shafts each terminating in a cell fan out from the main shaft but never
form a complete whorl so that the cells lie together in a group. In some instances the cells are
at different depths and the shafts leading to them leave the main shaft at different depths but
always forming a group. Within an excavated-cell there is a constructed mud-cell sealed at the
neck with a mud-plate. The section of the secondary shaft between the cell and the main shaft
is filled with soil and sealed off from the main shaft.

Method of construction of the nest, oviposition and provisioning

At the commencement of nesting a female may either initiate a new nest or enlarge the nest
from which she emerged. Water is required for nest construction. At an early stage in nesting
both nest initiators and nest enlargers construct a turret surmounting the excavation. At the
commencement of turret construction pellets are laid down around the shaft opening in such a
way that the turret will have the same inner diameter as the shaft, that is 5,5-7 mm (average 6,7
mm, sample of 10). The walls of the turret are approximately 1 mm thick, the external diameter
of the turret being 8-9 mm (average 8,8 mm, sample of 11). Pellets are added to the turret in
such a way that the turret soon curves over, the top of the curve being 11-15 mm (average 13
mm, sample of 11) in height. In the construction of new nests, after turret construction has been
completed, further pellets extracted in shaft sinking are discarded in a pellet dropping area
100-200 mm from the nest.

In the excavation of a new nest the initial diameter of the shaft is maintained to a depth of
35-84 mm (average 47 mm, sample of 12) after which the diameter is increased to 10-15 mm
(average 11,5mm, sample of 10). This diameter is maintained to a variable depth. In some
instances the expanded portion is no more than a “ bulb” beneath which the shaft returns to its
initial diameter. In others the increased diameter is maintained for several centimetres before
there is a return to the initial diameter and in still others the shaft continues at an increased
diameter for the remainder of its length. Having reached a depth of 85-115 mm (average 103,
sample of 12) there is a change in the direction of the excavation of the shaft which curves

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Nest

Fig. 10. Vertical plans of turrets and underground workings of C. rex Saussure (x 1). Hester Malan Nature Reserve,

20-21. X. 1987. For key to lettering see Table 1.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

outwards and has a diameter approximating that of the upper section. After a short distance the
diameter is increased to 14-17 mm and the excavation terminates in a round ended ovoid cell
about 32 mm in length. Within this cell a mud-cell is constructed having a wall 2,5-3 mm thick
and being carefully smoothed on the inside.

If a female is nesting in the maternal nest, she cleans out the shaft, constructs a turret, and
prepares for re-use the cell from which she herself emerged by cleaning the inside of her cocoon
which is left in position.

A new cell having been constructed or an old cell prepared for re-use oviposition takes
place. The egg is large , yellow and curved, 6,5 mm from tip to tip across the bow and is laid
loosely at the inner end of the cell.

Oviposition having been performed provisioning takes place. The provision, a mixture of
pollen and nectar, is a relatively dry and firm loaf of even diameter along its length.

The cell is sealed with a mud plate and the sub-horizontal shaft is then firmly packed with
soil until the sub-vertical shaft is reached, when it is sealed off with a mud plate.

Further cells terminate secondary shafts and are prepared in a similar fashion to the first cell
of a newly constructed nest. In a re-used nest only the cell from which the possessor of the nest
herself emerged is re-used, all other cells and the secondary shafts leading to them are freshly
excavated.

A sample of twelve nests was investigated (Table 1). Four were new nests. In all of these

Table I

Details pertaining to 12 nests of Ceramius rex Saussure investigated in the Flester Malan Nature

Reserve, Springbok on 20-21. x. 1987.

Nest

No.

Nest Status

Turret

No. of
cells

Nature of each cell,
cell contents

Wasp in nest

Old, reused

Present

A A A B/C B/C B/C F P

female

Old, reused

Present

A A A B/C F I

female

Old, reused

Present

B/C B/C F I

female

New nest

Present

E I

female

New nest

Present

E I

female

Old, reused

Rudimentary

A A A A B/C I

none

New nest

Present

none

Old, reused

Present

B/C B/C F I P

female

New nest

Present

female

Old, reused

Present

A A B/C I P

female

Old, reused

Present

A A E H

female

Old, reused

Present

A B/C B/C B/C E

female

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

B. Cell closed, containing pupa in cocoon.

C. Cell closed, containing pre-pupa in cocoon.

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

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

FI. Cell open, containing egg without provision.

I. Cell open, empty.

P. Parasitized.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

there was an open cell and in two a closed cell in addition. Eight nests were re-used of which two
possessed 4 cells, four 5 cells, one 6 cells and one 8 cells. In nests 1, 2, 8, 11 and possibly 12 more
than one of the cells had been re-used (previously re-used cells contain two cocoons, one within
the other). As it appears that only one female works a nest, other emergees leaving to start new
nests, this would indicate that these nests had been re-used more than once and were therefore
probably in their third year of use.

The cocoon is firmly attached to the cell walls except at the outer end which is capped with
meconium, this end is separated from the cell closure by a space about 3 mm long.

Associated insects

Three of the nests investigated each contained a cocoon with a small opening in the
meconium plug suggesting the emergence of parasites (Table 1). It was not possible to establish
what these might have been.

Group 7.

C. hispanicus Dusmet, C. moroccanus (G. Soika). C. spiricornis Saussure, C. beaumonti
(G. Soika), C. lusitanicus King (Richards, 1963, foraging record only), C. vechti Richards
(Richards, 1963, foraging record only), C. bischoffi Richards (Richards, 1963) and C. tuberculifer
Saussure (Giraud, 1871 and Ferton, 1901)

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret,
c-g. ?

h. A constructed mud-cell within an excavated-cell.

i. Leguminosae, Umbelliferae and Labiatae.

Group 8.

C. bicolor (Thunberg) (Gess and Gess, 1986), C. socius Turner (Gess and Gess, 1986 and
present paper), C. linearis King and C. capicola Brauns (Gess and Gess, 1980)

a. Nest excavated in non-friable soil.

b. Burrow surmounted by a mud turret.

c. Nest annual.

d. Nest with a relatively long main shaft.

e. First cell terminating the main shaft.

f. Succeeding cells terminating relatively long secondary shafts.

g. Secondary shafts sloping.

h. NO constructed mud-cell within an excavated-cell.

i. Forage plants, creeping and shrubby species of Mesembryanthemaceae.

Ceramius socius Turner

Geographic distribution

Ceramius socius is a western Cape species having been recorded from: the Clanwilliam
district; Worcester; the Touws River district; Constable; Matjesfontein; Robertson; Montagu; a
site 45 km east of Montagu; and Swellendam (Richards 1962; Gess, 1965 and 1968; Gess and
Gess, 1986). The present observations were made at Kransvlei in the Clanwilliam district.

GESS & GESS: KNOWLEDGE OE CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Fig. 11. Favoured water collecting point of C. socius Turner, wet sand near edge of dam, Kransvlei, 7-14. x. 1987.

Fig. 12. Aggregation of C. socius Turner at watering point, Kransvlei, 7-14. x. 1987, Actual length of females 13,5 mm.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988
Flowers visited

Gess and Gess (1986) recorded Ceramius socius males and females visiting Psilocaulon
acutisepalum (Berger) N. E. Br. (Mesembryanthemaceae) at Kransvlei in the Clanwilliam
district. During the period 7-14. x. 1987 plants in flower at various sites in the Clanwilliam district
were sampled for wasp visitors. C. socius was again found to be foraging only on flowers of
Psilocaulon acutisepalum and this only at Kransvlei in the vicinity of the nesting site.

Provision

Provision was obtained from 10 cells of C. socius. In all instances it was in the form of a firm
“pollen loaf” approximately 7 mm long and 4 mm in diameter. Pollen from the provision was
compared with that of Psilocaulon acutisepalum and found to be identical.

Water collection

Large numbers of females were observed to aggregate at one favoured spot on wet sand
near the edge of the dam at Kransvlei (Figs 11 and 12). The same spot and no other was
patronized during the period of seven days during which observations were made. The sand
became quite pitted from the wasps’ activities. Water for crop filling was extracted from the wet
sand.

Description of the nesting area

The nesting areas recorded by Gess and Gess (1986) were revisited by the authors in early
October 1987. There was no sign of the activity of C. socius on the Pakhuis Pass probably due
to the lack of water. The nesting site previously in use at Kransvlei had been abandoned in
favour of an extensive bare area on the opposite side of the road. This bare area had been
created during the construction of a dam and a farm road. The nests were in close proximity to
each other and extended over an area of approximately 50 paces by 3 paces. The nearest distance
between the dam and the nesting area was 35 paces.

Male behaviour

During the period 7-14. x. 1987 males of C. socius were common at Kransvlei both in the
nesting area and at the dam where the females were aggregating on the mud at a point some little
distance from the water’s edge. In the mornings the males were the first to appear in the nesting
area where they waited for the females which, about an hour later, appeared from their
nestswhere they had spent the night. Shortly after they emerge in the mornings the females
aggregate, a hundred or more at a time, at their selected point on the mud. The males join them
at this “swarming” point and there coupling takes place (Fig. 13). A male having gained a firm
hold on a female the pair flies off together. No interactions were observed between males and
females either in the nesting area or at the flowers on which, as mentioned, both sexes forage.
On a fine day swarming continues until the late afternoon.

Description of the nest

Gess and Gess (1986) were unable to make a full description of the nest as the only nests
located by them at the time though turreted had not yet reached the stage of cell excavation. A
full description is now possible.

The nest (Fig. 14) consists of a multi-cellular burrow surmounted by a curved tubular mud
turret, slightly flared above at the opening. The turret is constructed of mud pellets, rough on

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Fig. 13. Coupling male and female C. socius Turner at watering point, Kransvlei, 7-14. x. 1987. Actual length of females

13,5 mm.

the outside and smoothed on the inside. The turret and shaft opening are of equal diameters,.
The main shaft, of the same diameter as the entrance, descends sub-vertically before curving to
one side to end in a downward sloping to sub-horizontal cell. The main shaft is dilated for a short
distance at approximately half its depth to form a “bulb”. Secondary shafts diverge from the
main shaft at the level at which it departs from the sub-vertical. Each secondary shaft ends in an
excavated cell. Secondary shafts between the completed cell and the main shaft are filled withsoil
nd sealed off from the main shaft. Mud-cells are not constructed.

Method of construction of the nest, oviposition and provisioning.

No evidence was found of re-use of nests. A female at the outset of nesting initiates a
newnest. For this she requires water. At an early stage in shaft excavation turret construction
commences. Pellets are laid down around the shaft opening in such a way that the turret will
have the same inner diameter as that of the shaft, that is 4,5-5 mm (average of 4,5 mm, sample
of 7). The external diameter of the turret is 6,5-8 mm (average 7,2 mm, sample of 8) and the
height to which it rises before curving over and downwards is 8-11,5 mm (average 8 mm, sample
of 8). After turret construction has been completed further pellets extracted from the excavation
are discarded in a pellet dropping area at a short distance from the nest.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Fig. 14. Vertical plans of turrets and underground workings of C. socius Turner (x 1). Kransvlei, 7-14. x. 1987. For key

to lettering see Table 2.

In the sample of seven nests investigated shaft sinking had been continued downwards with
the diameter maintained at 4,5 mm for 17-44 mm after which it had been increased to 6-7,5 mm
and reduced again to 4,5 mm in such a way that a short “bulb” 10-13 mm long had been created.
Beneath the bulb the shaft had been continued downwards sub-vertically for 29-46 mm.

Below this depth the shaft curves away from the vertical, slopes gently downwards and ends
in an ovoid excavated-cell 11-14 mm (average 12,8 mm, sample of 9) in diameter at mid-length.
The walls of the cell are carefully smoothed and stabilised. No mud-cell is constructed within the
excavated-cell.

GESS & GESS: KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA; MASARIDAE)

After a cell has been excavated oviposition takes place. The egg which is laid loosely in the
cell is white, curved and 4-4,5 mm (average 4,1 mm, sample of 5) from tip to tip.

After oviposition has taken place provisioning commences. Successive loads of pollen and
nectar are collected in the crop, regurgitated and formed into a firm pollen loaf approximately
7 mm long and 4 mm in diameter.

After provisioning has been completed the sloping shaft between the cell and the vertical
shaft is filled with soil and sealed off.

Secondary shafts each terminating in a cell are prepared in a similar fashion. All the
secondary shafts leave the main shaft at approximately the same depth, however, they radiate
out in different directions so that the cells form a whorl rather than a group.

Seven nests were investigated (Table 2). Two had not yet reached the stage of cell
excavation, two were one-celled, two four-celled and the seventh six-celled.

Table 2

Details pertaining to seven nests of Ceramius socius Turner investigated at
Kransvlei, Clanwilliam district, from 7-14. x. 1987.

Nest

No.

Nest Status

Turret

No. of
cells

Nature of each cell, cells contents

New nest

Present

F F* G* F H F

New nest

Present

G H F G

New nest

Present

New nest

Present

—

New nest

Present

—

New nest

Present

New nest

Present

D F H D

Key: * neither egg nor larva found.

D. Cell closed, containing mature larva spinning cocoon.

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

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

H. Cell open, containing egg without provision.

I. Cell open, empty.

DISCUSSION

A fresh consideration of the ethology of the genus Ceramius as a whole seems appropriate.

The chosen nesting substrate for representatives of all species groups with the exception of
Group 2b is now known and is without exception non-friable to relatively non-friable soil, in all
instances with a high enough clay content to make it malleable when mixed with water. The
choice of nesting substrate by Ceramius seems to be that favoured by most ground nesting
masarids for which it has been established, Jugurtia confusa Richards (Gess and Gess, 1980),
Masarina familiaris Richards (Gess and Gess, 1988a), and Paragia tricolor Smith (Houston,
1984). The only species known to nest in friable soil is Quartinioides sp. which was found in 1985
nesting in friable beach sand at McDougal Bay, Port Nolloth, Namaqualand (Gess and Gess,
unpublished field notes).

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

Ceramius appears to show a preference for horizontally presented soil though some species,
C. lichtensteinii and C. jacoti, will nest in sloping ground, though in the case of the latter the nests
were on horizontal “ledges” on the sloping bank. Ceramius species have never been found
nesting in vertically presented soil, in marked contrast to M. familiaris which has been found
nesting only in vertically presented soil. The known choice of other species of ground nesting
masarids seems to be similar to that of Ceramius. J. confusa has been recorded nesting in
horizontally presented ground and P. tricolor in gently sloping ground.

The construction of a mud turret surmounting the burrow entrance is common to all species
of masarids nesting in non-friable soil for which nesting is known. All of these species construct
cylindrical turrets of the same diameter as that of the shaft. In the turrets of Ceramius there is
some variation in the degree of curvature and in the degree to which the interstices between
pellets are closed. There are, however, no striking differences in architecture between species
such as have been described for Dichragenia (Pompilidae) (Gess and Gess, 1976), Bembecinus
(Sphecidae) (Gess and Gess, 1975) and Parachilus (Eumenidae) (Gess and Gess, 1976a and
1988b). The diameter of the turret of Ceramius species is in proportion to the size of the wasp
(Gess and Gess, 1980) but is not constant for a species due to size variation between different
populations of a species as has been shown for C. lichtensteinii (Gess and Gess, present paper).

To date perennial nests have been recorded for most species which construct mud-cells
within excavated-cells but for none of the species which excavates cells but does not construct
mud-cells within them.

Attention has been drawn (Gess and Gess, 1986) to the correlation between the nature of
the cell and the length of secondary shafts. Based on a knowledge of the nesting of one species
each from Group 3 and Group 5 and three species from Group 8 the observation was made that
excavated cells in which mud-cells are not constructed (Group 8) seem to terminate long
secondary shafts and cells in which mud-cells are constructed (Group 3 and Group 5) seem to
terminate extremely short secondary shafts. The present observations concerning the nesting of
C. socius (Group 8) which excavates relatively long secondary shafts each terminating in an
excavated-cell in which a mud-cell is not constructed and of C. cerceriformis (Group 2) and C.
rex (Group 6) both of which excavate relatively short secondary shafts each terminating in an
excavated-cell in which a mud-cell is constructed add data to support the idea of the existence
of a correlation.

The length of and arrangement of secondary shafts and cells combined with the nature of
the turning “bulb”, in fact the architecture of the burrow itself rather than that of the turret,
characterize the species groups, there being considerable constancy of burrow plan within
groups and variation between groups. Within the limitations of present knowledge it appears
that Group 8 is distinct in having long secondary shafts and in the absence of a constructed cell
within the excavated-cell. Group 5 is distinct in that the main shaft is not terminated by the first
cell. Group 3 is distinct in that the cells are all excavated sub-vertically beneath the bulb. Group
2a and Group 6 are similar in that all the cells lie sub-horizontally in a group to one side of the
lower end of the main shaft, the first cell terminating that shaft, however, they are distinguished
by the nature of the “bulb”. The bulb in Group 2a is, when present, short and at mid-length of
the main shaft whereas that of Group 6 is relatively long and is positioned at or near the base
of the main shaft.

Although the genus Ceramius has been shown to forage on several families of plants,
Mesembryanthemaceae, Compositae and Leguminosae in the Afrotropical Region and

GESS & GESS; KNOWLEDGE OF CERAMIUS LATREILLE (HYMENOPTERA: MASARIDAE)

Leguminosae, Umbelliferae, Labiatae, Resedaceae and Plumbaginaceae in the Palaearctic Re-
gion , a considerable degree of oligolecty is exhibited at the species group and species levels . The four
species of Group 3 have been recorded foraging and the provision of two of these species has been in-
vestigated. In all instances the flowers utilized were of the family Compositae only. C. rex one of the
three species of Group 6 is recorded foraging on and provisioning with pollen from flowers of Com-
positae only. The four species of Group 8 have been recorded foraging on and provisioning with pol-
len from flowers of Mesembryanthemaceae only. The single species of Group 5 has been found to
provision with pollen of Mesembryanthemaceae only. It is therefore of particular interest that the
presently suggested division of Group 2 on morphological grounds is supported by differences in for-
aging preferences , the two species of Group 2a having been recorded foraging only on Mesembryan-
themaceae and C. clypeatus, one of the two species of Group 2b, only on Leguminosae.

That C. fonscolombei and C. bischoffi have been recorded visiting flowers of more than one
family is at present not given serious consideration as there is as yet no record of the nature of the pol-
len utilized. Records of casual plant visiting can be misleading. Although C. lichtensteinii has been
collected on numerous occasions from flowers of “mesems” and samples of provision from cells
were found to contain “mesem” pollen only the authors made a single record of this species visiting
flowers of Blepharis sp. (Acanthaceae). It is therefore considered that “mesems” are the habitual
forage plants of C. lichtensteinii and that the recorded visit to Blepharis sp. was casual in nature. The
latter flowers are visited by a very wide range of insects , apparently for their nectar . This opinion is in
keeping with that expressed by Cooper (1952) in a consideration of records of flower visiting by
Pseudomasaris in North America.

ACKNOWLEDGEMENTS

The authors wish to thank Mr Daan Smit of Kransvlei for permission to work on his land
and Mr Klaas van Zyl of the Cape Department of Nature and Environmental Conservation for
permission to work in the Hester Malan Nature Reserve.

Mrs Sue Dean and Mr Richard Dean of the Karoo Biome Research Station at Tierberg are
thanked for their hospitality.

Thanks are due to Mr Harold Gess for assistance in the field in the Oudtshoorn district and
for Figs 7 and 8; to Mr Robert Gess for assistance in the field in the Oudtshoorn district and at
Tierberg; to Mr Michael Struck for assistance in locating a good study area in the Hester Malan
Nature Reserve; and to Mrs Estelle Brink of the Albany Museum Herbarium and Mrs Sue Dean
for assistance with the identification of some of the forage plants.

Gratitude to the C.S.I.R. is expressed by F. W. Gess for running expenses grants (Main
Research Support Programme and National Programme for Ecosystem Research) for field work
during the course of which the present studies were made.

REFERENCES

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

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

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

Cooper, K. W. 1952. Records and flower preferences of masarid wasps. II. Politrophy or oligotrophy in Pseudomasarisl
(Flymenoptera: Vespidae). Amer. Midland Nat. 48; 103-liO.

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

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

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 1, AUGUST 1988

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

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

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

Gess, F. W. and Gess, S. K. 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.

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. 1980. 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. 1986. Ethological notes on Ceramius bicolor (Thunberg), C. clypeatus Richards,
C. nigripennis Saussure and C. socius Turner (Hymenoptera; Masaridae) in the Western Cape Province of South
Africa. Ann. Cape Prov. Mus. (nat. Hist.) 16 (7); 161-178.

Gess, F. W. and Gess, S. K. 1988a. A contribution to the knowledge of the taxonomy and the ethology of the genus
Masarina Richards (Hymenoptera; Masaridae). Ann. Cape Prov. Mus. (nat. Hist.) 16 (14); 351-362.

Gess, F. W. and Gess, S. K. 1988b. A contribution to the knowledge of the ethology of the genera Parachilus Giordani
Soika and Paravespa Radoszkowski (Hymenoptera; Eumenidae). Ann. Cape Prov. Mus. (nat. Hist.) 18 (3). In
press.

Houston, T. F. 1984. Bionomics of a pollen-collecting wasp, Paragia tricolor (Hymenoptera; Vespidae; Masarinae), in
Western Australia. Rec. West. Aust. Mus. 11 (2); 141-151.

Moll, E. J. , Campbell, B. M. , Cowling, R. M. , Bossi, L. , Jarman, M. L. and Boucher, C. 1984. A description of major
vegetation categories in and adjacent to the fynbos biome. South African National Scientific Programmes Report
83; i-vi. 1-29.

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

Richards, O. W. 1963. New species of Ceramius Latreille (Hymenoptera, Vespoidea) allied to Ceramius lusitanicus
Klug. Zodl. Meded., Leiden 38 (13); 213-220.

(

«i*4

ISSN 0570-1880

Annals of the_

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 2 31st August 1988

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 connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

Selected annotated bibliography of early developmental studies of African

freshwater fishes

J.A. CAMBRAY

(Albany Museum, Grahamstown. 6140. South Africa)
and

G.G. TEUGELS

(Ichtyologie Generale et Appliquee. Musee National d'Histoire Naturelle. 75231. Paris, France)

INTRODUCTION

The Class Osteichthyes (bony hshes). with over 20 000 known species, has more anatomical
diversity amongst the young than any other vertebrate class (Faber, 1985). On the African
continent 74 families and some 2 000 freshwater hsh are represented. Relatively little is known
about the early life histories of these species.

Over the last decade there has been in many countries an increasing interest in early life
history studies. However, for the African continent the literature is widely scattered and some
good manuscripts remain unpublished. The aim of the present paper is to provide a selected
annotated bibliography, which will clearly point out this important gap m the knowledge of the
African ichthyofauna. This annotated list of references is selective. The papers or manuscripts
which are included are on early development sensu stricto, where the identity of the species is
substantiated. Those which include only growth rates or age determinations and those concerned
with early stages solely for toxicity studies have not been included. As with any bibliography,
there may be omissions, especially of unpublished manuscripts. It is planned to update this
bibliography, possibly in hve years time and workers are therefore encouraged to assist by letting
the authors know of any omissions and future publications.

It is unfortunate that in all but a few papers (7 out of 153) mention is not made of the
repository of the developmental series on which the authors have worked. The Department of
Ichthyology at the Albany Museum has initiated a special collection of African freshwater fish
early life history stages. Reference collections would be most welcome. They would be
accessioned into the main collection. Many curators of adult fish collections are wary of
microscopic specimens stored in vials. Their maintenance and documentation are labour
intensive (Cohen, 1984). These early life history collections should be sent to organizations that
have an institutional (not personal) commitment to long-term collection storage. The authors

ANN, CAPE PROV. MUS. (NAT. HIST.) VOL. IS, PT. 2. AUGUST 1988

have tried to trace some of the collections, but have had little success. It is desirable that an
author include in his paper reference to the institution in which his collection has been lodged.

Early life history terminology can be confusing. In order to prevent further confusion, the
terms used in most of the present annotations are the same as those used in the papers to which
they refer. This point is noted to make readers aware of the presence of and reason for the
inconsistencies in the use of terms in the annotations. It is recommended that in future workers
should dehne clearly their use of early life history terminology or state which existing
terminology they have used. For examples, see Balon (1975). Kendall et al. (1984) and Snvder
(1981).

In the check-list of the freshwater fishes of Africa (CLOFFA) (Daget el al., 1984 and 1986).
early life history studies are usually not separated and are included under the terms biology or
reproduction. The study of fish eggs, embryos and larvae sometimes forms an independent
branch of fisheries biology, and it is suggested that it be given a separate listing in books of this
nature, CLOFFA, gives a good indication of how little is known about the African ichthyofauna.
For example out of a total of 292 Barbus species the biology and/or ecology of only 29 species
are/is known. The early development of only nine species (<3%) has been described. For the 81
African Labeo species a similar picture emerges. The biology and/or ecology of 19 out of 81
species are/is known and only four early life history studies have been completed. Possibly one
of the reasons why so few papers have been published on the early development of African
freshwater fishes is that the eggs of freshwater fishes, unlike adult fishes, are rather difficult to
collect. Once the developing eggs are collected, equipment and time are required to rear, collect
and preserve a well documented developmental series. In contrast to the pelagic eggs of many
sea fishes, many of the eggs of freshwater fishes are negatively buoyant and sink to the
substratum, or are deposited in crevices, or are attached to plants or rocks by an adhesive
membrane, or occur in the mouth of one parent. Many of the papers discuss development of eggs
obtained by artificial reproduction of species important for aquaculture, and rarely is
development followed under natural conditions, usually because of logistics and lack of
equipment.

It has only been possible to locate early developmental studies for 18 of the 74 fish families
in Africa. For phylogenetic studies the early stages offer a whole suite of characters not present
in the adults, and further studies of the early period of development will add to the knowledge
of fish interrelationships. Even within one species, the essence of development is change, and
the characters that identify one stage will often not even exist in a later stage, eg. Kupffer's
vesicle, finfolds, etc.

This bibliography should also be useful for those doing ecological studies in Africa. Early
life and juvenile stages of fishes are considered sensitive indicators of ecological perturbations,
such as pesticides, radioactive wastes, urban and industrial pollution and entrainment problems
in power plants (Alderdice, 1985). In addition the lack of understanding of the early life history
phase is still a major impediment to fish culture (Hempel, 1979), similarly to the management
of natural stocks. An understanding of the ecology of the early life history stages of fishes is
essential for answering certain practical questions with regard to exploitation of fish populations.

In the format for each reference author(s), date, title and source are included when
available. In the annotation a brief abstract is given, the presence of photomicrographs or
camera-lucida illustrations is noted and museum collection numbers are cited when possible.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STGDIES OF AFRICAN FRESHWATER FISHES

REFERENCES FOR INTRODUCTION

Alderdice. D. F. 1985. A pragmatic view of early life history studies of fishes. Trans. Am. Fish. Soc. 114: 445-451.
Balon. E. K. 1975. Terminology of intervals in fish development. J. Fish. Res. Bd Can. 32: 1663-1670.

Cohen. D. M. 1984. Ontogeny. Systematics. and Phylogeny. In: Moser. H. G. ed. Ontogeny and Systeniatics of Fishes.

Special Publication. Am. Soc. of Ichthyologists and Herpetologists. Lawrence: Allen, pp. 7-11.

Daget. J.. Gosse. J. P. and Thys van den Aldenaerde. D. F. E. 1984. Check-list of the freshwater fishes of Africa. Vol.
1. ORSTOM (PARIS). MRAC (TERVUREN). 41Upp.

Daget. J.. Gosse. J. P. and Titvs \an den Aldenaerde. D. F. E. 1986. Check-list of the freshwater fishes of Africa. Vol.

2 ISNB (Brussels). MRAC (Tervuren). ORSTOM (Paris) 52Hpp.

Faber. D. J. 1985. Book review. ELHS. A.F.S. newsletter 6{1)'. 18.

Hempel. G. 1979. Early life history of marine fish. Seattle: Univ. of Washington Press. 70pp.

Kendall. A. W. jr.. Ahlstrom. E. H. and Moser. H. G. 1984. Early life history stages of fishes and their characters. In:
Moser. H. G. ed. Ontogeny and systematics of fishes. Special Publication. .Am. Soc. ol Ichthyologists and
Herpetologists. Lawrence: .Allen, pp. 11-22.

Snyder. D. E. 1981. Contributions to a guide to the cypriniform fish larvae of the upper Colorado River System. U.S. Dep.
Inter. Bur. Land and Manage. Colo. Off. Biol. Sci Ser. 3. 81pp.

SELECTED PAPERS BY FAMILY
PROTOPTERIDAE

Agar, W. E. 1906. The development of the skull and visceral arches in Lepidosiren and
Protopteriis. Trans. R. Soc. Edinb. 45: 49-64. 16 figs. 3 pis.

The development of the skull and the visceral arches in Protopteriis sp. (= Protopteriis
annectens) from the Gambia River is described and illustrated with camera lucida drawings.

Agar, W. E. 1907. The development of the anterior mesoderm and paired fins with their nerves
in Lepidosiren and Protopteriis. Trans. R. Soc. Edinb. 45: 611-639.

Protopteriis sp. (= Protopteriis annectens) development is described as per title and
illustrated with camera lucida drawings.

Brien, P. 1959. Ethologie du Protopteriis dolloi (Boulenger) et de ses larves. Signification des
sacs pulmonaires des Dipneustes. Annls Soc. r. zool. Belg. 89: 9-48. 10 figs.

Eggs, embryos and larvae, from the hatching stage (19.6 mm) up to the fifteen day old
swimming larvae (25,1 mm) of Protopteriis dolloi. are drawn, described and compared to
Protopteriis annectens. The ontogeny of respiratory organs is included.

Brien, P. and Bouillon, J. 1959. Ethologie des larves de Protopteriis dolloi et etudes de leurs
organes respiratoires. Resultats scientifiques d’une mission zoologique CEMUBAC au
Stanley Pool, subsidiee par I'ULB et le MRCB. Annls Miis. r. Congo beige Ser. 8(71):
25-74. pis 7-11, 38 figs.

The development of Protopteriis dolloi f-rom the Malebo Pool (Zaire) is described and
compared to that of P. annectens. from the hatched larvae (19,6 mm) up to the 25,1 mm larva.
Illustrations are as in Brien, 1959.

Budgett, j. S. 1900. On the breeding habits of some West African fishes with an account of the
external features in development of Protopteriis annectens and a description of the larva of
Polypteriis lapradei. Proc. zool. Soc. Lond.: 835-6. Also in: Kerr, G. ed.. Budgett
Memorial Volume. 1907. Cambridge: Univ. Press.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 2. AUGUST 1988

Budgett, J. S. 1901a. On the breeding habits of some West African fishes, with an account of the
external features in development of Protoptems annectens and a description of the larva of
Polypterus lapradei. Trans, zool. Soc. Land. 16(2); 115-132, 5 hgs, 2 pis.

External features in the development of wild collected Protoptems annectens from the
Gambia River are illustrated using graphical reconstructions from the cleavage egg up to the
small adult with brief descriptions. Comparison of development of Protoptems with Lepidosiren
is included.

Budgett, J. S. 1901b. The habits and development of some West African fishes. Proc. Camb.
phil. Soc. 11: 102-104. Also in: Kerr, G. ed., Budgett Memorial Volume, 1907, Cambridge:
Univ. Press, pp. 141-142.

The eggs and larvae of Protopterus annectens from the Gambia River are briefly described.

Kerr, J. G. 1901. The development of Lepidosiren paradoxa, with a note upon the corresponding
stages in the development of Protopterus annectens (Protopteridae). Q. Jl. microsc. Sci.
2(45): 1-40, 4 pis.

Not seen.

Kerr, J. G. 1909. Normal plates of the development of Lepidosiren parado.xa and Protopterus
annectens. Norm Taf. EntwGesch. Wirbeltiere 3: 1-31.

Not seen.

Kerr, J. G. 1910. On certain features in the development of the alimentary canal in Lepidosiren
and Protopterus. Q. Jl microsc. Sci. 54: 483-518.

Not seen.

Pasteels, j. 1962. Resultats scientifiques des Missions zoologiques au Stanley Pool subsidiees par
le CEMUBAC (Universite Libre de Bruxelles) et le Musee Royal du Congo Beige
(1957-1958). XII. Gastrulation du Protoptems dolloi. Annls Mus. r. Afr. centr. Ser. 8(108):
173-183, 3 pis.

The gastrulation in Protopterus dolloi from the Malebo Pool in Zaire, is described in detail
and illustrated with camera lucida drawings and photographs.

SvENssoN, G. S. O. 1933. Freshwater fishes from the Gambia River (British West Africa): Results
of the Swedish expedition 1931. K. svenska VetenskAkad. Handl. 12(3): 1-102, 28 figs,
8 pis.

General data on eggs and larvae of Protopterus annectens are given.

POLYPTERIDAE

Abdel-aziz, I. A. 1957. Notes on the anatomy oi larva) Polvpterus. Proc. Egypt. Acad. Sci. 12:
79-87, 12 figs.

CAMBRAY & TEUGELS; BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

Anatomical features of larval Polypterus senegalus from Bahr El-Zaraf, Upper Nile
Province, Sudan, using serial sections and graphic reconstructions on digestive and urino-genital
systems are described. Illustrations of transverse sections are included.

Arnoult, J. 1962. Ponte naturelle suivie d’eclosions chez Polypterus senegalus senegalus
(Cuvier). C.r. Seanc. Acad. Sci. (Paris) 254: 2828-2829.

After Polypterus senegalus senegalus spawned in the Upper Volta, the eggs and early larval
stages were collected and described. There are no illustrations.

Arnoult, J. 1964. Comportement et reproduction en captivite de Polypterus senegalus (Cuvier).
Acta zooL, Stockh. 46: 191-199, 3 figs.

The development of Polypterus senegalus collected from the Upper Volta basin and reared
in an aquarium was followed. Photographs of the 41 hour old fertilized egg, the 54 hour old
larvae and the 20 day old juvenile are included.

Budgett, j. S. 1900. On the breeding habits of some West African fishes with an account of the
external features in development of Protopterus annectens and a description of the larva of
Polypterus lapradei. Proc. zool. Soc. Lond: 835-6.

Larvae of Polypterus lapradei (= Polypterus senegalus) are described.

Budgett, J. S. 1901a. On the breeding habits of some West African fishes, with an account of the
external features in development of Protopterus annectens and a description of the larva of
Polypterus lapradei. Trans, zool. Soc. Lond. 16(2): 115-132, 5 figs, 2 pis.

The larvae of Polypterus lapradei (= Polypterus senegalus) are described and illustrated
with drawings.

Budgett, J. S. 1901b. The habits and development of some West African fishes. Proc. Camb.
phil. Soc. 11: 102-104. Also in: Kerr, G. ed. Budgett Memorial Volume — 1907. Cambridge:
Univ. Press, pp. 141-142.

A larva of Polypterus, 30 mm length, from the Gambia River is described.

Budgett, J. S. 1902a. On the structure of larval Polypterus. Trans. Zool. Soc. Lond. 16(7):
315-340, 3 pis, 5 figs. Also in: Kerr, G. ed. Budgett Memorial Volume — 1907. Cambridge:
Univ. Press, pp. 141-142.

Transverse sections of a larva of Polypterus (30 mm) from the Gambia River, double
stained with haemotoxyline and eosine and drawn from reconstructions by Kerr’s method
(camera lucida) are described. An account is given of the structure of the cartilaginous skeleton
and subsequent development as studied in a number of specimens from 9 cm upwards. The
genital and excretory system in this larva are described.

Budgett, J. S. 1902b. On the anatomy of larval Polypterus. Rep. Br. Ass. Advmt Sci. 71: 693.

The osteology and the anatomy of a larval specimen of Polypterus (30 mm length) from the
Gambia River are briefly described. No illustrations are included.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 2, AUGUST 1988

Daget, J., Bauchot, M. L. and R. and Arnoult, J. 1964. Developpement du chondrocrane et
des arcs aortiques chez Polyptenis senegalus Cuvier. Acta zool., Stockh. 45: 201-244,
34 figs.

Organogenesis of the chondrocranium and the aortic arches in Polypterus senegalus from
4,5 mm up to 22,5 mm from Upper Volta is described. Graphic reconstructions based on
photographs are included in this paper.

Guitel, F. 1914. Contribution a I’etude d’une larve de Polvpterus senegalus Cuvier de 59 mm de
longueur. Archs Zool. exp. gen. 54(12): 411-437, 12 figs, 2 pis.

A 59 mm larvae of Polypterus senegalus from Lake Chad is described in detail (external
morphology) and compared to other species of the genus. Illustrations are included.

JoHNELS, A. G. 1954. Notes on fishes from the Gambia River. Ark. Zool. 6(17); 326-411, 19 figs.

The development of Polypterus senegalus{l) is briefly described and two photographs of a
10 mm TL larva are included.

JuNGERSEN, H. F. E. 1895. Ueber die Embryonalniere von Calamoichthys calabaricus (Smith)
von J. Lebedinsky. Zool. Zentbl. 2: 54-59.

Not seen.

Kerr, J. G. 1904. Note on the developmental material of Polypterus (Polypteridae) obtained by
the late Mr. J. S. Budgett. Rep. Br. Ass. Advmt Sci. 74: 604-605.

Not seen.

Kerr, J. G. 1907a. The development of Polypterus senegalus Cuv. In: Kerr, G. ed. Budgett
Memorial Volume — 1907. Cambridge: Univ. Press, pp. 195-284.

Developmental material of Polypterus senegalus, collected by Budgett, is investigated by
means of paraffin sections. Drawings and descriptions of the external development of early
cleavage eggs, embryos and larvae are given. Detailed internal features of development are
described and drawn, from gastrulation to larval fish, including development of mesoderm,
notochord, alimentary canal, cement organs, excretory organs, vascular system, cartilaginous
skeleton and the nervous system.

Kerr, J. G. 1907b. The development of Polypterus senegalus Cuv. Proc. R. phys. Soc. Edinb.
17(2): 73-5.

Not seen.

Nieuwenhuys, R., Bauchot, R. and Arnoult, J. 1969. Le developpement du telencephale d’un
poisson osseux primitif, Polypterus senegalus Cuvier. Acta zool., Stockh. 50: 101-125,
29 figs.

The development of the forebrain in Polypterus senegalus has been studied from the 5,5 mm
larvae up to a 28 mm juvenile. Photographs of transverse and horizontal sections are given for
the different stages examined. The forebrain in adult specimens of the same species is described.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV, STUDIES OE AERICAN FRESHWATER FISHES

SvENssoN, G. S. O. 1933. Fresh-water fishes from the Gambia River (British West Africa): results
of the Swedish expedition 1931. K. svenska VetenskAkad. Handl. 12(3): 1-102, 28 figs,
8 pis.

General data on eggs and larvae of Polypteriis lapradei (= Polyptems bichir lapradei) and
Polypterus senegalus are given.

CLUPEIDAE

Aigner, T. and Laurent, M. 1976. Note sur les alevins de Stolothrissa tanganicae et de
Limnothrissa miodon au Burundi. FAO, Fish. Res. Rep. BD1I73I020, working paper nr. 9.

Not seen.

Albaret, J. J. and Gerlotto, F. 1976. Biologie de I’ethmalose {Ethmalosa fimbriata Bowdich) en
Cote d’Ivoire. I. Description de la reproduction et des premiers stades larvaires. Doc.
scient. Cent. Rech. oceanogr. Abidjan. 7(1): 113-133, 12 figs.

The larval development of Ethmalosa fimbriata from Ebrie Lagoon, Ivory Coast, from the
unfertilized egg up to the vitelline vesicule resorption stage is described in detail. This paper
includes drawings and photographs.

Bainbridge, V. 1957. Eggs and larvae of Ethmalosa dorsalis. Cons. Scient. Afr., Symp. Luanda
1957.

The eggs and larvae of Ethmalosa dorsalis {— Ethmalosa fimbriata) are briefly described.

Bainbridge, V. 1960. The early life history of the Bonga, Ethmalosa dorsalis (Cuvier &
Valenciennes). J. Cons. Perm. Int. Expl. Mcr 26(1): 347-353, 2 figs.

The development of the eggs and larvae of Ethmalosa dorsalis (= Ethmalosa fimbriata)
from the Sierra Leone River, Sierra Leone, is described. The following stages are drawn: egg
two hours and fifteen hours after fertilization, newly hatched larva, yolk-sac larva twenty four
hours after hatching, larva 60 hours after hatching.

Bainbridge, V. 1962. The larvae of Pellonula vorax Gunther (Clupeidae) in Sierra Leone coastal
waters. Bull. Inst. fr. Afr. noire (A) 24 (1): 262-269, 4 figs.

Larvae of Pellonula vorax from the Sierra Leone River, Sierra Leone, covering the
complete size range (6 to 17,9 mm SL) are described externally. Camera lucida drawings are
included.

CoNAND, F. 1977. Oeufs et larves de la sardinelle ronde {Sardinella atirita) au Senegal:
distribution, croissance, mortalite, variations d’abondance de 1971 a 1976. Cah. ORSTOM,
oceanogr. 15(3): 201-214, 13 figs.

Six to 12 mm larvae of Sardinella aurita were collected using Bongo nets off the Senegalese
coast. These stages are briefly described.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 2, AUGUST 1988

CoNAND, F. and Fagetti, E. 1971. Description et distribution saisonniere des larves de Sardinelles
des cotes du Senegal et de la Gambie en 1968 et 1969. Cah. ORSTOM, oceanogr. 9(3):
293-318, 17 figs.

Morphological and biometrical observations on 5 to 15 mm larvae of Sardinella aurita and
S. eba (= S. maderensis) from the Senegalese coasts are given. For both species drawings of the
larval development are included (larvae smaller than 6 mm; larvae between 6 and 11 mm; larvae
between 11 and 15 mm; larvae between 16 and 20 mm).

Dessier, a. 1969. Note sur les stades larvaires et post larvaires d'llisha africana (Bloch, 1795)
(Pisces; Clupeidae). Cah. ORSTOM, oceanogr. 7(4): 21-25, 3 figs, 2 pis.

This paper includes descriptions and drawings of larval (5,7 up to 12,4 mm SL), postlarval
(15,5 up to 21,8 mm SL) and juvenile stages (31,5 mm SL) otilisha africana caught off the coasts
of Congo and Gabon.

JoHNELS, A. G. 1954. Notes on fishes from the Gambia River. Ark. Zool. 6(17): 326-411, 19 figs.
Biometrical features of the larvae of Pellonula afzeliusi, up to 30 mm TL, are given.

Marchal, E. G. 1967. Cle provisoire de determination des oeufs et larves des Clupeides et
Engraulides Guest- Africains. ORSTOM, Docums scient. Provis. 14, 4 pp., 10 figs.

This paper provides provisionary identification keys for Clupeidae and Engraulidae eggs
and larvae from the coasts of Senegal to Congo. Drawings of the egg and the 5,2, 9,2, 19,6 and
23,9 mm larvae of Sardinella aurita, the egg and the 10,4 and 19,2 mm larvae of Sardinella eba
(= S. maderensis), the egg and the 5,1 mm larva of Harengula rouxi, the egg of Ethmalosa
fimbriata and Ilisha africana are included.

SvENssoN, G. S. O. 1933. Fresh-water fishes from the Gambia River (British West Africa):
Results of the Swedish expedition 1931. K. svenska VetenskAkad. Handl. 12(3): 1-102,
28 figs, 8 pis.

General data on eggs and larvae of Pellonula vorax (= Sierrathrissa leonensis -I- Pellonula
spp.) are given.

Teugels, G. G. and Thys van den Audenaerde, D. F. E. 1979. A morphological and anatomical
study of Pellonula afzeliusi Johnels, 1954 and Sierrathrissa leonensis Thys, 1969 (Pisces;
Clupeidae). Rev. zool. afr. 93(3): 523-538, 10 figs.

Morphological and osteological comparisons between growth series of Pellonula afzeliusi
(16,2-58,6 mm SL) and Sierrathrissa leonensis (9,8-30,4 mm SL) from West Africa are given.
Camera lucida drawings of alizarin stained specimens are included.

Whitehouse, R. H. 1933. Report on fish eggs and larvae taken during 1931. Notes Mem. Fishery
Res. Dir., Cairo 4: 1-22, 20 figs.

Larval stages in Sardinella aurita from the Nile (2,3 mm up to 6,5 mm) are described.
Drawings of the 3,2 mm, 4,6 mm and the 5,9 mm larvae are included.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV, STUDIES OE AERICAN FRESHWATER FISHES

OSTEOGLOSSIDAE

Assheton, R. 1907. Report upon sundry Teleostean eggs and larvae from the Gambia River. In:
Kerr, G. ed.. Budget! Memorial Volume — 1907. Cambridge: Univ. Press, pp. 433-442,
6 figs.

This paper includes data from eight larval stages and drawings of the caudal fin development
of Heterotis niloticus from the Gambia River. A drawing and description of a yolk-sac larva as
well as drawings of transverse sections of forebrain and epiphysis are included.

The larva, one day after hatching, of Heterotis niloticus is drawn and described.

Budgett, J. S. 1901b. The habits and development of some West African fishes. Proc. Carnb.
phil. Soc. 11: 102-104.

This paper includes a brief description of and data on eggs and larvae of Heterotis niloticus.

Daget, j. 1957. Memoires sur la biologic des poissons du Niger Moyen. III. Reproduction et
croissance (VHeterotis niloticus Ehrenberg. Bull. Inst. fr. Afr. noire 19(1): 295-323, 3 figs,
2 pis.

The development of the eggs and the larvae of Heterotis niloticus from the Central Niger is
described. This paper includes detailed discussions and camera lucida drawings from stage I
(larva just after hatching — 7,5 mm TL) up to stage X (larva 8 days after hatching — 14 mm TL).

Daget, J. and D’aubenton, F. 1957. Developpement et morphologic du crane d'Heterotis
niloticus. Bull. Inst. fr. Afr. noire 19(3): 881-936, 31 figs.

The cranial development (neurocranium, splanchnocranium and dermal cranium) and the
morphology of Heterotis niloticus from the Central Niger, from stage I (hatching — 7,5 mm TL)
up to stage X (8 days after hatching — 14 mm TL) using longitudinal serial sections, stained after
Mallory are described. Illustrations of sections are included.

Olaniyan, C. I. O. and Zwilling, K. K. 1963. The suitability of Heterotis niloticus Ehrenbaum
(Osteoglossidae) as a fish for cultivation: with a note on its spawning behaviour. Bull. Inst,
fr. Afr. noire 25(2): 513-525.

A short account is given of the spawning behaviour of Heterotis niloticus under fish pond
conditions in Panyam, Northern Nigeria. Reference is made to Daget (1957) for the first week
of development. No illustrations are included.

SvENssoN, G. S. O. 1933. Fresh-water fishes from the Gambia River (British West Africa):
Results of the Swedish expedition 1931. K. svenska VetenskAkad. Handl 12(3): 1-102, 28
figs, 8 pis.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOT. 18, PT. 2. AUGUST 1988

General data on eggs and larvae of Heterotis niloticus are described, including the drawings
of three stages, 12,5, 14,0 and 27,0 mm.

NOTOPTERIDAE

SvENSSON, G. S. O. 1933. Fresh-water fishes from the Gambia River {British West Africa):
Results of the Swedish expedition 1931. K. svenska VetenskAkad. Handl. 12(3): 1-102, 28
figs, 8 pis.

This paper includes general data on eggs and larvae of Notopterus afer (= Papyrocranus
afer).

MORMYRIDAE

Daget, J. 1958. Alevins de Mormyrus nime C. & V.. Bull. Soc. zool. Fr. 83(2-3): 200-204, figs
1-3.

A description of external morphological development of Mormyrus rume from the Central
Niger, based on ten larvae from 10 mm (just after the resorption of the yolk sac) up to 26 mm
TL is outlined. Seven specimens are illustrated.

Fritsch, G. T. 1892. On the origin of the electric nerves in the torpedo, Gymnotus, Mormyrus
and Malapterurus. Rep. Br. Ass. Advmt Sci.\ 757-788.

Not seen.

JoHNELs, A. G. 1954. Notes on fishes from the Gambia River. Ark. Zool. 6( 17): 326-411, 19 figs.

Biometrical features of several larval and juvenile stages of Mormyrops deliciosus (18,2 to
430 mm SL) are tabulated. Detailed descriptions and camera lucida drawings of developmental
stages (from stage I, being the newly hatched embryo — 4,5 mm SL up to stage VII, being the
nine day old larva — 6,2 mm SL) of Hyperopisus bebe are included.

SzABO, T. 1960. Development of the electric organ of Mormyridae. Nature, Lond. 188(4752):
760-762.

Not seen.

Westby, G. W. M. and Kirschbaum, F. 1977. Emergence and development of the electric organ
discharge in the mormyrid fish, Pollimyrus isidori. I. The larval discharge. /. comp, physiol.
122: 251-271.

The development of the electric organ discharge in the mormyrid fish, Pollimyrus isidori is
described. Laboratory bred larvae were monitored. A photograph of an eight day old specimen
is included.

GYMNARCHIDAE

Assheton, R. 1907. The development of Gymnarchus niloticus. In: Kerr, G. ed. Budgett
Memorial Volume — 1907. Cambridge: Univ. Press, pp. 293-432, 79 figs.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

The external morphology of several developmental stages of Gymnarchus niloticus from the
Gambia River is described in detail and is illustrated from the egg to the juvenile period (fortieth
day). The development of the alimentary canal, the vascular system, the reproductive organs,
the excretory organs, the nervous system and the skeleton is described, including drawings of
sagittal and transverse sections of various stages.

This paper includes descriptions and drawings of the larvae of Gymnarchus niloticus.
between two and six days after hatching.

Budgett, J. S. 1901. The habits and development of some West African fishes. Proc. Camb. phil.
Soc. 11: 102-104. Also In: Kerr, G. ed. Budgett Memorial Volume — 1907 Cambridge: Univ.
Press, pp. 119-136, 2 pis.

Brief descriptions of the eggs and larvae of Gymnarchus niloticus from the Gambia River
are given.

Dahlgren, U. 1910. The origin of electric tissues in fishes. Am. Nat. 44: 193-202.

The ontogeny of the electric organ in Gymnarchus niloticus is described from the ninth day
embryo to the stage between the twelfth and the fourteenth day. No illustrations are included.
These specimens were collected by Budgett.

Dahlgren, U. 1914a. Origin of the electric tissues of Gymnarchus niloticus. Pubis Carnegie Instn
183: 159-194.

Not seen.

Dahlgren, U. 1914b. Embryonic history of the electric apparatus in Gymnarchus niloticus. Proc.
int. Congr. Zool. 9: 379.

Not seen.

Srivastava, C. B. L. and Szabo, T. 1972. Development of electric organs of Gymnarchus
niloticus (Fam. Gymnarchidae). 1 Origin and histogenesis of electroplates. J. Morph.
138(3): 375-386.

The authors used a fairly complete series of embryonic stages of Gymnarchus niloticus
collected from Dia farabe, Mali. Material included recently hatched free embryos and following
stages to complete absorption of the yolk. A table is included which indicates the developmental
characters for 17 stages. The development of the electric organs was studied and the origin and
histogenesis of an electroplate is described. One photograph of a fry of approximately 3,5 days
after hatching is included. The paper also includes photomicrographs of sagittal and transverse
sections of the tail at various developmental stages.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 2, AUGUST 1988

Srivastava, C. B. L. and Szabo, T. 1973. Development of electric organs of Gymnarchus
niloticus (fam. Gymnarchidae). J. Morph. 140(4): 461-465.

The development of the electric organ spindles of Gymnarchus niloticus was investigated
with regard to the exact time and place of origin. Photomicrographs of transverse and sagittal
sections of the tail at several stages of development are included.

SvENssoN, G. 1933. Fresh-water fishes from the Gambia River (British West Africa): Results of
the Swedish expedition 1931. K. svenska VetenskAkad. Handl. 12(3): 1-102, 28 figs, 8 pis.

General data on eggs and larvae of Gymnarchus niloticus are included.

HEPSETIDAE

Bertmar, G. 1959. On the ontogeny of the chondral skull in Characidae, with a discussion on the
chondrocranial base and the visceral chondrocranium in fishes. Acta zooL, Stockh. 40(2-3):
203-364, 85 figs.

The ontogeny of the chondrocranium of Hepsetus odoe from the Gambia River is described,
based on a large collection of embryological material consisting of seventy stages from the egg
up to a 45 mm juvenile, using graphical reconstructions based on photographs. The results are
compared to those for Alestes nurse (Characidae), Citharinus citharus (Citharinidae),
Distichodus brevipinnis (Distichodontidae), Distichodus rostratus (Distichodontidae). Nann-
aethiops unitaeniatus (Distichodontidae), Nannocharax ansorgii (Distichodontidae), Protopterus
annectens (Protopteridae), Polypterus senegalus (Polypteridae) and Heterotis niloticus (Osteo-
glossidae). The material examined, is housed in the Zootomical Institute, University of
Stockholm, Sweden.

Bertmar, G. 1962. On the ontogeny and evolution of the arterial vascular system in the head of
the African Characidean fish Hepsetus odoe. Acta zool., Stockh. 43: 255-294, 12 figs.

As per title on 70 specimens (6,5-45,0 mm) of Hepsetus odoe from the Gambia River, using
graphic reconstructions based on photographs. The series are housed in the Zootomical
Institute, University of Stockholm, Sweden.

This paper includes descriptions and drawings of the larvae of Sarcodaces odoe (= Hepsetus
odoe) just after hatching.

Budgett, J. S. 1901b. The habits and development of some West African fishes. Proc. Camb.
phil. Soc. 11: 102-104.

Brief descriptions and data of the eggs and larvae of Sarcodaces odoe (= Hepsetus odoe)
from the Gambia River are given.

SvENssoN, G. S. O. 1933. Fresh-water fishes from the Gambia River (British West Africa):
Results of the Swedish expedition 1931. K. svenska. VetenskAkad. Handl. 12(3): 1-102, 28
figs, 8 pis.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

The foam nest, eggs and larval behaviour of Hydrocyonoid.es odoe (= Hepsetus odoe) are
briefly described. These stages are not illustrated, but the author agrees with larva pictured by
Budgett (Kerr, 1907, Plate IX), and large adhesive organ described by Budget! (1901).

CHARACIDAE

Abdel- Aziz, I. 1961. The chondrocranium of Hydrocyon forskalii larva (9 mm). II. Branchial
arches. Proc. Egypt. Acad. Sci. 15: 65-69.

Osteological features of branchial arch development in 6,5 mm and 9,0 mm larvae of
Hydrocyon forskalii (= Hydrocynus forskalii) from the Nile are described. This paper is not
illustrated.

Bowmaker, a. P. M. 1973. An hydrobiological study of the Mwenda River and its mouth, Lake
Kariba. Unpubl. Ph.D. thesis, Univ. of Witwatersrand.

Ripe and running wild collected Hydrocynus vittatus were stripped of gametes.
Development was followed at 24-25 °C under laboratory conditions. Development is described
from cleavage egg, to a late larval stage at 35 days after fertilization. Illustrations of middle
gastrula, late gastrula, various embryonic stages, including free embryos and larval stages are
included. Larval behaviour and pigmentation is given. Specimens are lodged at the National
Museum and Monuments, Bulawayo, Zimbabwe and at Albany Museum, Grahamstown, South
Africa, (AMG/P 11187).

Durand, J. R. 1978. Biologic et dynamique des populations d'Alestes baremoze (Pisces;
Characidae) du bassin Tchadien. Trav. Doc. ORSTOM, 98: 1-332.

This paper includes the biology and population dynamics of Alestes baremoze from the
Tchad Basin, with a section on the reproduction and the early development of this species. The
ripe eggs, the larvae just after hatching, the 4 day old larva and the one month old larva are
described in detail. A 4 day old yolk-sac larva and a 30 day old juvenile are illustrated with
camera lucida drawings.

Durand, J. R. and Loubens, G. 1970. Observations sur la sexualite et la reproduction des Alestes
baremoze du bas Chari et du lac Tchad. Cah. ORSTOM, hydrobiol. 4(2); 61-81, 9 figs.

This paper includes a brief description of the eggs oi Alestes baremoze from Lake Chad and
Lower Chari Basin.

Durand, J. R. and Loubens, G. 1971. Developpement embryonnaire et larvaire A Alestes
baremoze. Cah. ORSTOM, hydrobiol. 5(2): 137-145, 8 figs.

The embryonic and larval development ot Alestes baremoze from Lake Chad is studied from
artificially fertilized eggs which were aquarium reared. Drawings of and data on the embryo in
segmentation, in gastrulation and just before hatching are included. Descriptions and drawings
of a 21 hour larva (3,5 mm TL), a 4 day larva (5,6 mm TL), a 15 day larva (15 mm TL) and a
juvenile of 30 days (17,5 mm TL) are given.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 2, AUGUST 1988

Neumann, M. 1968. Die Zucht von Phenacogrammus interruptus — ein Ratsel. Acjuar.-
ii.Terrar.-Z. 21: 136-139, 5 figs.

Descriptions and photomicrographs of the eggs of Phenacogrammus interruptus up to
twelve hours before hatching were based on aquarium reared fish.

CYPRINIDAE

Balon, E. K., Duyvene de Wit, J. J. and Holcik, J. 1962. The early stages of development of the
intersubfamiliar hybrid Puntius anoplus! Rhodens ocellatus (Cvprinidae). Zool. An:.
168(1-4): 102-112.

This paper includes one hgure and a brief description of Puntius anoplus (= Barbus
anoplus) protopterygiolarva (4,8 mm TL).

Barnard, K. H. 1943. Revision of the indigenous freshwater fishes of the S W Cape region. Ann.
S. Afr. Mus. 36(2): 101-262.

A brief description and morphometric and meristic table of Labeo capensis 16 mm to
450 mm TL are given. One drawing of a 16 mm TL late larvae is included. Lateral views of 9.5.
10,5, 12 and 15 mm TL larva of Labeo seeberi are illustrated. The author also gives some
morphometries and a short description. There is some doubt as to the proper identiheation of
these wild collected specimens. The lateral view of a 14 mm Barbus andrewi larval hsh is
illustrated and a morphometric-meristic table for fish from 12 mm to 525 mm TL is included.
There is a morphometric-meristic table of larval Barbus asperfrom 7-8 mm TL. The lateral view
of a 13 mm TL Barbus burchelli larval fish is illustrated. The lateral view of a 13 mm TL Barbus
callidus is illustrated and a morphometric and meristic table for fish from 11 mm to 93 mm TL
is included. Lateral views of a 15 mm TL larval hsh and a 23 mm juvenile of Barbus capensis are
illustrated. A morphometric and meristic table is included.

Cambray, j. a. 1983. Early development and larval behaviour of a minnow, Barbus anoplus
(Pisces; Cyprinidae). S. Afr. J. Zool. 18: 331-336.

This is the hrst known study of the development of an African barb, Barbus anoplus. Tables
and camera lucida drawings of development from fertilized ova to juvenile hsh are included.
Development was followed for 13 months. The paper includes egg and larval behaviour and a
description of the pigmentation of the various stages. The developmental series is housed at the
Albany Museum, Grahamstown, South Africa, (AMG/P 9259).

Cambray, J. A, 1985. Early ontogeny of Labeo capensis (Pisces: Cyprinidae). S. Afr. J. Zool. 20:
190-196.

The early development of Labeo capensis is described from wild-spawned, laboratory
reared specimens, which were followed for 5 months. Illustrations of early embryos (6), larval
hsh (7) and the juvenile stage are included. Meristic and morphometric tables are included for
specimens from 5 mm to 40 mm TL. A table is included comparing L. capensis developmental
stage sizes with other Labeo species. Specimens are lodged at the Albany Museum,
Grahamstown, South Africa, (AMG/P 10125).

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

Cambray, J. a. 1985. Early development of an endangered African barb, Barbus trevelyani
(Pisces: Cyprinidae). Rev. Hydrobiol. Irop. 18(1): 51-60.

Barbus trevelyani were artificially induced to spawn. Eggs were fertilized and their
development is recorded through the embryonic, larval and juvenile stages. Meristic 3-30 mm
TL and morphometric 3-38 mm TL tables are included. Camera lucida drawings of 8 larval
stages and 3 juvenile stages are included. Specimens are lodged at the Albany Museum,
Grahamstown, South Africa, (AMG/P 10135 and 10136).

Cambray, J. A. and Meyer, K. (in press). Early ontogeny of an endangered, relict, cold-water
cyprinid from Lesotho, Oreodaimon quathlambae (Barnard, 1938). Rev. Hydrobiol. trop.

The early development of Oreodaimon quathlambae is described from wild-collected
specimens. Illustrations of early embryos, larval fish and the juvenile stage are included. Meristic
and morphometric tables are included for specimens 5-40 mm TL. Developmental osteogenesis
is described. Specimens are lodged at the Albany Museum, Grahamstown, South Africa,
(AMG/P 11224).

Fryer, G. and Whitehead, P. J. P. 1959. The breeding habits, embryology and larval
development of Labeo victorianus Boulenger (Pisces; Cyprinidae). Revue Zool. Bot. afr.
59(1-2): 33-49, 24 figs.

The embryos and larvae of Labeo victorianus from Lake Victoria, from the fertilized egg up
to the juvenile period (37 days) are described in detail and illustrated with camera lucida
drawings.

Gaigher, I. G., Ntloko, M. M. and Visser, J. G. 1975. Reproduction and larval development of
Labeo umbratus (Pisces: Cyprinidae) in the Tyume River, Eastern Cape. J. Limnol. Soc.
sth. Afr. 1(1): 7-10.

Larval Labeo umbratus which were collected in the wild are briefly described. Four
illustrations, one of late embryonic, newly hatched, 3 and 12 days after hatching (20-23°C) are
included.

Groenewald, a. a. V. J. 1961. A progress report on the culture of Barbus holubi, the Vaal River
yellowfish, at the provincial fisheries institute, Lydenburg. Res. Rep. Prov. Fish. Inst. Tvl
Prov. Admin. Pretoria, 19pp.

Fertilized eggs (n=150) of Barbus holubi (= B. aeneus) at the sixteen celled stage were
collected from a breeding pond. The author gives a description of the eggs and larvae up to the
13 mm stage (19 days). Drawings of embryos and larvae are included. Larval behaviour is
described.

Mitchell, S. A. 1984. Further observations on the breeding behaviour of Labeo umbratus
(Smith) (Pisces: Cyprinidae). J. Limnol. Soc. sth. Afr. 10(1): 28-30.

This paper includes a brief description of egg and larval behaviour. No illustrations are
included.

Van der Merwe, F. J. 1981. Induced spawning of the common carp and Aischgrund carp

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 2, AUGUST 1988

{Cyprinus carpio) and the largemouth yellowlish (Barbus kimberleyensis). Water SA 7(2):
107-109.

A very brief account of hatching and larval behaviour of Barbus kimberleyensis is given.

Van der Westhuizen, T. F. 1974. Die ontogenese van die viscerocranium van Barbus holubi
Steindachner (Cyprinidae). M.Sc. thesis, Rand Afrikaans Univ.

The development of the viscerocranium of Barbus holubi (= B. aeneus) is described in
detail. Illustrations and descriptions for the following stages are included; 8 mm TL (3 days),
9,5 mm TL (12 days), 13,5 mm TL (27 days), 21 mm TL (45 days) and 24 mm TL (55 days).
Stages were microtomized and stained.

Van der Westhuizen, T. F. 1979. Die ontogenese van die neurocranium van Barbus holubi
Steindachner (Cyprinidae) met spesiale verwysing na die inkorporering van viscerale
elemente in die neurocranium. Unpublished Ph.D. thesis, Rand Afrikaans Univ.

The development of the neurocranium including the procartilagenous developmental stages
is described for Barbus holubi (= B. aeneus). Fourteen developmental stages were
microtomized at a thickness of 10 p.. Seven developmental stages were described using drawings
and reconstructions made from transverse sections; 7,5 mm TL (48h), 8,0 mm TL (3 days),
9,25 mm TL (6 days), 10,0 mm TL (10 days), 10,5 mm TL (20 days), 13,5 mm TL (27 days) and
21,0 mm TL (45 days). Special attention was given to the ontogenesis of the commissura
palatoquadrati.

Welcomme, R. L. 1969. The biology and ecology of the fishes of a small tropical stream. J. Zool.
158: 485-529, 18 figs.

Artificially fertilized eggs of Barbus kerstenni from the Kafunta River, Uganda are
described.

Wright, C. W. and Coke, M. M. 1975. The artificial propagation of Barbus natalensis. 2.
Hatching and early development. Lammergeyer 22: 42-48.

This paper includes a brief descriptive account of several egg and larval stages of Barbus
natalensis, no drawings accompanied the text. Development was followed for 68 days when
excessive siltation suffocated the larval fish. Important behavioural aspects are noted, such as
floating after the swim-bladder had inflated and burying themselves in the gravel substratum.

BAGRIDAE

Hirigoyen, j. P. and Petel, C. 1979. Nouvelles donnees sur la croissance en etang du poisson
Bagridae Auchenoglanis occidentalis Valenciennes, 1840. Not. Doc. Peches Piscic. C.T.F.
Nogent sur Marne 18: 20-26.

The larvae of Auchenoglanis occidentalis from the Bandama River (Ivory Coast) are briefly
described. This paper is not illustrated.

Ikusemiju, K. 1976. Distribution, reproduction and growth of the catfish Chrysichthys walkeri
(Gunther) in the Lekki Lagoon, Nigeria. J. Fish Biol. 8(6): 453-458.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

Laboratory incubated eggs of Chrysichthys walkeri (= probably C. auratiis) from Nigeria,
hatched after three days, with the yolk sac being completely absorbed on the seventh day. The
first day larva are about 8,6 mm TL. No other data on early ontogeny.

Pham, A. and Hirigoyen, J. P. 1979. Donnees preliminaires sur la reproduction provoquee de
Chrysichthys walkeri Gunther, 1899 (Poissons; Bagridae). Not. Doc. Peches Piscic C.T.F.
Nogentsur Marne 18: 10-19, 6 figs.

The first results of induced spawning of Chrysichthys walkeri (= C. maiirus) from Ivory
Coast are discussed. Larval development is briefly described for the first eight days. Photographs
of the eggs, the 18 hour old larvae and the 24 hour old larvae are included.

CLARIIDAE

Aboul-Ela, I. 1973. The embryonic and larval development of the Nile catfish, Clarias lazera
Cuv. et Val.. Proc. Seventh Arab. Sci. Cong., Cairo, Egypt.

Not seen.

Bowmaker, a. P. M. 1973. An hydrobiological study of the Mwenda River and its mouth. Lake
Kariba. Unpubl. Ph.D. thesis, Univ. of Witwatersrand.

The development of wild collected eggs and larvae of Heterobranchus longifilis? was
followed. Earliest stage described is late gastrula. Illustrations of late gastrula and 5,3, 7,4, 7,6
and 8,4 mm TL larvae are included, with a brief description.

Bruton, M. N. 1979. 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, 10 figs.

As per title on Clarias gariepinus from Lake Sibaya, South Africa. This paper includes
detailed descriptions and camera lucida drawings of the fertilized ova, the 33, 44, 56, 66 and 80
hour larvae, the ten and fourteen day fingerlings.

De kimpe, P. and Micha, J. C. 1974. First guidelines for the culture of Clarias lazera in Central
Africa. Aquaculture 4: 227-248, 13 figs.

Brief descriptions illustrated with photographs of larvae (24 hours/ 3 days/ 5 days after
hatching) obtained through induced spawning of Clarias lazera (= C. gariepinus) from Central
African Republic.

Greenwood, P. H. 1955. Reproduction in the catfish, Clarias mossambicus Peters. Nature,
Lond., 176: 516-518, 1 fig.

This paper includes the first brief description of the eggs, embryos and larvae (up to one
week) of Clarias mossambicus (= C. gariepinus) from Lake Victoria.

Greenwood, P. H. 1957. The reproduction of Clarias mossambicus Peters in Lake Victoria.
Pubis Cons, sclent. Afr. S. Sahara 25: 77-78.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 2, AUGUST 1988

Newly hatched larvae (3, 5-4,0 mm TL) and 10,0 to 12,0 mm larvae of Clarias mossambicus
(— C. gariepinus) from Lake Victoria are briefly described.

Hogendoorn, H. 1980. Reproduction et alevinage controles du poisson — chat africain Clarias
lazera (C. & V.). Not. Doc. Peches Piscic. C.T.F. Nogent sur Marne 20: 1-10.

The eggs, free embryo and larvae after resorption of the yolk sac of Clarias lazera (= C.
gariepinus) are briefly described. The eggs were obtained from fish artificially induced to spawn.

Hole, E. A. 1968. Notes of spawning behaviour of barbel Clarias gariepinus Burchell in
Rhodesia. Zool. afr. 3(2): 185-188.

A very brief description is given of the eggs, the newly hatched free embryos (3, 5-4,0 mm
TL) and the seven day old larvae (7, 0-9,0 mm TL) of Clarias gariepinus in Zimbabwe.

Janssen, J. 1985. Elevage du poisson — chat africain Clarias lazera (Cuv. & Val., 1840) en
Republique Centrafricaine. II. Alevinage en ecloserie. FAO, Pro). GCP/CAF/007/NET,
Doc. Techn. Nr. 21, 31pp.

The early development (egg, hatching, 12h, 24h, 36h, 48h, 6 days, 9-12 days) of Clarias
lazera (= C. gariepinus) from Central African Republic are described and illustrated.

JocQUE, R. 1975. Sur quelques essais de reproduction induite chez Clarias lazera et Clarias
senegalensis . PNUDIAVBIFAO-IVC 526 Rapp. Techn. 43: 17, 3 figs.

Clarias lazera (= C. gariepinus) and C. senegalensis (= C. anguillaris) from Lake Kossou,
Ivory Coast, were artificially induced to breed. Development of eggs and larvae is briefly
described. The eggs of C. senegalensis (= C. anguillaris) are illustrated (camera lucida drawing).

Pham, A. 1975. Donnees sur la production en masse d’alevins de Clarias lazera Val. (Pisces;
Clariidae). Annls. Univ. Abidjan, (E) 8(1): 139-145.

Clarias lazera (= C. gariepinus) from Lake Kossou (Ivory Coast) were artificially induced
to breed. The first fifteen days in the development of the larvae are very briefly discussed.

Weeks, D. 1984. Embryological and larval development of catfish Clarias gariepinus (Burchell)
(Pisces: Clariidae). Unpubl. BSc. project, Rhodes Univ. 20pp., 7 figs.

Clarias gariepinus were artificially induced to spawn. Development was followed from first
cleavage to metalarval phase. Early cleavage, blastula, gastrula, embryonic and protolarval,
mesolarval and metalarval stages are illustrated.

Zaki, M. I. AND Abdula, A 1983. The reproduction and development of Clarias gariepinus
(Clariidae) from Lake Manzala (Egypt). J. Ichthyol. 23(6): 48-58, 6 figs.

Detailed data are given on the structure of the eggs, nature .of spawning, embryonic and
larval periods of development of Clarias gariepinus in Lake Manzala, Egypt. Development is
followed from the mature egg before fertilization, up to the 39 day old juvenile. Illustrations of
19 stages of embryonic and 13 larval stages are included.

CAMBRAY & TEUGELS; BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

MOCHOKIDAE

Van Der Waal, B. C. W. 1986. Note on artificial fertilization and early development of
Synodontis nigromaculatus (Pisces: Mochokidae). S. Afr. J. Zool. 21(3): 269-271.

Eggs were obtained from a wild gravid female of Synodontis nigromaculatus in the
Okavango Swamps. Development was followed from activation to the mesolarval stage. 10 days
after fertilization. Photographs of several stages. 7h, lOh, 22h embryo, newly hatched prolarva
(35h), 54h prolarvae and 9,5 day old mesolarva are included. The developmental series is housed
at the Albany Museum, Grahamstown, South Africa, (AMG/P 11189).

CYPRINODONTIDAE

Avni, a. a. and Soin, S. G. 1974a. Adaptive characteristics of the embryogeny of
Nothobranchius guentheri in connection with existence in temporary tropical habitat. J.
Ichthyol. 14(5): 735-746, figs.

Nothobranchius guentheri from East African Coasts and Zanzibar are studied. Detailed
descriptions and drawings are included from the egg, soon after fertilization, up to the larval
stage, 3-4 days after hatching.

Avni, A. A. and Soin, S. G. 1974b. Adaptive peculiarities of Nothobranchius guentheri (Pfeffer)
embryogenesis in connection with life in periodically drying out tropical waters. Vopr.
ichtiologii 14: 846-858 (in Russian).

Not seen, after BATON 1977 and 1985. See AVNI and SOIN 1974a.

Kroll, W. 1984. Morphological and behavioural embryology and spontaneous diapause in the
African killifish Aphyosemion gardneri/Austrofundulus myersi developmental arrest
hatching adaptation. Environ. Biol. Fishes 11(1): 21-28.

Laboratory stock of Aphyosemion gardneri were bred in aquaria and the development of
the resulting eggs was followed. A brief account of the chronology of developmental events is
given from early cleavage to hatching. Eight photomicrographs depict early cleavage, epiboly
completion, organogenesis and a newly hatched free embryo. Diapause is discussed at certain
developmental points for annual killifishes.

Peters, N. 1963. Embryonale Anpassungen oviparer Zahnkarpfen aus periodisch austrock-
renden Gewasser. Int. Revue ges. Hydrobiol. Hydrogr. 48(2): 257-331.

Not seen.

Rasmussen, E. 1948. Spawning and early development of Epiplatvs chaperi. Aquar. J. 19(5):
23-25.

Not seen.

Terceira, a. C. 1977. Comparative observations on the early embryology of Nothobranchius
korthaiisae and Epiplatys sheljuzhkoi. With special reference to predicting the onset of
diapause. I. J. Amer. Killifish Ass. 10(10): 242-251.

Not seen.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 2, AUGUST 1988

Terceira, a. C. 1982. Observations comparatives sur le developpement embryologique primaire
de Nothobranchius korthausae et Epiplatys sheljushkoi. Killi Revue 6: 12-24 (French
translation), figs.

The eggs and the embryonic development of two Cyprinodontidae, Nothobranchius
korthausae from Sierra Leone and Epiplatys sheljushkoi from about 75 miles from
Dar-es-Salaam, Tanzania were studied using compound (for the eggs) and dissecting (for the
embryos) microscopes. The development was followed for about ten days. Camera lucida
drawings of the eggs and the embryo are included.

WouRMs, J. P. 1965. Comparative observations on the early embryology of Nothobranchius
taeniopygus (Flilgendorf ) and Aplocheilichthys pumilus (Boulenger) with special reference
to the problems of naturally occurring diapause in teleost fishes. Rep. E. Afr. Ereshw. Eish.
Res. Org. (1964): 68-73.

The developmental pattern of Nothobranchius taeniopygus from the Sio River, Kenya and
Aplocheilichthys pumilus from Lake Victoria are briefly described. The development of the eggs
was followed, using both dissecting and compound microscopes, for about 204 hours and 112
hours respectively. No drawings are included.

WouRMs, J. P. 1969. Comparative observations on the early embryology of Nothobranchius
taeniopygus (Hilgendorf) and Aplocheilichthys pumilus (Blgr.) with special reference to the
problem of naturally occurring embryonic diapause in teleost fishes. Rep. E. Afr. Ereshwat.
Eish. Res. Org. (1968): 68-73.

Not seen.

WouRMs, J. P. 1972. The developmental biology of annual fishes. II. Naturally occurring
dispersion and reaggregation of blastomeres during development of annual fish egg. J. exp.
Zool. 182: 169-200.

The dispersion and subsequent reaggregation of pre-embryonic blastomeres is discussed.
Several African annual fishes are compared to the development of Austrofundulus myersi. These
include Nothobranchius guentheri, N. palmquistii, Aphyosemion arnoldi, A. calliurum, A.
fallax, A. nigerianum, A. walkeri, A. sjoestedti and Roloffia occidentalis {= A. occidentale) .

Photomicrographs of the early and late blastula and of Epon sections to show dispersed and
reaggregated state of Austrofundulus myersi eggs are included.

CENTROPOMIDAE

Hopson, A. J. 1969. A description of the pelagic embryos and larval stages of Eates niloticus (L.)
(Pisces: Centropomidae) from Lake Chad, with a review of early development in lower
percoid fishes. Zool. J. Einn. Soc. 48(1): 117-134, 8 figs.

Eggs and larval development of Eates niloticus from Lake Chad are described in detail and
illustrated from the advanced embryo up to the juvenile period (13,0 mm TL). Twelve arbitrary
stages are defined to illustrate larval development.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

CICHLIDAE

Abu Gideiri, Y. B. 1969. The development of behaviour in Tilapia nilotica L.. Behaviour 34:
17-28, 4 figs.

Development and movements in embryos of Tilapia nilotica (= Oreochromis niloticus) were
observed through the egg membranes, using a binocular microscope, at regular intervals until
hatching. Some embryos were fixed for histological sections, stained by silver impregnation and
drawn with the aid of a camera lucida.

Arnold, M., Kriesten, K. and Peters, H. M. 1968. Die Haftorgane von Tilapia-Tdsven
(Cichlidae, Teleostei) Histochemische und elektronenmikroskopische Untersuchungen. Z.
Zellforsch. mikrosk. Anat. 91: 248-260.

This paper includes histochemical and electron microscopical investigations on cement
glands of larval Tilapia mariae and Tilapia nilotica {— Oreochromis niloticus), three to six day
old larvae.

Assheton, R. 1907. Report upon sundry Teleostean eggs and larvae from the Gambia River. In:
Kerr, G. ed. Budgett Memorial volume — 1907. Cambridge: Univ. Press, pp. 433-442, 6 figs.

Descriptions illustrated with drawings of sagittal sections of early cleavage egg and
embryonic stages in the development of Hyperopisus bebe (= Tilapia guineensis) from the
Gambia River. Specimens collected every few hours for 3 days and once a day for 3 additional
days by Budgett, Assheton notes stage too early for positive identification. Included are
descriptions of early cleavage stage of 114 blastomeres, late stage of cleavage, stage with
subgerminal cavity, stage after envelopment of the yolk mass by the blastoderm.

Baton, E. K. 1977. Early ontogeny of Labeotropheus Ahl 1927 (Mbuna, Cichlidae, Lake
Malawi), with a discussion on advanced protective styles in fish reproduction and
development. Environ. Biol. Fishes. 2(2): 147-176, 25 figs.

A detailed study of the ontogeny of Labeotropheus sp. (= Labeotropheus trewavasae or
L. fuelleborni) from Lake Malawi, based on observations in vivo and using alizarin stained
specimens. The form and the size of the eggs at several stages of development are discussed and
drawn. Photographs of stages in the cleavage, embryonic and eleutheroembryonic (= free
embryo) phases are included; furthermore drawings of the initial stages of epiboly, the first
stages in embryonic phase, the beginning of heart and vessel formation, of the stage shortly
before and shortly after hatching, of the eleutheroembryos and of the end of endogenous
nutrition. Eleutheroembryo develops without metamorphic larval stages directly into a juvenile.

Baton, E. K. 1981. Saltatory processes and altricial to precocial forms in the ontogeny of fishes.
Am. zool. 21: 573-593, figs.

This paper includes discussions and drawings of an eleutheroembryo of the mouthbrooding
cichlid Labeotropheus from Lake Malawi (see Balon 1977).

Baton, E. K. 1985. Early ontogeny of Labeotropheus Ahl, 1927 (Mbuna, Cichlidae, Lake
Malawi), with a discussion on advanced protective styles in fish reproduction and

ANN. CAPE PROV. MUS. (NAT, HIST.) VOL, 18. PT. 2. AUGUST 1988

development. In: Balon. E. K. ed. Earlv life histories of fishes. Dordrecht: Junk, pp.
207-236.

Duplicate of Balon (1977). see above.

Breder, C. M. Jr. 1943. The eggs of Bathygobius soporator (C.V.) with a discussion of the other
non-spherical teleost eggs. Bull. Bingham oceanogr. Coll. 8(3): 1-49, 16 figs.

This paper includes data and drawings of the eggs, of amongst others, Tilapia macrocephala
(= Sarotherodon melanotheron).

Budgett, J. S. 1901a. On the breeding habits of some West African fishes, with an account of the
external features in the development of Protopterus annectens, and a description of the larva
of Polypterus lapradei. Trans, zool. Soc. Lond 16(2): 115-136, 5 figs, 2 pis. (Also in: Kerr,
G. ed. Budgett Memorial Volume — 1907. Cambridge: Univ. Press, pp. 119-136, 2 pis.)

The eggs and larvae of Hyperopisus bebe (= Tilapia guineensis) from the Gambia River are
briefly described. Habitat drawings of eggs and larvae, and drawing of cement glands on larvae
are included.

Budgett, J. S. 1901b. The habits and development of some West African Ashes. Proc. Camb.
phil. Soc. 11: 102-104. (Also in: Kerr, G. ed. Budgett Memorial Volume — 1907. Cambridge:
Univ. Press, pp. 141-142.)

The eggs and larvae of Hyperopisus bebe (= Tilapia guineensis) from the Gambia River are
briefly described.

Daget, j. 1952. Observations sur une ponte de Tilapia zillii (Gervais), poisson de la famille des
Cichlides. Ann. Mag. nat. Hist. (12)5(51): 309-310.

The eggs and larvae of Tilapia zillii from Diafarabe, Central Niger system are briefly
described.

Donnelly, B. G. and Caulton, M. S. 1973. A note on the breeding behaviour and larval
development of Tilapia sparrmanii Smith (Pisces — Cichlidae). Unpubl. report, Rhodes
Univ. 16pp.

The developmental sequence after hatching was divided into eight stages, characterized by
specific morphological or behavioural characters, which are described. Eleven photomicro-
graphs of the egg and early larval to juvenile stages are included.

El Zarka, S. and Ezzat, A. 1972. Embryology and larval development of Tilapia galilaea Art.
C.l.E.S.N.Rapp. Proc. Verb. Reun. Monaco 1972, 20(4): 499-501, pi.

The embryology and larval development of Tilapia galilaea (= Sarotherodon galilaeus) from
the Nile are briefly described.

Fishelson, L. 1966. Comparative investigations of the development of some fish species of the
genus Tilapia (Cichlidae) . Unpubl. thesis, Hebrew Univ.

Not seen.

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

Fishelson, L. 1966. Untersuchungen zur vergleichenden Entwicklungsgeschichte der Gattung
Tilapia (Cichlidae, Teleostei). Zool. Jb. (Anat.) 83: 571-656.

The early development in Tilapia tholloni from the 22 hour old embryo to the 10 day old
larva is described with drawings and photomicrographs of the 2, 3, 4, 5 and 6 day stages.
Comparison is made to the development of T. macrocephala (= Sarotherodon rnelanotheron),
T. nilotica (= Oreochromis niloticus).

Francois, Y. 1958. Recherches sur I’anatomie et le developpement de la nageoire dorsale des
Teleosteens. Archs Zool. exp. gen. 97: 1-108.

The dorsal fin development in, amongst others, Tilapia monodi (= Oreochromis aureus) is
discussed. Drawings of transversal section of the dorsal fin region in 5.5 mm and 7 mm larvae are
included.

Imam, A. E. and ITashem, M. T. 1960. Notes on the breeding behaviour, embryonic and larval
development of Tilapia zillii Gervais. Notes & Mem. Hydrobiol. Fish. Dir., Cairo 51: 1-17,
7 figs.

The embryonic and larval development of Tilapia zillii from Lake Manzala (Egypt) are
discussed and stages in development are drawn, from the fertilized (1,4-1, 6 mm) eggs up to the
one month old fry (16 mm). The external morphology of the different stages is described.

Ismail, M. H. 1984. The postembryonic development of the chondrocranium in the Nile fish
Sarotherodon galilaeus Linnaeus (Teleostei; Cichlidae). 2. The postbuccal stages. Arab.
GulfJ. Sci. Res. 2(1): 221-238.

The development of the chondrocranium of two postbuccal stages of Sarotherodon galilaeus
(9 mm and 12 mm TL; 11/14 days after fertilization) is described and compared to that of other
bony fishes.

Jones, A. J. 1972. The early development of substrate-brooding cichlids (Teleostei: Cichlidae)
with a discussion of a new system of staging. J. Morph. 136: 255-272.

The gametic and the embryonic periods of Hemichromis bimaculatus and four South
American cichlids are described and illustrated with camera lucida drawings.

Jurgens, W. 1910. Ueber die Larven von Hyperopisus bebe Lac. Bl. Aquar.-u. Terrarienk. 21:
163-164.

This paper is based on the work of Budgett (1900) with descriptions and drawings of the
larvae of Hyperopisus bebe (= Tilapia guineensis).

Lanzing, W. j. R. 1976. A temporary respiratory organ in the tail of Tilapia mossambicus fry.
Copeia 1976: 800-802.

The temporary respiratory organ which occurs during the development of Tilapia
mossambica (— Oreochromis mossambicus) is described. Camera lucida drawings of the 5 day
old larvae are included.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 2, AUGUST 1988

McEwan, R. S. 1930. The early development of Hemichromis bimaculatus , with special
reference to factors determining the embryonic axis. J. Morph. 49: 579-619.

The early development in Hemichromis bimaculatus is described and drawn, in particular
from the fertilized egg up to 22 hours after spawning.

McEwan, R. S. 1940. The early development of the swimbladder and certain adjacent parts in
Hemichromis bimaculatus . J. Morph. 67: 1-40.

Not seen.

Peters, H. M. 1965. Ueber larvale Haftorgane bei Tilapia (Cichlidae, Teleostei) und ihre
Ruckbildung in der Evolution. Zool. Jb. (Zool. und Physiol.) 71: 287-300, 8 figs.

The cement glands in larvae of substrate breeding Tilapia {T. tholloni; T. mariae) are
described and illustrated (camera lucida drawings) and compared with the rudimentary cement
glands in larvae of mouth breeding Tilapia {Tilapia macrocephala = Sarotherondon
melanotheron; Tilapia mossambicus = Oreochromis mossambicus; Tilapia nilotica = Oreo-
chromis niloticus).

Peters, H. M. and Berns, S. 1982a. Die Maulbrutpflege der Cichliden. Untersuchungen zur
Evolution eines Verhaltensmusters. J. zool. Syst. Evolut.-forsch. 20: 18-52.

Not seen.

Peters, H. M. and Berns, S. 1982b. Larvophile und ovophile Maulbriiter. Tatsachen und
Informationen aus der Aquaristik 58: 19-22, 5 figs. (1 col.).

Not seen.

Rifaat, a., El-Din El-Zarka, S. and Ezzat, A. 1964. Tilapia fisheries investigation in Egyptian
lakes. 4. The embryology and larval development of Tilapia zillii (Gerv.) from Lake
Mariut, Egypt. Notes & Mem. Alexandria Inst. Hydrobiol. 72: 1-9, 18 figs.

The embryonic and larval development of Tilapia zillii from Lake Mariut, Egypt are
discussed from the fertilized ovum (1,6 mm) up to stage 21 (the 60 days post larva-13 mm). The
stages examined are illustrated with photographs and camera lucida drawings.

Shaw, E. S. 1956. Two weeks in father’s mouth. Nat. Hist., N.Y. : 152-153, figs.

Photomicrographs and descriptions of various stages in the development of Tilapia
macrocephala (= Sarotherodon melanotheron) from the zero hour egg to the fifteen day old
larvae are included (see also Shaw and Aronson 1954).

Shaw, E. S. and Aronson, I. R. 1954. Oral incubation in Tilapia macrocephala. Bull. Am. Mus.
nat. Hist. 103: 375-416, 4 figs, 13 pis.

Detailed descriptions and photomicrographs of the development in Tilapia macrocephala
(= Sarotherodon melanotheron) based on examinations in the living state, from the unfertilized
ovum (2, 0-3, 5 mm) up to the hatching stage (stage 24) (5,3 mm). Photomicrographs were made

CAMBRAY & TEUGELS: BIBLIOGRAPHY DEV. STUDIES OF AFRICAN FRESHWATER FISHES

of several serial sections, stained with Harris’ haematoxylin and eosin or a modification of the
Masson trichrome stain.

Welcomme, R. L. 1967. The relationship between fecundity and fertility in the mouthbrooding
cichlid fish Tilapia leucosticta. J. Zool. 151: 453-468, 6 figs.

The development, from egg up to the 11,0 mm larvae (17 days old) of Tilapia leucosticta ( =
Oreochromis leucostictus) reared in an artificial mouth is discussed using tables and drawings.

Welcomme, R. L. 1969. The biology and ecology of the fishes of a small tropical stream. J. Zool.
158: 485-529, 18 figs.

An account is given of the development of the mouthbrooder Hemihaplochromis multicolor
(= Pseudocrenilabrus multicolor) with tables and drawings of the larvae, five, six, eight, ten and
fourteen days after fertilization.

GOBIIDAE

Breder, C. M. Jr. 1943. The eggs of Bathygobius soporator (C.V.) with a discussion of other
non-spherical teleost eggs. Bull. Bingham oceanogr. Coll. 8(3); 1-49, 16 figs.

This paper includes data on and drawings of eggs of Gobius niger.

Tavolga, W. N. 1950. Development of the gobiid fish Bathygobius soporator. J. Morph. 87(3):
467-492.

Not seen.

ANABANTIDAE

Barnard, K. H. 1943. Revision of the indigenous freshwater fishes of the S.W.Cape region.
Ann. S. Afr. Mas. 36(2): 101-262.

A juvenile (= larval) Sandelia capensis is briefly described. Two larval fish are illustrated.

Berns, S. and Peters, H. M. 1969. The reproductive behaviour of Ctenopoma muriei and
Ctenopoma damasi (Anabantidae). Rep. E. Afr. Freshwat. Fish. Res. Org. (1968) : 44-49.

The developing eggs of C. muriei and C. damasi are briefly described. The egg, newly
hatched larva and a four day old larva are illustrated.

Morik, D. 1977. Vergleichende Untersuchungen zur Ethologie zweier Labyrinthfischarten,
Ctenopoma muriei (Boulenger, 1906) und Ctenopoma damasi (Poll, 1939), Anabantoidea,
Pisces). Unpubl. Ph.D. thesis. Univ. of Tubingen.

The cleavage egg, embryonic and free embryo phases are illustrated and photomicrographs
of the free embryo, larval and juvenile fish are included. Includes a brief description of
development including an histological study of swimbladder development.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 2, AUGUST 1988

Table 1.

Summary of number of papers dealing with the early development of African freshwater fishes.

FAMILY

No. of Papers

No. of Species

No.

Drawings

of species illustrated

Photographs

PROTOPTERIDAE

POLYPTERIDAE

CLUPEIDAE

OSTEOGLOSSIDAE

—

NOTOPTERIDAE

—

MORMYRIDAE

GYMNARCHIDAE

HEPSETIDAE

—

CHARACIDAE

CYPRINIDAE

BAGRIDAE

—

CLARIIDAE

MOCHOKIDAE

—

CYPRINODONTIDAE

CENTROPOMIDAE

—

CICHLIDAE

GOBIIDAE

—

ANABANTIDAE

, 2

TOTALS

153

/■

fE*

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing U/2 or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature 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 captions). 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 submitted on a separate sheet.

TABLES should be submitted 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 periodicals (4th ed.). Layouts
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.

BD8495 — CTP Book Printers, Cape

> ’

Ml /

•

***]

•.».

I'llkli ' ^

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 3 31st August 1988

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

A contribution to the knowledge of the ethology of the genera Parachilus
Giordani Soika and Paravespa Radoszkowski (Hymenoptera: Eumenidae)

in southern Africa

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

ABSTRACT

Ethological accounts are given for Parachilus capensis (Saussure), Parachilus major
(Saussure) and Paravespa (Gestrodynerus) mima Giordani Soika. The two species of Parachilus,
like Parachilus insignis (Saussure), divide their cells into an egg compartment and one or more
pantry compartments. The cells of Paravespa (Gestrodynerus) mima Giordani Soika are not
divided. Nest turrets of two distinct architectural forms are recorded for P. (G.) mima.

INTRODUCTION

The genus Parachilus Giordani Soika (1960) is restricted in its distribution to the
Afrotropical Region and includes ten species, six of which occur in southern Africa (Giordani
Soika, 1960 and 1985). The account of the nesting of Parachilus insignis (Saussure) (Gess and
Gess, 1976) is to date the only account of nesting for this genus. The surprising discovery that
P. insignis divides its cells into two compartments, an egg compartment and a pantry
compartment led the authors to seek for nests of other species of the genus Parachilus and of
closely related genera in order to establish whether this behaviour is restricted to P. insignis or
to the genus Parachilus, or is of wider occurrence. Two nests of Parachilus capensis (Saussure)
were discovered in the summer of 1976/77 but no further nests of this species or of any other
species of Parachilus were found until seven nests of Parachilus major (Saussure) were
discovered in October 1987.

The genus Paravespa Radoszowski is composed of two sub-genera, Paravespa Radoszowski
which is restricted to Palaearctic Asia, and Gestrodynerus Giordani Soika which is restricted to
the Afrotropical Region (Giordani Soika, 1960). Eleven species of Gestrodynerus have been
described and of these four have been recorded from southern Africa (Giordani Soika, 1960 and
1985). A relatively large number of nests of Paravespa (Gestrodynerus) mima Giordani Soika
was located during December 1987 and seventeen of these were investigated. As Paravespa is
related to Parachilus (Carpenter, pers. comm.) a comparison between the nesting of
representatives of these two genera is of interest.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

ETHOLOGICAL ACCOUNTS
Parachilus capensis (Saussure)

Geographical distribution

Parachilus capensis is known from the Cape Province from: Somerset East, Klaver,
Kamieskroon, Oliphants River between Citrusdal and Clanwilliam (Giordani Soika, 1960);
Somerset East, Matjesfontein, Ceres and Camps Bay (British Museum (N.H.) collection); and
the farms Thursford, Hilton, Lynton, Vlakwater and Vaal Vlei (Mosslands) in the Grahamstown
district, Alicedale (New Years Dam), and the Ouberg Pass (27 km NE of Montagu) (Albany
Museum collection).

Description of the nesting area

A nesting area of Parachilus capensis was located in the valley of the New Years River on
the farm Hilton, 18 km WNW of Grahamstown (33° 19' S, 26° 32' E) in the Eastern Cape
Province. The farm receives an annual rainfall of 356 mm. The vegetation of the area as a whole
is Acocks’ Veld Type 37, False Karroid Broken Veld (Acocks, 1953 and 1975). That in the
immediate vicinity of the nesting sites is more specifically dwarf scrub. Pentzia incana (Th.)
O.Ktze and Chrysocoma tenuifolia Berg, (both Compositae) predominate. The nests were
excavated in fine grained non-friable clayey soil in a bare area below a water furrow (Gess and
Gess, 1976: Plate 2). (For a full description of the area as a whole see Gess, 1981).

Identification of the prey

A total of 12 prey caterpillars of Parachilus capensis was obtained, 11 from a single sealed
cell and 1 from a female which was transporting it to her nest. Two species of Psychidae
(bagworms) were represented. On comparison these were found to be indistinguishable from the
two species recorded as utilized by Parachilus insignis (Gess and Gess, 1976). Nine of the
caterpillars were of the species with the pale ochreous, brown-spotted head and three were of the
species with the rufous head. All bore sting lesions anterior to the legs on the prothorax and
metathorax and three in addition bore sting lesions anterior to the legs on the mesothorax. No
sting lesions were found on the abdomen.

Provenance of prey

The species with the pale ochreous, brown-spotted head has been found feeding on Pentzia
incana and that with the rufous head on a yellow daisy-flower (? Gazania sp.) (Compositae) at
Hilton (Gess and Gess, 1976).

Description of the nest, egg and distribution of the provision

The nest of Parachilus capensis consists of a subterranean burrow surmounted by a mud
turret (Fig. 1). Only two nests have been found ( Nest 1 on 10. xi. 1976 and Nest 2 on 3.1.1977).
The turrets of these two nests were constructed from loosely cemented mud pellets with no
attempt having been made at smoothing. The turrets were funnel-shaped. The inner diameter at
the base of each turret was equal to that of the opening of the respective main shaft, 6 mm in
Nest 1 and 7 mm in Nest 2, and that at the lip was c 15 mm and 20 mm respectively.

Nest 1 consisted of a short vertical shaft leading to two curved branches which descended
to a depth of 36 mm. One of these branches was sealed off from the main shaft by a mud plug.
The entire sealed off portion, 31 mm in length, constituted a cell divided by three mud plates

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

Nest

Fig. 1. Vertical plans of turrets and underground workings of nests of Parachilus capensis (Saussure) (x 1). e-egg

compartment; p-pantry compartment.

into four compartments. The innermost compartment, the egg compartment, contained an egg
and two caterpillars. The egg was yellow, slightly curved, 2,83 mm long and 0,92 mm wide at
mid-length. It was attached to the cell wall by a filament. The second, third and fourth
compartments, all pantry compartments, contained three, four and two caterpillars respectively
(see Table 1).

Table 1.

Analysis of the provision from the closed cell of Nest 1 of Parachilus capensis (Saussure).

Part of nest

No. of prey
constituting
provision

Mass of
provision
mg

closed egg compartment

123

closed pantry compartment 1

200

closed pantry compartment 2

208

closed pantry compartment 3

186

combined pantry compartments 1 -1- 2 -1- 3

594

The other branch of the burrow was clearly the second to be excavated. Though empty it
was undoubtedly a second cell as the nest builder was captured bringing in prey.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

The discarded pellets, 5 mm in diameter, had been dropped in a pellet dropping area
60 mm in diameter and 300 mm from the nest entrance.

Nest 2 had clearly been usurped by a sphecid wasp as it contained 2 grasshoppers and had
been filled in with earth and debris.

Reaction to rain damage

The turret of Nest 1 which was still being provisioned was destroyed by rain during the
period that the nest was under observation. The wasp made no attempt to rebuild it.

Parachilus major (Saussure)

Geographical distribution

Parachilus major is known from the southern and western Cape Province from:
Matjesfontein and Willowmore (Giordani Soika, 1960); Matjesfontein and Worcester (British
Museum (N.H.) collection); Augusfontein (Calvinia), Zebra (Oudtshoorn), Venterstad Region,
Swart Doringrivier (Namaqualand) and Hester Malan Nature Reserve (Springbok, Namaqua-
land) (Albany Museum collection).

Description of the nesting area

A nesting area of Parachilus major was located in the Hester Malan Nature Reserve. This
reserve lies 12 km east of Springbok in the Carolusberg, in the region of Namaqualand termed
Namaqualand Klipkoppe which is characterized by rocky hills and Eindoorn granite domes and
receives an annual rainfall of 100-200 mm. The area in which the nests of P. major were located
was an east facing slope between the bed of the Droedap River and the rocky hill tops. The
vegetation of the reserve is classified Acocks’ Veld Type 33, Namaqualand Broken Veld
(Acocks, 1953 and 1975; and van Rooyen, Theron and Grobbelaar, 1979). That of the nesting
area is characterized by dwarf shrubs. The dominant plants are Mesembryanthemaceae and
perennial Compositae. The nests were sited in sloping ground in bare patches between plants
(Fig. 2). The soil in which they had been excavated although coarse and sandy is non-friable and
when moistened is malleable.

Water collection

Parachilus major, at the time of the investigation in the Hester Malan Nature Reserve,
15-21. X. 1987, was collecting water for nest construction from a puddle, in a rut in the road, fed
by a trickle seeping out from the bank above the road. In 1966 this species was collected when
it was found in large numbers visiting residual saline pools in the bed of the Swart Doringrivier
between Bitterfontein and Garies (F. W. Gess and W. H. R. Gess, 2-3. x. 1966).

Identification of the prey

A total of 54 caterpillars was recovered from nine cells of Parachilus major. Two species of
Psychidae (bagworms) were represented. On comparison these were found to be indistinguish-
able from the two species taken from the nests of Parachilus insignis and Parachilus capensis at
Hilton. Fifty were of the species with the pale ochreous, brown-spotted head and four were of
the species with the rufous head. The two species did not differ from each other in size and taken
together both species ranged in length from 7,3-16,0 mm (average 11,5 mm, sample of 52). As
a general rule each caterpillar exhibited at least one sting lesion on the underside of each thoracic
segment, in all cases positioned anterior to the legs. The incidence of stings on the prothorax.

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

Fig. 2. Nesting area of Parachilus major (Saussure) in the Hester Malan Nature Reserve, 18.x. 1987. The nesting sites were
in bare patches. The water source was the puddle in road rut in foreground.

mesothorax and metathorax was 100%, 90% and 84% respectively. Some caterpillars also had
sting lesions on the first or first and second abdominal segments. The incidence of stings on these
two segments was 25% and 11% respectively. Whereas sting lesions on the abdomen were in
some instances present in specimens lacking lesions on the mesothorax or metathorax, in others
they were additional to those on the three thoracic segments.

Prey provenance

It seems likely that in the Hester Mala^ Nature Reserve the species with the pale ochreous,
brown-spotted head feeds on the same plant as it does at Hilton, that is on Pentzia incana, which
is present in the vicinity of the nesting sites of Parachilus major. The species with the rufous
head, which feeds on yellow daisy-flowers at Hilton, presumably feeds on one or more of the
daisy-flowered species which are not uncommon in the vicinity of the nesting sites of P. major.

Description of the nest

The nest of Parachilus major consists of a subterranean burrow surmounted by a mud turret
(Fig. 3). The turret which is constructed from mud pellets is a vertical or sloping tube with the
diameter at the base equal to that of the shaft and that at the distal opening the same or
somewhat greater. The subterranean burrow consists of a short sloping main shaft terminating
in a sub-vertical cell and additional sub-vertical to sloping cells terminating secondary shafts each
nearer to the burrow entrance than that preceding it. The cell diameter is equal to or only slightly
in excess of that of the shaft, the sides are parallel and the ends truncate. Each cell is divided into

ANN. CAPE PROV, MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

Nest

Fig. 3. Vertical plans of turrets and underground workings of nests of Parachiliis major (Saussure) (X 1). e-egg

compartment; p-pantry compartment.

two compartments, a lower egg compartment and an upper pantry compartment separated by a
thin mud partition.

Method of construction of the nest, oviposition and provisioning

Water which is required for nest excavation is fetched in the crop from a nearby pool. A
female whilst filling her crop stands on the mud at the edge of the water.

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARAVESPA

The turret is constructed early in nest excavation. Pellets are laid down around the shaft
initial in such a way that the base of the turret is of the same diameter as that of the shaft, that
is 7, 0-8,0 mm (average 7,6 mm, sample of 5). Pellets are added in such a way that the final
diameter is somewhat greater than the initial diameter. In the sample the difference ranged from
1 to 3,5 mm. The pellets used in turret construction are closely packed and no interstices are left
open. Only the inside of the turret is smoothed. After turret construction is completed shaft
excavation continues, further pellets from the excavation being discarded.

The main shaft, which initially slopes downwards at a constant angle, is terminated by a cell
which is excavated at a much steeper angle. The diameter of the cell differs only marginally,
0,5 mm, if at all, from that of the shaft which gives rise to it. The inner end of the cell is truncate.
The wasp clearly rotates in a constant direction whilst excavating as the end wall bears a
noticably spiral pattern.

After a cell has been excavated oviposition takes place. An egg, bright yellow, only slightly
curved, 3,5 mm from tip to tip and 1 mm wide at mid-length is attached to the cell wall by a
filament 0,5 mm in length. Provisioning then takes place. A small number of prey is placed in
the cell and a thin mud plate is then constructed completing the egg compartment. Hunting then
continues and the rest of the provision is stored in the upper part of the cell, the pantry
compartment (Table 2).

Table 2.

Analysis of the provision from eight cells of Parachilus major (Saussure).

Part of nest

Size of sample

No. of prey
constituting
provision
(range)

No. of prey
constituting
provision
(average)

Closed egg compartment

1-2

1,8

Closed pantry compartment

3-9

6,7

Complete cells

5-11

8,7

Provisioning having been completed the cell is sealed with a mud plug and firmly packed
earth. Successive cells terminate successive steeply sloping secondary shafts each one being
nearer to the burrow entrance than that preceding it so that the cells lie close together and more
or less parallel to each other.

The total length of a cell, that is egg compartment plus pantry compartment, ranged from
21-30 mm. The eight egg compartments measured were all 8 mm in length. The variation in cell
length is therefore due to a variation in the length of the pantry compartment. Of the eight cells
for which the diameter was measured all but one had a diameter of 7,5 mm.

Seven nests were located and excavated. Two of these which lacked turrets were old nests
from which wasps had emerged. One was one-celled and the other three-celled. The remaining
five nests were surmounted by turrets. Two had not yet reached the stage of cell excavation, one
was one-celled and two were four-celled. From such a small sample it is not possible to
determine the maximum number of cells likely to be excavated. The excavation of each cell
nearer to the nest opening than the one preceding it would, however, seem to preclude the

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

excavation of a large number of cells unless additional cells are excavated from other shafts
lateral to the main shaft.

Paravespa (Gestrodynerus) mima Giordani Soika

Geographic distribution

Paravespa (Gestrodynerus) mima is known from the Cape Province from: Willowmore
(Giordani Soika, 1960); Doornberghoek (Karoo Region), Port Alfred, Cradock, near Lake
Mentz, Andries Vosloo Kudu Reserve, Bloutoring, the farms Frischgewaagd and Onverwacht
(Oudtshoorn district) and Tierberg (Prince Albert district) (Albany Museum collection) and
from the Orange Free State from: Chicago (Lindley district) (Albany Museum collection).

Description of the nesting area

Two nesting areas of Paravespa (G.) mima were located, one at Onverwacht, Oudtshoorn
district and the other at Tierberg, Prince Albert district. That at Onverwacht (33° 37' 35"S,
22° 14' 18"E) which lies immediately inland of the Outeniqua Mountains receives an annual
rainfall of 240 mm and that at Tierberg (33° 42'S, 22° 16' 24"E) which lies inland of both the
Outeniqua Mountains and the Swartberg receives only 170 mm. Both areas lie in Acocks’ Veld
Type 26, False Karroid Broken Veld (Acocks, 1953 and 1975) and are characterised by low
growing dwarf scrub with shrubs mainly along the watercourses and at Tierberg also on the
koppies. The nest sites (Figs 4 and 5) are in bare areas in relatively close proximity to water. The
soil in which the nests had been excavated was semi-non-friable to non-friable with a high
enough clay element to make it malleable when mixed with water.

Daily flight pattern and sheltering behaviour

On a hot sunny day the period of activity of Paravespa (G.) mima seems to be from
mid-morning to late afternoon. At night, before and after the period of daily activity, and during
periods of inactivity resulting from the sky being overcast females were observed to shelter in
their nests head uppermost a short distance below the ground surface (Fig. 6).

Plant visiting

No flower visiting records were obtained. However, a nesting female which was being
observed flew from her nest to some low bushes of Pteronia sp. (Compositae) which were in
flower. She did not appear to be interested in the flowers but rather to be hunting. Her search
was unfortunately unsuccessful and she flew further afield.

Male behaviour

One male only was observed in the nesting area at Tierberg. He was seen to be making a
regular inspection of all the nests in the area. After the nests had been covered with glass jars
in order to capture the females when they left their nests the male continued his visits
undeterred. The presence of a female in such a jar caused the male to become extremely
agitated.

Identification of the prey

Somewhat in excess of fifty prey caterpillars were recovered from six cells of Paravespa (G.)
mima (see Table 3). Of these caterpillars slightly over thirty were in a good state of preservation,
the remainder being partly eaten, shrivelled or otherwise damaged. At least four species were

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

Fig. 4. Nesting area of Paravespa (G.) mima Giordani Soika at Onverwacht, 6.vii.l987. The nest site was in the bare area

in the middle distance.

represented. The key to larvae of Lepidoptera adapted from Forbes (1923) by Brues, Melander
and Carpenter (1954) was used in an attempt to identify the caterpillars to familial level. All the
specimens were tentatively determined as belonging to the Noctuidae though a considerable
element of doubt remains.

The most commonly represented species, Species A, was relatively short and broad, and
pale-coloured with a virtual absence of any markings. Prolegs were absent from abdominal
segments 3 and 4 but were well developed on 5, 6 and 10 and bore a longitudinal band of
uni-ordinal crochets. This was the only species recovered from Nest 3 and Nest 8 but was also

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

Fig. 5. Nesting area of Paravespa (G.) mima Giordani Soika at Tierberg, 26. xi. 1987-5. xii. 1987. The nesting sites were in

close proximity to the figure in the middle distance.

Fig. 6. Turret of Nest 14 of Paravespa (G.) mima Giordani Soika to show the sheltering position of the female (x 1).

represented by two specimens in Nest 12. Species B was of more usual proportions than Species
A. It was pale-coloured with, however, an anterior and a posterior dark transverse stripe
dorsally on the majority of the segments. It too had prolegs only on abdominal segments 5, 6 and

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

10 and uni-ordinal crochets arranged in a longitudinal band. It was the only species present in
Nest 1 (at Onverwacht) but was also represented in Nest 2 and Nest 12 (both at Tierberg) by two
and three specimens respectively. Species C bore fleshy papillae on the thorax and abdomen. It
too was pale-coloured. Prolegs were present on abdominal segments 3-6 and 10 and bore
uni-ordinal crochets arranged in a longitudinal band. The species was represented in Nest 2 and
Nest 12 by one and two specimens respectively. Species D was non-papillate and pale-coloured.
Prolegs and crochets were as in Species C. The species was represented in Nest 2 and Nest 12
by three and one specimens respectively.

Table 3

Provision of Paravespa (G.) mima Giordani Soika from five nests.

Nest

No.

Locality

Cell

Number
of prey

Size of prey mm

Identity of
prey

Onverwacht

8,2-11,3 av. 9,5

Sp. B.

Tierberg

5,3-9,6 av. 6,9

Sp. B. (2)
Sp. C. (1)
Sp. D. (3)

Tierberg

c 12

not measured —
condition poor

Sp. A.

5,3-11,7 av. 7,9

Sp. A.

Tierberg

? of 4

not measured —
condition poor

Sp. A.

Tierberg

7,1-11,5

Sp. A. (2)
Sp. B. (3)
Sp. C. (2)
Sp. D. (I)

As is evident above, some cells were found to be provisioned with a single species of
caterpillar, others with a mixture of up to four species. Caterpillar lengths ranged between
5,3-11,7 mm. Several sting lesions were present on each prey, distributed on the underside of
thoracic segments 1-3 and abdominal segments 1-4. In Nest 3 (Cell 1) some of the caterpillars
were in a semi-pupated state.

Prey provenance

No indication has, as yet, been found of what are the forage plants of the caterpillars preyed
upon by Paravespa (G.) mima. The appearance of at least the most commonly utilized
caterpillar. Species A, suggests that it occurs in protected situations.

Description of the nest

The nest of Paravespa (G.) mima consists of a multicellular subterranean burrow
surmounted by a mud turret (Figs 7-12). The turret which is constructed from mud pellets is

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

Fig. 7. Tubular turret of Nest 12 of Paravespa (G.) mima Giordani Soika (x 1). Note discarded pellets to left of turret.

either in the form of a funnel or a curved tube. The subterranean burrow consists of a relatively
long vertical main shaft terminating in a cell, and several secondary shafts each of which after
leaving the main shaft at a gentle angle curves downwards steeply to end in a vertical cell. All
the cells lie at the same depth. The cell diameter is equal to or only slightly in excess of that of
the shaft. The sides of a cell are parallel and the end is truncate. There is no subdivision of the
cell into compartments.

Method of construction of the nest, oviposition and provisioning

Water which is required for nest excavation is fetched in the crop. At Tierberg water was
being collected from pools in the river bed.

The turret is constructed early in nest excavation (Figs 13 and 14). Pellets are laid down
around the shaft initial in such a way that the base of the turret will have the same inner diameter
as that of the shaft, that is 6, 5-9,0 mm (average 7,7 mm, sample of 19). Initially pellets are laid
down regularly so that the base of the turret is in the form of a vertical cylinder, smoothed on
the inside and rough on the outside. As turret construction proceeds, however, the manner in
which the pellets are laid down changes in one of two ways. Either the diameter is rapidly
increased to form a stalked funnel with an almost horizontal rim (Figs 8 and 10) with a distal
diameter of 17-31 mm (average 22,9 mm, sample of 9) or the original diameter is maintained
and the turret curves over to one side and is continued sub-horizontally so that the lower lip
which is somewhat longer than the upper lip extends 30-43 mm (average 34,5 mm, sample of 4)
from the shaft opening (Figs 7 and 9). The pellets used in the construction of funnel-shaped
turrets are closely packed and no interstices are left open whereas those used in the construction

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

Fig, 8. Funnel-shaped turret of Nest 14 of Faravespa (G.) mima Giordani Soika from above (x 1).

Fig. 9. Tubular turret of Nest 12 of Faravespa (G.) mima
Giordani Soika from the side (x 1).

Fig. 10. Funnel-shaped turret of Nest 1 of Faravespa (G.)
mima Giordani Soika from the side (x 1,8).

of curved turrets are only closely packed on the underside of the tube, a large number of
interstices being left open on the upper side.

After the completion of the turret shaft sinking continues vertically and further pellets
extracted from the excavation are discarded either in a distinct pellet dropping area in close
proximity to the nest (Fig. 7) or randomly at some distance from the nest. This difference in
technique seems to be an individual variation. One female observed at Tierberg was away from
the nest for 31-36 seconds when fetching water and between fetching each load of water
extracted 2-3 pellets and in addition used some water and spent some time in stabilizing the shaft

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 3, AUGUST 1988

Nest

Fig. 11. Vertical plans of turrets and underground workings of Paravespa (G.) mima Giordani Soika (x 1).

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARAVESPA

Nest

Fig. 12. Vertical plans of turrets and underground workings of Paravespa (G.) mima Giordani Soika (x 1); Crosshatching

represents leaf cells of Megachile stellarum Cockerell.

walls. The shaft walls are smoothed with water so that their surfaces which would otherwise be
coarse and gritty due to the nature of the substrate become as though plastered. Whilst the shaft
is relatively shallow the wasp, head down in the shaft, can be observed moving up and down and
rotating whilst performing this smoothing.

ANN. CAPE PROV. MUS. (NAT, HIST.) VOL. 18, PT. 3, AUGUST 1988

Fig. 13. A female Paravespa (G.) mima Giordani Soika constructing the turret of Nest 14 (x 1,6).

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARAVESPA

Fig. 14. A female Paravespa (G.) mima Giordani Soika constructing the turret of Nest 14 (x 1,6).

- A./i ^

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

The first cell when ready for use is not distinct from the main shaft as a cell is only marginally
greater in diameter than the shaft leading to it and as it has parallel sides and the distal end is
sub-truncate. The average diameter of a provisioned cell is greater than that of the shaft leading
to it by only 1,1 mm (range 0-1 mm, sample of 11).

After oviposition and provisioning has taken place the cell is sealed with a barely
discernable mud plate and the portion of the shaft close to the cell is filled with earth for a short
distance and then a mud seal is constructed between it and the rest of the shaft. Immediately
above this seal a secondary shaft is excavated (Fig. 11: Nest 3). This shaft curves to one side for
a short distance before turning sub-vertically downwards, the sub-vertical portion constituting a
cell.

Nineteen nests were investigated. Of these nests three were one-celled, two two-celled and
one four-celled. Of the remaining 13 nests nine had been usurped by megachilid bees before the
completion of a cell, a tenth had been usurped by an Ammophila species, probably A. braunsi
(Turner) and the remaining three neither had provisioned Paravespa (G.) mima cells nor had
they been usurped (Table 4).

Nests in which a cell or cells had been constructed varied in depth from 102-125 mm.

Final sealing of nests was not observed.

Table 4.

Details pertaining to 19 nests of Paravespa (G.) mima Giordani Soika. Nest 1 investigated at
Onverwacht, Oudtshoorn, 9.xii.l986 and nests 2-19 investigated at Tierberg, Prince Albert

district, 27.xi. 1987-5. xii. 1987.

Nest No.

Nature of
turret

No. of
cells

Cell no.

Nature of cell contents

funnel

3 caterpillars

funnel

fully grown Paravespa (G.)
mima larva in cocoon

6 caterpillars and
Paravespa (G.) mima larva

funnel

c 12 caterpillars and
Paravespa (G.) mima larva

14 caterpillars and
Paravespa (G.) mima larva

funnel

—

caterpillar and egg of
Ammophila sp.

funnel

—

funnel

—

funnel

—

petal cell of Megachile
aliceae

funnel

fully grown Paravespa (G.)
mima larva in cocoon

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

Nest No.

Nature of
turret

No. of
cells

Cell no.

Nature of cell contents

several caterpillars, con-
dition poor

3 leaf cells of Megachile
stellarum

curved

—

curved

leaf cells of Megachile stel-
larum

curved

—

petal cell of Megachile
aliceae

curved

8 caterpillars and
Paravespa (G.) mima larva

curved

—

funnel

—

leaf cell of Megachile stel-
larum

curved

—

2 leaf cells of Megachile
stellarum

funnel

—

leaf cell of Megachile stel-
larum

—

4 leaf cells of Megachile
stellarum

funnel

—

2 leaf cells of Megachile
stellarum

curved

—

3 leaf cells of Megachile
stellarum

Reaction to rain damage

A heavy downpour in which nest turrets were damaged occurred one afternoon during the
period when Paravespa (G.) mima was being observed at Tierberg. The next day a wasp was
observed repairing the turret of Nest 14, the rim of which had been damaged by the rain. Turrets
of other nests had been completely washed away but as these nests had been usurped by
megachilid bees before the storm it is not known what the response of P. (G.) mima would be
to such extensive damage. It is, however, of interest that several “double turrets” were found.
It is possible that the building of a second rim within the first may in some instances be a response
to damage to the original turret.

Cocoon

The fully grown larva of Paravespa (G.) mima spins a cocoon which is closely bonded to the
cell walls and entirely fills the cell. The cocoon is capped with the head capsules of the prey
caterpillars which separate it from the mud plate closing the cell.

Associated insects

Megachilid bees had usurped a high percentage of the nests of Paravespa (G.) mima
investigated at Tierberg between 26. xi. 1987 and 5.xii.l987. Of the eighteen nests investigated
nine, that is 50 %, had been usurped. All but one of these nests, the four-celled Nest 8, had been
invaded before nest provisioning by P. (G.) mima . These nests had not been abandoned by the

ANN. CAPE PROV. MUS, (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

wasps before take over by the bees. Wasps which were actively working on their nests were seen
to be severely harassed by the bees resulting in their becoming confused and eventually
abandoning their nests. The activities of the megachilid bees had therefore resulted in a total loss
of nesting effort on the part of the original nest owners.

Two species of megachilid bees were involved, Megachile stellarum Cockerell in seven
instances and Megachile aliceae Cockerell in two instances. M. stellarum has previously been
recorded nesting in burrows of Dichragenia pulchricoma (Arnold) (Gess and Gess, 1974:
204-206, Fig. 8), Parachilus insignis (Gess and Gess, 1976: 98) and Ceramius lichtensteinii
(Klug) (Gess and Gess, 1980: 78) at Hilton near Grahamstown, Eastern Cape Province. In all
instances the bee’s cells, constructed from pieces cut from leaves, were arranged serially. In the
nests of Paravespa (G.) mima , one to five leaf cells were constructed, in all instances positioned
at the inner end of the excavation. Towards the outer end of the excavation a seal (Fig. 16: Nest
16) was constructed, like the cells, from leaf pieces. Pollen from the provision was examined and
on comparison it was found to be a mixture derived from Pentzia incana and “mesems”.

Megachile aliceae has previously been recorded nesting in burrows of Parachilus insignis
(Gess and Gess, 1976: 98) at Hilton near Grahamstown, Eastern Cape Province and in burrows
of Ceramius nigripennis Saussure (Gess and Gess, 1986: 173-174) at Mesklip, near Springbok,
Namaqualand. The cells of this bee are constructed from pieces of petal. At Mesklip and
Tierberg the petals used were orange. The cells of this species seem to follow the shape of the
cavity, those constructed in cells of Ceramius nigripennis being “pot” shaped and those
constructed in burrows of P. (G.) mima being cylindrical. In the nests of P. (G.) mima a lower
provisioned cell and an upper “cell” filled with earth had been constructed and in addition in one
nest an earth fill had been added in the shaft above.

One of the nests of P. (G. ) mima (Fig. 15 and Fig. 16: Nest 4) was found to have been filled
to the mouth with fine gravel. On investigationn the underground workings were found to
consist of a main shaft of depth 132 mm. The bottom of this shaft had been filled with gravel
(particle size up to 5 mm) to a depth of 56 mm, above which had been placed a noctuid
caterpillar to which an egg had been attached. The caterpillar occupied a cell 20 mm long sealed
above with very fine gravel which filled the remainder of the shaft and the turret. It seems
probable that this was a nest of Ammophila braunsi (Turner) as this is the only wasp nesting in
pre-existing cavities in the ground known to the authors to provision with noctuid caterpillars.
The nesting of A. braunsi was investigated by the authors at Hilton in the summer of 1983/1984.
There it was found to nest in pre-existing cavities, the burrows of Parachilus insignis, and to
provision its cells with noctuid caterpillars. A. braunsi was collected in the nesting area of P. (G.)
mima at Tierberg where it was not uncommon.

DISCUSSION

The division of the cell into an egg compartment and a pantry compartment was first
described for Parachilus insignis (Gess and Gess, 1976). It was noted by Gess and Gess that this
behavioural character marked off the nesting of this species from that known at that time for
other Eumenidae . The present investigations of the nests of Parachilus capensis and Parachilus
major show that the division of the cell into an egg compartment and a pantry compartment or
compartments is not a behavioural character exclusive to the species P. insignis but is of wider
occurrence within the genus Parachilus. Whether or not it is characteristic of the genus as a

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARAVESPA

Fig, 15. Turret of Nest 4 of Paravespa (G.) mima Giordani Soika filled with gravel by the nest usurper, Ammophila sp.

(X 1).

whole awaits the discovery of the nesting of the other species of the genus, however, it seems
likely that it will be found to be so. The present investigation of the nesting of Paravespa
(Gestrodynerus) mima adds a species of an additional genus to the list of those known not to
divide their cells.

The investigations of the nesting of P. capensis , P. insignis and P. major show that these
species are separable from each other on characters of nest architecture (Figs 1, 17 and 3). The
turrets of the three species are distinct. That of P. insignis is little more than a collar, that of
P. capensis is a vertical funnel with the diameter of the distal opening considerably greater than
that of the base (1:2,5 and 1:3 in the two nests discovered) and that of P. major is a sub-vertical
to sloping tube with the diameter of the distal opening not markedly greater than that of the base
(not more than 1:1,5 in the sample of seven) so that it cannot be considered to be a funnel.

P. insignis excavates burrows having apparently no more than one or two cells (sample of
35) whereas burrows of P. major with up to four cells have been found (sample of 7). As only
two nests of P. capensis have been located a statement on cell numbers is premature. The single
completed cell of P. capensis differed markedly from those of the other two species, which have
a single pantry compartment per cell, in having three pantry compartments.

The main shaft in all three species terminates in the first cell, however, P. major (Fig. 3)
differs from P. capensis (Fig. 1) and P. insignis (Fig. 17) in the gradient of both the shaft and the
cell. The main shaft of the first descends gently and then curves downwards steeply so that the
gradient of the cell is steeper than that of the shaft whereas in the other two species the main
shaft descends vertically and then terminates in a sloping cell. Additional cells in burrows of P.
m.ajor terminate secondary shafts in gradient similar to that of the first cell and each leaves the

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

Nest

Fig. 16. Vertical plans of turrets and underground workings of Paravespa (G.) mima Giordani Soika (xl); Nest 4 usurped
by Ammophila sp. and Nest 16 by Megachile stellarum Cockerell.

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

main shaft nearer to the shaft opening than that preceeding it in such a way that the cells come
to lie with their long axes more or less parallel to each other. The secondary shaft/second cell in
burrows of P. capensis and of P. insignis by contrast leaves the main shaft in such a way that it
diverges from the first cell, that is the upper ends of the cells are close together and the lower
ends widely spread to form an inverted “V”.

Fig. 17. Vertical plans of three nests of Parachilus insignis (Saussure) (x 1) to show : 1. nest turret 2. a single celled sealed
nest and 3. a two celled sealed nest. Hilton. 29.x. 1973-27. xii. 1973. e-egg compartment; p-pantry compartment.

Psychidae appear very rarely to be taken as prey by aculeate wasps. No instances are given
by Iwata (1976) and only a single record is listed by Krombein etalia (1979) for the relatively well
studied North American fauna. This concerns a species of Eumenidae, Stenodynerus
fundatiformis fundatiformis (Robertson) the nesting of which was studied in Florida by
Krombein (1964). Prey recovered from the nests consisted of both lepidopterous and
coleopterous larvae, the Lepidoptera being represented by one species of each of four families
including Psychidae. It was suggested that the wasps were exploring a particular microhabitat
and that the species preys upon larvae which seek protection by enclosing themselves in various
ways, for example in cases or rolled leaves. The taking of psychid caterpillars by the three species

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 3, AUGUST 1988

of Parachilus would, from the above, appear to represent a highly unusual prey preference, all
the more so as prey selection seems to be absolutely restricted to these caterpillars. Remarkable
too is the fact that all three wasps appear to prey upon the same two caterpillar species.

Though the nesting biology of only three of the ten species of Parachilus Giordani Soika has
been studied, the consistency shown by P. capensis, P. insignis and P. major with regard to the
subdivision of the cells into egg and pantry compartments and the taking of larval Psychidae for
provisioning the cells suggests that the genus is a natural one and clearly distinguishes it from
other ground-nesting odyneroid genera including Paravespa Radoszkowski, exemplified in the
present study by P. (Gestrodynerus) mima.

The construction by P. (G.) mima of turrets of two distinct architectural styles is remarkable
and that the two forms should have been constructed by wasps belonging to the same population
is even more remarkable. It is of note that the single male which was patrolling the nesting area
visited nests of both forms indiscriminately confirming that the builders of these nests were truly
of one species.

ACKNOWLEDGEMENTS

The authors wish to thank: Mr T. C. White of the farm Hilton, near Grahamstown, for
his much appreciated kindness over the years in allowing them free access to his land; Mr
Klaas van Zyl of the Cape Department of Nature and Environmental Conservation for
permission to work in the Hester Malan Nature Reserve, Springbok; Mrs Sue Dean and Mr
Richard Dean of the Karoo Biome Research Station at Tierberg, near Prince Albert, for
their hospitality.

Mr Robert Gess is thanked for his assistance with the investigation of Paravespa mima. Mr
Harold Gess is thanked for Fig. 10.

REFERENCES

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

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

Gess, F. W. 1981. Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the vicinity
of Grahamstown, South Africa. Arm. Cape Prov. Mus. (nat. Hist.) 14 (1) 1-80.

Gess, F. W. and Gess, S. K. 1974. An ethological study of Dichragenia pulchricoma (Arnold) (Hymenoptera:
Pompilidae), a southerrt 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. 1976. An ethological study of Parachilus insignis (Saussure) (Hymenoptera: Eumenidae) in
the Eastern Cape Province of South Africa. Ann. Cape Prov. (nat. Hist.) 11 (5): 83-102.

Gess, F. W. and Gess, S. K. 1980. 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. 1986. Ethological notes on Ceramius bicolor (Thunberg), C. clypeatus Richards,
C. nigripennis Saussure and C. socius Turner (Hymenoptera: Masaridae) in the Western Cape Province of South
Africa. Ann. Cape Prov. Mus. (nat. Hist.) 16 (7): 161-178.

Giordani Soika, A. 1960. Notulae vespidologicae. XIV-XV-XVI. Atti Soc. ital. Sci. nat. 99 (4): 361-409.

Giordani Soika, A. 1987. Nuovo contributo alia conoscenza degli Eumenidi afrotropicali (Hymenoptera). Boll. Mus. civ.
St. nat. Venezia 36: 109-214.

GESS & GESS: CONTRIBUTION TO KNOWLEDGE OF PARACHILUS AND PARA VESPA

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.

Krombein, K. V. 1964. Results of the Archbold Expeditions. No. 87. Biological notes on some Floridian wasps
(Hymenoptera, Aculeata). Am. Mus. Novit. 2201: 1-27.

Krombein, K. V., Hurd, P.D., Smith, D.R, and Burks, B.D. 1979. Catalog of Hymenoptera in America North of Mexico.
Washington, D.C: Smithsonian Institution Press.

Van Rooyen, M. W., Theron, G.K. and Grobbelaar, N. 1979. Phenology of the vegetation in the Hester Malan Nature
Reserve in the Namaqualand Broken Veld. Jl S. Afr. Bot. 45 (3): 279-293.

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing U/i or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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

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

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

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.

BD8540 — CTP Book Printers, Cape

SMITHSONIAN INSTITUTION

3 9088 01429 5505

ISSN 0570-1880

6>I3

Annals of th^

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mas. (nat. Hist.)
Volume 18 Part 4 28 July 1989

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

New species of the genus Celonites Latreille (Hymenoptera: Masaridae)

from South Africa.

F. W. GESS

(Albany Museum, Grahamstown)

ABSTRACT

Descriptions are given of four new species of southern African Celonites Latreille, namely
bergenwahliae and wahlenbergiae from the Clanwilliam District of the western Cape Province
and davidi and peliostomi from Namaqualand.

INTRODUCTION

The writing of the present paper is occasioned by the need to provide names for three
hitherto undescribed species of Celonites included in a paper on flower visiting by masarid wasps
in southern Africa to be published shortly by S. K. Gess and F. W. Gess. The opportunity is
taken to describe also a species for which no ethological data are as yet available. All four species
belong to the assemblage termed by Richards (1962) the “Group of C wheeled Brauns” which
includes all the known South African species of the genus and is characterized by the marked
separation of the lateral lamellae of the propodeum from the median part of that body segment.

With the exception of one male paratype and one female paratype of each of the following
three species, C. bergenwahliae , C. peliostomi and C. wahlenbergiae, deposited in the Museum
of Comparative Zoology, Harvard University, Cambridge, Massachusetts, all the type material
is in the Albany Museum, Grahamstown.

TAXONOMIC DESCRIPTIONS
Celonites peliostomi sp. nov.

Female (Figs 1 and 3)

Black-, a spot on each side of face above ocular sinus and end of V-shaped raised frontal
keel, very occasionally a small spot or streak on upper edge of each arm of frontal keel, posterior
margin of pronotum, humeral angles, an elongate spot on each prepectus, base and apex of each
tegula, almost entire upper and lower faces of propodeal lamellae, elongate wedge-shaped
postero-lateral markings on tergites 1-5, sometimes a weakly defined median posterior
triangular marking on tergite 5, small spots at distal ends of fore-femora and on proximal ends

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 4, JULY 1989

of tibiae of all legs, yellowish-white; mandibles (other than at base), underside of antennal clubs,
greater part of tegulae, hind margin of scutellum medially, metanotum medially, transverse
bands between black basal bands and pale-coloured postero-lateral markings on tergites 1-5,
sternites, most of tibiae and tarsi, reddish-brown.

Wings lightly browned.

Length 6,5-7,5 mm; length of fore wing 4,7-5,0 mm; hamuli 7.

Head (Fig. 1), pronotum, mesopleura, mesonotum, scutellum and dorso-lateral areas of
propodeum coarsely and densely punctured; punctures on sides of pronotum, mesonotum, sides
of scutellum and upper regions of mesopleura tending to form longitudinal striae.

Clypeus with a V-shaped raised keel which starts near but below antennal sockets and has
its point on the midline a little below centre of disc; frons above antennae with a transverse
V-shaped raised keel which starts in upper third of ocular sinuses and has its point on the midline
at level of upper margin of antennal sockets.

Scutellum medially convex, strongly raised above level of mesonotum and with a wide
crenulate anterior furrow.

Propodeal lamella (Fig. 3) of each side inclined at about 45 degrees to horizontal, narrow,
bluntly pointed distally, with outer edge largely smooth and disc translucently punctured,
separated from the median part of the propodeum by a proximally straight slit; median part of
propodeum with ventro-lateral area bordering slit on each side markedly angled at edge and
smoothly surfaced and shining, with angular tubercle situated dorsally on each side moderately
strong and with general posterior surface (including concave declivity) longitudinally (vertically)
striate.

Gastral tergites closely, finely and somewhat superficially punctured, weakly shining, with
their posterior margins mostly smooth; tergites 1-5 with at least some of the posterior outer
angles (most commonly those of tergites 4 and 5) moderately projecting; tergite 6 with median
part roundly produced, emarginate before sides. Gastral sternites shining; sides and
postero-lateral corners of sternite 2, a broad transverse proximal band on and postero-lateral
corners of sternites 3-5, and all of sternite 6 (excepting a smooth region along midline) with fine
shallow punctures.

Male (Figs 2, 4 and 5)

Colouration very similar to that of female differing only with regard to the yellowish-white
markings on the head. These markings are; a large irregularly-shaped spot (absent in female) on
disc of clypeus basally, a spot on each side of face within (not above as in female) ocular sinus;
usually (not only occasionally as in female) a small spot or streak on upper edge of each arm of
V-shaped frontal keel.

Length 7,2-7, 6 mm; length of fore wing 4,4-4, 7 mm; hamuli 6-7.

Structure much like that of female differing most noticeably with respect to the following:
antennal club both longer and wider with individual segments less discernible and with three
sensory depressions beneath; eyes closer below; clypeus narrower and with V-shaped raised keel
almost obliterated, only its ends indistinctly indicated; frons with V-shaped raised keel weak
medially; tergites with posterior outer angles more strongly projecting; tergite 7 compared to
tergite 6 of female with median part much less rounded, subtruncate, and with lateral
emarginations deeper (due to stronger development of posterior outer angles).

GESS: GENUS CELONITES LATREILLE (HYMENOPTERA; MASARIDAE)

Figs 1 and 2. Celonites peliostomi sp. nov.: frontal view of head of female (Fig. 1) and of male (Fig. 2) (both x 25).

ANN. CAPE PROV. MUS. (NAT, HIST.) VOL. 18, PT. 4, JULY 1989

Figs 3, 4 and 5. Celonites peliostomi sp. nov.: dorsal view of right half of propodeum of female showing lamella (Fig. 3)
(X 65); subdorsal view (Fig. 4) and subventral view (Fig. 5) of genitalia of male (both x 100).

GESS: GENUS CELONITES LATREILLE (HYMENOPTERA: MASARIDAE)

Genitalia (Figs 4 and 5); parameres gently tapering towards their ends which are rounded
and entire; they and volsellae with scattered fine hairs.

Material examined: Cape Province: Namaqualand, Springbok, Hester Malan Nature Reserve,
15-21. X. 1987 (F. W. and S. K. Gess), Holotype female. Allotype male, 37 female Paratypes and
2 male Paratypes (all in flowers of Peliostomum virgatum E. Mey ex Benth., Scrophulariaceae);
same locality, dates and collectors, 5 female Paratypes (all in flowers of Aptosimum spinescens
(Thunb.) Weber, Scrophulariaceae); same locality, dates and collectors, 1 male Paratype (in
flower of Aptosimum lineare Marl, and Engl., Scrophulariaceae); same locality, dates and
collectors, 1 female Paratype (Malaise trap), same locality, 10-12. x. 1988 (F. W. and S. K.
Gess), 3 male Paratypes, same locality and dates (D. W. Gess), 1 male Paratype; [Cape
Province:] Namaqualand, [Springbok] 2917 DB, Hester Malan N[ature] R[eserve], 30.x. 1987
(M. Struck), 1 female Paratype (on Peliostomum virgatum).

Etymology: The name, in the genitive singular, is formed from the generic name of the plant,
Peliostomum virgatum E.Mey ex Benth. (Scrophulariaceae), in the flowers of which the wasp
was most commonly found foraging for nectar or nectar and pollen, and serves to draw attention
to the floral association.

C. peliostomi sp. nov. may be grouped with C. capensis Brauns and C. humeralis Richards
in that the separation of the propodeal lamellae from the median part is by a deep, narrow,
straight slit, not by a spiral slit ending in a circular emargination into which an extension of the
median part of the segment projects strongly. However, in its possession of raised frontal and
clypeal keels it is similar to C. clypeatus Brauns and C. andrei Brauns.

Celonites wahlenbergiae sp. nov.

Female (Figs 6 and 10)

Black', a narrow band along posterior margin of pronotum, variably developed
postero-medial spots on tergites 2-4 (and occasionally also in a reduced size on 5),
yellowish-white', distal half of mandibles, most of pronotum, whole of tegulae, at least extreme
hind margins of scutellum and metanotum laterally (occasionally greater part of scutellum and
also middle of metanotum), occasionally outer portions of propodeal lamellae, most of tergites
1-3 and occasionally part of tergite 4, whole of sternite 1 and sides of sternites 2 and 3, extreme
distal ends of femora and whole of tibiae and tarsi of all legs, reddish-brown.

Wings lightly browned.

Length 7,3-8,3 mm; length of fore wing 4,7-5, 3 mm; hamuli 7-8.

Head (Fig. 6) with puncturation of frons fine and very sparse on a microscopically
longitudinally aciculate surface and constrasting with fine but denser puncturation on more
strongly longitudinally aciculate to finely rugose surface of clypeus and moderate and dense
puncturation of vertex. Pronotum, mesopleura, mesonotum, scutellum and dorso-lateral areas
of propodeum coarsely and densely punctured; punctures on upper regions of mesopleura with
a tendency to form not very noticeable striae.

Clypeus unmodified, its disc evenly convex and without any indication of a keel; frons with
a feeble transverse prominence above antennae.

Scutellum medially convex, strongly raised above level of mesonotum and with a wide
crenulate anterior furrow.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 4, JULY 1989

Figs 6 and 7. Celonites wahlenbergiae sp. nov.: frontal view of head of female (Fig. 6) and of male (Fig. 7) (both x 25).
Figs 8 and 9. Celonites bergenwahliae sp. nov.: frontal view of head of female (Fig. 8) and of male (Fig. 9) (both x 25).

GESS: GENUS CELONITES LATREILLE (HYMENOPTERA: MASARIDAE)

Propodeal lamella (Fig. 10) of each side wide, broadly truncate distally, with outer edge
gently convex, separated from the median part of the propodeum by an inwardly curving slit
ending in a relatively large circular emargination; lateral projection of the ventral margin on
each side of the median part of the propodeum with its hind edge transverse and its point
narrowly rounded, as wide as long, and projecting across opening of curved slit at level of end
of lamella.

Gastral tergites coarsely and densely punctured, with their posterior margins mostly
smooth; tergites 1-5 with posterior outer angles moderately projecting; tergite 6 with median
part roundly produced, very weakly emarginate before sides and with margin nowhere angular.

Gastral stemites shiny; sternite 2 with fine punctures scattered rather sparsely over surface;
sternites 3-5 with close, moderate punctures in a broad transverse proximal band and on
postero-lateral corners and with sternite 6 with similar punctures covering entire surface other
than for smooth region along midline.

Male (Figs 7 and 11)

Colouration similar to that of female, differing most noticeably in the basically black (not
reddish-brown) pronotum, in the reduction in the amount of reddish-brown on the tergites, and
in the greater extent of yellowish-white markings.

Black-, whole of clypeus and labrum, occasionally a small median spot on face directly above
clypeus, occasionally a spot within each ocular sinus or whole ocular sinus, an antero-medially
expanded narrow band along posterior margin of pronotum and a spot on humeral angles (these
markings occasionally meeting and fusing), occasionally a dot medially at apex of scutellum,
postero-medial spots on tergites 1-5, yellowish-white-, distal half of mandibles, whole of tegulae,
usually extreme hind margins of scutellum and metanotum laterally, tergite 1 (other than
anterior declivity), most of tergite 2, usually a postero-medial band on tergite 3 (i.e. usually not
postero-lateral areas), whole of sternite 1 and antero-lateral areas of sternite 2, extreme distal
ends of femora and whole of tibiae and tarsi of all legs, reddish-brown.

Length 6, 5-7, 3 mm; length of fore wing 4, 2-4, 8 mm; hamuli 7-8.

Structure much like that of female differing most noticeably with respect to the following:
antennal club wider with individual segments less discernible and with three sensory depressions
beneath; eyes closer below; clypeus narrower; frons with transverse prominence above antennae
less feeble especially medially; tergites with posterior outer angles more strongly projecting;
tergite 7 compared to tergite 6 of female with median part more widely rounded and with lateral
emarginations much better developed (due to strong development of posterior outer angles).

Genitalia (Fig. 11); parameres wide and emarginate at their ends, furnished with long and
strong inwardly directed curved hairs; each volsella with a subapical transverse band of short
strong hairs; transverse hair bands on the right and left volsellae together forming a straight
transverse band.

Material examined: Cape Province: Clanwilliam District, Clanwilliam Dam, 14.x. 1987 (F. W.
and S. K. Gess), 1 female Paratype (on ground) and 2 male Paratypes (in flowers of
Wahlenbergia sp. A, Campanulaceae); same locality, 3-7. x. 1988 (F. W. and S. K. Gess),
Holotype female. Allotype male, 3 female Paratypes (1 in flower of Wahlenbergia sp. A,
Campanulaceae) and 2 male Paratypes; same locality and dates (D. W. Gess), 4 female
Paratypes and 6 male Paratypes; Clanwilliam District, 5 km W of Clanwilliam, road to
Graafwater, 12.x. 1987 (F. W. and S. K. Gess), 2 male Paratypes (on ground); same locality.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 4, JULY 1989 •

Figs 10 and 11. Celonites wahlenbergiae sp. nov.: dorsal view of right half of propodeum of female showing lamella
(Fig. 10) (x 65); subventral view of genitalia of male (Fig. 11) (x 100).

Figs 12 and 13. Celonites bergenwahliae sp. nov.: dorsal view of right half of propodeum of female showing lamella
(Fig. 12) (X 65); subventral view of genitalia of male (Fig. 13) (x 100).

GESS: GENUS CELONITES LATREILLE (HYMENOPTERA: MASARIDAE)

5-6. X. 1988 (F. W. and S. K. Gess), 1 female Paratype and 1 male Paratype; same locality and
dates (D. W. Gess), 1 female Paratype and 3 male Paratypes.

Etymology: The name, in the genitive singular, is formed from the generic name, Wahlenbergia
(Campanulaceae), of the plants in the flowers of which the wasp was commonly found foraging
for nectar or nectar and pollen, or near which the wasp was commonly found resting on the
ground, and serves to draw attention to the floral association.

Celonites bergenwahliae sp. nov.

Female (Figs 8 and 12)

Black', sometimes portions (especially medially) of a very narrow band along posterior
margin of pronotum, sometimes variably developed but always insignificant postero-medial
spots on tergites 2-4, yellowish-white-, distal half of mandibles, whole or portions of a very
narrow band along posterior margin of pronotum, whole of tegulae, extreme hind margin of
scutellum laterally, middle of metanotum, usually outer portions of propodeal lamellae, most of
tergites 1-3 and occasionally part of tergite 4, whole of sternite 1, most of sternite 2,
postero-lateral corners of sternite 3, extreme distal ends of femora and whole of tibiae and tarsi
of all legs, reddish-brown.

Wings lightly browned.

Length 7, 5-8, 2 mm; length of fore wing 5, 0-5, 3 mm; hamuli 7.

Head (Fig. 8) with puncturation of frons moderate and fairly close on a finely longitudinally
rugose surface and contrasting with finer but denser puncturation on a more strongly
longitudinally rugose surface of clypeus and coarser and denser puncturation on vertex.
Pronotum, mesopleura, mesonotum, scutellum and dorso-lateral areas of propodeum coarsely
and densely punctured; punctures on upper regions of mesopleura with a tendency to fdrm not
very noticeable striae.

Clypeus unmodified, its disc evenly convex and without any indication of a keel; frons with
a feeble transverse prominence above antennae.

Scutellum medially convex, strongly raised above level of mesonotum and with a wide
crenulate anterior furrow.

Propodeal lamella (Fig. 12) of each side wide, broadly truncate distally, with outer edge
gently convex, separated from the median part of the propodeum by an inwardly curving slit
ending in a relatively small circular emargination; lateral projection of the ventral margin on
each side of the median part of the propodeum with its hind edge not transverse but directed
anteriorly at 45 degrees and its point narrowly rounded and projecting across upper part of slit
well anterior to level of end of lamella.

Male (Figs 9 and 13)

Colouration similar to that of female, differing most noticeably in the possession of
yellowish-white markings on the head and in the greater amount of that colour on the pronotum.

Black', whole of clypeus and labrum, occasionally one or two small spots within each ocular
sinus, an antero-medially expanded narrow band along posterior margin of pronotum,
occasionally a spot on humeral angles, occasionally small and diffuse postero-medial spots on
tergites 2-5 or fewer (sometimes these spots are totally absent), yellowish-white-, distal half of
mandibles, whole of tegulae, extreme hind margin of scutellum laterally, tergite 1 (other than

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 4, JULY 1989

anterior declivity), usually most of tergite 2, sometimes whole of tergite 3 or sometimes this
tergite with only a laterally attenuated postero-medial transverse band (sometimes even this
lacking), very occasionally median and lateral patches on tergite 4, extreme distal ends of femora
and whole of tibiae and tarsi of all legs, reddish-brown.

Length 6,2-7, 3 mm; length of fore wing 4,0-4, 6 mm; hamuli 6-8.

Structure much like that of female, the differences between the sexes being the same as
those given above for C. wahlenbergiae sp. nov.

Genitalia (Fig. 13); parameres wide and emarginate at their ends, furnished with long and
strong inwardly directed curved hairs; each volsella with a longitudinal band of short strong hairs
near inner margin; longitudinal hair bands on right and left volsellae opposing one another and
together forming a double longitudinal band.

Material examined: Cape Province: Clanwilliam District, Klein Alexandershoek (32° 20' 20" S,
18° 46' E), 8-13. X. 1987 (F. W. and S. K. Gess), 1 female Paratype and 3 male Paratypes (all on
ground); same locality, 6.x. 1988 (F. W. and S. K. Gess), Holotype female. Allotype male, 1
female Paratype and 6 male Paratypes (all in flowers of Wahlenbergia sp. B, Campanulaceae);
same locality and date (D. W. Gess), 1 female Paratype and 2 male Paratypes; Clanwilliam
District, 5 km W of Clanwilliam, road to Graafwater, 12.x. 1987 (F. W. and S. K. Gess), 1 male
Paratype (on ground).

Etymology: The name, in the genitive singular, is formed from an anagram of the generic name,
Wahlenbergia (Campanulaceae), of the plants in the flowers of which the wasp was commonly
found foraging for nectar or nectar and pollen, or near which the wasp was commonly found
resting on the ground, and serves to draw attention to both the floral association and the present
wasp species’ close resemblance to C. wahlenbergiae sp. nov.

C. bergenwahliae sp. nov. is superficially very similar to C. wahlenbergiae sp. nov. but is
readily distinguishable in the female by the colour of the pronotum and scutellum (black, not
reddish-brown), in both sexes by the puncturation of the frons (moderate and fairly close as in
Figs 8 and 9, not fine and very sparse as in Figs 6 and 7) and by the form of the lateral projections
of the ventral margin of the median part of the propodeum (Fig 12 as compared with Fig. 10),
and in the male by the genitalia (Fig. 13 as compared with Fig. 11).

Celonites davidi sp. nov,

Male (Fig. 14)

Black', whole of clypeus, proximal half of labrum (distal half is unpigmented and
translucent), spot on proximal half of mandibles, large median patch on frons above clypeus,
patch entirely filling ocular sinus and extending upwards and downwards along portion of inner
eye margin and mesad towards median frontal patch (but not meeting the latter), most of
pronotum, narrow elongate marking margining anterior edge of mesopleura opposite pronotal
lobes, a spot on tegulae anteriorly, a small spot on antero-lateral corners of scutellum opposite
tegular ends, greater part of upper and lower faces of propodeal lamellae, postero-medial spots
on tergites 1-6, elongate wedge-shaped postero-lateral markings on tergites 1-3, small spots on
distal end of fore femora and proximal end of fore tibiae, yellowish-white-, antennae (other than
for 3 basal segments), distal half of mandibles, pronotal lobe and oblique longitudinal band on
each side of pronotum, most of underside of pronotum, most of tegulae, most of scutellum.

GESS: GENUS CELONITES LATREILLE (HYMENOPTERA: MASARIDAE)

Fig. 14. Celonites davidi sp. nov.: frontal view of head of male to show colour pattern (x 25).

median part of metanotum, tergites (other than for above indicated pale markings), most of
sternite 2 and parts of sternite 3 laterally, distal part of femora and whole of tibiae (other than
for pale spots on fore legs) and tarsomeres 1-4 , reddish-brown; coxae, trochanters, proximal
part of femora and fifth tarsomeres, sternites (other than indicated above), dark brown.

Wings very lightly browned.

Length 6,5 mm; length of fore wing 3,8 mm; hamuli 6.

Head with surface of clypeus, frons and vertex shagreened, with puncturation of frons
limited to area immediately adjacent to anterior ocellus where sparse and with that of vertex
somewhat coarser and denser. Clypeus and frons unmodified, without any indication of keels.

Thorax and abdomen moderately to coarsely punctured; punctures weakest and least dense
on pronotum, most pronounced and densest on mesonotum; punctures on scutellum and upper
regions of mesopleura forming weak and very strong longitudinal striae respectively.

Propodeal lamella of each side wide, broadly truncate distally, with outer edge gently
curved, separated from median part of propodeum by an inwardly curved slit ending in a
relatively large circular emargination; lateral projection of ventral margin on each side of the
median part of the propodeum with its hind edge transverse and its point bluntly rounded, wider
than long, and projecting across opening of curved slit at level of end of lamella.

Abdominal tergites with posterior outer angles only moderately projecting; tergite 7 with
hind margin gently curved and with small lateral emarginations.

Female unknown.

Material examined: Cape Province: Namaqualand, Anenous (29° 14' 30" S, 17° 34' 45" E),
11-13. X. 1988 (D. W. Gess), Holotype male (on ground).

Etymology: The name, in the genitive singular, is formed from the name of the collector of the
present specimen, Mr David W. Gess, in recognition of his enthusiastic and discriminating
collecting over the years.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 4, JULY 1989

ACKNOWLEDGEMENTS

The author wishes to thank Mr Udo Martinussen of Klein Alexandershoek for permission
to work on his land and Messrs Klaas van Zyl and Adriaan Oosthuizen of the Cape Department
of Nature and Environmental Conservation for permission to work in the Hester Malan Nature
Reserve.

Thanks are due also to Mr Robin Cross of the Electron Microscopy Unit, Rhodes
University, Grahamstown, for producing the scanning electron micrographs reproduced in Figs
1-13.

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

REFERENCE

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

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing I5 or 2). In exceptional cases typed manuscripts will be accepted. These 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 speciflc names should be underlined to indicate italics. All other marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

BD9961 — CTP Book Printers, Cape

SMITHSONIAN INSTITUTION LIBRARIES

3 9088 01429 5513

Annals of the.

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 5 28 July 1989

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

Flower visiting by masarid wasps in southern Africa (Hymenoptera:

Vespoidea: Masaridae)

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

CONTENTS

Abstract 95

Introduction 96

Flower visiting records 97

Ceramius Latreille 97

Jugurtia Saussure 109

Masarina Richards 112

Celonites Latreille 114

Quartinia Ed. Andre 123

Quartinioides Richards 125

Quartiniella Schulthess 131

Discussion 131

Acknowledgements 133

References 133

ABSTRACT

Flower visiting records are given for 69 species belonging to six of the seven genera of
southern African masarid wasps: Ceramius (14 spp.), Jugurtia (4 spp.), Masarina (3 spp.),
Celonites (8 spp.), Quartinia (10 spp.) and Quartinioides (30 spp.).

The records, so far assembled, indicate that southern African masarids are most commonly
associated with Mesembryanthemaceae (51%) and Compositae (28%), those species visiting
flowers of other families such as Campanulaceae (12%), Scrophulariaceae (5%), Leguminosae
(Papilionatae) (3%) and Liliaceae (2%) being the exceptions.

It is demonstrated that the majority of species exhibit fidelity to particular families or even
genera of plants and that some species, at least, are probably of importance as pollinators to the
plants which they visit.

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

INTRODUCTION

Studies of flower visiting by aculeate Hymenoptera have been concerned in the main with bees.
General works on pollination such as Percival (1969), Proctor and Yeo (1973), Faegeri and van der
Pijl (1979), Real (1983), and Barth (1985) have few references to flower visiting by aculeate wasps.
Most aculeate wasps are, however, regular flower visitors, both adult male and female depending
largely on nectar for nourishment. Though adult aculeate wasps are nectar feeders the larvae of the
majority of species feed on spiders or insects supplied by their mothers. The Masaridae are
exceptional in that the majority of species provision their larvae, bee-like, with pollen and nectar.
Furthermore though short tongues are characteristic of the majority of wasps most masarids have
long tongues (Figs 1 and 2), some considerably longer than the wasps’ length from the frons to the
tip of the abdomen, giving them the potential to obtain nectar from a wide range of flower forms
including those in which the nectaries are inaccessible to short tongued wasps. They do not,
however, get even a passing mention in Percival, Real, and Faegeri and van der Pijl and are
mentioned only briefly in Proctor and Yeo (pages 367-368) and Barth (pages 33 and 61). Sufficient
pollen, strategically placed is carried externally by masarid wasps to make them potential pollinators.
Pollen for provisioning is ingested and is carried mixed with nectar in the crop. Richards (1962) in
his world revision of the Masaridae reviewed the literature on flower visiting by these wasps and
concluded that “The higher masarids are so closely attached to particular kinds of flowers that the
subject cannot be omitted from any serious study of the group though our knowledge is still very
incomplete and inaccurate. It may well be possible in the future to relate the structure of some of the
genera to that of the flowers they visit and to the methods they use in exploiting them.” Since then
one such study has been conducted, the subjects of the study being a North American masarid,
Pseudomasaris vespoides (Cresson), and flowers of the genus Penstemon (Scrophulariaceae)
(Torchio, 1974). It has been stated in “A preliminary synthesis of pollination biology in the Cape
flora” that Masaridae “are probably important floral visitors in southern Africa” (Whitehead,
Giliomee and Rebelo in Rebelo, 1987), however, no indication is given of masarid/flower
associations. The present authors have, during the past two decades, in the course of their studies
of aculeate wasps and bees in the Cape Roral Region (more especially in the karroid areas of this
region), kept records of flower visiting by wasps and bees (Gess and Gess, Catalogue of flower visits
by aculeate wasps, unpublished). Some of these records have been published, those concerned with
Masaridae being in Gess (1968, 1973 and 1981) and in Gess and Gess (1980, 1986, 1988a and 1988b).
In the light of the current interest in pollination it seems useful to gather together all the available
records of flower visiting by masarid wasps in southern Africa and those as yet unpublished and to
assess the degree of fidelity exhibited by these wasps and their possible effectiveness as pollinators.

The family Masaridae {sensu Richards, 1962) is constituted of three sub-families, the
Euparagiinae, Gayellinae and Masarinae. All three sub-families are represented in the New
World but only the Masarinae are represented in the Old World. It is, however, in the Old
World and in southern Africa in particular that the greatest speciation has occurred. Seven
genera, Ceramius Latreille, Jugurtia Saussure, Masarina Richards, Celonites Latreille, Quartinia
Ed. Andre, Quartinioides Richards and Quartiniella Schulthess are represented in southern
Africa. Masarina, Quartinioides and Quartiniella are endemic to this region.

In the following account the genera are considered in order of decreasing body length which
is associated with increasing relative tongue length (Table 1) (Quartinia and Quartiniella being
the notable exceptions) and not following a systematic or phylogenetic sequence.

The flower visiting records are presented in tabular form necessitating the use of
abbreviations. These are:

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

B - blue; O - orange; Pi - pink; Pu - purple;

PuPi - purplish pink; V - violet; W - white;

Y - yellow; WY - cream.

F - female; M - male.

digits - represent numbers of specimens captured;
m - represents many observations of visits to flowers;

p - pollen from provision representing an unknown number of visits to
flowers.

CDM - C. D. Michener; CFJG - C. F. Jacot Guillarmod;

DWG - D. W. Gess; EMCC - E. McC. Callan; FWG - F. W. Gess;

HWG - H. W. Gess; JGHL - J. G. H. Londt; MS - M. Struck;

OWR - O. W. Richards; RET - R. E. Turner; RWG - R. W. Gess;

SKG - S. K. Gess; WHRG - W. H. R. Gess.

In order to be consistent with earlier papers on Masaridae by Gess and Gess the names
Compositae and Leguminosae are used rather than the presently favoured alternative names
Asteraceae and Fabaceae. It should be noted that the plant here referred to as Aspalathus
spinescens Thunb. lepida (E. Mey.) Dahlgren (det. J. Vlok, confirmed E. Brink) was previously
referred to as Aspalathus desertorum Bol. (det. E. Brink following Bayliss BRI 618) (Gess and
Gess, 1986, 1988a and 1988b).

Although full locality details are recorded on most specimen labels the localities are given
in the tables by district, expressed as the name of the nearest town, as it is more informative for
the purposes of this publication to group the localities.

FLOWER VISITING RECORDS

Colours:

Sex:

Numbers:

Collectors:

Ceramius Latreille

The genus Ceramius occurs in two widely separated geographical regions in the Old World,
one being the extreme south west of the Afrotropical Region and the other that portion of the
Palaearctic bordering on the Mediterranean Sea. In the Afrotropical Region the genus is in the
main restricted to the Cape Province where it is found in Namaqualand, the South Western
Cape, the Little Karoo, the southern parts of the Great Karoo and in the Eastern Cape as far
east as the Great Fish River. Outside the Cape Province, one species, C. damarinus Turner, is
endemic to the Kaokoveld and Ovamboland in Namibia (S. W. A.), and one Eastern Cape
species, C. capicola Brauns, has been recorded from two localities (Kroonstad and Thaba Nchu)
in the Orange Free State. The areas favoured by these wasps are in the main characterised by
a predominantly winter rainfall and low, semi-arid vegetation. In southern Africa Ceramius
favours those parts of the Karroid and False Karroid areas (Acocks, 1953 and 1975) which lie
within the winter rainfall and spring/autumn rainfall regions. That is those karroid areas which
fall within the Cape Floral Region (sensu Bond and Goldblatt, 1984) as defined in Rebelo
(1987). The flight period in southern Africa is from September to March, the period for
individual species being somewhat more limited.

The genus Ceramius has been divided on morphological characters into eight species groups
(Richards, 1962; Gess and Gess, 1986) one of which has recently been sub-divided (Gess and

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Fig. 1. Ceramiits clypeatus Richards, an example of a relatively short tongued masarid wasp. Above: dorsal view of head
with tongue extended (x 14). Below: short length of tongue (glossa) with the two halves separated (x 500).

GESS & GESS: FLOWER VISITING BY MAS ARID WASPS IN SOUTHERN AFRICA

Fig. 2. Quartinioides laeta (Schulthess), an example of a relatively long tongued masarid wasp. Above: dorsal view with
tongue extended (x 14). Below; short length of tongue (glossa) with the two halves separated (x 500).

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Gess, 1988b). Six of these species groups are represented in southern Africa. Compared with
other southern African masarids Ceramius spp. are medium to large wasps, females ranging in
length from 11 mm (C. bicolor (Thunb.)) to 22 mm (C. rex Saussure). Ceramius species are
relatively short tongued for masarid wasps (Table 1 and Fig. 1).

Table 1.

Body length, tongue length, and tongue length : body length for some southern African masarids.

GENUS SPECIES

SEX

AVERAGE

BODY

AVERAGE

TONGUE

AVERAGE TONGUE LENGTH

LENGTH

AVERAGE BODY LENGTH

Ceramius Latreille

bicolor (Thunberg)

10,83

2,96

0,27

10,50

2,92

0,28

braunsi Turner

17,28

4,70

0,27

capicola Brauns

10,90

2,54

0,23

clypeatus Richards

15,43

2,98

0,19

15,48

3,18

0,21

lichtensteinii (Klug)

17,78

5,56

0,31

17,83

5,54

0,31

nigripennis Saussure

14,86

4,08

0,27

rex Saussure

20,86

5,83

0,28

Jugurtia Saussure

braunsi (Schuithess)

9,92

3,69

0,37

braunsiella (Schuithess)

11,17

4,11

0,37

confusa Richards

10,17

4,00

0,39

10,08

4.23

0,42

Masarina Richards

familiaris Richards

10,09

3,54

0,35

8,80

3,28

0,37

mixta Richards

8,85

3,71

0,42

7,45

2,92

0,39

Celonites Latreille

capensis Brauns

8,89

5,71

0,64

8,75

5.04

0,58

clypeatus Brauns

8,80

5.68

0,66

7,63

4,96

0,65

peliostomi Gess

6,76

4,73

0,70

7,08

4,17

0,59

wahlenbergiae Gess

7,47

4,29

0,57

6,88

3,40

0,49

bergenwahliae Gess

7,56

4,28

0,57

6,56

3,35

0,51

Quartinia Ed. Andre

parcepunctata Richards

5,53

2,25

0,42

Quartinioides Richards

laeta (Schuithess)

3,69

4,88

1,32

sp. F

3,94

5,40

1,37

sp. M

4,20

1,76

0,42

Nesting is in burrows excavated with the aid of water in non-friable soil in horizontal or
sloping but not vertical ground. As these wasps tend to nest in close proximity to their natal nests

100

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

they form nesting aggregations. These aggregations may themselves be extensive or may though
small be abundant within an area so that where these wasps occur in large numbers there may
be several thousand nests in close proximity (Gess and Gess, 1988b). The burrows are
multicellular and are surmounted by a mud entrance turret (Gess and Gess, 1980, 1986 and
1988b). Each larva is provisioned with pollen and nectar presented in the form of a single firm
loaf. Pairing at water seems to be most common for Ceramius species, however, C. clypeatus
which has never been observed at water has been observed to pair on the forage plant, the male
descending on a foraging female.

The only flower visiting records for Palaearctic Ceramius species seem to be few and casual
and do not indicate any particular preferences (Richards, 1962 and 1963). Flower visiting records
for Afrotropical Ceramius species, that is for Ceramius species in southern Africa, are listed in
Table 2.

It should be noted that records in which digits are given in the column “number” are derived
from samples of insects collected and are therefore in no way indicative of numbers of observed
instances. This is particularly relevant in the cases of C. capicola Brauns, C. linearis Klug, and
C. lichtensteinii (Klug) which, being species very familiar to the authors and common in the
Grahamstown district, have been regularly observed by them foraging on “mesems” in large
numbers. Similarly the sight of C. lichtensteinii foraging in large numbers on Sphalmanthus cf.
bijliae (N.E.Br.) L.Bol. at Tierberg in the Prince Albert district in December 1987 was noted
with interest but only a small number of voucher specimens was collected. C. bicolor
(Thunberg), C. clypeatus Richards, C. socius Turner and C. jacoti Richards have similarly been
observed foraging in greater numbers than the numbers of specimens collected would indicate.
Such observations of large numbers of individuals visiting particular flowers are indicated in
Table 2 by an “m” in the “numbers” column.

The results of analyses of pollen from pollen loaves, by comparing pollen obtained from
these loaves with that from flowers growing in the vicinity of the nests, have been included in
Table 2 and are indicated as visits to flowers of a particular species by a female but instead of a
digit in the column “number” there is a “p” indicating “provision”. In the cases of C. clypeatus
Richards and C. braunsi Turner for which nests have not been found the pollen analysed was
taken from the crops of female wasps. These records are also indicated in Table 2 by a “p” in
the column “number”. Apart from indicating fidelity in flower visiting for the purpose of
collecting pollen for provisioning the pollen analyses serve to supplement and support flower
visiting records. This is of particular interest for species such as C. rex Saussure and
C. nigripennis Saussure for which flower visiting records are very few and for C. braunsi which
has been recorded from forage plants of two families.

The records demonstrate that species and species groups within the genus Ceramius exhibit
marked fidelity to flowers of a single family indicating that pollen and nectar are being obtained
from the same plants. A possible exception is C. braunsi which has been recorded from both
composite flowers and the flowers of Aspalathus spinescens Thunb. subsp. lepida (E.Mey.)
Dahlgren (Leguminosae: Papilionatae). Pollen from the crop of a female captured on
Aspalathus flowers was found to be entirely composite, indicating a fidelity to Compositae when
pollen gathering. Visits to Aspalathus flowers seem to have been solely for obtaining nectar.
Records are too few to establish whether visits to flowers other than composites for obtaining
nectar are habitual for C. braunsi. There are occasional records of casual visiting of flowers of
another family by C. lichtensteinii (Group 5) which has been collected on Blepharis

101

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

(Acanthaceae) and Senecio pterophorus (Compositae) and by C. capicola (Group 8) which has
been collected on Berkheya sp. (Compositae). However, pollen from pollen loaves of eight
Ceramius species, C. nigripennis Saussure, C. jacoti Richards, C. lichtensteinii (Klug), C. rex
Saussure, C. bicolor (Thunberg), C. socius Turner, C. linearis Klug and C capicola Brauns, and
from the crop of one species, C. clypeatus Richards, has never been found to contain pollen from
mixed families (Gess and Gess, 1980, 1986, 1988b and present paper; Table 2). It is therefore
considered that “mesems” are the habitual forage plants of C. lichtensteinii and C. capicola and
that visits to other plants are casual in nature. This opinion is in keeping with that expressed by
Cooper (1952) in a consideration of the records of flower visiting by Pseudomasaris in North
America.

The flower visiting pattern of southern African Ceramius derived from the records at
present available is:

Group 2a — Mesembryanthemaceae;

Group 2b — Leguminosae: Papilionatae;

Group 3 — Compositae;

Group 4 — Mesembryanthemaceae;

Group 5 — Mesembryanthemaceae;

Group 6 — Compositae;

Group 8 — Mesembryanthemaceae.

Gess (1965) when discussing the distribution of the genus Ceramius in southern Africa, at a
time when no flower visiting records were available, stated that “it is likely that the flowers visited
by Ceramius will prove to be low-growing Compositae and mesembryanthemums (Aizoaceae)
which following the winter rains, are such a striking feature of the semi-desert areas inhabited by
Ceramius in South Africa.” This prediction has been upheld for the species occurring in karroid
areas, of the 14 species for which records are available eight species are associated with
Mesembryanthemaceae and five with Compositae. The exception, C. clypeatus, which forages on
Aspalathus seems to be associated with Macchia (Fynbos) (Acocks, 1953 and 1975), more
particularly Dry Mountain Fynbos and Mesic Mountain Fynbos (Moll et al. , 1984). It is notable that
the “mesems” favoured by Ceramius species are most commonly white, pink or cream coloured.
Dark pink to cerise, red and yellow “mesems” do not seem to be favoured. It is also of note that the
composite flowers favoured by Ceramius species most commonly have entirely yellow or orange
flower heads.

The posture of the wasp when gathering pollen or nectar on a “mesem” flower or a
composite capitulum is to a large degree dependent on the relative sizes of the flower or
capitulum and the wasp visitor. A flower or capitulum of greater diameter than the length of the
wasp may readily be alighted upon (Fig. 3), however, when the diameter of the flower or
capitulum is considerably smaller than the length of the wasp a more specialised technique is
required. The flower or capitulum is grasped with the second and third pairs of legs and the
abdomen is curved down and under to act as a balance.

The pollen gathering method employed by a Ceramius species was most clearly determined
for C. braunsi Turner which was observed during the first week of October 1988 collecting pollen
from flowers of Arctotis laevis Thunberg and Athanasia trifurcata (L.) L. (both Compositae) on
a slope above the Clanwilliam Dam. The short curved fore legs were held beneath the wasps as
they rotated them one over the other to agitate the anthers and draw the pollen towards the

102

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Fig. 3. Ceramius braunsi Turner, collecting pollen, on a capitulum of Arctotis laevis Thunberg (Compositae) above the

Clanwilliam Dam, October 1988.

mouth for ingestion. The wasp whilst thus engaged receives a coating of pollen on its
undersurface (Fig. 4) and this is carried by it to the next capitulum which it visits.

Ceramius clypeatus Richards, when alighting on the small pea flowers of Aspalathus
spinescens Thunb. lepida (E.Mey.) Dahlgren (Leguminosae: Papilionatae) grasps the flower with
the second and third pairs of legs and curves the abdomen down beneath the flower aiding its
balance (Figs 5 and 6). The wasp always holds the alae whilst inserting its tongue at the base of
the standard to reach the nectary. Whilst it is thus engaged the carina is forced open and the
essential parts curve upwards to make contact with the front legs, which are held folded beneath
the wasp, and with the prosternum. A considerable amount of pollen is deposited on these hairy
surfaces (Fig. 7) and as the wasp invariably positions itself in this manner it is ideally suited to
transfer pollen from one flower to the stigma of another. During the second week of October
1987 and the first week of October 1988 Ceramius clypeatus with Masarina familiaris Richards
(Masaridae) were found to be the commonest insects working the large number of Aspalathus
spinescens lepida bushes on a hillside above the Clanwilliam Dam. Furthermore their daily
period of foraging activity was remarkably long, being from 9.30 am to 5.30 pm. One individual
of C. clypeatus alone was observed to visit forty flowers in a single foraging excursion.

When collecting pollen for provision Ceramius clypeatus grasps the alae with its second and
third pair of legs and balances itself in much the same manner as it does when alighting on a
flower preparatory to imbibing nectar. It ingests pollen directly from the anthers.

103

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Fig. 4. Ceramius braunsi Turner. Above: ventral view of anterior half of wasp showing pollen of Arctotis laevis Thunberg
(Compositae) on hairy underside (x 11). Below left: part of prosternum and base of front legs (x 30). Below right: boxed

area x 150.

104

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Fig. 5. Ceramius clypeatus Richards withdrawing from a
flower of Aspalathus spinescens Thunb. subsp. lepida
(E.Mey.) Dahlgren (Leguminosae: Papilionatae) above
the Clanwilliam Dam, October 1988.

Fig. 6. Simplified diagrammatic representation of Ceramius
clypeatus Richards (legs omitted) in nectar drinking
position on flower of Aspalathus spinescens Thunb. subsp.
lepida (E.Mey.) Dahlgren (Leguminosae: Papilionatae).

Table 2.

Flower visiting records for Ceramius Latreille in southern Africa.

WASP

SPECIES

FORAGE PLANT

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Ceramius Group 2a
C. cerceriformis Saussure

Mesembryanthemaceae

Mesembryanthemum
sensu law

Garies

FWG&WHRG

7/8. X. 85

Mesembryanthemum L.
M. crystallinum L.

Willowmore

CFJG

3LX.67

Psilocaulon N.E.Br.

P. acutisepalum (Berger)
N.E.Br.

WPi

Springbok

FWG&SKG

1.X.85

C. peringueyi Brauns

Mesembryanthemaceae

Psilocaulon N.E.Br.

P. acutisepalum (Berger)
N.E.Br,

WPi

Vredendal

FWG&SKG

30.ix.85

105

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Group 2b

C. clypeatiis Richards

Leguminosae

Aspalathus L.

A. spinescens Thunb.

subsp. lepida (E.Mey)
Dahlgren

Clanwilliam

FWG&SKG

26.ix.85

A. spinescens Thunb.

subsp. lepida (E.Mey)
Dahlgren

Clanwilliam

FWG&SKG

28.ix.85

A. spinescens Thunb.

subsp. lepida (E.Mey)
Dahlgren

Clanwilliam

FWG&SKG

7-14.X.87

A. spinescens Thunb.

subsp. lepida (E.Mey)
Dahlgren

Clanwilliam

SKG

7-14.X.87

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

FWG&SKG

3-7.X.88

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

DWG

3-7.X.88

A. spinescens Thunb.

subsp. lepida (E.Mey.)

FWG.SKG

Group 3

Dahlgren

F&M

Clanwilliam

&DWG

3-7.X.88

C. nigripennis Saussure

Compositae

Dimorphotheca Vaill.

ex. Moench.

D. sinuata DC
Pentzia Thunb.

P. suffruticosa (L.)

Springbok

SKG

9.X.85

Hutch, ex. Merxm.

Springbok

FWG&SKG

15-21.X.87

Berkheya Ehrh.
B. sp.

Springbok

FWG&SKG

15-21. X.87

B. fruiticosa (L.) Ehrh.

Springbok

14-15.X.87

Hirpicium Cass.

H. alienatus (Thunb.)

Druce

Springbok

30.X.87

C. jacoti Richards

Compositae

Pteronia L.

P. incana (Burm.) DC
Senecio L.

Barrydale

CFJG

1.X.67

S. rosmarinifolius L.f.

Oudtshoom

FWG.SKG,

HWG&RWG

7-12.xii.85

S. rosmarinifolius L.f.

Oudtshoom

SKG

7-12.xii.87

C. toriger Schulthess

Compositae

“blue rayed”

Clanwilliam

CDM

19.ix.66

C. braunsi Turner

Compositae

Clanwilliam

FWG&SKG

7-13. x.87

Athanasia L.

A. trifurcata (L.) L.

Clanwilliam

FWG&SKG

3-7.X.88

A ret Otis L.

A. laevis Thunb.

Clanwilliam

FWG&SKG

3-7.X.88

A. laevis Thunb.
Pentzia Thunb.

Clanwilliam

DWG

3-7.X.88

P. sp

—

Clanwilliam

DWG

3-7.X.88

composite

—

Clanwilliam

SKG

3-7.X.88

106

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

WASP FORAGE PLANT

SPECIES FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Leguminosae

Aspalathus L.

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

FWG&SKG

3-7.X.88

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

DWG

3-7.X.88

Group 4

C. beyeri Brauns

Mesembryantheraaceae

“mesem"

Grahamstown

FWG

16.i.69

Sphalmanthus N.E.Br.
5. cf. bijliae (N.E.Br.)

WPi

Pr. Albert

FWG,

26.xi.87-

L.Bol.

SKG&RWG

5.xii.87

Group 5

C. lichtensteinii (Klug)

Mesembryanthemaceae

Aridaria N.E.Br.
A. sp.

Grahamstown

FWG&SKG

7.xi.72

Mesembryanthemum L.
M. aitonis Jacq.
Ruschia Schwant.

Grahamstown

FWG

16.i.69

R. sp.

—

Grahamstown

FWG

ll.xii.68

R. sp.

—

Grahamstown

FWG

8.i.69

R. sp.

Grahamstown

FWG

30.xi.70

R. sp.

PuPi

—

Alicedale

FWG

2.xii.70

R. sp.

PuPi

—

Alicedale

JGHL

2.xii.70

R. sp.

—

Grahamstown

SKG

"raesem"

PuPi

Grahamstown

29.xi.79

“mesem”

PuPi

Grahamstown

26.X.77

“mesem”

Grahamstown

DWG

6.i.81

“mesem”

Grahamstown

FWG

Li. 81

FWG

30.xi.81

“mesem”

F&M

Kommadagga

FWG&SKG

l.xii.85

“mesem”

F&M

Kommadagga

FWG&SKG

l.xii.85

“mesem”

F&M

Kommadagga

FWG&SKG

l.xii.85

“mesem”

WPi

F&M

Grahamstown

FWG&SKG

xii.85-i.86

Sphalmanthus N.E.Br.

S. cf. bijliae (N.E.Br.)

WPi

Pr. Albert

FWG, SKG

26.xi.87-

L.Bol.

&RWG

5.xii.87

Compositae

(it was noted at the

time that there were

no mesems in

flower) Senecio L.

5. pterophorus DC

Grahamstown

29.xi-

2.xii.79

Acanthaceae

Blepharis Juss.

B. capensis (L.f.) Pers.

Grahamstown

FWG&DWG

15.i.81

B. capensis (L.f.) Pers.

Grahamstown

FWG&DWG

3.ii.81

B. capensis (L.f.) Pers.

Waterford

FWG&RWG

25.xi.87

Group 6

C. rex Saussure

Compositae

Berkheya Ehrh.

B. spinosissima (Thunb.)

Wind.

Springbok

FWG

15-21.X.87

B. spinosissima (Thunb.)

Willd.

Springbok

SKG

15-21.X.87

107

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

WASP FORAGE PLANT

SPECIES FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

“composite” including

Springbok

SKG

15-21.X.87

B. spinosissima (Thunb.)

Wind.

Group 8

C. bicolor (Thunberg)

Mesembryanthemaceae

Psilocaulon N.E.Br.

P. acutisepalum (Berger)

N.E.Br.

WPi

F&M

Klawer

FWG&WHRG

14/15.X.65

P. acutisepalum (Berger)

N.E.Br.

WPi

F&M

Klawer

FWG&SKG

27.ix.85

P. acutisepalum (Berger)

N.E.Br.

WPi

Klawer

FWG&SKG

27.ix.85

P. acutisepalum (Berger)

N.E.Br.

WPi

Klawer

SKG

29.ix.85

P. acutisepalum (Berger)

N.E.Br.

WPi

Springbok

SKG

4.x. 85

“mesem”

Springbok

SKG

4.x. 85

“mesems”

—

Clanwilliam

CDM

19.ix.66

C. socius Turner

Mesembryanthemaceae

Psilocaulon N.E.Br,

P. acutisepalum (Berger)

N.E.Br.

WPi

F&M

Clanwilliam

FWG&SKG

28.ix.85

“mesem”

Montagu

FWG

3.xii.86

“mesem”

Montagu

RWG

3.xii.86

“mesem”

Montagu

SKG

3.xii.86

“mesem”

Touws River

FWG

4.xii.86

“mesem”

Montagu

FWG&SKG

4.xii.86

P. acutisepalum (Berger)

N.E.Br.

WPi

F&M

Clanwilliam

FWG&SKG

7-14. X. 87

P. acutisepalum (Berger)

N.E.Br.

Clanwilliam

SKG

7-14.X.87

C. linearis Klug

Mesembryanthemaceae

Aridaria N.E.Br.

A. sp.

Grahamstown

FWG&SKG

17.X.72

A. dyeri L.Bol.

Alicedale

FWG

2.xii.70

A. plenifolia (N.E.Br.)

Steam

Alicedale

JGHL

2.xii.70

A. plenifolia (N.E.Br.)

Steam

Alicedale

FWG

16.xii.71

Malephora N.E.Br.

M. sp.

Grahamstown

FWG&SKG

26.X.72

Mesembryanthemum L.

M. aitonis Jacq.

Grahamstown

FWG

30.xii.71

M. aitonis Jacq.

Grahamstown

FWG

28.xi.82

Ruschia Schwant.

R. sp.

PuPi

—

Alicedale

JGHL

2.xii.70

R. sp.

—

Grahamstown

JGHL

5.xii.69

Drosanthemum Schwant.

D. floribundum (Hw.)

Schwant.

Grahamstown

29.xi.76

108

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

WASP FORAGE PLANT

SPECIES FAMILY GENUS AND SPECIES SEX LOCALITY DATE

D. ftoribundurn (Hw.)

Schwant.

“mesem"

‘■mesem”

“mesems”

“mesem”

C. capicola Brauns

Mesembryanthemaceae

Aridaria N.E.Br.

A. plenifolia (N.E.Br.)
Steam

Mesembryanthemum L.
M. aitonis Jacq.
Mestoklema N.E.Br.

M. tuberosum (L.)
N.E.Br.

Ruschia Schwant.

R. sp.

Drosanthemum Schwant.
D. ftoribundurn (Haw.)

Schwant.

“mesems”

Compositae

Berkheya Ehrh.

B. sp.

Grahamstown

SKG

10.xii.74

Grahamstown

FWG

13.i.81

Grahamstown

FWG

30.xi.81

F •

Grahamstown

FWG&SKG

22.X.81

F&M

Grahamstown

FWG&SKG

F&M

—

Kommadagga

FWG&SKG

l.xii.85

—

Kommadagga

FWG&SKG

l.xii.85

—

Kommadagga

FWG&SKG

l.xii.85

—

Alicedale

FWG&JGHL

2.xii.70

Grahamstown

FWG

6.ii.69

PuPi

—

Grahamstown

FWG

6.ii.69

PuPi

Grahamstown

FWG

18.ii.69

Grahamstown

FWG

27.xi.-

ll.xii.68

Grahamstown

FWG

8-16.i.69

Grahamstown

FWG

12.xi.-

22.xii.69

Grahamstown

FWG

30.xi.70

Grahamstown

FWG

19.xii.71

—

Grahamstown

JGHL

4.xii.69

PuPi

—

Alicedale

JGHL

2.xii.70

Grahamstown

SKG

—

Grahamstown

FWG&SKG

—

Kommadagga

FWG&SKG

l.xii.85

—

Kommadagga

FWG&SKG

l.xii.85

—

Kommadagga

FWG&SKG

l.xii.85

Hofmeyr

DWG

17.xi.87

Thaba Nchu

CJFG

l.xii.52

(Orange Free
State)

Jugurtia Saussure

Jugurtia is an Old World genus occurring in the Palaearctic Region, bordering the
Mediterranean and extending eastwards into Armenia and south western Persia, and in the
Afrotropical Region, in Nigeria and southern Africa (Richards, 1962). In southern Africa its
distribution parallels that of Ceramius. Similarly the flight period in this region is, like
that of Ceramius, from September to March, the period for individual species being more
limited.

Compared with other southern African masarids Jugurtia species are medium sized,
ranging in length from 7-11 mm. The relative tongue length in Jugurtia is greater than in
Ceramius, however, the actual tongue length is similar (Table 1).

109

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Fig. 7. Ceramius clypeatus Richards. Above: ventral view of anterior half of wasp showing area of impact with anthers of
Aspalathus spinescens Thunb. lepida (E.Mey.) Dahlgren (Leguminosae: Papilionatae) (x 12). Below left: prostemum and
base of front legs (x 30). Below right: boxed area x 150.

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Nesting of only one species, Jugurtia confusa Richards, has been investigated (Gess and
Gess, 1980). This wasp nests in burrows excavated with the aid of water in horizontal ground.
Like Ceramius, Jugurtia forms nesting aggregations. The burrows are multicellular and
are surmounted by a mud entrance turret. Each larva is provisioned with pollen and nectar
presented in the form of a single loaf.

The only flower visiting records for Palaearctic Jugurtia seem to be those of Bequaert (1940
in Richards, 1962) from Algeria. Richards states that “Jugurtia as far as the scrappy records go
is not attached to any particular family”. However, it can be seen from the flower visiting records
for Jugurtia species from southern Africa listed in Table 3 that preferences similar to those of
Ceramius are exhibited, Mesembryanthemaceae and Compositae appearing to be favoured. The
single record of a male J. confusa on Acacia karroo may be considered as casual visiting. There
is no evidence from pollen loaves examined that this species provisions with any pollen other
than that obtained from “mesem” flowers (Gess and Gess, 1980).

The available records are at present too few for it to be possible to note any colour
preferences for Jugurtia species.

Table 3.

Flower visiting records for Jugurtia Saussure in southern Africa.

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COL-

LECTOR

DATE

J. confusa Richards

Mesembryanthemaceae

Drosanthemum Schwant.
D. parvifolium (Haw.)

Schwant.

Grahamstown

SKG

8-xii.76

—

Grahamstown

—

Leguminosae; Mimosoidea

Acacia Mill.

A. karroo Hayne.

Grahamstown

FWG

10.ii.77

J. braunsi (Schulthess)

Mesembryanthemaceae

“mesem”

Springbok

FWG&SKG

1.x. 85

Drosanthemum Schwant.
D.sp.

Springbok

SKG

15-21.X.87

Compositae

Senecio L.

S. sp.

Springbok

FWG&SKG

10-12. X. 88

J. braunsiella (Schulthess)

Compositae

Lasiospermum Lag.

L. bipinnatum (Thunb.)

Druce

Grahamstown

FWG&SKG

12.X.77

Pteronia L.

P. paniculata Thunb.
Senecio L.

Grahamstown

FWG&SKG

27.X.72

S. rosmarinifolius L.f.

Oudtshoom

FWG&RWG

7-12.xii.86

S. rosmarinifolius L.f.
Felicia Cass.

Oudtshoom

SKG

F. sp.

Springbok

SKG

15-21. ix. 87

J. polita Richards

Compositae

Senecio L.
S. sp.

Cradock

OWR

25.ix.52

(Richards, 1962)

Ill

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Masarina Richards

The genus Masarina is endemic to southern Africa and includes only four described
species (Gess and Gess, 1988a). In distribution it seems to be centred in the South Western
Cape, only one species, M. familiaris Richards, extending to the east as far as Willow-
more.

Compared with other southern African masarids Masarina species are medium sized,
ranging in length from 7,0-11,5 mm. The tongue length is, for masarids, moderate (Table 1).

Nesting is known only for M. familiaris (Gess and Gess, 1988a). There seems to be a
tendency for an individual wasp to nest in close proximity to her natal nest and therefore for
the development of nesting aggregations though at the two nesting sites where the
investigations were made the numbers of nests were not great. It is possible that the time of
the study, which was opportunistic, was not at the peak nesting period. The nests are
burrows excavated with the aid of water in non-friable vertically presented soil. The burrows
are multicellular and a downwardly curved mud turret is constructed at the entrance. The
provision which is composed of pollen and nectar is moist and sticky and although it forms
a single mass it does not form a discrete loaf as does the drier provision of Ceramius and
Jugurtia.

Flower visiting records are available for three species of Masarina, M. familiaris, M. mixta
Richards and M. strucki Gess. These records are given in Table 4. All the records for
M. familiaris are for yellow flowered Aspalathus species (Leguminosae: Papilionatae). Pollen
from the provision of M. familiaris was found to be exclusively of the Aspalathus type (Gess and
Gess, 1988a).

During the second week of October 1987 and the first week of October 1988 Masarina
familiaris and Ceramius clypeatus were found to be the commonest insects working the large
number of Aspalathus spinescens Thunb. subsp. lepida (E.Mey.) Dahlgren (Leguminosae:
Papilionatae) bushes on a hillside above the Clanwilliam Dam. The daily period of foraging
activity was remarkably long, being from 9.30 am to 5.30 pm. This wasp adopts a completely
different stance on the small “pea flowers” from that adopted by the larger wasp C. clypeatus
(Figs 5 and 6). Instead of alighting on the alae it alights on the standard in such a way that it faces
downwards towards the centre of the flower (Figs 8 and 9). When imbibing nectar the wasp
inserts its tongue into the flower at the base of the standard to reach the nectary causing the
Carina to open and the essential parts to curve upwards to come firmly into contact with the frons
of the wasp (Fig. 9) so that it receives a considerable load of pollen (Fig. 10). As the wasp always
positions itself in the same manner it is ideally suited to transfer pollen from one flower to the
stigma of another.

When collecting pollen for provision M. familiaris ingests it directly from the anthers.

Although M. familiaris was observed in abundance on flowers of Aspalathus spp. in the
second week of October 1987 and the first week of October 1988 and samples of 43 and 57
specimens were taken only one instance of Masarina mixta Richards visiting Aspalathus flowers
was recorded. However, in October 1988 M. mixta was found to be a not uncommon visitor to
flowers of Wahlenbergia sp. A, a sample of 24 specimens having been taken. More records are
required, however, a possible preference for Wahlenbergia is indicated.

As only one record is available for M. strucki no comment can be made.

112

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Fig. 8. Masarina familiaris Richards on flower of Aspa-
lathus spinescens Thunb. subsp. lepida (E.Mey.) Dahlgren
(Leguminosae: Papilionatae).

Fig. 9. Simplified diagrammatic representation of Masarina
familiaris Richards (legs omitted) in nectar drinking
position on flower of Aspalathus spinescens Thunb. lepida
(E.Mey.) Dahlgren (Leguminosae: Papilionatae).

Table 4.

Flower visiting records for Masarina Richards in southern Africa.

WASP

FORAGE PLANT

SPECIES

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COL-

LECTOR

DATE

M. familiaris Richards

Leguminosae: Papilionatae

Aspalathus L.

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

FWG&SKG

8-13.X.87

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Paleisheuvel

FWG&SKG

8-13. X.87

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

FWG&SKG

3-7.X.88

A. spinescens Thunb.

subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

DWG

3-7.X.88

113

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

WASP

FORAGE PLANT

SPECIES

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COL-

LECTOR

DATE

A. spinescens Thunb.
subsp. lepida (E.Mey.)
Dahlgren

F&M

Clanwilliam

FWG.SKG

&DWG

3-7.X.88

A. vulnerans Thunb.

Clanwilliam

FWG&SKG

8-I3.X.87

A. sp.

Clanwilliam

FWG&SKG

8-I3.X.87

A. sp./spp.

Clanwilliam

SKG

8-I3.X.87

M. mixta Richards

Campanulaceae

Wahlenbergia
W. sp. A

Clanwilliam

DWG

3-7.X.88

W. sp. A

Clanwilliam

FWG&SKG

3-7.X.88

Leguminosae: Papilionatae

Aspalathus L.

A. spinescens Thunb.
subsp. lepida (E.Mey.)
Dahlgren

Clanwilliam

FWG&SKG

8-I3.X.87

M. strucki Gess

Sterculiaceae

Hermannia L.

H. disermifolia Jacq.

Springbok

20.viii.85

Celonites Latreille

Celonites is an Old World genus occurring in the Palaearctic Region in the countries
bordering the Mediterranean Sea, northwards to Switzerland and southern Germany and
eastwards to Transcaspia and south western Persia, and in the Afrotropical Region in north east
Africa and the Cape Province of South Africa (Richards, 1962). In the Cape Province its
distribution seems to be similar to that of Ceramius and Jugurtia: Namaqualand, the South
Western Cape, the Little Karoo, the southern parts of the Great Karoo and the Eastern Cape,
no further east than the Great Fish River and north to Aliwal North on the Orange River. One
species, C. promontorii Brauns has been recorded from Thaba Nchu (Orange Free State)
(collector C.F.Jacot Guillarmod recorded in Richards, 1962). Collecting records indicate a flight
period in the Cape Province of October to December.

Compared with other southern African masarids Celonites species are medium sized ranging
in length from 7-11,5 mm. The tongue length is relatively long (Table 1).

Little is known concerning the nesting behaviour of Celonites due undoubtedly to the
cryptic nature of its nests. Brauns (1913) recorded that Celonites andrei Brauns constructs fragile
mud cells in groups attached to each other lengthwise and situated on the underside of stones
and on twigs. Only one group of Celonites cells has been discovered by the authors. This group,
attached to a stem of a Peliostomum plant, consists of three robust mud cells cemented together
lengthwise and enclosed in a mud envelope.

Richards (1962) lists flower visiting records for seven Palaearctic species of Celonites and
one Afrotropical species. He comments that '"Celonites does not seem to be attached to any one
family of plants but nearly all records so far have been made very casually”. A few additional
records are given in Richards (1969), which might indicate some preference for Labiatae. Flower
visiting records for seven southern African species are given in Table 5. Records for C. capensis
Brauns certainly do not indicate any specificity in flower visiting. Those available for C. wheeled

114

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Fig. 10. Masarina familiaris Richards. Above: dorsal view of head showing area of impact with anthers of Aspalathus
spinescens Thunb. lepida (E.Mey.) Dahlgren (Leguminosae: Papilionatae) (x 20). Below left: area of impact x 30. Below

right: boxed area x 150.

115

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Brauns and C. promontorii Brauns are too few for comment. The sample of specimens of
C. wahlenbergiae Gess and C. bergenwahliae Gess taken from Wahlenbergia flowers is small,
however, during intensive collecting at various sites in the Clanwilliam district during the periods
8-14. X. 1987 and 3-7. x. 1988 these wasps were seen to be relatively common in association with
flowering Wahlenbergia spp. (samples of 28 specimens of C. wahlenbergiae and 17 specimens
C. bergenwahliae were taken) but were found no where else indicating a probable association.
Records for C. andrei Brauns, C. clypeatus Brauns and C. peliostomi Gess indicate a clear
preference by these species for blue/violet flowers of Aptosimum spp. (Scrophulariaceae) (Figs
11 and 12) and the purple/violet flowers of the closely related Peliostomum virgatum
(Scrophulariaceae) (Figs 13, 14 and 15). Particularly noteworthy is the record of 38 females of
C. peliostomi collected in flowers of Peliostomum virgatum in the Hester Malan Nature Reserve,
Springbok (FWG and SKG, 15-21. x. 87) and the record of 24+ females of C. clypeatus collected
in flowers of Aptosimum depressum 28 km from Grahamstown (FWG, SKG, DWG and RWG,
13.x.- 3.xii.81). Also of note is the record of the latter species collected in flowers of
Peliostomum virgatum in the Hester Malan Nature Reserve, showing a constancy of preference
between two very widely separated populations.

Flowers of Aptosimum spinescens (Thunberg) Weber in the Hester Malan Nature Reserve,
Springbok were observed by the authors for an hour, from 10.30-11.30 am on 20.x. 87 and visiting
insects were recorded. During this period 12 instances of Celonites entering flowers were recorded
and one instance of an Anthophora, which was of too great a diameter to enter these flowers,
hovering at the mouth of a flower. A sample of six specimens of Celonites was taken as was the
Anthophora. No other insects approached the flowers. The Celonites taken were 5 females of
C. peliostomi and 1 female of C. andrei. On the same day flowers of Peliostomum virgatum
E.Mey were observed from 12.30-1.15 pm. During this period 22 instances of Celonites entering
flowers were recorded. No other insects approached the flowers. A sample of 16 specimens was
taken. In all during the period 15-21. x. 87 a sample of 45 Celonites entering flowers of P. virgatum
was taken. These were 38 female and 3 male C. peliostomi, one female C. andrei and one female
and two male C. clypeatus. No other insects were observed in, on or approaching these flowers
although they were observed at all times of the day. Flowers of Aptosimum depressum in the
Grahamstown district observed on various occasions for lengthy periods were found to be visited
only by C. clypeatus and the very small Quartinioides tarsata Richards.

Pollen from the crops and backs of two females of C. peliostomi, one from a flower of
A. spinescens and one from a flower of P. virgatum was examined. All the pollen was of the same
type and when compared with that from the two flowers was found to match. Similarly pollen
from the crop of a female C. clypeatus was compared with pollen from Aptosimum depressum
which it was found to match. These records are indicated in the relevant table with a “p”, as
were provision records for the other genera.

The flowers of Aptosimum species and Peliostomum species are gullet flowers (Figs 16 and
17). The corolla is very narrow in the basal region protecting the nectaries from all but minute
or long tongued visitors. The greater part of the corolla tube is wider but again restricts the size
of visitors wishing to enter the flower. There are four stamens in two pairs, a pair with relatively
long filaments and a pair with relatively short filaments. The shorter pair of stamens is sterile in
some species of Aptosimum. The anthers are adpressed in pairs and positioned dorsally in the
flower which has its long axis horizontal. The style lies in a grove dorsally and when fully

116

GESS & GESS: FLOWER VISITING BY MAS ARID WASPS IN SOUTHERN AFRICA

Fig. 12. Aptosimum depressum Burch, ex Benth. (Scrophulariaceae).

117

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Fig. 13. Peliostomum virgatum E.Mey. ex Benth. (Scrophulariaceae).

Fig. 14 and 15. Peliostomum virgatum E.Mey. ex Benth. (Scrophulariaceae).

118

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Fig. 16. Simplified diagrammatic representation of longitudinal section of flower of Aptosimum depressum Burch, ex Benth.
(Scrophulariaceae) and profile of Celonites clypeatus Brauns (legs and wings omitted) to demonstrate flower/wasp fit.

extended projects at the mouth of the flower. The tip curves downwards so that the stigmatic
surface is downwardly presented.

Celonites spp. when entering these flowers, especially those of Peliostomum, fit snugly so
that they brush against the stigma and also receive a pollen load, particularly on the hind end of
the thorax (Fig. 18). It is even likely that in entering a flower with ripe but not yet dehisced
anthers these wasps trigger the dehiscence of the anthers. Having reached the base of the wide
portion of the corolla tube they are well able to reach the nectaries. The narrow basal portion

119

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

of the corolla tube of P. virgatum is 4 mm long, that of A. spinescens 4,5 mm long and that of
A. depressum is 5 mm long. The tongue of C peliostomi which visits P. virgatum and
A. spinescens is 4,8-5 mm long and that of C. clypeatus which visits P. virgatum and
A. depressum is 5,8 mm long.

Barth (1985: 61, Fig. 16) gives a short description illustrated with line drawings of the alleged
pollen collecting apparatus of Celonites abbreviatus (Villers) and the manner in which it is used.
The story is ingenious but unfortunately when the authors examined a specimen of
C. abbreviatus they found that there were no “button-ended collecting bristles on the front
surface” of the head. The frons is hairy but the hairs taper towards their tips which are curved.

It seems most likely that as has been observed for Ceramius and Quartinioides (present
paper) and for a pollen ingesting paracolletine bee (Houston, 1981) pollen is simply drawn
towards the mouth by the front legs and that it is then ingested. Pollen from dehisced anthers of
Aptosimum and Peliostomum flowers would be freely available for collection from the “floor”
of the horizontal corolla tube. Similarly as the flowers of the Wahlenbergia spp. visited by
Celonites spp. are erect pollen from dehisced anthers would be freely available for collection
being retained within the cup-like corolla. The description of pollen collecting by Celonites in
Barth, disregarding the inaccurate description of the hairs, is surely related to grooming
behaviour rather than deliberate pollen collection.

10mm

Fig. 17. Simplified diagrammatic representation of longitudinal section of flower of Peliostomum virgatum E.Mey. ex Benth.
(Scrophulariaceae) and profile of Celonites peliostomi Gess (legs and wings omitted) to demonstrate flower/wasp fit.

120

GESS & GESS; FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Fig. 18. Celonites peliostomi Gess. Left: dorsal view of posterior end of thorax and anterior end of abdomen showing
deposition of pollen (x 30). Right: boxed area x 150.

Table 5.

Flower visiting records for Celonites Latreille in southern Africa.

WASP

SPECIES

FORAGE PLANT

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COL-

LECTOR

DATE

C. andrei Brauns

Scrophulariaceae

Aptosimum Burch.

A. spinescens (Thunb.)
Weber

Springbok

SKG

15-21. X.87

Peliostomum Benlh.

P. virgatum E.Mey ex
Benth.

Springbok

FWG&SKG

15-21. X.87

C. peliostomi Gess

Scrophulariaceae

Aptosimum Burch.

A. spinescens (Thunb.)
Weber

Springbok

FWG&SKG

15-21.X.87

A. spinescens (Thunb.)
Weber

Springbok

FWG&SKG

15-21.X.87

A. linear e Marloth &
Engl.

Springbok

SKG

15-21.X.87

121

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

WASP

FORAGE PLANT

SPECIES

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COL-

LECTOR

DATE

Peliostomum Benth.

P. virgatum E.Mey ex
Benth.

■ F

Springbok

FWG&SKG

I5-21.X.87

Springbok

FWG&SKG

15-21.X.87

P. virgatum E.Mey ex
Benth,

Springbok

FWG&SKG

I5-21.X.87

P. virgatum E.Mey ex
Benth.

Springbok

-.X.1987

C. clypeatus Brauns

Scrophulariaceae

Aptosimum Burch.

A. depressum Burch, ex

Benth.

24-1-

Grahamstown

FWG.SKG

13.x-

DWG&RWG

3.xii.81

A. depressum Burch, ex

Benth.

Grahamstown

FWG&SKG

22-30.X.81

A. depressum Burch, ex
Benth.

Peliostomum Benth.

P. virgatum E.Mey ex

Grahamstown

FWG&SKG

30.X.81

Benth.

Springbok

FWG&SKG

15-21.X.87

Springbok

SKG

15-21.X.87

C capensis Brauns

Scrophulariaceae

Phyllopodium Benth.
P. cuneifolium (L.f.)

Benth,

Grahamstown

DWG

9-14.xii.82

Boraginaceae

Ehretia P.Br.

E. rigida (Thunb.)

Druce

Grahamstown

FWG&SKG

26.X.77

Geraniaceae

Pelargonium L'Herit.
P. mvrrhifolium (L.)

L'Herit.

Oudtshoom

CFJG

10.X.72

Compositae

Berkheya Ehrh.

B. sp.

Riebeek East

FWG

22.xi.82

B. sp.

Oudtshoom

FWG

9-12.xii.86

C. wheeleri Brauns

Compositae

Berkheya Ehrh.

B. cf. spinosa (L.f.)

Pr. Albert

FWG&SKG

26.xi-

Druce

5.xii.87

C. promontorii Brauns

Compositae

Berkheya Ehrh.
B. sp.

Thaba Nchu

CFJG

Lxi-52

(OFS)

(Richards, 1962)

B. cf. spinosa (L.f.)

Pr. Albert

SKG

26. xi.'

Druce

5.xii.87

Senecio L.

S. rosmarinifolius L.f.

Oudtshoom

FWG

7-8.xii.86

C. wahlenbergiae Gess

Campanulaceae

Wahlenbergia Schrad. ex

Roth.

W. sp. A

Clan william

FWG&SKG

14.x. 87

W. sp. A

Clanwilliam

FWG&SKG

3-7.X.88

W. sp. A

Clan william

DWG

3-7.X.88

122

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COL-

LECTOR

DATE

C. bergenwahliae Gess

Campanulaceae

Wahlenbergia Schrad. ex
Roth.

W. sp. B

Clanwilliam

FWG&SKG

6.x. 88
6.x. 88

Quartinia Ed. Andre

The genus Quartinia Ed. Andre is an Old World genus occurring in the Palaearctic Region
bordering the Mediterranean Sea and extending eastwards into Asiatic Russia and India, and in
the Afrotropical Region in the South Western Cape, Southern to South Eastern Cape,
Namaqualand and Namibia (S. W. A.) (Richards, 1962 and label data, Albany Museum). In
numbers it is a relatively large genus.

Compared with other southern African masarids Quartinia species are small ranging in
length from 2,5-6, 5 mm. The tongue is moderately long. It is folded only once on itself as the
glossal sac is relatively long.

There are no nesting accounts for Quartinia. Of interest, therefore, is the observation that
Quartinia species have been observed “making burrows in the ground” (C. F. Jacot Guillarmod,
pers. com.).

In the Palaearctic Region six species have been collected on Compositae and two species on
Chenopodiaceae (Richards, 1962). Flower visiting records for the Afrotropical Region are given
in Table 6. Seven of the ten species for which records are available have been recorded from
Mesembryanthemaceae only, one species from Aizoaceae, one from Mesembryanthemaceae
and Campanulaceae and one from Campanulaceae only. Records of particular interest are those
for species which were visiting Wahlenbergia spp. (Campanulaceae). Patches of low growing
Wahlenbergia sp. A at the Clanwilliam Dam were observed for insect visitors during the period
3-7. X. 88. The most common visitors were Quartinia spp. and Quartinioid.es spp. The plants
which were in full flower were approximately 10 cm high. The deep violet flowers which are held
erect are 5 mm high with the upper part of the corolla tube 1,6 mm wide and the lower part
0,8 mm wide (Fig. 19). In the bud and the newly opened flower the receptive surfaces of the
three stigmatic lobes are closely adpressed to form a knob-like tip to the style. The upper two
fifths of the style are hairy particularly at the lower end where the hairs are short and robust and
form a distinct collar (Fig. 19a). The anthers dehisce introrsely before the bud opens. When the
flower opens the pollen coats the upper part of the style being supported by the collar and giving
the whole a club-like appearance. After the flower has been open some little while the hairs
supporting the pollen disappear and the pollen falls being retained within the corolla. The style
lobes then separate presenting their receptive surfaces (Fig. 19b). The wasps when visiting the
flowers alight on the outwardly curved corolla lobes before entering so that when they enter a
newly opened flower their dorsal surfaces brush against the pollen clad style and pollen is lodged
principally on the hind end of the thorax and the anterior end of the abdomen. When a wasp
comes from such a newly opened flower and then enters a flower in which the stigmatic lobes
have spread out pollen will be transferred from it to the stigma. Quartinia spp. are particularly
well suited to effect pollination as they fit the flowers very snugly (Fig. 19c).

123

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

Table 6.

Flower visiting records for Quartinia Ed. Andre in southern Africa.

WASP

SPECIES

FORAGE PLANT

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Q. persephone Richards

Mesembryanthemaceae

Psilocaulon N.E.Br.

P. acutisepalum (Berger)

Clanwilliam/

N.E.Br.

Campanulaceae

Wahlenbergia Schrad.
ex Roth.

Klaver

FWG&SKG

27.ix.65

IV. sp. A

Clanwilliam

DWG

3-7.X.88

W. sp. D

Clanwilliam

DWG

5-6.X.88

Q. parcepunctata Richards

W. sp. D

Campanulaceae

Wahlenbergia Schrad.
ex. Roth.

Clanwilliam

FWG&SKG

5-6.X.88

W. sp. A

Clanwilliam

DWG

3-7.X.88

W. sp. A

Clanwilliam

FWG&SKG

3-7.X.88

W. sp. D

Q. ochraceopicta Schulthess

Mesembryanthemaceae

Clanwilliam

FWG&SKG

5-6.X.88

Q. punctulatum Schulthess

“mesem”

Mesembryanthemaceae

Mesembryanthemum L.

Aus (Namibia)

RET

(Turner,

1939)

M. crystallinum L.

—

Aus (Namibia)

RET

(Turner,

1939)

M. crystallinum L.

—

Matjesfontein

RET

(Turner,

1939)

Q. media Schulthess

M. crystallinum L.
Mesembryanthemaceae

Pr. Albert
Road

RET

(Turner,

1939)

Mesembryanthemum

crim-

son

—

Worcester

RET

(Turner,

1939)

Q. vegipunctata Schulthess

Q. atra Schulthess

Mesembryanthemaceae

RET

(Turner,

1939)

(Turner.

Q. sp. A

Mesembryanthemum

Mesembryanthemaceae

Polymita N.E.Br.

P. albiflora (L.Bol.)

RET

1939)

Q. sp. B

L.Bol.

Mesembryanthemaceae

Prenia N.E.Br.

P. sladeniana (L.Bol.)

Springbok

6.xi.87

L.Bol.

—

Springbok

17.X.87

124

GESS & GESS: FLOWER VISITING BY MAS ARID WASPS IN SOUTHERN AFRICA

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Q. jocasta Richards

Aizoaceae

Galenia L.

G. filiformis (Thunb.)

N.E.Br. —

Springbok

3.xi.87

Fig. 19. (a and b). Simplified diagrammatic representations of longitudinal sections of two flowers of Wahlenbergia sp. A:
(a) freshly opened with stigmatic lobes closely adpressed and pronounced collar of hairs and (b) open for some time with
stigmatic lobes separated and collar of hairs withered away. (c). Simplified diagrammatic representation of Quartinia
parcepunctata Richards (legs and wings omitted) to demonstrate flower/wasp fit.

Quartinioides Richards

The genus Quartinioides Richards is endemic to southern Africa where it is recorded from
the South Western Cape, the Southern Cape, the Eastern Cape, Namaqualand, Namibia
(S. W. A.), Lesotho (Basutoland) and Bulawayo, Zimbabwe (Southern Rhodesia) (Richards,
1962). In numbers of species it is a relatively large genus. Richards (1962) lists 38 species but
states that he expects that there are many more species. The species here listed by letter from
A-T represent species which could not be determined and must include many undescribed
species. Clearly a revision of the genus is required, however, for the present purpose, that of
determining the flower visiting pattern of the genus, names are not essential.

125

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 5. JULY 1989

Figs 23-25. Quarlimokk’s Icteia (Schulthcss) in flower of Stoebena sp. (Mcscmbryanthcmaccae) at Aggcncys,

Bushmanland. October 1988.

Figs 26-28. QiHininioides sp. M in flower of Wtililenhergiu sp. C (Campanulaceae) at Anenous. Namaqualand, October

1988.

126

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

Compared with other southern African masarids Quartinioides species are small,
2, 5-5,0 mm in length. The greatest relative tongue length (1,37 x body length) is found in
Quartinioides (Table 1 and Fig. 2). As the glossal sac is relatively short the tongue when
retracted is folded several times upon itself. That not all species of Quartinioides have
remarkably long tongues is demonstrated by (2- sp. M in which the tongue is less than half the
body length (Table 1). This remarkable variation in tongue length seems to be related to the
nature of the flowers visited, the two remarkably long tongued species listed being visitors to
“mesems” and the relatively shorter tongued species being a visitor to Wahlenbergia.

There are no nesting accounts for Quartinioides . Of interest, therefore is the observation of
Quartinioides sp. H excavating burrows in sand in the coastal dunes at McDougal Bay near Port
Nolloth on the west coast (Gess and Gess, 1985, unpublished field notes). The nests as might be
expected were not surmounted by entrance turrets. Pairing on flowers has been observed.

Flower visiting records for 30 species of Quartinioides are listed in Table 7. There appears
to be, as also noted for Ceramius and Jugurtia, a marked association between these wasps and
flowers of the families Mesembryanthemaceae (18 spp.) (Figs 20-25) and Compositae (7 spp.),
favoured colours seeming to be light shades for the Mesembryanthemaceae and yellow to orange
for Compositae. Also of interest is a possible preference by some species for Wahlenbergia spp.
(Campanulaceae) (3 spp.) (Figs 26-28), a preference which is also indicated for some species of
Celonites (Table 5) and Quartinia (Table 6). Of particular interest is the association of
Q. antigone Richards with Aloe striata (Liliaceae), 24 females and 5 males being indicative of
more than casual visiting.

When collecting nectar from “mesem” flowers Quartinioides insert themselves deeply into
the flower being visited, however, when collecting pollen from these and composite capitula they
stand on their second and third pairs of legs on the “surface” of the flower or capitulum with the
first pair of legs beneath them being rotated in such a manner that the anthers are agitated and
the pollen is drawn forwards for ingestion in a similar manner to that observed for Ceramius
braunsi. When visiting Wahlenbergia flowers they alight on the lip of the corolla and then enter.
Being smaller than Quartinia they are able to turn around in the flower and may therefore
emerge head first. Although they do not “fit” the flowers as snugly as Quartinia, through their
activities in the flowers they probably serve effectively as agents transfering pollen from one
flower to the stigma of another.

Quartinioides are generally the commonest visitors to the plants which they patronize and
this combined with their usually being very numerous makes them potentially of importance as
pollinators to their forage plants.

Table 7.

Flower visiting records for Quartinioides Richards in southern Africa.

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Q. antigone Richards

Liliaceae

Aloe L.

A. striata Haw.

PiO

Pr. Albert

FWG

26. xi-

5.xii.87

127

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

T rsr' A T TT\/ LULLEL-

LOCALITY TOR

DATE

Q. helichrysi Richards

Compositae

Helichrysum Mill.

H. fruticans (L.) D.Don.

Lesotho

CFJG

28.xii.48

H. fruticaris (L.) D.Don.

—

Lesotho

CFJG

28-31.xii.48

(in Richards, 1962)

Q. metallescens

(Schulthess)

Compositae

Gazania Gaertn.
G.sp.

G. linearis (Thunb.)

—

Lesotho

CFJG

3.xi.48

Druce

Lesotho

CFJG

9.xi.48

Helichrysum Mill.
//. sp.

Lesotho

CFJG

9-17.xi.52

(in Richards, 1962)

Q. senecionis Richards

Compositae

Senecio L.

S. laevigatus Thunb.

OFS

CFJG

l.xii.52

5. laevigatus Thunb.

—

OFS

CFJG

l.xii.52

(in Richards, 1962)

Aster L.

A. niuricatus Thunb.

Lesotho

CFJG

12.xii.54

A. muricatus Thunb.

Lesotho

CFJG

12.xii.54

(in Richards, 1962)

Gazania Gaertn.
G. sp.

Lesotho

CFJG

13.xi.48

(in Richards. 1962)

Q. poecila (Schulthess)

Compositae

Berkheya Ehrh.
B. sp.

Namibia

RET

(Turner. 1939, in Richards, 1962)

Q. basuto Richards

Compositae

Gazania Gaertn.

G. linearis (Thunb.)

Druce

Lesotho

OWR

29.ix.52

(in Richards, 1962)

Aster L.

A. muricatus Thunb.

Lesotho

CFJG

17.xi.52

Q. propinqua (Schulthess)

Compositae

Hirpicium Cass.
H. echinus Less.

Aegenevs

FWG&SKG

14.x. 88

Q. sp. G

Compositae

Berkheya Ehrh.

B. cf. spinosa (L.f.)

Pr Albert

SKG

26. xi-

Druce

5.xii.87

Q. sp. I

Compositae

“daisy*’

Springbok

FWG&SKG

10-12.X.88

Mesembryanthemaceae

Prenia N.E.Br.
P. pollens

Springbok

27.X.87

(Ait.) N.E.Br.
Drosanthemum Schwant.
D. sp.

Anenous

FWG&SKG

11-13.X.88

128

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

WASP

: PLANT

SPECIES

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

. tarsata Richards

Mesembryanthemaceae

Delosperma N.E.Br.
D. acuminatum L.Bol.

Grahamstown

CFJG

24.iv.64

Drosanthemum Schwant.
D. hispidum (L.)
Schwant.

Grahamstown

EMCC

18.X.52

D. hispidum (L.)
Schwant.

Grahamstown

EMCC

10.X.53

Scrophulariaceae

Aptosimum Burch.

A. depressum Burch, ex
Benth.

(in Richards, 1962)
Grahamstown FWG&SKG

13-30.X.81

. sp. A

Mesembryanthemaceae

Drosanthemum Schwant.
D. sp.

Bitterfontein/

SKG

14.X.87

. sp. B

Mesembryanthemaceae

Drosanthemum Schwant.
D. sp.

Garies

Bitterfontein/

SKG

14.X.87

. sp. C

Mesembryanthemaceae

Drosanthemum Schwant.
D. hispidum (L.)
Schwant.

Garies

Springbok

FWG&SKG

15-21.X.87

. laeta (Schulthess)

Mesembryanthemaceae

Drosanthemum Schwant.
D. hispidum (L.)
Schwant.

Springbok

FWG&SKG

15-2LX.87

Psilocaulon N.E.Br.

P. acutisepalum (Berger)
N.E.Br.

WPi

Springbok

FWG&SKG

15-21.X.87

Stoeberia Dinter &
Schwant.

5. sp.

275

Aggeneys

FWG&SKG

14.X.88

5. sp.

Aggeneys

DWG

14.X.88

Prenia N.E.Br.

P. sladeniana
(L.Bol.) L.Bol.

Springbok

17.x. 87

. sp. E

Mesembryanthemaceae

Drosanthemum Schwant.
D. hispidum (L.)
Schwant.

Springbok

FWG&SKG

15-21. X.87

. sp. F

Mesembryanthemaceae

Sphalmanthus N.E.Br.
S. cf. bijliae (N.E.Br.)
L.Bol.

WPi

209

Pr. Albert

FWG.SKG

26. xi-

Psilocaulon N.E.Br.
P. cf. articulatum
(Thunb.) Schwant.

Pr. Albert

&RWG

FWG.SKG

5.xii.87
26. xi-

&RWG

5.xii.87

129

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

WASP

SPECIES

FORAGE PLANT

FAMILY GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Q. sp. H

Mesembryanthemaceae

Drosanthemum Schwant.
D. sp.

Port NoHoth

FWG&SKG

2.X.85

D. sp.

Port Nolloth

FWG&SKG

11.X.88

D. sp.

Port Nolloth

DWG

11.X.88

Q. sp. K

Mesembryanthemaceae

“mesem '

Willowmore

CFJG

4.X.71

Q. sp. J

Mesembryanthemaceae

“mesem ”

Oudtshoom

FWG

7-8.xii.86

“mesem”

Oudtshoom

SKG

7-8.xii.86

“mesem”

Oudtshoom

FWG

7-8.xii.86

Q. capensis (Kohl)

Mesembryanthemaceae

Mesembryanthemum

Cape Town

RET

(Turner,

Q. niveopicta (Schulthess)

Mesembryanthemaceae

Mesembryanthemum

—

Mossel Bay

RET

1939)

(Turner,

1939)

(Turner,

Q. signata (Schulthess)

Mesembryanthemaceae

Mesembryanthemum

RET

1939)

(Turner.

Q. sp. O

Mesembryanthemaceae

Stoeberia Dinter &
Schwant.

S. sp.

Ag&eneys

FWG&SKG

1939)

14.X.88

S. sp.

Aggeneys

DWG

14.x. 88

Prenia N.E.Br.

P. pallens (Ait.) N.E.Br.

Springbok

27.X.87

Polymita N.E.Br.

P. albiflora (L.Bol.)
L.Bol.

Springbok

31.X.87

Q. sp. P

Mesembryanthemaceae

Prenia N.E.Br.

P. pallens (Ait.) N.E.Br.

Springbok

27.X.87

P. sladeniana (L.Bol.)
L.Bol.

Springbok

I7.X.87

Q. sp. Q

Mesembryanthemaceae

Stoeberia Dinter &
Schwant.

S. sp.

Aggeneys

DWG

14.x. 88

Q. sp. R

Mesembryanthemaceae

Stoeberia Dinter &
Schwant.

S. sp.

Aggeneys

DWG

14.X.88

Q. sp. T

Mesembryanthemaceae

Drosanthemum Schwant.
D. sp.

Anenous

FWG&SKG

11-13.X.88

Q. sp. M

Campanulaceae

Wahlenbergia Schrad.
W. sp. C

Anenous

FWG&SKG

11-13.X.88

W. sp. C

Anenous

DWG

11-13.X.88

130

GESS & GESS: FLOWER VISITING BY MAS ARID WASPS IN SOUTHERN AFRICA

WASP

FORAGE PLANT

SPECIES

FAMILY

GENUS AND SPECIES

COL-

OUR

SEX

NUM-

BER

LOCALITY

COLLEC-

TOR

DATE

Q. sp. N

Campanulaceae

Wahlenbergia Schrad.
W. sp. A

Clanwilliam

FWG&SKG

3-7.X.88

W. sp.

Clanwilliam

FWG&SKG

5-6.X.88

Q. sp. S

Campanulaceae

Wahlenbergia Schrad.
W. sp. A

Clanwilliam

DWG

3-7.X.88

Quartiniella Schulthess

The genus Quartiniella Schulthess is endemic to southern Africa where it is recorded from
the South Western Cape; Southern Cape in the Little Karoo east to Willowmore; and in Namibia
(S. W. A.) (Richards, 1962 and label data, Albany Museum). In numbers it appears to be a small
genus.

Quartiniella falls within the same size range as Quartinioides from which, however, it differs
markedly in being relatively short tongued.

Nothing seems to be known of the nesting of Quartiniella.

Flower visiting records for Quartiniella seem to be almost entirely lacking. According to
Turner (1939) two species “are found on Athanasia sp. (Compositae) and do not visit
Mesembryanthemum ’ ’.

DISCUSSION

Masarid wasps differ from all other wasps in that they depend on flowers not only for nectar
for adult nourishment but also for both nectar and pollen for the nourishment of the young. It
is therefore important to bear in mind that, though visits by males are solely for imbibing nectar
for their own nourishment, visits by females to particular flowers may be for this purpose or to
collect nectar and/or pollen for provisioning.

Generally speaking wasps are attracted in large numbers and great diversity to whatever
suitable plants may be in flower, extremely popular are plants with small, white, cream or yellow
flowers presented in heads. Amongst those plants especially favoured in southern Africa are
Acacia spp. (Leguminosae: Mimosoidea), Maytenus spp. (Celastraceae), Asclepias spp.
(Asclepiadaceae) and the roadside weed Foeniculum vulgare A. W. Hill (Umbelliferae) (Gess
and Gess, Catalogue of flower visits by aculeate wasps, to date including in excess of 4 000
records for circa 420 species, unpublished). It is noteworthy that masarid wasps are almost
entirely absent from samples of wasps from such plants.

From the available flower visiting records it appears that southern African masarids as a
whole are most commonly associated with Mesembryanthemaceae (51%) and Compositae
(28%), those species visiting flowers of other families such as Campanulaceae (12%),
Scrophulariaceae (5%), Leguminoseae (Papilionatae) (3%) and Liliaceae (2%) being the
exceptions. Furthermore, that individual species on the whole show extreme fidelity to flowers
of a single family of plants suggests that both nectar and pollen are generally obtained from the
same source. A possible exception is Ceramius braunsi which, as already discussed, appears to
provision entirely with composite pollen but has been observed taking nectar from Aspalathus

131

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

flowers. Casual nectar gathering by a species which normally exhibits marked fidelity has been
noted for Ceramius lichtensteinii .

Clearly the considerable fidelity shown by masarids in their choice of provision is indicative
of the importance of particular families or even genera of plants to these wasps. What is more
difficult to assess is the importance of the wasps as pollinators to the plants which they visit.

Certainly where masarids are abundant they are probably the commonest and most
dependable visitors to certain species of relatively small to medium flowered light coloured
“mesems” although bees are not uncommon visitors to these flowers. It is of interest that these
“mesem” flowers are not favoured by monkey beetles of the genus Anisonyx (Scarabaeidae;
Hopliini) which, however, are particularly attracted to “mesem” flowers of the darker shades of
purplish pink to magenta, not favoured by masarids. Some at least of the composite flowers
visited by masarid wasps are also visited by other insects, non-masarid wasps, bees, flies, beetles
and butterflies, which may pollinate these flowers. Generally speaking they are, however,
probably less dependable visitors than the masarids where these are abundant. A notable
exception is Fidelia braunsiana Friese (Hymenoptera: Apoidea: Fideliidae) which is restricted to
the genus Berkheya (Compositae) (Whitehead, 1984). Certainly the dependence of certain
masarid wasps on the flowers of Mesembryanthemaceae and of others on Compositae and the
nature of their behaviour in and on these flowers makes them ideally suited to being dependable
pollinators.

Evidence is building up to suggest that Wahlenbergia spp. are of importance to some species
of the smaller masarids. Furthermore, when the frequency of and nature of their visits is
compared to that of other insect visitors the impression is gained that masarids may prove to be
of importance in the pollination of these flowers.

Collecting data and observations indicate that amongst those species which do not visit
Mesembryanthemaceae, Compositae or Campanulaceae there are species which are probably of
paramount importance to the plants which they patronize. Indications are that amongst these
relationships there is a considerable degree of mutualism. Most notable are the relationships
between Aptosimum and Peliostomum (both Scrophulariaceae) and some Celonites spp.
A. spinescens, A. lineare, A. depressum and Peliostomum virgatum appear to be most
commonly, indeed almost exclusively, to be visited by Celonites spp. Furthermore it has been
demonstrated that Celonites obtaining nectar from these flowers are ideally suited to trigger the
dehiscence of the anthers and to transfer pollen from one flower to another.

Also of particular interest are the relationships between Aspalathus (Leguminosae:
Papilionatae) and Ceramius clypeatus and Masarina familiaris. Although Aspalathus spinescens
lepida is visited by other insects including honey bees C. clypeatus and M. familiaris appear, in
the Clanwilliam district at least, to be the commonest and most dependable visitors. Honey bees,
for example, are not restricted to Aspalathus and therefore, if there is in the vicinity some plant
in flower which is equally attractive or more attractive to them, they will not necessarily visit the
Aspalathus flowers. Furthermore honey bees, when visiting the Aspalathus flowers, do not adopt
a set stance such as do C. clypeatus and M. familiaris. It has been demonstrated that C. clypeatus
and M. familiaris obtaining nectar from Aspalathus flowers are ideally suited to trigger the
dehiscence of the anthers and to transfer pollen from one flower to another.

The present account is in its nature a preliminary review of flower visiting by masarid wasps
in southern Africa. It is the authors’ intention, however, to investigate further the role that
masarid wasps play in the pollination of the flowers which they visit.

132

GESS & GESS: FLOWER VISITING BY MASARID WASPS IN SOUTHERN AFRICA

ACKNOWLEDGEMENTS

The authors wish to thank the late Miss Grace Britten, of the Albany Museum Herbarium,
Mrs Estelle Brink of the Albany Museum Herbarium and Mrs Sue Dean of the Karoo Biome
Research Station at Tierberg, Prince Albert, for assistance with the identification of some of the
forage plants. Mr Jan Vlok of Saasveld Forestry Research Centre and Mr Peter Phillipson of
Rhodes University are thanked respectively for determining Aspalathus spinescens Thunb.
subsp. lepida (E.Mey.) Dahlgren and for confirming the generic identity of the Wahlenbergia
spp.

The authors would also like to thank Mr Michael Struck of Hamburg, West Germany, for
making available his flower visiting records for some of the specimens which he submitted to
F. W. Gess for determination.

Thanks are due also to Mr Robin Cross of the Electron Microscopy Unit, Rhodes
University for producing the scanning electron micrographs.

Gratitude to the C.S.I.R. is expressed by F. W. Gess for a running expenses grant for field
work during the course of which most of the flower visiting records presented here were
accumulated.

REFERENCES

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

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

Barth, F. G. 1985. Insects and Flowers: The Biology of a Partnership. London: Allen and Unwin.

Brauns, H. 1913. Dritter Beitrag zur Kenntnis der Masariden (Hym.) von Siidafrika. Entomol. Mitteilungen 2 (7/8):
193-209.

Cooper, K. W. 1952. Records of flower preferences of masarid wasps. II. Politrophy or oligotrophy in Pseudomasaris?
(Hymenoptera: Vespidae). Amer. Midland Nat. 48: 103-110.

Faegri, K. and van der Pul, L. 1979. The priciples of pollination biology. Oxford: Pergamon.

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

Gess, F. W. 1968. Further contribution to the knowledge of the South African species of the genus Ceramius Latreille
(Hymenoptera: Masaridae). Novos Taxa ent. 57: 29-33.

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 (6): i09-122.

Gess, F. W. 1981. Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the vicinity
of Grahamstown, South Africa. Ann. Cape Prov. Mus. (nat. Hist.) 14 (1): 1-80.

Gess, F. W. and Gess, S. K. 1980. 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. 1988a. A contribution to the knowledge of the taxonomy and ethology of the genus Masarina
Richards (Hymenoptera: Masaridae). Ann. Cape Prov. Mus. (nat. Hist.) 16 (14): 351-363.

Gess, F. W. and Gess, S. K. 1988b. A further contribution to the knowledge of the ethology of the genus Ceramius
Latreille (Hymenoptera: Masaridae) in the southern and western Cape Province of South Africa. Ann. Cape
Prov. Mus. (nat. Hist.) 18 (1): 1-29.

Houston, T. F. 1981. Alimentary transport of pollen in a paracolletine bee (Hymenoptera: Colletidae). Aust. ent. Mag.
8 (4): 57-59.

Moll, E. J., Campbell, B. M., Cowling, R. M., Bossi, L., Jarman, M. L. and Boucher, C. 1984. A description of major
vegetation categories in and adjacent to the fynbos biome.' South African National Scientific Programmes Report
83: i-iv, 1-29.

Percival, M. S. 1969. Floral Biology. Oxford: Pergamon.

Proctor, M. and Yeo, P. 1973. The pollination of flowers. London: Collins.

Real, L. 1983. Pollination Biology. London: Academic Press.

Rebelo, a. G. 1987. Introduction. In: Rebelo, A. G. ed, A preliminary synthesis of pollination biology in the Cape flora.
South African National Scientific Programmes Report 141: 1-5.

133

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 5, JULY 1989

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

Richards, O. W. 1963. New species of Ceramius Latreille (Hymenoptera, Vespoidea) allied to Ceramius lusilanicus Klug.
Zodl. Meded., Leiden 38 (13): 213-220.

Richards, O. W. 1969. Records of Masarid wasps with descriptions of two new species of Quartinia Ed. Andre
(Hymenoptera). J. nat. Hist. 3: 79-83.

Torchio, P. F. 1974. Mechanisms involved in the pollination of Penstemon visited by the masarid wasp, Pseudomasaris
vespoides (Cresson) (Hymenoptera: Vespoidea). The Pan Pacific Entomologist 50 (3); 226-234.

Turner, R. E. 1939. Notes on the masarid wasps of the genus Quartinia. Ann. Tvl. Mus. 20 (1): 1^.

Whitehead. V. B. 1984. Distribution, biology and flower relationships of fideliid bees of southern Africa (Hymenoptera,
Apoidea, Fideliidae). S. Afr. J. Zool. 19 (2): 87-90.

Whitehead, V. B., Giliomee, J. H. and Rebelo, A. G. 1987. Insect pollination in the Cape flora. In: Rebelo, A. G. ed,
A preliminary synthesis of pollination biology in the Cape flora. South African National Scientific Programmes
Report 141: 52-82.

134

INSTRUCTIONS TO AUTHORS

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

Only generic and specific names should be underlined to indicate italics. All other marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

BD8764 — CTP Book Printers, Cape

>'(pl3

ISSN 0570-1880

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 6 28 July 1989

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

Morphometries, moult and taxonomy of the Nectarinia afralNectarinia
chalybea complex of South African double-collared sunbirds

P. LLOYD
and

A.J.F.K. CRAIG

(Department of Zoology, Rhodes University, Grahamstown 6140)

ABSTRACT

Five morphological characters were measured on museum specimens of two closely related
double-collared sunbird species, Nectarinia afra (Linnaeus) and Nectarinia chalybea (Linnaeus),
and their wing moult was examined. The mensural data were analysed and compared
statistically, to assess the validity of the races currently recognised for each species. We conclude
that the races N. c. albilateralis Winterbottom and N. c. capricornensis (Roberts) are not
separable, and that the races N. c. chalybea (L.), N. c. subalaris (Reichenow), N. a. afra (L.)
and N. a. saliens Clancey are based on rather arbitrary grounds. Moult data support the
recognition of two races in the case of N. chalybea, but not for N. afra. Clinal variation in size
occurs in both species, but it is not the same in each species, nor the same for all characters.
Further field study of habitat preferences, seasonal movements, and possible physiological
differences is needed.

INTRODUCTION

The Greater Double -collared Sunbird Nectarinia afra (Linnaeus) and the Lesser
Double-collared Sunbird N. chalybea (Linnaeus) are widely sympatric in South Africa. Both
species occur in a variety of habitats from the Southwestern Cape to the Northern Transvaal
(Maclean, 1985).

Currently two races of N. afra and four races of N. chalybea are recognised on the basis of
distribution and morphological characters (Clancey, 1980). These races are:

Nectarinia afra afra (Linnaeus) (W. Cape to E. Cape)

N. a. saliens Clancey (Transkei to Transvaal)

Nectarinia chalybea chalybea (Linnaeus) (S. and S.W. Cape)

N. c. albilateralis Winterbottom (W. Cape coast and Karoo)

N. c. subalaris (Reichenow) (E. Cape to Natal)

N. c. capricornensis (Roberts) (Transvaal)

135

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

Mayr (1969) defines a subspecies (synonym of race) as ‘an aggregate of phenotypically
similar populations of a species, inhabiting a geographical subdivision of the range of a species,
and differing taxonomically from other populations of the species’. This definition implies that
subspecies should be allopatrically distributed, and distinguishable from each other on the basis
of one or more characters. A potential problem for the subspecies concept is the variable criteria
used by different taxonomists to justify subspecific separation of slightly differentiated local
populations (Mayr, 1969). More recently, Mayr (p. 595 in Wiens, 1982) stated that ‘the primary
use of subspecies is as a sorting device in collections’. Other contributors to this discussion felt
that subspecies are still useful in the study of geographical variation in birds, but much more
critical assessment was required before subspecies were named. The validity of these races was
assessed with such considerations in mind.

METHODS

In total 359 specimens of N. chalybea and 295 specimens of N. afra were examined.
Measurements (in mm) of wing (maximum cord), bill, tarsus, tail and red breast-band width
were taken using a wing rule and vernier calipers. Damaged parts of specimens were not
measured. The data obtained were analysed using the following elements of the BMDP
statistical package: basic data analysis (programme BMDP ID), means cluster analysis
(programme BMDP KM), and comparison of two groups with t-tests (programme BMDP 3D).
First, sexual dimorphism was described for each species separately, and the sexes of each species
were then compared using BMDP ID. Secondly, male birds were grouped into “races”
according to the geographical distributional limits set out in Clancey (1980), and these
populations analysed and compared using BMDP ID and 3D. Thirdly, all male specimens were

Table 1.

Summary of the measurements of Nectarinia afra and Nectarinia chalybea.

Nectarinia afra

Nectarinia chalybea

CHARACTER

SEX

NO.

BIRDS

MEAN

RANGE

NO.

BIRDS

MEAN

RANGE

Bill

Male

203

29,7

1,4

19,6-33,0

248

23,1

1,6

19,5-28,0

Female

27,0

1,2

23,6-29,5

20,9

1,4

16,2-23,3

Breast-band

Male

200

20,9

2,5

14,0-29,5

208

10,6

2,6

6,0-19,0

Tarsus

Male

213

16,6

0,8

14,0-19,5

261

15,2

0,7

14,0-17,5

Female

15,6

0,6

14,0-17,0

14,5

0,7

12,5-16,2

Wing

Male

216

65,3

2,2

59,0-71,5

270

55,3

1,8

48,5-60,0

Female

59,3

1,9

54,3-65,5

50,5

2,1

45,5-58,0

Tail

Male

212

55,2

3,1

48,5-65,0

265

47,6

3,1

39,5-57,5

Female

47,5

2,6

42,0-55,5

40,2

3,4

27,0-49,5

136

LLOYD & CRAIG: MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

assigned to clusters on the basis of their measurements, using BMDP KM. The number of
clusters was specified (two for N. afra, four for N. chalybea), but not the mid-point of any
cluster. Finally the males were grouped into arbitrary allopatric geographical entities, and the
character means of the group members calculated.

Moult of the primary remiges was recorded for each specimen for which date and locality
information was available, using the 0-5 scoring system of Newton (1966). For males, the extent
of breeding plumage and body moult was also noted. The moult sample includes 35 N. afra and
14 N. chalybea caught while bird-ringing in the Eastern Cape.

RESULTS

The male is larger than the female in all respects in both species (Table 1). There is a wide
range in breast-band width in the males of both species. There is, however, little overlap in
breast-band widths between the two species. The breast-band of N. afra is roughly twice as wide
as the breast-band of N. chalybea, and the mean values of all measurements differ between
species in both sexes, although there is some overlap.

Tables 2 and 3 compare the measurements of the races, according to the geographical
divisions of Clancey (1980). Since females constituted a small part of the sample in both species,
statistical comparisons are restricted to males.

Table 2.

Comparative measurements of the races of Nectarinia chalybea.

CHARACTER

RACE

SEX

NO.

BIRDS

MEAN

RANGE

Bill

chaly

male

22,3

1,3

20,5-27,0

female

20,2

1,3

19,0-22,0

albil

male

22,3

1,1

19,5-25,1

female

19,8

1,4

16,2-23,0

subal

male

24,1

1,5

20,4-27,0

female

21,5

1,0

19,0-23,3

capri

male

24,2

1,0

22,5-26,0

female

21,0

0,9

20,0-22,0

Breast-

chaly

male

8,7

1,7

6,0-12,0

band

albil

male

9,1

1,8

6,2-13,7

subal

male

11,7

2,2

7,5-16,0

capri

male

13,6

2,0

10,0-19,0

Tarsus

chaly

male

15,1

0,5

14,0-16,0

female

14,4

0,7

13,3-15,5

albil

male

15,1

0,7

14,0-17,5

female

14,4

0,9

12,5-16,2

subal

male

15,3

0,7

14,0-16,5

female

14,5

0,5

13,5-16,0

137

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

CHARACTER

RACE

SEX

NO.

BIRDS

MEAN

RANGE

Tarsus

capri

male

15,4

0,8

14,0-17,0

female

14,6

0,7

13,7-16,0

Wing

chaly

male

54,7

1,8

48,5-57,0

female

49,2

2,2

45,5-52,0

albil

male

55,1

1,6

50,5-59,0

female

50,0

1,9

46,0-55,0

subal

male

55,7

1,8

50,0-59,0

female

50,8

1,9

46,0-58,0

capri

male

55,1

1,7

52,0-59,0

female

51,0

2,4

47,0-56,0

Tail

chaly

male

47,3

2,4

43,5-52,0

female

41,5

3,4

39,5-48,0

albil

male

46,8

2,9

39,5-52,4

female

39,8

3,6

32,5-49,5

subal

male

48,3

3,3

41,0-57,5

female

39,9

3,5

27,0-48,9

capri

male

48,4

2,8

42,0-54,3

female

40,4

3,1

35,5-46,5

Table 3.

Comparative measurements of the races of Nectarinia afra.

CHARACTER

RACE

SEX

NO.

BIRDS

MEAN

RANGE

Bill

afra

male

152

29,8

1,4

19,6-33,0

female

27,2

1,1

24,5-29,5

saliens

male

29,2

26,5-32,2

female

26,2

1,4

23,6-28,0

Breast-

afra

male

142

21,2

2,5

15,5-29,5

band

saliens

male

20,2

2,4

14,0-26,0

Tarsus

afra

male

155

16,6

0,7

14,0-19,5

female

15,6

0,6

14,0-17,0

saliens

male

16,7

0,8

14,5-18,2

female

15,5

0,7

14,5-16,5

Wing

afra

male

156

64,9

2,0

60,0-71,5

female

58,8

1,7

54,3-62,5

saliens

male

66,4

2,4

59,0-70,0

female

60,9

2,0

58,0-65,5

138

LLOYD & CRAIG: MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

CHARACTER

RACE

SEX

NO.

BIRDS

MEAN

RANGE

Tail

afra

male

153

54,6

2,7

48,5-63,0

female

47,2

2,5

42,5-55,5

saliens

male

56,9

3,6

49,5-65,0

female

48,4

3,0

42,0-55,0

In N. chalybea the races chalybea and albilateralis are remarkably similar in all character
means, especially with respect to bill length and breast-band width. These two races have a
markedly smaller bill and breast-band than do the races subalaris and capricornensis . Tarsus
length appears to vary little in different populations. No real trends between races are evident
for wing and tail means.

For N. afra the race afra has a broader breast-band, while the race saliens has longer wings
and tail. The two are similar in respect of bill and tarsus length.

If these races are compared using grouped characters, N. c. chalybea and N. c. albilateralis
are the only races of N. chalybea that are not significantly different from each other. The two
races of N. afra are significantly different from each other overall.

From Table 4 it can be seen that tail length is a very conservative character in N. chalybea,
with one significant difference between any of the races compared. N. c. chalybea and
albilateralis have no characters which differ significantly, while subalaris and capricornensis are
strikingly different only in respect of breast-band width; subalaris is significantly different from
both chalybea and albilateralis in most characters.

Table 4.

Statistical comparison of males of the races of Nectarinia chalybea on the basis of single

characters.

RACE

CHARACTER

chaly

albil

subal

Bill

Breast-band

albil

Tarsus

Wing

Tail

Bill

<0,001

Breast-band

<0,001

subal

Tarsus

<0,05

Wing

<0,05

Tail

<0,01

Bill

<0,001

Breast-band

<0,001

capri

Tarsus

<0,05

Wing

<0,05

Tail

139

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

For N. afra. Table 5 shows that the two races differ with regard to breast-band width, wing
and tail length. Again tarsus length does not vary significantly between different populations.

Table 5.

Statistical comparison of males of the races of Nectarinia afra on the basis of single characters.

RACE

CHARACTER

saliens

Bill

Breast-band

<0,01

afra

Tarsus

Wing

<0,001

Tail

<0,001

Regional variation

There is a clear increase in bill, wing, tarsus and tail length and breast-band width of
N. chalybea from the Southwestern Cape to the Transvaal, with a discontinuity in Natal-
Transkei, where the largest mean values are found (Table 6). Except for breast-band width.

Table 6.

Character means of male Nectarinia chalybea according to region. (Sample size indicated in

brackets).

REGION

BILL

BREAST-

BAND

TARSUS

WING

TAIL

W. Cape

21,8

7,9

15,1

54,7

46,3

(40)

(27)

S. Cape

22,6

8,7

15,2

55,3

47,7

(24)

(15)

PE area

22,8

10,4

15,1

55,3

47,3

(31)

(26)

E. Cape

23,6

11,4

15,1

55,2

47,3

(52)

(38)

Transkei

24,5

12,1

15,4

56,2

49,7

(23)

(20)

Natal

25,3

12,1

15,9

56,6

49,8

(10)

Transvaal

24,2

13,6

15,4

55,1

48,4

(28)

(27)

140

LLOYD & CRAIG: MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

there is a very slight increase in character means between the Southern and Eastern Cape, with
a much larger increase between the Eastern Cape and Natal-Transkei areas, and then a smaller
decrease in character means between the Natal- Transkei and Transvaal areas.

There is an increase in tail length from the Cape through to the Transvaal in N. afra (Table
7). However, Transvaal birds are smaller than those from Natal and the Transkei in respect of
breast-band width and tarsus length, while bill length is greatest in the Transkei, and wing length
in Natal.

Table 7.

Character means of male Nectarinia afra according to region. (Sample size indicated in

brackets).

REGION

BILL

BREAST-

BAND

TARSUS

WING

TAIL

S. Cape

28,5

20,4

15,7

63,7

54,8

(24)

(22)

PE area

29,8

20,3

16,3

64,2

54,4

(28)

(25)

E. Cape

29,8

21,5

16,7

65,3

54,4

(99)

(90)

Transkei

30,2

20,6

16,9

55,4

(10)

Natal

29,0

20,7

17,0

68,0

57,1

(13)

(12)

Transvaal

29,0

19,7

16,5

66,1

58,2

(23)

Cluster analysis

For N. chalybea, members of clusters 3 and 4 (Table 8) have similar mean bill lengths and
breast-band widths, which are distinctly smaller than those of clusters 1 and 2. There is no
observable trend in the mean lengths of tarsus, wing and tail. In the case of N. afra, birds
assigned to cluster 1 have on average longer wings and tails, but there is little difference in the
means for bill and tarsus length, or breast-band width.

The distribution of cluster members for N. chalybea is shown in Fig. 1. Clearly, the
members of clusters 1 and 2 form a distinct group in the Transvaal and Natal, whilst the members
of clusters 3 and 4 form a distinct group in the Western and Southern Cape. The two clusters
within each of these groups cannot be separated geographically, since the members show an
almost random distribution within their region. In the Eastern Cape and Transkei the two
groups of clusters intergrade. There are also singletons within the Cape which are assigned to
northern clusters.

141

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

Table 8.

Mean measurements of the groups formed by the cluster analysis.

CLUSTER

BILL

BREAST-

BAND

TARSUS

WING

TAIL

N. chalybea

24,8

13,8

15,4

56,9

50,8

23,6

12,1

15,2

54,6

47,3

22,4

8,6

15,0

55,8

50,1

22,4

8,8

15,2

54,6

44,5

N. afra

29,4

21,1

16,8

66,7

58,4

29,8

20,8

16,5

64,5

53,3

There is a large degree of overlap in the distribution of cluster members of N. afra (Fig. 2).
There is, however, a general tendency for birds assigned to cluster 1 to be more numerous in the
Transvaal and Natal, whilst the opposite is true for the Transkei and Cape where members of
cluster 2 are more abundant.

The timing of primary moult

The moult data for N. chalybea are shown in Table 9. The records are not evenly distributed
over the year, and the bulk of the material is from the Southwestern or Eastern Cape. It appears
that birds from the Southern and Western Cape start the moult earlier than birds from the
Eastern Cape, and Natal and Transvaal birds appear to follow the same schedule as those from
the Eastern Cape. Skead (1967) indicated that breeding was more seasonal in the Western Cape
than in the Eastern Cape, and this is supported by unpublished data gathered for the Atlas
schemes in these areas. In the Western Cape breeding ends in October, whereas in other regions
there are records for early summer as well as winter, and breeding at all seasons is possible.

For N. afra the majority of records are again from the Eastern Cape (Table 10). In this case
there is no indication of regional differences in the timing of moult, nor is there a clearly defined
breeding season for any part of the bird’s range (Skead, 1967; Tarboton et al., 1987; unpubl.
data).

In both species, males in full breeding plumage have been recorded throughout the year.
Birds in partial breeding plumage may be sub-adult males, acquiring adult plumage for the first
time. The months in which such specimens have been collected are in accordance with the known
breeding times for the areas concerned. Males and females moult at the same time of the year.

DISCUSSION

The two species differ significantly with respect to most single characters, and are certainly
separable on the basis of pooled characters. Tarsus length seems to be least variable in both
species. There is little or no overlap in bill lengths of both species, and although there is wide
intraspecific variability in breast-band width, there is little interspecific overlap, with N. afra
having a red band approximately twice as wide as that of N. chalybea. The commonly-used

142

LLOYD & CRAIG; MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

Fig. 1. Map of South Africa showing the geographical distribution of Nectarinia chalybea members of the
computer-generated clusters based on morphological similarity.

143

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

Fig. 2. Map of South Africa showing the geographical distribution of Nectarinia afra members of the computer-generated

clusters based on morphological similarity.

144

LLOYD & CRAIG: MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

Table 9.

The timing of primary moult in Nectarinia chalybea.

MONTH

TOTAL

NO. BIRDS
(E. CAPE)

140

% OF TOTAL
IN MOULT

NO. BIRDS
(S.W. CAPE)

% OF TOTAL
IN MOULT

NO. BIRDS
(TVL, NTL,
TRANSKEI)

% OF TOTAL
IN MOULT

Table 10.

The timing of primary moult in Nectarinia afra.

MONTH

TOTAL

NO. BIRDS
(CAPE
PROV.)

209

% OF TOTAL
IN MOULT

NO. BIRDS
(TVL, NTL,
TRANSKEI)

% OF TOTAL
IN MOULT

identification guide of a twofold difference in bill size and red breast-band for these two species
(Skead, 1967), therefore applies in the general sense.

The races of Nectarinia chalybea

Winterbottom (1963) distinguished the race N. c. albilateralis solely on the basis of a whiter
flank coloration in the males. However, Clancey and Irwin (1978) state that this colour
difference is due to salt-bleaching and wind borne sand abrasion affecting the birds breeding on
the white sand dunes of the Western Cape. If this is the case, why did Winterbottom not note
such colour differences in the females as well? Clancey and Irwin (1978) justify the recognition
of the race on ‘the plain rump in males and the lack of a yellowish tinge to the lower ventral
surface in freshly moulted males’. The relevance of plumage differences in these regions for the

145

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

birds themselves is questionable; it seems likely that breast-band and song characters are the key
features in both male-male and male-female interactions. Winterbottom (1963), in the original
description of N. c. albilateralis , makes no mention of these colour differences. He concludes
that the female N. c. albilateralis is not distinguishable from N. c. chalybea, whereas Clancey and
Irwin (1978) state that the female is paler and less olive above (Light Brownish Olive versus
Buffy Olive in N. c. chalybea), and ‘distinctly less greenish yellow’ below, being ‘pale greyish
olive buff’. The publications of Winterbottom (1963) and Clancey and Irwin (1978) are the only
papers dealing with the taxonomy of N. c. chalybea, and it would seem that confusion about the
true colour differences exists. We badly need a system of colour measurement which allows for
independent assessment and comparison of colour differences.

Winterbottom (1963) cites the measurements of 19 male N. c. albilateralis and Clancey and
Irwin (1978) 12 males. None of these measurements is significantly different from those of
N. c. chalybea, and in this study a statistical comparison of the two races also found no significant
differences in any of the characters measured (Table 4).

Clancey and Irwin (1978) also cite a wider breast-band as a difference (7-8 mm for
albilateralis versus 10 mm for chalybea). The data from Table 2 certainly do not support this,
since here albilateralis has a slightly wider breast-band on average. It would seem therefore that
there is no quantitative morphological basis for the separation of chalybea and albilateralis . This
conclusion is further supported by the regional comparisons in Table 6, which indicate the
occurrence of a dine, defined by Mayr (1969) as a character gradient. Winterbottom (1963)
found a simple dine from north to south in culmen length for the species as a whole, whilst the
variation in wing length was from a maximum in the central part of the range (Natal and
Transkei) to minima north and south. Exactly the same trends are found in this study (Table 6).
From Winterbottom’s results, it can be seen that the distributions of both N. c. albilateralis and
N. c. chalybea lie along this dine. There will therefore be no independent variation in character
measurements, thus reducing the likelihood of individuals of both subspecies being different at
the same point along the dine. The implications of clinal variation for the taxonomy of a species
is discussed later. Clancey and Irwin (1978) divided chalybea and albilateralis distributionally on
the basis of habitat, with albilateralis occupying the arid karroid scrub region, and chalybea the
wetter mountains and coastal region of the Southern and Southwestern Cape. Our data suggest
that this division is artificial in morphological terms. This implies that the two subspecies
proposed by Clancey (1978) are not allopatrically distributed, which, according to Mayr (1969)
strongly indicates a wrong usage of the term ‘subspecies’. This is discussed more fully later. The
validity of the subspecies N. c. albilateralis is therefore highly questionable.

Clancey (1975) decided to resurrect N. c. capricornensis (Roberts) as a minor subspecies on
the basis of its smaller size. Previously, capricornensis was treated as synonymous with subalaris.
The main subspecific difference from N. c. subalaris (Reichenow) is a shorter wing (Clancey and
Irwin, 1978). TTiey give measurements of 12 males and three females of subalaris and 11 males
and nine females of capricornensis. In this small and select sample, wing lengths show almost no
overlap between subspecies for both sexes; male culmen lengths are similar, but with those of
subalaris tending to be slightly larger; female culmen lengths show virtually no overlap; male tail
lengths show a fairly large overlap, but subalaris has a longer tail on average; and female tail
lengths are very similar, but with subalaris again slightly longer. It is probably correct to say that
subalaris is larger than capricornensis , but in the light of this study using a much larger sample
size, the significance of the size differences is questionable (Table 2). Male subalaris have slightly

146

LLOYD & CRAIG; MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

larger tarsus, wing and tail means than male capricornensis, but the comparative analyses using
t-tests show no significant difference in tarsus and tail lengths (Table 4). Wing lengths are
marginally significantly different. There is certainly a far greater overlap in measurements than
is suggested by Clancey and Irwin’s (1978) data. A pooled comparative analysis does find a
significant difference between the two races, but this result depends primarily on the wider
breast-band of capricornensis . The data on the females (Table 2) come from a smaller sample,
and indicate a similarity in all measurements with no trends in size differences evident.

Clancey and Irwin (1978) state that N. c. subalaris differs taxonomically from the nominate
race N. c. chalybea in being larger, with a longer bill. The data obtained from this study support
these size differences. In both males and females (Table 2), subalaris has a larger mean for all
characters (except female tail length). The comparative analysis of males (Table 4) shows,
however, that tarsus and wing length are not significantly different statistically.

The results of Table 6, which are supported by a similar study by Winterbottom (1963), are
strong evidence for the existence of a dine in size variation of N. chalybea. This brings the
validity of subspecies in N. chalybea as a whole into question. It can be seen from the table that
there is a regular dine in breast-band measurements, with an increase in mean size from the
Southern Cape through to the Transvaal. A similar pattern is observed for bill, tarsus, wing and
tail measurements but with the largest means occurring in Natal. Mayr (1969) states that ‘when
the geographic variation of a species is clinal, it is usually inadvisable to recognise subspecies,
except possibly for the two opposite ends of the dine when they are very different or separated
by a pronounced step’. If one calculates the differences between the means for adjacent
geographical regions in Table 6, there is a pronounced step in the differences for all characters
except breast-band width between the Natal-Transkei and Eastern Cape groups. At this point,
the geographical plot of the four clusters generated by the computer (Fig. 1) proves useful.
Clusters 1 and 2 form a distinct group in the Transvaal and Natal-Transkei, whilst clusters 3 and
4 form a distinct group in the Southern and Western Cape. The two groups of clusters intergrade
in the Transkei and Eastern Cape regions. It is quite possible that, should the breast-band
measurements be removed from the cluster analysis, a sharper distinction between the two
groups may be found in the intergrade regions. Table 6 shows breast-band width to be very
similar in these two areas but quite different in the adjacent Southern Cape and Transvaal areas.

The moult data also suggest that N. chalybea can be divided into two populations with
respect to the timing of wing-moult. Birds from the Southern and Western Cape start the moult
as early as October, while those from the Eastern Cape start moulting in December (Table 9).
However, the boundary between the two “moult regions” is not clearly defined on the basis of
the present data. Birds from Knysna (34°02'S, 23 °03'E) may belong to the Eastern Cape, on the
strength of a specimen from Knysna which is moulting in April, much later than is typical for
Southern Cape birds from areas such as Mossel Bay. It appears that the division on the basis of
moult will lie well to the east of the morphological boundary between the groups defined above.
It is of course not certain that moulting seasons are genetically determined rather than
responsive to environmental conditions, and although this may be the case in some species (c/.
Craig, 1988) it should not be assumed to constitute a general rule. The Eastern Cape does form
a transitional zone in many respects, and this may explain why a change in moult timing occurs
at this point rather than further east in a climatically more predictable zone.

It has been argued above that albilateralis is not morphologically separable from chalybea,
and that capricornensis is not separable from subalaris. However, in mensural characters, both

147

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

subalaris and capricornensis are significantly different from both chalybea and albilateralis
(Table 4). Furthermore, the division in the distribution between subalaris and albilateralis occurs
in the Eastern Cape, where a pronounced step in the dines of bill, tarsus, wing and tail length
occurs. The moult data also divide the species into two groups. We propose that the number of
subspecies be reduced from four to two, through the fusion of albilateralis with chalybea and the
fusion of capricornensis with subalaris. Indeed, prior to 1963, only two subspecies of N. chalybea
were recognized, namely N. c. chalybea and N. c. subalaris (Clancey, 1962).

The races of Nectarinia afra

Clancey (1962), in his description of N. a. saliens as a new subspecies of N. afra, cited both
colour and size differences to support his decision. He stated that the male saliens differs from
afra ‘in having the non-metallic lower ventral surface lighter and more yellowish in freshly
moulted dress, less dark and brownish (Ecru Olive as against Citrine Draby. The female saliens
is described as ‘less dark and greyish and markedly reticulate over the throat and breast, and
more saliently greenish or yellowish over the medial lower ventral surface less grey (about
Barium Yellow abdominally)’. Clancey (1962) also stated that saliens is larger than afra. From
a sample of 14 saliens and 63 afra he presented measurements that are not only highly
significantly different, but also show no overlap whatsoever between the two subspecies. Taken
at face value, such measurements would strongly support the division of N. afra into two
subspecies. On the other hand, the results of this study, using a larger sample size, show a very
wide overlap in character measurements (Table 3). Most of the specimens studied, however,
came from the Eastern Cape area (Fig. 2), a region where Clancey (1962) claimed that the two
subspecies intergrade. Table 7 suggests a possible dine in different character measurements,
with bill length and breast-band width decreasing from the Cape to the Transvaal, tarsus length
being very constant, and wing and tail length increasing from the Cape to the Transvaal. The
dines for wing and tail length, and bill and breast-band, thus occur in opposite directions. As
already mentioned, Mayr (1969) states that it is inadvisable to divide a species into subspecies
along a dine, except possibly for the two opposite ends of the dine when they are different or
separated by a pronounced step. Clancey’s (1962) samples are definitely very different between
the two subspecies. The comparative analysis of males in this study also finds a significant
difference between the two races (Table 5). There may, therefore, be grounds for the
subdivision of N. afra into two subspecies. On the other hand, the dines appear to be rather
smooth, with no pronounced step (Table 7) at the area proposed by Clancey (1962, 1980) as the
boundary between the two subspecies, namely the Transkei-Eastern Cape border. Again, the
distributional plot of two clusters generated by the computer (Fig. 2) may prove useful. This
shows a very wide overlap in the distribution of clusters 1 and 2, and serves as further evidence
to suggest the dines in the characters of N. afra are very smooth. Moult data (Table 10) do not
suggest any regional variation in the timing of wing-moult, although the samples for Natal and
the Transvaal are small. Further study is required to resolve the problem, but we would question
whether the recognition of subspecies in N. afra is justified.

So far, the ‘significant differences’ between the populations of birds discussed have been
based on the grouping of individuals from allopatric areas, and then comparing the averages
obtained for each group. This study has shown the occurrence of great variation within each of
these groups, and the existence of dines in character measurements. The borders of the areas
used in the analysis of dines were arbitrary. The delimitation of the distributional range and the
description of each subspecies seems to have involved a critical sample size never greater than

148

LLOYD & CRAIG: MORPHOMETRICS, MOULT AND TAXONOMY OF NECTARINIA COMPLEX

20 (Clancey, 1962, 1975; Clancey and Irwin, 1978; Winterbottom, 1963). The probability of such
small, and often obviously select samples being adequate to describe geographical variation is
slight. The subspecies currently recognised for N. afra and N. chalybea thus appear to be poorly
defined units. After all, if a dine exists through the distribution of a species, it can be split up
into a variable number of populations using any boundaries, which are subsequently found to
separate ‘significantly different’ groups.

TV. chalybea shows a wide tolerance of habitat, being found from the dry karroid scrub of
the Southwestern Cape to wet forests in northern areas of its range. Despite this wide tolerance,
Skead (1967) reports that where the N. chalybea overlaps with N. afra in the dry Valley Bushveld
and Macchia of the Southern and Eastern Cape, the former inhabits forest and seldom ventures
into the more open thornveld which is the favoured habitat of N. afra. The latter is reported to
avoid forests, only venturing into forest edges and narrow strips of dune forest that are bushier
and less well-treed. Tarboton et al. (1987) repoft that, in the Transvaal, N. chalybea is found in
montane forest and its edges, whilst N. afra occurs in more open, hilly country, especially where
aloes are concentrated. These differences in habitat preference are also supported by Maclean
(1985), and Hall and Moreau (1970). In the Western Cape, where N. afra and the Black Sunbird
N. amethystina do not occur, N. chalybea prefers the macchia scrub to the forest (Skead, 1967).
During the course of field observations in the Eastern Cape the two species were only observed
together in forest-edge situations and in suburban gardens. The latter is an artificial habitat, and
with large numbers of cultivated flowers, nectar will usually be more abundant than in the
adjacent natural habitats.

It is possible that N. chalybea, owing to its smaller size, has been out-competed in the
habitats occupied by its larger relatives, especially the very closely related N. afra, and become
adapted to a different ecological habitat. Since the African sunbirds commonly puncture the
base of flowers to obtain nectar which they cannot otherwise reach (Skead, 1967; pers. obs.),
differences in bill size are unlikely to be associated with different flower preferences. However,
the importance of insect food to sunbirds, and the role of bill morphology in prey capture, need
further study. Variations in bill size follow a similar geographical trend in both these species (see
Tables 6 and 7), and the present data provide no suggestion that morphological changes can be
ascribed to interspecific interactions in sympatry.

ACKNOWLEDGEMENTS

We thank P.E. Hulley for critical discussion; H.D. Jackson of the Natural History Museum
of Zimbabwe (Bulawayo), A.C. Kemp of the Transvaal Museum, J. Mendelsohn of Durban
Museum, C.J. Vernon of East London Museum, B.R. Wilmot of Albany Museum, and
D. Hamerton of the South African Museum for access to material in these collections. The study
was supported by a research grant from Rhodes University and a special projects grant from
Johnson and Johnson Ltd.

REFERENCES

Clancey, P.A. 1962. On the status of the taxon Nectarinia afra afra (Linnaeus). Durban Museum Novit. 6(15): 187-189,
Clancey, P.A. 1975. Endemic birds of the Transvaal Montane forests. Durban Museum Novit. 10(12): 151-180,
Clancey, P.A. 1980. S.A.O.S. checklist of southern African birds. Johannesburg: Southern African Ornithological
Society.

Clancey, P.A. and Irwin, M.P.S. 1978. Species limits in the Nectarinia afralN. chalybea complex of African Double-
collared Sunbirds. Durban Museum Novit. 11(20): 331-350.

149

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 6, JULY 1989

Craig, A.J.F.K. 1988. The timing of moult, morphology, and an assessment of the races of the Redwinged Starling. Bonn,
zool. Beitr. 39: 347-360.

Hall, B.P. and Moreau, R.E. 1970. An atlas of specialion in African passerine birds. London: British Museum (Natural
History) .

Maclean, G.L. 1985. Roberts' birds of southern Africa. Cape Town: John Voelcker Bird Book Fund.

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

Newton, I. 1966. The moult of the bullfinch Pyrrhula pyrrhida. Ibis 108; 41-67.

Skead, C.J. 1967. The simbirds of Southern Africa also the sugarbirds, the white-eyes and the Spotted Creeper. Cape Town:
Balkema.

Tarboton, W.R., Kemp, M.I. and Kemp, A.C. 1987. Birds of the Transvaal. Pretoria: Transvaal Museum.

Wiens, J.A. (Ed.) 1982. Forum: avian subspecies in the 1980’s. Auk 99: 593-615.

WiNTERBOTTOM, J.M. 1963. Systematic notes on birds of the Cape Province. 23. Nectarinia chalybea. Ostrich 34: 155-156.

150

INSTRUCTIONS TO AUTHORS

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

Only generic and specific names should be underlined to indicate italics. All other marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

BK0020 — CTP Book Printers, Cape

:'x

}

ISSN 0570-1880

^6? 1 3

KJI4

Annals of th^

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 7 31st August 1989

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

Notes on nesting behaviour in Bembix bubalus Handlirsch in southern
Africa with the emphasis on nest sharing and reaction to nest parasites

(Hymenoptera: Sphecidae)

SARAH K. GESS and F.W.GESS

(Albany Museum, Grahamstown)

ABSTRACT

Data on the nesting behaviour of Bembix bubalus Handlirsch are presented. Of particular
note is an instance of nest sharing as this appears to be the first record of nest sharing by a
progressive provisioner in the Sphecidae. An instance is recorded of a wasp provisioning, in
addition to its own larva, the larvae of a sarcophagid ffy which had parasitised the nest.

INTRODUCTION

Due to the current interest both in nest sharing and in response to parasitism it seems useful
to publish this account of the nesting of Bembix bubalus despite its preliminary nature,
particularly as opportunities for such studies are not as common in southern Africa as might be
supposed.

The present paper is based on a preliminary investigation of three days duration carried out
during a short visit to the Oudtshoorn district in December 1986. A follow up investigation was
planned for January 1987 but due to unforeseen problems a second visit to the area in the
summer of 1986/1987 did not come about. The site was revisited in December 1987, that is in the
following summer. Unfortunately, whereas there had been a thousand or more individuals of
Bembix bubalus nesting there, when the preliminary observations were made, there were then
fewer than a dozen wasps attempting unsuccessfully to excavate nests. The sand had become so
destructured, due to drought and trampling, that their efforts were in vain. This decline in
nesting activity was particularly disappointing as, in the course of their investigations of the
nesting of wasps in the arid areas of southern Africa during the past two decades, the authors had
not previously found any species of Bembix nesting in aggregations of more than about a dozen.
Nesting aggregations of several hundred (Evans, 1957) or thousand (Bequaert, 1932) have,
however, been reported for some species of elsewhere in the world.

THE NESTING OF BEMBIX BUBALUS HANDLIRSCH
Description of the nesting area

The nesting aggregation covered an area, approximately 300 m^, of level friable sandy soil

151

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 18, PT. 7, AUGUST 1989

Fig. 1. Nesting area of Bemhix bubaliis Handlirsch at Onverwacht, Oudtshoorn district.

sparsely vegetated. The site lies on the flood plain of the Kammanassie River at Onverwacht
(33° 37' 35"S, 22° 14' 18"E) to the south of Oudtshoorn in the Little Karoo.

This area situated immediately inland of the Outeniqua Mountains lies in a rain shadow and
consequently receives an annual rainfall of only 240 mm. Rain may fall during any season of the
year, however, spring and late summer are the wettest periods. The soils above the flood plain
are relatively coarse grained and are of the Cretaceous Enon Formation. Those of the flood plain
on which the nesting site of B. bubaliis is situated are light coloured, finer textured and are of
diverse provenance having been carried down from further east by the river. The area lies within
Acocks’ Veld Type 26, False Karroid Broken Veld (Acocks, 1953, 1975 and 1988). The area is
characterized by dwarf scrub, with a noticeable succulent element, and with taller shrubs mainly
along the water courses (Fig. 1).

Nest excavation and temporary closure

The nests were excavated in the usual Bembix manner: that is the two fore-legs, equipped
with sand rakes, were repeatedly swept back in unison whilst the abdomen was synchronously
raised and lowered allowing the soil to be shot out behind the wasp.

The spoils of excavation were drawn back to some little distance from the nest entrance
where they accumulated to form a tumulus up to 65 mm in diameter. Throughout nest
excavation and provisioning the tumulus was not dispersed by the wasp.

152

GESS & GESS: NOTES ON NESTING BEHAVIOUR IN BEMBIX BUBALUS HANDLIRSCH

Fig. 2. Female Bembix bubalus Handlirsch in flight carrying her prey, a bombyliid fly, held ventral side up close beneath

her.

Nests were temporarily sealed with sand when wasps were away from them and when wasps
were within them but not actively working, for example in cloudy weather or at night.

During nest excavation miltogrammine flies were observed stationed on perches, such as
nearby twigs, in close proximity to nests. From their behaviour it was clear that they were
monitoring the wasps’ nesting activities.

Male behaviour

Males were common in the nesting area during the period of the investigation 10-12. xii. 86
when nesting by a thousand or more females was in full swing. They were seen to fly rapidly low
over the ground frequently changing direction and patrolling the entire area of the nesting
aggregation.

At the time of the second visit to the study site on 8. xii. 87 when there were fewer than a
dozen females attempting to nest only one male was observed. This male was actively interested
in all the females. Each time he spotted a female he advanced towards her with a markedly high
pitched buzzing flight, came above her and tapped her with his abdomen. Only one female was
seen to accept his advances, that is to permit him to grasp her and to fly off with her. Actual
copulation was not observed.

153

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 18, PT. 7, AUGUST 1989

Identity of the prey, carriage of the prey and satellite flies

Seventy eight prey were obtained, 24 from females captured flying with prey and 54 from
nests (Table 1). The latter category was made up of complete prey and prey in the form of
recognizable remnants. All prey were flies, the vast majority (91%) being flower-visiting flies of
the families Bombyliidae and Syrphidae, suggesting that B. bubalus seeks prey at flowers.

Although actual prey capture was not witnessed large numbers of wasps were observed
flying with prey. In many instances these wasps were accompanied by satellite flies. The prey fly
is held ventral side up, close beneath the wasp by her middle-pair of legs when she is in flight
(Fig. 2). On nearing her nest the wasp moves the fly back (Fig. 3) so that her approach to the
nest is very “tail heavy”. She alights at the concealed nest entrance and immediately clears the
sand with her fore-legs whilst standing on her hind-legs (Fig. 4) and continuing to hold the prey
with her middle-legs. Opening of the nest and entry into it are rapid so that little opportunity is
given to the satellite flies for larviposition.

Of interest was the observation of a female which due to some disturbance dropped her
prey, a large syrphid, and without hesitation returned, dived down and picked up her prey
without alighting.

Description of twelve nests, the nature of their contents and a record of the presence of three wasps
in a nest

A sample of twelve nests was excavated. Nine of these were unicellular and of a typical basic

Fig. 3. A female Bembix bubalus Handlirsch in flight carrying her prey, a syrphid fly (Eristalinus taeniops (Wied.)), and

preparing to land.

154

GESS & GESS: NOTES ON NESTING BEHAVIOUR IN BEMBIX BUBALUS HANDLIRSCH

Table 1.

Prey of Bembix bubalus Handlirsch taken with wasps in flight and from the nests excavated at
Onverwacht, Oudtshoorn district, 9-12 .xii.86.

Identity of prey

no. taken
with wasp
in flight

no. taken from nest

no. total

7STRATIOMY1DAE

‘.’Genus and species

—

TABANIDAE

Chrysops obliquifasciata Macquart

—

BOMBYLllDAE

Bombvlius discoideus E.

Bombylius ornalus Wied.

—

Bombylius ?sp.

—

Exoprosopa sp. A

—

Exoprosopa sp. B

—

Exoprosopa

—

Sysioechus sp. A

—

Svsioechus sp. B

—

Villa sp. A

—

Villa sp. B

'.’Genus and species A

—

'.’Genus and species B

SYRPHIDAE

Allograpta calopus Wied.

—

Eristalinus taeniops (Wied.)

—

Eristalis tenax (L.)

'.’Genus and species A

—

'.’Genus and species B

MUSCIDAE

Muse a sp.

SARCOPHAGIDAE

'.’Genus and species

—

TACHINIDAE

'.’Genus and species A

—

‘.’Genus and species B

'.’Genus and species C

155

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 18, PT. 7, AUGUST 1989

Fig. 4. A female Bembix bubalus Handlirsch holding her prey, a syrphid fly (Eristalinus taeniops (Wied.)), with her
middle-legs whilst standing on her hind-legs and opening her nest with her fore-legs.

Bembix nest pattern (Evans, 1957 and Gess, 1986), that is, with a short sloping entrance burrow
dipping down to end in a spur and giving rise just above the spur to a secondary shaft ending in
a large ovoid cell (Fig. 5). Of the three remaining nests one was two-celled, one three-celled and
the third four-celled (Fig. 6). Shaft diameter was 9-9,5 mm, cell diameter 12,5-14 mm and cell
depth 130-170 mm.

Six of the unicellular nests each contained a wasp larva with one or two prey flies and fly
fragments. Two were empty. The ninth contained a wasp larva, three large sarcophagid
maggots, two small sarcophagid maggots, eleven partially eaten prey flies and fly remains.

Only one cell in each of the two-celled nests and in the three-celled nest contained a wasp
larva and prey flies.

In the four-celled nest two of the cells contained cocoons and fly remains, the third a large
wasp larva and fly remains, and the fourth a small wasp larva, fly remains and a freshly
introduced prey fly. One female had been observed taking a fly into the nest and a female,
presumed to be this female, had been captured when she was leaving the nest. The nest upon
excavation was found to contain two additional females. Each was positioned facing outwards
within a cell containing a feeding larva which she appeared to be guarding.

It is noteworthy not only that three females should have been present in the nest but that,
that being so, there were only two wasp larvae being actively provisioned.

156

GESS & GESS: NOTES ON NESTING BEHAVIOUR IN BEMBIX BUBALUS HANDLIRSCH

Fig. 5. Plan of a unicellular nest of Bembix bubalus Handlirsch investigated at Onverwacht, Outshoorn district,

10-12. xii. 86

Fig. 6. Plan of the four-celled nest of Bembix bubalus Handlirsch investigated at Onverwacht, Oudtshoorn district,
10-12. xii. 86: a-cell containing wasp-cocoon and fly remains; b-cell containing large wasp-larva and fly remains; c-cell
containing small wasp-larva, entire fly and fly remains.

157

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 18, PT. 7, AUGUST 1989
I

Visits to nests by more than one female

It was observed that in several instances nests were being visited by more than one female.
Furthermore when two females had entered a nest neither was evicted. During the short period
of the study, three days, during which provisioning of nests was being actively pursued no case
of two or more females taking prey into a nest was noted. It is therefore not known whether any
one nest was being provisioned by more than one female.

Provisioning mode

In common with the majority of species of Bembix, B. bubalus is clearly a progressive
provisioner, that is the larva is provided with fresh prey throughout its development.

Voltinism

B. bubalus is clearly bi-voltine or multivoltine as an adult emerged before the end of the
86/87 summer season from a cocoon obtained from one of the cells of the four-celled nest.

DISCUSSION

It seems of interest to consider whether the findings for B. bubalus are of any particular
significance in a consideration of the ethology of Sphecidae and of solitary aculeate wasps as a
whole.

The points to be considered are: the variation in the number of cells; the presence of three
wasps in one of the nests in relation to the development of presocial behaviour; and the wasp’s
reaction to the presence of sarcophagid maggots in her nest.

The variation in the number of cells in itself is not of particular note. Tsuneki (1956), in
Japan found that, though the nest of Bembix niponica F. Smith is typically unicellular, in some
nesting aggregations about 4% of the nests were compound and in one exceptional aggregation
the majority of nests contained two or more cells. Further, observations on Bembix nubilipennis
Cresson in North America (Evans, 1966) suggest that that species when bi-voltine may make
multi-cellular nests in spring and unicellular nests in autumn.

What is, however, of note is that the four-celled nest of B. bubalus was occupied by three
wasps. It would appear therefore that B. bubalus practises facultative nest sharing. The nearest
recorded approach to this condition in Bembix seems to be the observation by Evans (1966), in
North America, that in Bembix amoena Handlirsch nesting aggregations, where nests are in very
close proximity, one nest entrance may serve more than one nest. However, in that species the
burrows diverge immediately beneath the ground surface so that there is only entrance sharing,
not nest sharing as exhibited by B. bubalus.

Observations that two wasps could enter a nest without either being evicted and that one
nest was found to be occupied by three wasps is indicative of a breakdown in territoriality, which
is a pre-condition to nest sharing and co-operative behaviour.

Relatively few studies of Sphecidae provide evidence of co-operative behaviour. An
interesting example of differential aggression is exhibited by the mass provisioning nyssonine
Sphecius speciosus Drury (Pfennig and Reeve, 1989) which tolerates the intrusion of large near
neighbours into the nest but only when no prey cicada is exposed. Examples of actual nest
sharing have been recorded for species in the sub-families Sphecinae (Brockmann and Dawkins,
1979; Brockmann, Grafen and Dawkins, 1979; Eberhard, 1972 in Evans, 1977), Pemphredon-

158

GESS & GESS: NOTES ON NESTING BEHAVIOUR IN BEMBIX BUBALUS HANDLIRSCH

inae (Matthews, 1968 in Evans, 1977), Crabroninae (Bowden, 1964; and Evans, 1964 and
Peters, 1973 both in Bohart and Menke, 1976) and Philanthinae (Alcock, 1975; Evans, 1973;
Evans and Hook, 1982a, 1982b, and 1986; Hook, 1987). All are for mass provisioners, that is all
the provision required for the development of the larva is generally provided before the hatching
of the egg. The present record for B. bubalus is therefore probably the first for nest sharing by
a nyssonine and is also probably the first example of nest-sharing by a progressive provisioner to
be recorded for the Sphecidae. That B. bubalus is a progressive provisioner is of particular
interest as it therefore exhibits a combination of para-social behaviour, that is adult females of
the same generation associate in a common nest, and sub-social behaviour, that is the larvae are
cared for by a parent for some time after hatching. This behaviour combination has otherwise,
for wasps, been recorded only in the Vespoidea. Furthermore as B. bubalus is at least bi-voltine
the possibility exists that wasps of two generations may come into contact in a nest.

At this point it is of interest to consider the accounts of Alcock (1975), Evans and Hook
(1982a, 1982b and 1986) and Hook (1987) of nest sharing by some species of Cerceris
(Philanthinae). Some evidence was obtained of differentiation in the roles of the females
associated with the nest. Some appeared to be provisioners and others non-provisioners.
Females of both types showed mandibular wear, suggesting that both were involved in preparing
cells, also both types had ovaries containing well-formed oocytes, which suggested that both laid
eggs more or less regularly. Non-provisioners were considered to perform an important function
as guards but no suggestion was made as to what factors determine whether a female is a
provisioner or not.

The fact that in the four-celled B. bubalus nest three females were present but only two cells
were being provisioned and that both of these were being guarded leads one to ask whether there
is here too some degree of division of labour or co-operation between the females.

Finally it is of interest to consider the response by B. bubalus to nest parasites. It has been
suggested that gregarious nesting may be a “selfish herd” response to parasites and Wcislo
(1984) sees parasites as important agents of selection for the maintenance of aggregations and
thus for more advanced social levels as well. Evans and Hook (1986) in their study of Cerceris
were satisfied that there is tittle doubt of the importance of guards in protecting the nest contents
from invasion by ants and mutillids. On the other hand they found that flies which operate at the
nest entrance are not deterred by guards, in fact, the delay sometimes caused by the guard to a
prey-laden wasp entering the nest actually enhanced the attack by satellite flies which larviposit
on the prey at the nest entrance. As Cerceris is a mass provisioner the introduction of fly maggots
into the cell results not only in the consumption of the available prey but also of the wasp larva.
Sphex decipiens Kohl (Sphecinae), the nesting of which was studied in the Grahamstown district
by Gess and Gess (unpublished field notes, 1985/1986), which is a mass provisioner suffers a high
incidence of loss of provision and larvae to the ravages of fly maggots. Evans (1966) suggested
that in the case of progressive provisioners the fly maggots would most probably be detected and
destroyed, and goes as far as to say that in the genus Bembix progressive provisioning has led to
parasitism by miltogrammine flies being reduced to virtually zero (Evans, 1977).

This might well be the case in species which clean out their nests. The presence of fly
remains in cells of B. bubalus containing cocoons indicates that this species does not clean out
its cells. It was therefore of interest to discover its reaction to the presence of fly maggots in its
cells. One single-celled nest was found to have been so parasitised. The cell which was still being
actively provisioned contained a large wasp larva, three large fly maggots and two small fly

159

ANN. CAPE PROV. MUS. (NAT HIST.) VOL. 18, PT. 7, AUGUST 1989

maggots. The wasp had not as one might have expected destroyed the fly maggots or abandoned
the nest but was provisioning the maggots as well as her own larva. In doing so the wasp
undoubtedly saved her larva from destruction by the maggots, however, she had succeeded in
rearing only one wasp offspring when she might have reared several had she not been
provisioning the maggots. Had she been a mass provisioner her energies would not have been
wasted in this way. This seems a curiously disadvantageous result of what would otherwise
appear to be an advanced behaviour pattern.

Clearly investigation of many more nests is required before any more definite statements
can be made concerning: the incidence of nest sharing: the nature of and degree of co-operation
between wasps sharing a nest: and the incidence of and response to nest parasitism.

ACKNOWLEDGEMENTS

The authors wish to thank A. W. Hook and W. Wcislo for reading and commenting on an
earlier version of the manuscript.

H. W. Gess is thanked for taking the photographs reproduced as figures 2, 3 and 4.
Gratitude is expressed to the C.S.I.R. by F. W. Gess for a running expenses grant 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. hot. Surv. S. Afr. 29: i-iv, 1-192.

Acocks, J. P. H. 1975. Veld Types of South Africa. Mem. hot. Surv. S. Afr. 40: i-iv, 1-128.

Acocks, J. P. H. 1988. Veld Types of South Africa. Mem. hot. Surv. S. Afr. 57: i-x, 1-146.

Alcock, j. 1975. Social interactions in the solitary wasp Cerceris simplex (Hymenoptera; Sphecidae). Behaviour 54 (1-2):

142-152.

Bequaert, j. 1932. The Tabanidae of the American Museum Congo Expedition, 1909-1915. Amer. Mus. Nov. 539: 1-19.
Bohart. R. M. and Menke, A. S. Sphecid wasps of the world. Berkeley; University of California Press.

Bowden, J. 1964. Notes on the biology of two species of Dasyproctus Lcp. and Br. in Uganda (Hymenoptera: Sphegidae).
J. Etu. Soc. sth. Afr. 26 (2): 425-437.

Brockmann, H. j. and Dawkins, R. 1979. Joint nesting in a digger wasp as an evolutioary stable preadaptation to social
life. Behaviour 71 (3-4): 203-245.

Brockmann, H. J., Grafen, A. and Dawkins, R. 1979. Evolutionarily stable nesting strategy in a Digger Wasp. J. theor.
Biol. 77: 473-496.

Evans, H. E. 1957. Comparative ethology of digger wasps of the genus Bembix. Ithaca, New York: Comstock Publishing
Associates.

Evans, H. E. 1966. The comparative ethology and evolution of the sand wasps. Cambridge, Massachusetts: Harvard
University Press.

Evans, H. E. 1973. Burrow sharing and nest transfer in the digger wasp Philanthus gibbosus (Fabricius). Anim. Behav.
21 (2); 302-308.

Evans, H. E. 1977. Extrinsic versus intrinsic factors in the evolution of insect sociality. BioScience 27 (9): 613-617.
Evans, H. E. and Hook, A. W. 1982a, Communal nesting in Australian Cerceris Digger Wasps. In: Breed, M. D,,
Michener, C. D. and Evans, H. E. eds. The Biology of Social Insects. Boulder, Colorado: Westview Press.
Evans, H. E. and Hook, A. W. 1982b. Communal nesting in the Digger Wasp Cerceris australis (Hymenoptera:
Sphecidae). Aust. J. Zool. 30: 557-568.

Evans, H. E. and Hook, A. W. 1986. Nesting behaviour of Australian Cerceris Digger Wasps, with special reference to
nest reutilization and nest sharing (Hymenoptera: Sphecidae). Sociobiology 11 (3): 275-302.

Gess, F. W. 1986. Three new species of southern African Bembix, a new synonomy, and biological notes on other species
of the genus (Hymenoptera: Sphecidae: Nyssoninae). Ann. Cape Prov. Mus. (nat. Hist.) 16 (6): 137-160.
Hook, A. W. 1987. Nesting behaviour of Texas Cerceris Digger Wasps with emphasis on nest reutilization and nest sharing
(Hymenoptera: Sphecidae). Sociobiology 13 (2): 93-118.

Pfennig, D. W. and Reeve, H. K. 1989. Neighbor recognition and context-dependent aggression in a solitary wasp,
Sphecius speciosus (Hymenoptera: Sphecidae). Ethology 80: 1-18.

Wcislo, W.T. 1984. Gregarious nesting of a digger wasp as a “selfish herd” response to a parasitic fly (Hymenoptera;
Sphecidae; Diptera: Sarcophagidae). Behav. Ecol. Sociobiol. 15: 157-160.

160

p|!^^r’>' ?t<' •SJf lip/ii’’^-'^''' ■ '.^•i’'" ,'

-*' P-' ■ *1 •-•«' -V r. /„

ShLk!. -u ^.’f^'^^r^ .i-t'/ .1, • . , i. 1

SSMf-ittSJiS-L'F ■•&'•. .Jfcirt. Vt.. 'JfUu -.

V ‘V-

^ju' •'■ .: ••■•■ ■ !
■’'■■ tt- -yf -.

V'WC , V

:i, •:■»•?,••■ -/■ :'r ;,Ti¥

/■

''r^:4''-.^r^''

iV:! ■- '

•C' .5

,Vi

■^ .

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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. I23-I29.

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.

BK0273 — CTP Book Printers, Cape

-niv

ik''

'"v> V"

^ o,'^ ^}f

PBi' ••*'«ii»-i4 *i- • “■•■v/•.1^r■ ■■ .•vT*''^*--i:. . V. ». V ■ ^«:.'L.-,.. -^ .^ . . . ':^„v‘. -V ‘

'UX'i

-'<tv

!,,y.o

r. •.■■-• »•'•' ■:•• *'4^--l!^M -:;^^^itl'‘U >*• l**t rS . vin

i- ^

■7 .'"^r .•.

: ' '^,<4i

-» ^’\ -

"'"•Vi

i>V >1

7*. /jS

ISSN 0570-1880

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 8 30th April 1990

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

Early ontogeny and notes on breeding behaviour, habitat preference and
conservation of the Cape kurper, Sandelia capensis (Pisces: Anabantidae)

J. A. CAMBRAY

(Albany Museum, Grahamstown, 6140, South Africa)

ABSTRACT

The development of the eggs, free embryos and larvae of Sandelia capensis is described
from laboratory-reared specimens.

The eggs had 1,0-1, 4 mm diameters, adhesive egg envelopes, narrow perivitelline spaces
and single oil globules 0,62 mm in diameter. The eggs were negatively buoyant.

The free embryos hatched at a premature ontogenetic stage and were 3,0-3, 6 mm NL.
They attached themselves to objects using the adhesive surface on the dorsum of their heads. At
4,7 mm NL the oil globule began to migrate dorsally and divided into a right and a left globule
and during that period the swimbladder inflated. These asymmetrical oil globules disappeared at
between 6, 3-6, 9 mm NL. The swimbladder began to extend posteriorly at 6,25 mm SL and
reached maximum posterior penetration to the caudal peduncle at 14,4 mm SL. The larval fishes
commenced feeding at 5,35 mm NL. Pelvic buds formed between 6, 9-7, 2 mm SL. The larval
fish transformed to juveniles by 13,0 mm SL.

Notes on the breeding behaviour, habitat preference and conservation of S. capensis are
included.

INTRODUCTION

The early life history of 5. capensis is poorly known, in fact there have been few studies on
the early development of African anabantids (Cambray and Teugels, 1988).

Sandelia capensis is one of the freshwater fish species which is endemic to the Cape Fold
Mountain region. The natural distribution of S. capensis is confined to the Cape Province where
it occurs in most lowland and middle reaches of Cape rivers from the Coega River in the Eastern
Cape to Verlorevlei and the Berg River in the South-Western Cape (Jubb, 1965; Gaigher et al.,
1980). Hamman et al. (1984) reported an introduced population which had become established
in a tributary of the Olifants River System in the Western Cape Province.

Sandelia capensis is capable of tolerating a wide variety of water conditions, both physical
and chemical (Harrison, 1952). It is found in the upper reaches of clear, swift-flowing rivers, in
intensely peat-stained acidic rivers, alkaline rivers, in heavily silted rivers, and muddy still

159

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

waters. Harrison (1952) also noted that it survives in very stagnant back-waters and cut-offs
which are avoided by Barbus species and the exotic predators, trout and black bass.

The presence of Sandelia, which belongs to the primary division of freshwater fishes (see
Hofmeyr, 1966) being completely isolated in the southern tip of Africa, is a zoogeographical
puzzle (Jubb, 1967). The two anabantoid species, Sandelia capensis (Cuvier in C. and V., 1831)
and S. bainsii Castelnau 1861 are unusual elements of the Cape ichthyofauna. The members of
this family, which is considered to be the most primitive of the anabantoid families (Liem, 1963),
are mainly distributed in tropical latitudes (Skelton, 1986). Liem (1963) considered that
S. capensis was derived from the genus Ctenopoma that occurs in tropical Africa. However,
Skelton (1986) considered that the phylogenetic relationships of the two Sandelia species
required further study according to cladistic principles. S. Norris (in litt.) is examining the
interrelationships of the African Anabantidae. At present the anabantid genus, Sandelia, is a
biogeographically uninformative element in the southern African fauna (Skelton, 1986).

METHODS

Adult S. capensis were collected from a feral population in the Thomas Baines Nature
Reserve near Grahamstown.

Sandelia capensis spawned in the aquarium, and developing embryos were collected from
four separate spawning acts and each spawning was put into a separate rearing aquarium. The
embryos were collected for observation, drawing and photographing at various stages of
development. The early larval stages were fed live rotifers and sieved daphnia. Older larval fish
were fed daphnia and a balanced flake food. The temperature in the rearing tanks was
maintained at between 22 and 24 °C and one-third of the water was replaced every five days. The
pH of the water was neutral to slightly alkaline (7, 0-7, 2).

Specimens were illustrated with the aid of a camera lucida and a binocular dissecting
microscope. Eleven morphometric and seven meristic characters were measured or counted.
Meristic characters included counts of preanal and postanal myomeres and caudal (principal and
secondary), dorsal, pectoral and pelvic fin rays and spines.

Measurements of the smaller specimens were taken using an ocular micrometer in a
dissecting microscope and larger specimens were measured with dial callipers (accuracy
0,05 mm). Measurements were made at least six months after preservation. All measurements
in the text are from preserved material unless otherwise noted.

DEFINITION OF TERMS USED.

Lengths of specimens are reported as notochord length (NL) or standard length (SL) in mm
unless otherwise stated. Total length (TL) is measured from the snout tip, through the horizontal
body axis, to the end of the caudal finfold or a perpendicular to the end of the longest caudal fin
ray. Standard length (SL) is measured from the snout tip, through the horizontal body axis to the
end of the median bones at the caudal base. Standard length measurement is not accurate until
the horizontal alignment of the median hypural bones is completed. Notochord length (NL) is
measured from the snout tip of the notochord before its dorsal flexion, and afterwards
perpendicular to the horizontal body axis through the tip of the upturned notochord (Berry and
Richards 1973). Eye diameter (ED) is the maximum width of the eye measured on the horizontal
axis. Head length is the horizontal distance from the snout to the posterior edge of the opercle.
Preanal length (PL) is the distance from the snout to the origin of the anal fin, (in small larvae

160

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

before differentiation of the anal fin, the measurement was taken from the snout to the posterior
edge of the anus). Head depth (HD) is the length of a vertical line immediately posterior to eye.
Body depth at pectoral fin (BDp) is the depth of the body at pectoral fin base. Body depth at
anus (BDa) is the depth of body at posterior edge of anus. Head width (HW) is the width of head
immediately posterior to eyes. Body width at pectorals (BWp) is the width of the body at base
of pectorals. Preanal myomeres are all those segments of which their bordering myosepta are at
least partly anterior to the anus including one segment anterior to the first myoseptum. Postanal
myomeres are all those segments posterior to preanal myomeres including a urostylar segment
(Fuiman, 1982). Myomeres were counted with the aid of polarizing filters. Myomeres of large
opaque specimens were not counted because of lack of clarity.

Embryos and larvae were preserved in a 5% phosphate-buffered formalin solution after
being tranquillized with benzocaine to relax specimens and reduce curvature during fixation.

The surface of the egg envelope was viewed, up to a magnification of 6 000 x , using a JEOL
JSM 840 Scanning Electron Microscope operating at 10 kV. The eggs had been preserved in 5%
phosphate buffered formalin. The eggs were prepared for SEM work by standard procedures,
alcohol dehydration and critical point drying. The samples were mounted on aluminium stubs
and coated with a thin layer of gold in a sputtering device. Several eggs were also prepared using
the Cryo technique, which involves freezing the eggs in liquid nitrogen, coating with gold and
then viewing.

A population of S. capensis in the Wit River, a clear, slightly acidic (pH 6, 6-6, 9), Cape Fold
Mountain tributary of the Gamtoos River System, was observed and early life history stages
were collected. The habitat of these early stages was recorded.

All specimens have been catalogued (AMG/P 12245) in the Ichthyological Collection of the
Albany Museum, Grahamstown.

RESULTS

FERTILIZATION

The exact time of activation was not known. Eggs were removed from the male’s nest and
transferred to a rearing aquarium, after they had been fertilized naturally in the breeding tank.

EGG DESCRIPTION

Ripe unshed eggs were pale yellow in colour with a mean diameter of 1,1 mm
(SD = 0,05 mm, n = 25). Water-hardened S. capensis eggs were strongly adhesive and
negatively buoyant. The outer egg envelope remained strongly adhesive until the free embryos
hatched. The water-hardened eggs were 1,28 mm (SD = 0,08 mm, n = 31) in diameter and were
initially crystal clear in colour, mainly round, as only 9 out of 31 measured eggs were slightly
elliptical. Therefore the swelling of the egg membrane was minimal (0,18 mm) or 14% of the
diameter of the shed fertilized egg. A single oil globule measuring 0,62 mm (SD = 0,04 mm,
n = 31) was suspended in the yolk (Fig. la). When the eggs were artificially released from their
adhesive contact the oil globule was not large enough to give positive buoyancy to the eggs.
Surface tension was sufficient to keep the eggs buoyant but if the surface tension was broken the
eggs sank.

The egg envelope remained highly adhesive until the free embryos hatched. Some of the
free embryos remained connected to the broken egg envelope (Fig. 5a) which could be seen on
the aquarium gravel to which the developing embryo had been attached.

161

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

Fig. 1. Several early embryonic stages of Sandelia capensis. All ages given are calculated from time of first egg collection,
which is the egg shown in (a), (a) Blastodermal cap, age 0 h. (b) Equatorial position, age 3,5 h. (c) Age 4,5 h. (d) Neural
keel, age 6,5 h. AH the eggs are between 1,0 and 1,4 mm in diameter, bc-blastodermal cap; nk-neural keel; og-oil globule.

162

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

The ultrastructure of the egg envelope of S. capensis was viewed up to a magnification of
6 OOOx and only a faint ultrastructural patterning was observed (Fig. 3). The egg envelope of S.
capensis was therefore fairly smooth with round markings and a central raised structure. The
Cryo treated material gave the best results.

EMBRYONIC DEVELOPMENT

On the 24-01-1985 between 14h00 and 15h00 one male S. capensis started to chase a
female. At this time petri dishes were put in the nesting area of the male and at IVhOO, water
temperature 23 °C, several eggs were found in one of the petri dishes. The eggs were examined
and found to be in a many-celled blastodermal cap stage (Fig. la). The cells were opaque and
the yolk was clear.

The blastoderm was in equatorial position after 3,5 h (Fig. lb). The neural keel was evident
after 4,5 h (Fig. Ic) and after 6,5 h the neural keel was a dominant feature of the developing
embryo (Fig. Id). The oil globule within the yolk made it very difficult to orientate the
developing embryo for observation and photographing. After 14 h from collection time the optic
cups and Kupffer’s vesicle had formed (Fig. 2a). Somites and pigment were first observed after
16,5 h (Fig. 2b). The black pigment became a dominant feature of the developing embryo after
this stage and obscured some of the embryonic development. The pigment developed as large
irregular ‘blotches’ on the embryonic head and over the surface of the yolk (Fig. 2c). After 22 h
the eye lens was forming and the somites almost reached to the tail tip (Fig. 2c). Muscular
contractions were first seen after 24 h (the heart-beat stage), also the otic placode had just
started to form (Fig. 2d).

At 26 h the circulatory system had developed over the surface of the yolk and under the
notochord. After 30 h otoliths were observed and approximately 20% of the tail was free from
the yolk-sac, the heart beat was 112 (SD = 4, n = 3). After 32 h the tail overlapped the head
and the embryo was now actively twisting and revolving in the egg envelope.

After 40,5 h the first free embryo hatched (3,1 (3,0) mm TL (NL), alive) (Fig. 4a). When
the free embryos were released from their adhesive attachment they could swim for
approximately 15 seconds in a spiralling upward movement. After this active period they
stopped and passively sank and readhered to either the glass or the aquarium gravel. The
heart-beat had lowered from 123 (SD = 6, n = 3) just prior to hatching to 104 (SD = 10,
n = 3) at hatching. After 42,5 h, seven of the 16 developing embryos had hatched.

FREE EMBRYOS AND LARVAL S. CAPENSIS

Lengths given are those at which the initial formation of selected structures occurred.
AL = alive, P = preserved measurements. After 40,5 h at between 22-23 “"C the first free
embryo hatched (3,1 (3,0) mm TL (NL) AL) (Fig. 4a). Including the results from all spawnings
newly hatched free embryos were 3,47 mm NL (L) (3,0-3, 6 mm NL, SD = 0,15, n = 13), with
a single oil globule 0,65 mm in diameter (SD = 0,08) positioned antero — ^ventrally in the yolk
sac. The dorsum of the head was adhesive (Fig. 4b). When the free embryo became attached to
an object possibly the buoyancy of the oil globule resulted in the ventral surface of the free
embryo being uppermost (Fig. 5a). The newly hatched free embryos had 8-9 preanal and 17-18
postanal myomeres (Table 1). The head remained deflected ventrally over the anterior margin
of the yolk sac (Figs 5b and c) until c 4,6 mm NL (L) and during this period the oil globule
migrated posteriorly (Fig. 6a). The pectoral fin buds were present at 4,35 mm NL and the
swimbladder first inflated at c 4,8 mm NL (P) (Fig. 6b). Functional mouthparts formed at

163

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

Fig. 2. Several early embryonic stages of Sandelia capensis. All ages given are calculated from time of first egg collection,
which is the egg shown in Fig. 1(a). (a) Optic cups forming, age 14 h. (b) and (c) Somite formation and commencement
of pigmentation, age 16,5 h. (d) Heart-beat stage, age 24,5 h. All the eggs are between 1,0 and 1,4 mm in diameter, el-eye

lens; oc-optic cup; p-pigment; s-somite.

164

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

c 4,8 mm NL (P). The bluish-gray coloured oil globules divided into left and right globules at
c 4,7-4,92 mm NL. These globules moved dorsally (Figs 9a-e) and came to lie in a position
flanking the developing swimbladder (Figs 6c, 7a and 9e). The paired oil globules were
asymmetrical (Fig. 9d). The oil globules disappeared at between 6, 3-6,9 mm NL. The yolk was
absorbed at c 5, 1-5,3 mm NL and the larval fish first commenced feeding at c 5,4 mm NL
(alive), (5,35 mm NL preserved). Notochord flexion commenced at c 6,2 mm NL (P) (Fig. 7b)
and was completed at c 6,9 mm SL (P). The caudal finfold exhibited a transitory dorsal lobe
(Fig. 8a) which formed at c 6,9 mm NL (P) and this lobe disappeared at c 6,9 mm SL (P) when
the caudal fin was now rounded (Fig. 8b).

The incipient dorsal fin margin was partially differentiated at 6,5 mm NL (P) and
completely differentiated at 8,2 mm SL (P). The dorsal fin origin was over myomeres 3-4. The
first dorsal rays formed at 7,0 mm NL (P). Theincipient anal fin margin was partially
differentiated at 6,9 mm NL (P) and completely differentiated at c 8,6 mm SL (P). Anal and

Fig. 3. Egg envelope ultrastructure of Sandelia capemis. Scale bar = 1 jum.

165

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

Fig. 4. Post hatching developmental stages of Sandelia capensis. All measurements are for live fish, (a) Newly hatched free
embryo, 3,1 (3,0) mm TL (NL). (b) Two recently hatched free embryos, (top 3,5 (3,4) mm TL (NL) and 3,7 (3,6) mm
TL (NL)). a-adhesive material; np-no pigment; og-oil globule.

166

Meristic characters of 118 Sandelia capensis free embryos, larvae and juveniles used in the present developmental

description.

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

Pectoral fin
rays

0-9

7- 10

10- 14

8- 14

9- 13

11- 14

10- 13

11- 12
11-13

13
12

Pelvic fin
rays

fin buds
fin buds
0-3
3-6
3-5
3-5
6

i + 5
i + 5
i + 4-5
i + 5
i + 5

Anal fin
rays

0-8

8-14

14-16

vii-viii + 7-9
vii-viii + 8-9
vii-viii -1- 8
vii-viii -1- 8
vii-viii + 7-8
vii-viii -1- 8
viii + 8
viii -1- 8
viii + 8
viii -1- 8

Dorsal fin
rays

0-6

4-18

15-22

xii-xiii -1- 7-8
xii-xiii -1- 8-9
xiii + 8-9
xiii + 8-9
xii-xiv -1- 8
xi-xiii 4- 8-9
xiii -1- 9
xiii -1- 8
xiii 4- 8
xiii 4- 8

Caudal fin rays

Ventral

secondary principal

7-9

7- 8
8
8
8
8
8
8
8

8- 9
9

8-9

0-2

2-3

2- 3
3

3- 4
3-4
3-4

Dorsal

secondary principal

7-8

2-3

2- 4

3- 4
3-4

Myomeres

Total

25-26

23-27

25-28

25- 28

26- 28
27-28

Postanal

17-18

15-17

14- 17

15- 17
14-15

opaque

Preanal

8-9
7-10
1 8-11
10-12
10-11
12-14
opaque
opaque
opaque
opaque
opaque
opaque
opaque
opaque
opaque
opaque
opaque

Length interval

NL or SL
(mm range)

3 (3, 4-3,8)

4 (4, 0-4, 9)

5 (5, 1-5, 8)

6 (6,2-6,9)

7 (7,0-7,3)NL

7 (7,0-7,8)SL

8 (8,0-8,6)

9 (9,0-9,95)

10 (10,1-10,75)

11 (11,0-11,9)

12 (12,18-12,71)

13 (13,0-13,9)

14 (14,37-14,4)

15 (15,24)

16 (16,03-16,8)
18 (18,06)

19 (19,13-19,24)

167

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

168

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

Fig. 5. Post hatching developmental stages of Sandelia capensis. All measurements are for live fish, (a) In situ dorsal view
of a 3,1 mm TL free embryo, (b) Free embryo, age 53, 5h, 4,1 (4,0) mm TL (NL). (c) Free embryo, 4,1 (4,0) mm TL

(NL). h-^ead; o-otolith.

169

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

Fig. 6. Post hatching developmental stages of Sandelia capensis. All measurements are for live fish, (a) Free embryo, 4,7
(4,6) mm TL (NL). (b) Larval fish, 5,0 (4,9) mm TL (NL). (c) Larval fish 5,2 (5,0) mm TL (NL). og-oil globule;

sb-swimbladder.

170

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

Fig. 7. Post hatching developmental stages of Sandelia capensis. All measurements are for live fish, (a) Larval fish 5,8
5,7) mm TL (NL). (b) Flexion, larval fish 7,3 (7,0) mm TL (NL). cfr-caudal fin ray; f-flexion; og-oil globule;

sb-swimbladder.

171

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

Fig. 8. Post hatching developmental stages of Sandelia capensis. All measurements are for live fish, (a) Transitory caudal
fin lobe, larval fish, age 21,2 days (23 °C), 7,7 (6,2) mm TL (SL). (b) Late larval fish, 8,8 (6,9) mm TL (SL). tl-transitory

lobe of the caudal fin.

172

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

dorsal fin rays commenced branching at 8,6 mm SL (P). The pelvic buds formed at c 6, 9-7, 2
mm SL (P) and the first pelvic rays were seen in a 9,2 mm SL (P) specimen. The entire finfold
was absorbed at 11,74 mm SL (P). Squamation began between 9,0-9,5 mm SL (P).

OIL GLOBULE DESCRIPTION

Each early free embryo had only one oil globule (Figs 4a and 9a). This oil globule and its
position (Fig. 4b) are probably the reasons why free embryos lie with their ventral surfaces
positioned upwards (Fig. 5a). The bluish oil globule begins to move posteriorly (Fig. 6a) and
then dorsally at c 4,7 mm NL (P) (Figs 9b and c) and by 4,95 mm NL (P) there are distinct left
and right oil globules (Fig. 9d). This movement of the oil globule coincides with the inflation of
the swimbladder (Fig. 6b). In some specimens it was observed that the oil globule on one side
divided into an upper and lower globule (Fig. 9e). There appeared to be variability in the size
of the oil globules and the right side is usually the larger of the two (Fig. 9d). Oil globules
disappeared in specimens between 6,3 and 6,9 mm NL (P), a stage at which the swimbladder was
well developed.

SWIMBLADDER DEVELOPMENT

The swimbladder first inflated at c 4,8 mm NL (P) (Fig. 6b). At 6,25 mm NL (P) (estimated
6,3 mm SL) the swimbladder started to project posteriorly and at c 8,7 mm SL (P) (Fig. 10a) the
swimbladder had completed its extension over the gut and now commenced to extend
posteriorly under and parallel to the vertebral column. This posterior elongation of the
swimbladder is completed at c 14,4 mm SL (Fig. 10b) when the swimbladder touches the hypural
plates. The development of the swimbladder for several specimens of different ages can be seen
in Fig. 11. The relationship of this posterior prolongation of the swimbladder relative to
notochord or standard length is shown in Fig. 12.

PIGMENTATION

Pigmentation commences on the embryonic yolksac (Fig. 2c). The pigmentation has the
form of irregular spots which occur randomly over the yolksac surface and on the developing
embryo (Fig. 2d). Pigment on the newly hatched free embryo occurs on the yolksac and body
except for the posterior tip of the body, 0,8 mm from the posterior tip of the caudal finfold
(Fig. 4a). The pigment is very variable in size and shape. This pigment pattern is also visible on
a free embryo of 4,1 mm TL (Fig. 5c).

At 4,7 mm TL the fish still have a clear unpigmented posterior tip to their bodies. The eye
is heavily pigmented and the rest of the body is covered with black pigment spots, but no pattern
is obvious (Fig. 6a).

At 5,2 mm TL the dorsal half of the body is yellow and some black pigment now occurs on
the posterior tip of the body. There is also pigment on the dorsal and ventral finfolds. There is
still no distinct pattern, however, pigment spots are now relatively smaller and darker, especially
noticeable on the dorsum of the head (Figs 6b and 6c).

At 7,7 mm TL the pigment spots are finer and the entire body is peppered with these
pigment spots (Fig. 8a). On the dorsal fin the first large dark pigment spot is evident at the
posterior of the dorsal fin and the rays in the dorsal fin are outlined with pigment at 10,0 mm TL
(7,9 mm SL). The finer peppering continues until at c 12 mm TL the dorsal to ventral body

173

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

Fig. 9. Oil globule movement during the development of S. capensis. All measurements refer to preserved material, (a)
Position of the single oil globule in a recently hatched free embryo, 4,75 (4,6) mm TL (NL). (b) Movement of the oil
globule dorsally, left and right views of a 4,9 (4,7) mm TL (NL) free embryo, (c) Dorsal and ventral views of the migrating
oil globule of a 4,9 (4,7) mm TL (NL) free embryo, (d) Two separate asymmetrical oil globules in a 5,1 (4,95) mm TL
(NL) individual, (e) The oil globule on the left has dispersed into two separate globules in a 6,6 (6,4) mm TL (NL) fish,
og = oil globule, sb = swimbladder, arrows in (b) indicate movement of the oil globule.

174

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

Fig. 10. (a) Close-up of the beginning of the posterior prolongation of the swimbladder parallel to the vertebral column
in a 11,0 (8,78) mm TL (SL) S. capensis. (b) Close-up of the maximum posterior prolongation of the swimbladder in a 18,2
(14,4) mm TL (SL) S. capensis. All measurements are from live material. Arrows indicate the posterior tip of the posterior

prolongation.

175

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

stripes begin to appear which are characteristic of the juvenile fish as shown in Fig. 11 for a
13,8 mm TL (10,8 mm SL) specimen.

EARLY FREE EMBRYO AND LARVAL BEHAVIOUR

After hatching some of the free embryos remained attached for up to 10 h to the same rocks
to which the eggs had adhered (Fig. 5a). Other free embryos were active and swam upwards to
the surface and then sank passively. After this period when the rocks were lifted the embryos
would come off and rapidly swim away. Three days post-hatch, at a size of c 5 mm TL, the
embryos had moved away from the rocks (substratum) they had been attached to since hatching.
Some of the free embryos hid between the gravel on the substratum of the nursery tanks, others
were seen against the side of the tank hanging vertically with the dorsum of their heads in contact
with the glass. Some fish swam ventral side up and only stopped when their heads touched an
object in the tank. Others swam to the surface and ‘butted’ at the surface until they attached
themselves to the surface tension. As soon as the surface tension was broken the young fish sank
in a spiralling fashion. After four days from hatching the young fish started to feed on brine
shrimp. The larval fish remain motionless, then take short (2-5 cm) dashes. Once they were near

Fig. 11. Posterior prolongation of the swimbladders in several different age groups of S. capensis, from the top, 13,8
(10, 8), 9,7 (7,3), 8,8 (6,9) mm TL (SL). All measurements are from live material. Arrows indicate the posterior tip of

the posterior prolongation.

176

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

% SL

^ -

(35

<»

(35

Swi mb ladder inflates

•• ••

• •

- :

Fig. 12, The relationship of the distance from the tip of the posterior prolongation of the swimbladder to the tip of the
notochord or the posterior edge of the hypural plates to notochord or standard length during the development of the

swimbladder in S. capensis.

Ill

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

a prey item they manoeuvred within a few millimetres, maintained position, observed the prey,
then darted forward and if successful swallowed the item. At this stage the larval fish have a
swimbladder and can maintain themselves parallel to the surface of the water, and swim and
behave in a manner similar to that of adults.

The arcing behaviour, as defined in Tooker and Millar (1980), which is characteristic of
S. capensis adults (pers. obs.), was first observed in 5,0 mm NL fish, when the swimbladder had
just inflated and the oil globule was moving from its ventral position to a position flanking the
swimbladder (Fig. 9d). They arc (lateral bending) their caudal area and maintain a position
while sculling with the large pectoral fins. When live prey items were given to the larval fish they
swam within a few millimetres of the item, arc, observed and then dashed forward and tried to
swallow the prey item. In many cases when the prey did not move the young fish ignored the item
and swam away. The adult male was still guarding the young at this stage. After approximately
6,7 mm TL the larval fish were mainly on or near the substratum. Aggression was first noted
when the fish were 15,0 mm TL. They turned a darker colour and chased siblings, in some cases
killing them. The 15,0 mm TL specimen was able to kill fish of 10,9 and 13,8 mm TL.

FIELD OBSERVATIONS

Field observations of a population of 5. capensis in the Wit River showed that fish were
actively breeding during the middle of November 1988 when water temperatures were around
20-23 °C. Several small (c 15 cm TL) males, in breeding colours, were chasing females in small
circles, however, no actual spawning was observed. The darkened males readily chased away
any other sexually active males and, in this case, the chase was usually in a straight line until the
chased fish found cover. Chase distances were usually less than three metres.

Diving observations revealed that the larger males, some as big as 20 cm TL, were
occupying territories between tree roots where the river had eroded the bank. The tree roots as
well as the open gravel area occupied by the smaller males were examined for eggs but no eggs
were found. The following month (19-xii-1989) young-of-the-year S. capensis were collected
amongst the submerged roots of Bushman’s rhubarb {Gunnera perpensa). These fish were
between 7,8 (7,4) and 13,48 (10,8) mm TL (SL) (n=17) (P). The smallest was a larval fish and
flexion had just commenced. Minimum-maximum water temperatures between the two
observational periods were 16,5-20,5 °C. These temperatures were recorded at a water depth of
80 cm with a min. -max. thermometer which was left in the water between monthly sampling
periods. At the next sampling period (22-i-1989) young-of-the-year S. capensis in this area
ranged in size from 8,2 (7,0)-15,44 (12,22) mm TL(SL) (n=19) (P). The range in size and
developmental stages would indicate that several separate spawnings had taken place during the
summer months. The largest fish (15,44 mm TL) still had a remnant of a preanal finfold.

DISCUSSION

EARLY LIFE HISTORY

Very little work has been done on the early life history of S. capensis and other African
Anabantidae (Cambray and Teugels, 1988). So there are very few data for comparative
purposes. Siegfried (1963) reported that Mr. G. Reinhardt observed that ‘eggs’ (= free
embryos) of S. capensis hatch after approximately 35 hours at a temperature of 24 °C and that
the young are free swimming 4 days after hatching. In the present study the exact time of
activation was not established. However, from an early multi-celled blastodermal cap stage to

178

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

hatching took 40,5 h at 22-23 °C. During the summer months (November, December and
January) Barnard (1943) collected and examined ‘juveniles’ (= larval fish) of S. capensis from
8 mm upwards. Barnard (1943) included drawings of 8 and 10 mm larval S. capensis.

Mayekiso (1986) found that the ova of S. bainsii are released when their wet diameter is
between 0,99 and 1,24 mm. For S. capensis the diameter at release is between 1,0 and 1,4 mm.

Since the eggs of S. capensis were highly adhesive the oil globule may have acted as a
buoyancy organ and possibly had very little of a functional role to play until the embryo hatched
and had mobility. The adhesive organ on the dorsum of the head functioned to keep the embryo
in one place. If disturbed, the free embryo could move to another site and the possible buoyancy
of the oil globule would be of importance at that stage. During the development of S. capensis
the oil globule migration and its subsequent division would indicate that it was functioning as a
temporary swimbladder. The movement and division of the oil globule occurred before
swimbladder inflation. Initially the free embyros were positioned with their ventral surfaces
upwards. The oil globule migrated (initially posteriorly then dorsally) and then divided. The two
oil globules then lay in a position where the swimbladder was forming. At this stage the
orientation of the free embryo changed to dorsal surface upwards which was followed by
swimbladder inflation.

The development of the posterior prolongations of the swimbladder in S. capensis as
described here was similar to that described for Ctenopoma muriei (Morike, 1977: Figs 10 and
11). The posterior prolongations extended as far posteriorly as the caudal skeleton (Fig. 10b).

The pigmentation of the early life history stages of S. capensis was unlike that of the
co-occurring cyprinid minnow species, which characteristically have stellate melanophores (pers.
obs.). Another very good character which could be used to separate S. capensis eggs and early
free embryonic stages from cyprinids was the presence of the oil globule in the Cape kurpers. In
addition the small size and adhesive egg envelope were also good characters to separate these
eggs. The heavy pigmentation on the yolk-sac and on the developing embryo would also be
characters to separate this species from other co-occurring species. Free embryos and larval fish
could easily be separated from other co-occurring fish species by the shape of the swimbladder,
large eyes and the distinctive caudal fin shape with its transitory dorsal lobe.

BREEDING BEHAVIOUR

Harrison and du Plessis (1947) were the first to describe the breeding behaviour of
S. capensis. Their observations were made on fish breeding in a reservoir. The darkly coloured
males were ‘hostile to each other’ and chased others from their territory. The nuptial embrace
was observed and described as relatively simple and was accompanied by a whirling movement.
Harrison and du Plessis (1947) observed one S. capensis guarding a small stump of dead wood
and roots which had yellowish eggs adhering to them. All the nests were at a depth of 61-76 cm.
The authors carefully examined one nest. They did not find any definite construction and
observed that the eggs were distributed haphazardly over an area of approximately 0,092 m^
(1 ft^). Siegfried (1963) noted that the spawning bed measured approximately 30 cm in diameter,
which agrees with the present aquarium study, and that the male defended an area with a radius
of about 50 cm from the centre of the spawning bed. Harrison and du Plessis (1947) found eggs
adhering to stones, dead twigs, pieces of wood, roots and weeds, which would indicate that they
had not been placed. In the present study the same random pattern of egg laying was also
observed and most of the eggs were attached to aquarium gravel. The male alone guards the nest
(Siegfried, 1963, and the present study). The nest area did not appear to be cleaned in any way

179

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

although there was a shallow depression in the aquarium gravel possibly caused during the
spawning act. The eggs, however, were not only found in the depression but also outside this
area. This is unlike some of the guarding Ctenopoma, such as C. intermedium, which have
bubble nests and place the eggs in the nest (pers. obs.). Using the reproductive guilds as put
forward by Balon (1975, 1985), 5. capensis belongs in the guarder (B) ethological section, the
ecological group is that of the substratum choosers (B. 1), and it is a rock or plant tender and
therefore in the reproductive guild B. 1.3 (lithophils) or B. 1.4 (phytophils) (Balon, 1985).

Harrison (1952) recorded small (6,4 mm) S. capensis from the Berg River in December
1933. In early summer in the Hex River, ‘similar’ young Sandelia were observed (Harrison,
1952). On December 5 1933, Harrison (1952) collected small S. capensis (12,7-25,4 mm long
(0,5-l,0 in.)) and suggested that they had been spawned from July to September. From this data
set Harrison (1952) speculated that S. capensis, may spawn in rivers of the western area in the
spring after the floods abate whereas the species in seasonal vleis may spawn earlier when the
winter rains restore suitable water conditions. In De Hoop Lake, Siegfried (1963) studied the
reproductive cycle of S. capensis, and noted that the females underwent ‘incomplete’ spawning
(= serial) at repeated intervals over a prolonged period throughout the months of spring and
summer, with two spawning peaks, one during mid-spring and one during mid-summer.
Siegfried (1963) also suggested that older fish bred earlier in the season than first-year
individuals. In the Wit River it has been recorded that S. capensis breed throughout the summer
months (this study). The aquarium study verified the serial breeding habit of the Cape kurper.
S. bainsii have been recorded to spawn repeatedly during the summer breeding season
(Mayekiso, 1986).

Field observations of a population of S. capensis in the Wit River showed that fish were
actively breeding during the middle of November 1988 with water temperatures around
20-23 °C. Several small (c 15 cm TL) males, in breeding colours, were chasing females in small
circles, however, no actual spawning was observed. The darkened males readily chased away
any other sexually active males. Diving revealed that the larger males, some as big as 20 cm TL,
were occupying territories between tree roots where the river had eroded the bank. The tree
roots were examined for eggs, as was the open gravel area of the smaller males, but no eggs were
found. The following month young-of-the-year S. capensis were collected amongst the
submerged roots of Gunnera perpensa. These fish were between 7,4-10,8 mm SL.

HABITAT

Sandelia capensis is confined to the South Coastal Drainage Basins in the Cape Province.
This species can tolerate a wide variety of both physical and chemical water conditions
(Harrison, 1952). In a recent study Scott and Hamman (1988) suggested that in De Hoop Vlei,
a southern Cape Coastal Lake, the major limiting factor for the survival of S. capensis was
salinity. High salinity levels of up to 607oo eliminated S. capensis from certain sites in De Hoop
Vlei and when salinity levels were lowered recruitment took place from the freshwater springs
in the northern half of the vlei (Scott and Hamman, 1988). Hofmeyr (1966) determined that
S. capensis is fairly tolerant of sodium chloride and has a median tolerance limit of 10 000
minutes in a 15,6 7oo sodium chloride solution. However, he suggested that S. capensis could
not tolerate undiluted sea water. No work has been done on the salinity tolerance of the early
developmental stages of this species. Since the early life stages have the least mobility their
tolerance of salinity needs to be established.

In the Wit River all the larval and juvenile S. capensis were collected from under mats of

180

CAMBRAY: EARLY ONTOGENY OF THE CAPE KURPER SANDELIA CAPENSIS

vegetation. The adults were collected from this habitat but also from more open areas devoid of
aquatic vegetation.

Although S. capensis does not have a well-developed labyrinthine organ (Liem, 1963;
Peters, 1978), as do the species of Ctenopoma, it is a fairly hardy species (Harrison, 1952).
Cambray (1978) suggested that the reduction in the supra-branchial organ in S. capensis was
accompanied by an increase in the gill respiratory area. S. capensis also lacks the well-developed
series of spines on the edge of the gill-cover which allows more tropical forms such as
C. multispinis to migrate over land when temporary pools dry out. It is therefore suggested that
S. capensis are not as well adapted to low oxygen conditions as are many of the Ctenopoma
species. The lack of adaptation to low oxygen levels was also evident in the early life history of
the Cape kurper. The S. capensis eggs spawned on the substratum would need adequate oxygen
to survive. In comparison some of the Ctenopoma species have overcome low oxygen substratum
areas by using bubble nests to keep their eggs and free embryos at the air/water interface.

CONSERVATION

At several localities known populations of S. capensis have been exterminated by the
introduced exotic predator, black bass {Micropterus salmoides and M. dolomieui) . Harrison
(1952) reported that the once abundant S. capensis population in Paarde Vlei was exterminated
by largemouth bass. Harrison (1952) also reported the drastic reduction of S. capensis in the
Berg River by the introduced smallmouth bass. In the larger pools of the Baviaanskloof River,
a tributary of the Gamtoos River system, S. capensis have been exterminated by black bass, and
the only remaining populations of Cape kurpers were found in the shallow upper reaches of this
system (pers. obs.). The guarding S. capensis males would be very easily preyed upon by a large
predator as the Cape kurpers try to defend their nests.

Adult S. capensis are known to be lurking predators (Bruton et al., 1982) which feed on
aquatic insects, crustaceans (Siegfried, 1963) and small fish. The lurking behaviour was seen in
fish as small as 5 mm NL when they fed on rotifers. In the Wit River population cannibalism has
been found.

The early life history stages are known to be the most critical stages in the development of
a fish. The early stages of the Cape kurper are guarded by the parental male who chases all other
fish including the parental female away from the nest area. Later in their development the larval
fish require cover. In the Wit River a population of S. capensis was seen to increase after the area
became part of the Cockscomb Nature Reserve. Goats and cattle were removed from the area.
These animals had previously grazed the aquatic vegetation thereby removing the habitat for
larval and juvenile Cape kurpers. Man induced changes in the physical and chemical conditions
of the body of water, the lack of suitable cover for larval fish, and the introduction of large exotic
predaceous fish species are probably the main threats to populations of this species.

ACKNOWLEDGEMENTS

The present paper is contribution number one in a series of publications, Baviaanskloof
aquatic studies. Support for these studies is being provided by the Albany Museum and by a
FRD research grant.

The author would like to thank the Director of the Albany Museum who allowed him to
conduct this research. He is grateful to Eve Cambray who read and commented on several of
the drafts.

181

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 8, APRIL 1990

REFERENCES

Balon, E. K. 1975. Reproductive guilds of fishes: a proposal and definition. Fish. Res. Board Can. 32: 821-864.

Balon, E. K. 1985. Additions and amendments to the classification of reproductive styles in fishes. In: Balon, E. K. ed.,
Early life histories of fishes. Junk, Dordrecht, pp. 59-72.

Barnard, K. H. 1943. Revision of the indigenous freshwater fishes of the S. W. Cape region. Ann. S. Afr. Mns. 36:
101-262.

Berry, F. H. and W. J. Richards 1973. Characters useful to the study of larval fishes. In: Pacheco, A. L. ed.. Proceedings
of a workshop on egg, larval, and juvenile stages of fish in Atlantic Coast estuaries. Tech. Publ. No. 1, NMFS
Mid. Atl. Coast Fish. Cent., Highlands, N. J. , pp. 48-65.

Bruton, M. N. , P. B. N. Jackson and P. H. Skelton 1982. Pocket guide to the freshwater fishes of southern Africa. Cape
Town: Centaur.

Cambray, j. a. 1978. A contribution to the character phytogeny of the Anabantidae: with particular reference to the
respiratory organs of Sandelia bainsii. Unpublished hons project, Rhodes University, Grahamstown.

Cambray, J. A. and G. G. Teugels 1988. Selected annotated bibliography of early developmental studies of African
freshwater fishes. Ann. Cape Prov. Mus. (not. Hist.) 18 (2): 31-56.

Fuiman, L. a. 1982. Correspondence of myomeres and vertebrae and their natural variability during the first year of life
in yellow perch. In: Bryan, C. F. , J. V. Conner, and F. M. Truesdale eds. The Fifth Ann. Larval Fish Conf,
Louis. Coop. Fish. Res. Univ., LSU, Baton Rouge, LA, pp. 56-59.

Gaigher, I. G. , K. C. D. Hamman and S. C. Thorne 1980. The distribution, conservation status and factors affecting the
survival of indigenous freshwater fishes in the Cape Province, Koedoe 23: 57-88.

Hamman, K. C. D. , S. C. Thorne and P. H. Skelton 1984. The discovery of the Cape kurper, Sandelia capensis (Cuvier
in C. & V. 1831) in the Olifants river system (Western Cape Province). The Naturalist 28 (1): 24-26.

Harrison, A. C. 1952. The Cape kurper (with notes on the rockey of the eastern province). Piscator No. 23: 82-91.

Harrison, A. C. and S. S. Du Plessis 1947. Notes on the spawning of the Cape kurper {Sandelia capensis). Melville Dam,
Oudtshoorn, November 16, 1946. Investigational Report No. 4, Cape Dept of Nature Conservation, Cape
Town, pp. 14-16.

Hofmeyr, H. P. 1966. The salinity tolerance of some Eastern Province fish in relation to their known distribution.
Unpublished M.Sc. thesis, Rhodes University, Grahamstown.

JuBB, R. A. 1965. Freshwater fishes of the Cape Province. Ann. Cape Prov. Mus. 4:1-72.

JuBB, R. A. 1967. The freshwater fishes of Southern Africa. Cape Town: Balkema.

Liem, K. F. 1963. The comparative osteology and phylogeny of the Anabantoidei (Teleostei, Pisces). Illinois Biol.
Monographs, no. 30, Urbana.

Mayekiso, M. 1986. Some aspects of the ecology of the Eastern Cape Rocky (Pisces: Anabantidae) in the Tyume River,
Eastern Cape, South Africa. Unpublished M.Sc. thesis, Rhodes University, Grahamstown.

Morike, D. 1977. Vergleichende Untersuchungen zur Ethologie zweier Labyrinthfischarten, Ctenopoma muriei
(Boulenger 1906) und Ctenopoma damasi (Poll 1939) (Anabantoidea, Pisces). Unpublished Ph.D. thesis.
University of Tubingen, Tubingen.

Peters, H. M. 1978. On the mechanism of air ventilation in Anabantoids (Pisces: Teleostei). Zoomorphologie 89:93-123.

Scott, H. A. and K. C. D. Hamman 1988. Recent fish records from De Hoop Vlei, a southern Cape coastal lake. Bontebok
6: 30-33.

Siegfried, W. R. 1963. Observations on the reproduction and feeding of the Cape kurper Sandelia capensis (C. and V.)

in De Hoop Lake, Bredasdorp. Investigational Report No. 3, Cape Dept of Nature Conservation, Cape Town.

Skelton, P. H. 1986. Distribution patterns and biogeography of non-tropical southern African freshwater fishes.
Palaeoecology of Africa 17: 211-230.

Tooker, C. P. and R. J. Millar 1980. The ontogeny of agonistic behaviour in the blue gourami, Trichogaster trichopterus
(Pisces, Anabantoidei). Anim. Behav. 28: 973-988.

182

INSTRUCTIONS TO AUTHORS

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

Only generic and specific names should be underlined to indicate italics. All other marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

BK1041 — CTP Book Printers, Cape

^15

ISSN 0570-1880

Annals of th^

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 9 28th May 1990

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS; 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

A fourth contribution to the knowledge of the ethology of the genus
Ceramius Latreille (Hymenoptera: Vespoidea: Masaridae) in southern

Africa.

F. W. GESS and S. K. GESS*

(Albany Museum, Grahamstown)

*(The order of names is alphabetical and joint authorship should be understood. The
same applies to previous papers on the ethology of Hymenoptera by the same authors.)

ABSTRACT

Ethological accounts including nest structure and flower visiting records are given for four
species of Ceramius Latreille, C. clypeatus Richards, C. micheneri Gess, C. toriger Schulthess
and C. braunsi Turner. In addition, first flower visiting records are given for C. coffer Saussure
and C. metanotalis Richards. Notes are given on usurpation of nests of C. braunsi by Megachile
(Eutricharaea) aliceae Cockerell (Megachilidae) and the association with this bee of the parasitic
bee Coelioxys (Coelioxys) recusata Schulz (Megachilidae).

In the discussion the ethological data are used to clarify the species grouping within the
genus Ceramius.

INTRODUCTION

The present paper is the fourth in a series of publications (Gess and Gess, 1980, 1986 and
1988a) dealing with the ethology of southern African species of the genus Ceramius Latreille
(Hymenoptera: Masaridae).

The genus Ceramius is represented in southern Africa by 19 species belonging to six of the
eight species groups suggested by Richards (1962) and revised by Gess and Gess (1986 and
1988a).

It would seem to be desirable to have made ethological studies of all the species and then
to have presented these as a study of the ethology of the southern African species as a whole.
However, due to lack of certainty associated with finding many of the species, let alone their
nests, it has been the authors’ policy to publish ethological information as it becomes available
and there by to gradually piece together an understanding of the genus and the species groupings
within the genus.

In the previous papers accounts were given of the nests of nine species belonging to five
groups — Group 2a: C. cerceriformis Saussure; Group 3: C. nigripennis Saussure and C. jacoti

183

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

Richards; Group 5: C. lichtensteinii (Klug); Group 6: C. rex Saussure; Group 8: C. bicolor
(Thunberg), C. linearis Klug, C. capicola Brauns and C. socius Turner. In the present paper
accounts are given for a further four species — Group 2b: C. clypeatus Richards; Group
uncertain: C. micheneri Gess; Group 3: C. toriger Schulthess and C. braunsi Turner.

As Ceramius provisions its young with pollen and nectar flower visiting records are of
importance. Gess and Gess (1989) presented flower visiting records for 14 species. In the present
paper records are given for a further two species, C. caffer Saussure and C. metanotalis Richards
and additional records are given for some species for which previous records were extremely
scant.

The paper is structured on the species groupings and the data presented are used to discuss
the characterization of the species groups.

ETHOLOGICAL ACCOUNTS

Group 2b

Group 2b is constituted of Ceramius clypeatus Richards and C. richardsi Gess.

Geographic distribution

Both species appear to be restricted to the area lying between the Olifants River Mountains
in the south and Namaqualand in the north.

Ethology

An account of flower visiting by C. clypeatus in the Clanwilliam District was presented by
Gess and Gess (1988a and 1989). Further observations on flower visiting and an investigation of
nesting were made in the Clanwilliam District during the period 16-20. x. 1989.

Nothing is known of the nesting or flower visiting behaviour of C. richardsi.

Description of nesting area

Two nesting areas of C. clypeatus were located in the Clanwilliam district, a brief account
of the soils and vegetation of which was given in Gess and Gess (1988a). Both areas are situated
above and to the east of the Clanwilliam Dam.

One is a levelled sparsely vegetated stoney area in the grounds of the Clanwilliam Dam
resort. The soil is compacted sand with sufficient clay to make it malleable when mixed with
water. At the time of the study water was available from a roadside trickle. Most of the nests
were partially concealed either amongst stones or under bushes. The nests were aggregated in
small groups, for example five nests were grouped within a radius of 15 cm.

The other area is a sparsely vegetated slope above the old Olifants River Valley Road above
Caleta Cove (3Z14'20"S, 18’55'45"E), on the east bank of the Clanwilliam Dam (Fig. 1). The soil
is extremely hard compacted sand with sufficient clay to make it malleable when mixed with
water. Water was available from a roadside pool (Fig. 7). The nests were not concealed nor
grouped but occurred singly in bare areas between bushes.

Plants visited

Gess and Gess (1988a and 1989) established that C. clypeatus is a common visitor to the
flowers of Aspalathus spinescens Thunb. subsp. lepida (E. Mey) Dahlgren (Leguminosae:
Papilionatae) in the Clanwilliam district.

184

GESS & GESS; ETHOLOGY OF CERAMIUS LATREILLE

Fig. 1. Old Olifants River Valley Road to the east of the Clanwilliam Dam; nesting area of Ceramiits clypeatus Richards,
C. micheneri Gess and C. braunsi Turner in vicinity of bare patch in middle distance, above Caleta Cove.

Further investigation in October 1989 confirmed this association and in addition revealed
that C. clypeatus is also a common visitor to flowers oi Aspalathus linearis (Burm. /.) Dahlgren
(Rooibos Tea). A sample (5 $ $ , F. W. and S. K. Gess, 16.x. 1989) was taken in the grounds of
the Rooibos Co-op, Clanwilliam. Above Caleta Cove C. clypeatus was found to be visiting a
third species of Aspalathus, A. pulicifolia Dahlgren. A. pulicifolia was growing in an area of hard
compacted sandy soil not favoured by A. spinescens or A. linearis which seem to favour looser
sandy soil. A sample was taken (2 9 $ and 2 d d, F. W. and S. K. Gess; 1 9 , D. W. Gess; all
19.x. 1989).

It is of note that all three forage plants are Aspalathus species. All other plants in flower
were sampled for insect visitors. None was found to be visited by this wasp.

The posture of C. clypeatus on all three species of Aspalathus is constant and is as described
by Gess and Gess (1989: 103 and figs 5, 6 and 7) for this wasp on A. spinescens subsp. lepida.
That is briefly, the wasp, when alighting on one of the small pea flowers, grasps the alae with the
second and third pairs of legs and curves the abdomen down beneath the flower aiding its
balance whilst it imbibes nectar from the base of the standard or consumes pollen directly from
the anthers.

185

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

Provision

The cell provision of C. clypeatus is of the typical Ceramius type being a firm relatively dry
nectar and pollen loaf (Fig. 2).

Samples of pollen were taken from the crop of a female and from the provision from four
cells from the nests of four further females from the Clanwilliam Dam resort and from the
provision from one cell from a nest from Caleta Cove. In each sample the pollen was found to
be all of one kind. Pollen from all the samples was of the “Aspalathus type” (triangular sided
and with each side 25 p) matching that of Aspalathus spinescens subsp. lepida from the
Clanwilliam Dam and that of Aspalathus pulicifolia from Caleta Cove.

Water collection

Females of C. clypeatus were observed collecting water from a roadside trickle at the
Clanwilliam Dam resort and from a small pool above Caleta Cove. In all cases water was being
imbibed on the wet soil at the edge of the water source. No females were seen to alight on the
water surface.

Male behaviour

Male C. clypeatus were not seen at water or in the nesting areas. They were, however,
observed visiting flowers co-incident in time with the females and several instances of a male
mounting a female were observed.

Description of the nest

The nest consists of a multi-cellular subterranean burrow surmounted by a curved tubular
mud turret (Figs 3 and 4a). The turret is constructed of mud pellets roughly smoothed on the
outside and well smoothed on the inside. Some interstices are left open distally. The turret and
shaft opening are of equal diameters. The main shaft descends sub-vertically and for the greater
part of its length is of the same diameter as the entrance. Near the lower end of the shaft there
is a short wider section forming a “bulb” below which the shaft continues with a diameter
equalling that of the upper section of the shaft. The main shaft at its base curves outwards to
form a short lateral shaft terminating in a cell which lies sub-horizontally. Within an
excavated-cell there is a constructed mud-cell sealed at the neck with a mud-plug. The section

Fig. 2. Provision, firm nectar and pollen loaf, from a cell of
Ceramius clypeatus Richards (x3).

Fig. 3. Mud turret surmounting burrow of Ceramius
clypeatus Richards (xl,5).

186

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

Fig, 4. Vertical plans of turrets and underground workings of nests of Ceramius clypeatus Richards (a) and C. micheneri

Gess (b) (X 1).

187

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

of the secondary shaft between the cell and the main shaft is filled with soil and is sealed off from
the main shaft. Successive cells lie to one side of the shaft in a group.

Method of construction of the nest, oviposition and provisioning

Water is required for nest construction. The turret is constructed at an early stage in shaft
sinking. At the commencement of turret construction pellets are laid down around the shaft
opening in such a way that the turret will have the same inner diameter as the shaft. The walls
of the turret are approximately 1 mm thick. Pellets are added to the turret in such a way that it
soon curves over. After turret construction is completed further pellets extracted in shaft sinking
are discarded at some distance from the nest.

After the construction of the first cell is completed oviposition takes place and is followed
by provisioning. The provision, a mixture of pollen and nectar, is in the form of a relatively dry
firm loaf which only partially fills the cell.

The completed provisioned cell is sealed with a mud-plug constructed just within the mouth
of the cell and having its outer face concave. The sub-horizontal shaft is then firmly packed with
soil until the sub-vertical shaft is reached, when it is sealed off with a mud plate.

Further cells terminate secondary shafts and are prepared in a similar fashion to the first.

A sample of twelve nests was investigated. All were new nests, that is in no instance had a
wasp reused her maternal nest. Seven nests were single-celled, one was four-celled and the
remaining four were nests which had not yet reached the stage of cell excavation. Nest
measurements are given in Table 1.

Table 1.

Measurements of nests of Ceramius clypeatus Richards.

Range

(mm)

Average

(mm)

Sample size

Height of turret

9-13

10,8

Diameter of shaft

4.5-4,5

4,5

Diameter of bulb

7-9,5

8,3

Diameter of excavated-cell

8-9

8,5

Length of shaft above bulb

42-68

56,3

Length of bulb

10-15

12,5

Total length of vertical shaft

67-85

78,3

Length of excavated-cell

20-21 (approximate)

Group: uncertain

Placement of Ceramius micheneri Gess into any particular species group is uncertain. The
affinities of the species, judged purely from consideration of morphological characters, were
stated by Gess (1968) to be with the species of the Ceramioides group of species (C. cerceriformis
Saussure, C. peringueyi Brauns, C. clypeatus Richards and C. richardsi Gess) and possibly with
the group of four species comprising C. nigripennis Saussure, C. toriger Schulthess, C. braunsi
Turner and C. jacoti Richards— that is, with groups 2 and 3 of Richards (1962) and Gess and Gess
(1986 and 1988a). For some reason, at present no longer clear, Gess and Gess (1986 and 1988a)
tentatively assigned C. micheneri to Group 3 rather than Group 2. On ethological grounds this
assignation is now recognised as having been incorrect (see Discussion).

188

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

Geographic distribution

C. micheneri is known only from the Olifants River Valley between Citrusdal and Klawer.
It has been recorded previously from Citrusdal and from Olifants River between Clanwilliam
and Klawer (Gess, 1968 and 1973).

The present observations were made in the Clanwilliam District above Caleta Cove (Fig. 1)
during the period 16-20. x. 1989.

Ethology

Nothing was previously known of the flower visiting and nesting behaviour of C. micheneri.
The present account indicates in both flower association and nest structure affinities with Group
2b.

Description of nesting area

Three nests were located on a sparsely vegetated slope above the old Olifants River Valley
Road above Caleta Cove (Fig. 1). The soil is extremely hard compacted sand with sufficient clay
to make it malleable when mixed with water. Water was available from a roadside pool (Fig. 7).
The nests were not concealed nor grouped but occurred singly in bare areas between bushes.

Plants visited

C. micheneri has not previously been recorded visiting flowers. It is here recorded as visiting
the small yellow pea flowers of Aspalathus pulicifolia (Leguminosae: Papilionatae) at Caleta
Cove. A sample was taken (2 ? 9 and 4d d , F. W. Gess and S. K. Gess, and 2 9 9 , D. W. Gess,
all 19-20.X.1989).

All other plants in flower were sampled for insect visitors. None was found to be visited by
this wasp.

Provision

Provision from a nest of C. micheneri investigated at Caleta Cove was examined. It was in
the form of a relatively dry nectar and pollen loaf. The pollen was found to be all of one kind
and to match that of A. pulicifolia.

Water collection

Females and males of C. micheneri were observed on wet soil at the edge of a pool at Caleta
Cove. None was seen to alight on the water surface.

Male behaviour

Male C. micheneri were observed in company with females at water and visiting the forage
plant, however, no instances of pairing were observed. A male was found together with a female
in her nest.

Description of the nest

The three nests investigated all consisted of a subterranean burrow surmounted by a low
vertical cylindrical mud turret (Fig. 4b) constructed from mud pellets cemented together and
smoothed on the inside. The inner diameter equalled that of the descending sub-vertical main
shaft.

One nest was in an early stage of construction and lacked a cell. In the other two nests the
main shaft at its base curved outwards to terminate in an ovoid cell lying sub-horizontally. At
some little distance above the base of the vertical shaft it was widened to form a “bulb”.

189

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

Within each excavated-cell had been constructed a mud-cell with walls of somewhat less
than 1 mm in thickness and smoothed on the inside.

Nest measurements are given in Table 2.

Table 2.

Measurements of nests of Ceramius micheneri Gess.

Range

(mm)

Average

(mm)

Sample size

Height of turret

3,7

Diameter of shaft

3,5^, 0

3,7

Diameter of bulb

5-6

5,5

Diameter of excavated-cell

5-6

5,5

Length of shaft above bulb

75-85

80,0

Length of bulb

12-12

12,0

Total length of vertical shaft

95-100

97,5

Length of excavated-cell

16-17

16,5

Group 3

Group 3 is constituted of four species, Ceramius braunsi Turner, C. jacoti Richards,
C nigripennis Saussure and C. toriger Schulthess.

Geographic distribution

Group 3 as a whole has a relatively wide distribution in the western and southern Cape,
however, the distribution of each of the four species is distinct.

C. braunsi has been recorded from Vanrhynsdorp lying below and to the west of the
Bokkeveldberge, from the Olifants River Valley and the western foothills of the Cederberg,
south of the Tankwa Karoo and the Hex River Mountains at Worcester and in the east from the
southern Great Karoo at Willowmore (Richards, 1962; Gess, 1965, 1968 and 1973; Gess and
Gess, 1988a and present paper).

C. jacoti Richards is also a southern species but appears to be more restricted in distribution
than C. braunsi, all records being from east of Worcester, that is from Hex River in the west to
Oudtshoorn in the Little Karoo in the east (Richards, 1962; Gess, 1965; Gess and Gess 1988a
and present paper).

C. nigripennis seems to be a Namaqualand species being particularly characteristic of the
Namaqualand Klipkoppe. It has been recorded from the Swart Doringrivier north to Nababeep,
from the Hester Malan Nature Reserve, 12 km east of Springbok in the Carolusburg and from
the mountainous area to the west and southwest of Springbok (Richards, 1962; Gess, 1965 and
1968; Gess and Gess, 1986 and 1988a; present paper).

C. toriger has been recorded from the escarpment formed by the Bokkeveldberge to the
east of the Knersvlakte, the Skuinshoogte Pass, 15 km north of Nieuwoudtville on the road to
Loeriesfontein, eastwards to Calvinia, and southwards through the Tankwa Karoo lying to the
east of the Cederberg to Karoopoort, 43 km ENE of Ceres on the road to Sutherland (Richards,
1962; Gess 1965 and 1968; present paper).

Ethology

Accounts of flower visiting by C braunsi, C. jacoti and C. nigripennis were given by Gess
and Gess (1988a and 1989; 1988a and 1989; and 1986 and 1989, respectively). A flower visiting
record for three male C. toriger was given in Gess (1968). Further flower visiting records for
C. braunsi, C. nigripennis and C. toriger are given in the present paper.

190

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

Accounts of nesting by C. jacoti and C. nigripennis were given by Gess and Gess (1986).
Accounts of nesting by C. braunsi and C. toriger are given in the present paper.

Plants visited

The only flower visiting records for C. toriger previously available were for three males “on
blue-rayed Compositae” (Gess, 1968) indicating a possible association with Compositae.

This association is supported by the observation of both males and females visiting the
flowers of Pteronia cf. divaricata (Berg.) Less, and Berkheya fruticosa (L.) Ehrh. (both
Compositae) in the Skuinshoogte Pass, 3-8. x. 1989. Samples were taken (visiting the yellow
flowers of P. divaricata, 2 ? 5 and 5 cJ c3, D. W. Gess; visiting the yellow flowers of B. fruticosa,
2 ? 9 , D. W. Gess, and 1 9 and 2 c? <3 , F. W. Gess and S. K. Gess). In addition one female was
collected foraging on Athanasia trifurcata (L.) L. in Karoopoort, 3.xii.l989 (S. K. Gess).

All other plants in flower were sampled for insect visitors. None was found to be visited by
this wasp.

Gess and Gess (1989: 102-103 and figs 3 and 4) recorded C. braunsi as visiting primarily
Compositae, Athanasia trifurcata (L.) L., Arctotis laevis Thunb. and Pentzia sp. and occasion-
ally Aspalathus spinescens subsp. lepida (Leguminosae: Papilionatae). Pollen gathering on
the capitula of Arctotis laevis was described. Visits to A. spinescens appeared to be for nectar
only.

During the present study C. braunsi was again found commonly on the capitula oi Athanasia
trifurcata and Arctotis laevis to which it was by far the most common visitor. Samples were
taken {Athanasia trifurcata, 45 9 9 and 1 c? , F. W. Gess and S. K. Gess, and 4 9 9 , D. W. Gess,
all Caleta Cove and Arctotis laevis, 2 9 9 , F. W. Gess and S. K. Gess, and 2 9 9 , D. W. Gess,
all Clanwilliam Dam resort). Although the insects visiting Aspalathus species were sampled no
visits by C. braunsi were observed. All other plants in flower were sampled for insect visitors.
None was found to be visited by this wasp.

Plant visiting records for C. nigripennis are all for members of the family Compositae:
Dimorphotheca sinuata DC., Pentzia suffruticosa (L.) Hutch, ex Merxm., Berkheya fruticosa
(L.) Ehrh. and Hirpicium alienatus (Thunb.) Druce (Gess and Gess, 1988a and 1989).

The fidelity of this species to Compositae is supported by new flower visiting records
resulting from further plant sampling.

Hester Malan Nature Reserve, Springbok, 10-1 1.x. 1989:

Pentzia suffruticosa (L.) Hutch, ex Merxm. 1 9, F. W. and S. K. Gess; 1 9, D. W. Gess
Berkheya fruticosa (L.) Ehrh. 3 9 9,1 d, F. W. and S. K. Gess; 1 c3, D. W. Gess
Hirpicium sp. 2 9 9, F. W. and S. K. Gess; 2 (?(?, D. W. Gess
Arctotheca calendula (L.) Levyns 1 <3, D. W. Gess

Nababeep, 12-13. x. 1989:

Berkheya fruticosa (L.) Ehrh. 3 9 9 , 1 c3 , F. W. and S. K. Gess; 1 (3, D. W. Gess

Ceramius toriger Schulthess and Ceramius braunsi Turner

The present observations concerning C. toriger were made in the Skuinshoogte Pass, 15 km
north of Nieuwoudtville on the road to Loeriesfontein and between Nieuwoudtville and Calvinia
during the period 3-8. x. 1989, and in Karoopoort, 43 km ENE of Ceres on the road to Sutherland
on 3.xii.l989.

191

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

Those concerning C. braunsi were made in the Clanwilliam District, at the Clanwilliam
Dam, at the resort and at Caleta Cove, and at Zeekoevlei, c 20 km west of Clanwilliam on the
road to Graafwater during the period 16-20. x. 1989.

Descriptions of nesting areas

A nesting area of C. toriger was located in the Skuinshoogte Pass. The vegetation is
probably closest to Acocks Veld Type 28, Western Mountain Karoo (Acocks, 1953 and 1957).
The nests were situated on a steep slope 50 m from a small farm dam (Fig. 5). The slope was
mostly eroded down to the underlying shale, soil remaining around the bases of old woody
shrubs and as derelict termite heaps. It was these islands of soil which had been used by the
wasps as nesting sites (Fig. 6). The nests were either solitary or in groups, the largest group being
of five nests in an area 50 cm in diameter.

A nesting area of C. braunsi was located on a sparsely vegetated slope above the old
Olifants River Valley Road above Caleta Cove (Figs 1 and 7). The vegetation of this area is
categorized by Moll etal (1984) as a “Mosaic of Dry Mountain Fynbos and Karroid Shrublands”.
The soil is extremely hard compacted sand with sufficient clay to make it malleable when mixed
with water. Water was available from a roadside pool. The nests were grouped in aggregations
of 20 or more. Sloping ground was favoured (Fig. 8). Water erosion indicated that the
aggregations had existed at the same sites over a considerable number of years.

Provision

The provision of both species is of the typical Ceramius type being a firm relatively dry
pollen loaf.

Provision from four nests of C. toriger investigated in the Skuinshoogte Pass was examined.
The pollen was found to be all of one kind and to match that of Berkheya fruticosa.

The pollen from samples taken from the provision and from crop contents of C. braunsi was
all of the spiny composite type.

Water collection

The females of all four species in Group 3 alight on the water surface when imbibing water.
Large numbers of both males and females of C. toriger were observed at water at two small
farm dams and at puddles in a stream bed in the Skuinshoogte Pass and at two small farm dams
between Nieuwoudtville and Calvinia. The females alighted on the surface of the water and
never at the edge on the mud. Activity at water lasted from mid-morning to mid-afternoon.

Gess and Gess (1988a) recorded females of C. braunsi as having been collected on very wet
sand, that is supersaturated sand covered with a film of water, at the edge of a dam.

During the present study both females and males were observed on the water surface of the
pool at Caleta Cove.

Fig. 5. Skuinshoogte, 15 km north of Nieuwoudtville on the road to Loeriesfontein; nesting area of Ceramius toriger
Schulthess.

Fig. 6. Island of soil; a nesting site of Ceramius toriger Schulthess; mud turret surmounting a nesting burrow in fore-
ground.

192

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

193

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

194

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

Male behaviour

Male C. toriger were observed in company with females visiting the forage plant and at
water. The males “waited” in large numbers on the ground at some distance from the water
being visited by females, however, no instances of pairing were observed.

Ten nests of C. toriger were investigated, all in the late afternoon in damp weather when the
builders were sheltering in their nests. A male was found in company with a female in two of
these nests.

Male C. braunsi were observed in company with females at water and visiting flowers,
however, no instances of pairing were observed. No males were observed in association with
nests.

Description of the nests

The nests of C. toriger and C. braunsi consist of a subterranean burrow surmounted by a
curved tubular mud turret (Figs 9, 10 and 11). The turret is constructed of mud pellets cemented
together and well smoothed on the inside so that few interstices remain. The main shaft of the
burrow is short and vase-shaped, having the upper part of the same diameter as that of the turret
and the lower part, the “bulb”, of a diameter up to two and a half times greater depending upon
the number of cells present. From the “bulb” extend one or more very short sub-vertical
secondary shafts each terminating in an excavated cell within which is a constructed mud-cell the
neck of which extends into the secondary shaft. All completed cells are sealed and the secondary
shaft between the sealed cell and the main shaft is filled with tightly packed earth and its opening
to the main shaft is sealed with a mud plate.

Method of construction of nest, oviposition and provisioning

At the commencement of nesting a female may either initiate a new nest or enlarge the nest
from which she emerged.

Water is required for nest construction. At an early stage in nesting both nest initiators and
nest enlargers construct a turret surmounting the main shaft. At the commencement of turret
construction pellets are laid down around the shaft opening in such a way that the turret will have
the same inner diameter as the shaft.

After turret construction has been completed further mud pellets extracted in shaft sinking
are discarded in close proximity to the nest.

The first mud cell having been carefully smoothed on the inside oviposition takes place and
is followed by provisioning.

No egg was obtained for C. toriger. That of C. braunsi was as is usual for Ceramius laid loose
in the bottom of the empty cell. The single egg obtained was curved, yellow, 6 mm from tip to
tip across the bow and 1,5 mm in diameter. The provision is in the form of a relatively dry firm
nectar and pollen loaf which partially fills the cell.

The completed provisioned cell is sealed, the shaft is firmly packed with soil until the “bulb”
is reached and the opening is then sealed off with a mud plate.

Further cells are prepared in a similar fashion to the first.

Samples of 10 C. toriger and eight C. braunsi nests were investigated.

Fig. 7. Area above Caleta Cove, Clanwilliam Dam; bare slope, nesting site of Ceramius braunsi Turner; roadside pool,
water source for C. clypeatus Richards, C. micheneri Gess and C. braunsi.

Fig. 8. Part of an aggregation of nests of C. braunsi Turner.

195

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

Fig. 9. Mud turret surmounting burrow of Ceramius toriger
Schulthess (xl,5).

Fig. 10. Mud turret surmounting burrow of Ceramius
braunsi Turner (xl,5).

Of the C. toriger nests three were new nests which were surmounted by turrets but which
had not yet reached the level of the bulb, and seven were old reused nests surmounted by newly
constructed turrets (Table 3). Nest measurements are given in Table 5.

Table 3.

Details pertaining to 10 nests of Ceramius toriger Schulthess investigated in the Skuinshoogte
Pass, 15 km north of Nieuwoudtville on the road to Loeriesfontein.

Nest No.

Nest Status

Turret

No. of cells

Nature of each cell, cell contents

Remarks

New

Present

—

5 C. 1.

Old, reused

Present

A B/C B/C E F

9 C. t.

Old, reused

Present

A A A A A I X

—

Old, reused

Present

A A I

—

Old, reused

Present

A B/C B/C I

5 C. t.

Old, reused

Present

A B/C B/C E F

$&(3 C. i.

Old, reused

Present

A A B/C F

9 C. t.

Old, reused

Present

A A A B/C B/C F

9&S C. t.

New

Present

—

New

Present

—

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

B. Cell closed, containing pupa in cocoon.

C. Cell closed, containing pre-pupa in cocoon.

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

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

I. Cell open, empty.

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

C. t. = Ceramius toriger Schulthess

All eight nests of C. braunsi were surmounted by turrets, seven nests were newly excavated
and one was an old reused nest (Table 4). Of particular note was the reuse of a cell containing
a cocoon from which a wasp had emerged. Such reuse of an old cell had not previously been
recorded for Group 3. Nest measurements are given in Table 5.

196

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

E D D

Fig. 11. Vertical plans of turrets and underground workings of three nests of Ceramius toriger Schulthess (above) and of
three nests of C. braunsi Turner (below) (x 1). For key to lettering see tables 3 and 4.

197

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

Table 4

Details pertaining to 8 nests of Ceramius braunsi Turner investigated above Caleta Cove,

Clanwilliam.

Nest No.

Nest Status

Turret

No. of cells

Nature of each cell, cell contents

Remarks

New

Present

2 C. b.

New

Present

H D

2 C. b.

New

Present

2 C. b.

New

Present

EDO

2 C. b.

New

Present

M. (E. ) a.
seal in shaft

New

Present

M. (E. ) a.
2 in nest

New

Present

M. (E. ) a.
seal in shaft

Old, reused

Present

AEDDF

—

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

D. Cell closed, containing mature larva spinning cocoon.

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

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

H. Cell open, containing egg without provision.

I. Cell open, empty.

Z. New cell containing bee cell.

C. b. = Ceramius braunsi Turner

M. (E.) a. = Megachile (Eutricharaea) aliceae Cockerell

Table 5.

Measurements of nests of Ceramius toriger Schulthess and Ceramius braunsi Turner.

Range (mm)
C. t. C. b.

Average (mm)
C. i. C. b.

Sample
C. t.

size
C b.

Height of turret

7-10

9-11

8,0

10,3

Diameter of shaft

4-5

4,6

4,5

Diameter of bulb

15-22

16-23

17,3

19,3

Diameter of excavated-cell

9-9

10-11

10,3

Length of shaft above bulb

7-20

11-59

13,2

23,7

Length of bulb

15-22

16-23

17,3

19,4

Total length of main shaft

26-36

30-78

30,5

43,3

Length of excavated-cell

18-19

18,5-20

18,2

19,5

Associated insects

The leaf cutting bee Megachile (Eutricharaea) aliceae Cockerell (Megachilidae) and the nest
parasite Coelioxys (Coelioxys) recusata Schulz (Megachilidae) were observed in attendance on
nests of C. braunsi. A sample was taken; male and female M. (E. ) aliceae and female C. recusata
were found to be present.

Three of the eight nests of C. braunsi investigated were found each to contain a petal-cell
(Fig. 12). A further nest thought to be that of C. braunsi also contained a petal-cell. All of the
nests were new and one-celled. Two of the nests were closed with a final bee seal but in each of
the other two, in which the petal-cells were still being constructed, a female M. (E.) aliceae was
found in the nest.

198

GESS & GESS; ETHOLOGY OF CERAMIUS LATREILLE

'i Vi*i

■ft .

' 4. '

'* The bee’s flask-shaped petal-cells had

C, ■" * » Li r** been constructed within the wasp’s cells in such

a way that the bee’s cells entirely filled the
wasp’s cells. Three of the petal cells were
constructed from pink petals and the fourth
from yellow petals. The pink petals matched
those of a pink flowered Pelargonium species
growing close by, which was found to have
petals cut off and which was observed to be
visited by M. (E.) aliceae. The yellow “petals”
appeared to be cut from composite ray florets.

The provision was syrupy in nature and
contained mixed pollen of three types. The
pollen was compared with that of the plants in
flower in the nesting area. One of the pollens
matched that of a yellow flowered Homeria sp.
(Iridaceae), another was of the spiny composite type and the third, small and thin walled, was
possibly “mesem”.

In the two nests in which a final bee seal had been constructed, the sealed petal-cell had
been surmounted by a mud-seal, the shaft above the cell had been filled with earth and a final
seal had been constructed across the shaft entrance beneath the mud turret. The final seal had
been constructed from a layer of petals covered with a layer of mud.

M. (E.) aliceae cells have previously been recorded from burrows of Parachilus insignis
(Saussure) (Eumenidae) at Hilton, Eastern Cape Province (Gess and Gess, 1976), Paravespa
(G.) mima Giordani Soika (Eumenidae) at Tierberg, Prince Albert, southern Cape Province
(Gess and Gess, 1988b) and C. nigripennis at Mesklip, 20 km south of Springbok, Namaqualand
(Gess and Gess, 1986). In all instances the bee’s cells were orientated vertically.

The presence of C. (C.) recusata Schulz is of interest as Coelioxys is known to be a parasite
in nests of Megachile.

Fig. 12. Petal cell of Megachile (Eutricharaea) aliceae
Cockerell (Megachilidae) constructed in a cell of Ceramius
braunsi Turner. (xl,5).

Group 6

Group 6 is constituted of three species, Ceramius caffer Saussure, C. metanotalis Richards
and C. rex Saussure.

Geographic distribution

The distributions of C. caffer, C. metanotalis and C. rex are poorly known, however,
available collecting records indicate very distinct and restricted distributions within the Western
Cape: C. caffer to the south of the Olifants River Mountains; C. metanotalis north of these
mountains but south of Namaqualand; and C. rex in the area of Namaqualand termed
Klipkoppe.

C. caffer was described from “Cape of Good Hope” (Saussure, 1855). It is recorded from
Stellenbosch, 1888 (1 6 in Richards, 1962 and a further 3 6 6 and 2 $ $ in Gess, 1965) and 1908
(5 9 ? in Richards, 1962 and a further 26 9 ? in Gess, 1965). Two further records (Albany
Museum Collection) are presented in the present paper: Tulbagh, xi.l947, 1 9 , J. G. Theron and
8 km ENE of Ceres at the western end of the Theronsberg Pass, 29. xi. 1989, 1 9 , F. W. Gess and
S. K. Gess.

199

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

C. metanotalis was originally described from Calvinia, Nieuwoudtville and “Capland”
(Richards, 1962); additional material from Bulshoek, Klawer-Clanwilliam was noted by Gess
(1965). Additional records are here presented. These are from 30 km north of Clanwilliam on
the N7 road to Klawer above the Bulshoek Dam, 17.x. 1989; 2 9 $ , F. W. and S. K. Gess; 3 6 3,
D. W. Gess.

An account was given in Gess and Gess (1988a) of the search for C. rex which until it was
found nesting, not uncommonly, in the Hester Malan Nature Reserve in October 1987 had been
known only from three specimens: the type specimen, a female labelled Cape Colony, which
was described by de Saussure in 1855; a female collected at Garies in Namaqualand in 1931
(S. A. M. Staff); and a male collected at Garies in 1970 (Dr and Mrs H. Townes). Due to the
limited knowledge of this species it seems useful to record additional collecting records. These
records represent samples not sightings which were more numerous. All are for theiarea known
as Namaqualand Klipkoppe.

Hester Malan Nature Reserve, Springbok, 10-11. x. 1989: 29 9, F. W. and S. K. Gess.

W end of Wildeperdehoek Pass, 29° 56' 21"S, 17° 37' 30"E, 14.x. 1989: 2 9 9, F. W. Gess and

S. K. Gess; 1 9, D. W. Gess.

Nababeep, 12-13.X.1989: 1 cJ, D. W. Gess.

Ethology

An account of flower visiting, water collection and nesting by C. rex was given by Gess and
Gess (1988a). A further record of flower visiting for C. rex and first records for C. capensis and
C. metanotalis of are presented in the present paper.

Flower visiting

Flower visiting records are few, however, they all support an association between Ceramius
Group 6 and Compositae.

Gess and Gess (1988a and 1989) recorded a female C. rex visiting Berkheya spinosissima
(Thunb.) Wind and supported this record with analyses of nest provision and crop contents. In
the present paper one record of a male visiting flowers of Pteronia sp. A (Compositae) is
presented (Nababeep, 12-13. x. 1989, D. W. Gess).

Only one record of flower visiting by C. caffer has been obtained. This is for a female
visiting flowers of Berkheya carlinifolia (DC.) (Compositae) (18 km ENE of Ceres at the
western end of the Theronsberg Pass, 29. xi. 1989, F. W. and S. K. Gess).

C. metanotalis is recorded foraging on Athanasia trifurcata (L.) L. (30 km N of Clanwilliam
on the N7 road to Klawer above the Bulshoek Dam, 17.x. 1989, 2 9 9 , F. W. & S. K. Gess and
2 dd, D. W. Gess).

All other plants in flower in the areas where these wasps were collected were sampled for
insect visitors. No others were found to be visited by these wasps.

Water collection

Water collection by C. rex was observed in the Hester Malan Nature Reserve, Springbok,
by Gess and Gess (1988a). The wasps collected water from small pools in a river bed. This
behaviour has been confirmed by further observations in the Hester Malan Reserve and also at
the western end of the Wildeperdehoek Pass to the south west of Springbok.

One male C. metanotalis was collected at a small earthen dam above the Bulshoek Dam in
close proximity to the forage plants indicating that this was probably the water source for females
nesting in the area.

200

GESS & GESS: ETHOLOGY OF CERAMIUS LATREILLE

DISCUSSION

Gess and Gess (1986 and 1988a) proposed the use of ethological characters in conjunction
with morphological characters in defining species groups in Ceramius. Attention was drawn to
the considerable constancy of nest plan exhibited within the species groups and the considerable
differences between groups. It was suggested that these similarities and differences are such that
they may be sufficient for it to be possible to place a nest in a species group.

The discoveries of nests of C. braunsi and C. toriger have made it possible to add Group
3 to those groups (5 and 8) for which the nest characters of a group as a whole can be defined.
It was suggested (Gess and Gess, 1988a) that Group 3 may prove to be distinct in that the
cells are all excavated sub-vertically beneath the bulb. At that time, however, the nests of
only two species (C. jacoti and C. nigripennis) of the four assigned to this group were known.
In the present paper the nests of the remaining two species (C. braunsi and C. toriger) are
described. That they also excavate their cells sub-vertically and that no species from another
group has been found to exhibit this characteristic adds support to the suggestion.

Gess and Gess (1988a) further drew attention to the apparent constancy in choice of forage
plant family within species groups, three groups and one sub-group seeming to obtain nest
provision from Mesembryanthemaceae only, two groups from Compositae only and one
sub-group from Leguminosae only. At the time records for all species in a group were available
for groups 3, 5 and 8.

The presently added forage plant records for C. toriger and C. braunsi support the
contention that the characteristic forage plant family of Group 3 is the Compositae.

The present first forage plant records for C. caffer and C. metanotalis with those previously
recorded for C. rex (Gess and Gess, 1988a) make known the preferred forage plant family of all
the species of Group 6, confirming a constant choice of plants of the family Compositae by this
group.

Gess and Gess (1988a) on the basis of the morphology of the clypeus and nature of the
forage plants divided Group 2 into two sub-groups. Group 2a being constituted of
C. cerceriformis Saussure and C. peringueyi Brauns and Group 2b being constituted of
C. clypeatus and C. richardsi. At that time the nest structure was known only for
C. cerceriformis. In the present paper the nest structure of C. clypeatus is described. It is very
similar to that of C. cerceriformis suggesting a close relationship between the species of the two
sub-groups despite their differences in morphology and choice of provision.

As already stated in the ethological account above, morphologically C. micheneri does
not readily fall into one of the eight recognized species groups as it shows morphological
similarities with both Group 2 and Group 3. The nest structure and nature of the forage plant
recorded in the present paper, however, indicate that it does not belong to Group 3 nor
indeed to any of the other recognised species groups with the possible exception of Group
2. One of the morphological characters in which C. micheneri differs from the four species
currently forming Group 2 is the rounded rather than spinose or at least bluntly angular
propodeum. That this difference in propodeal form need not be of any great significance is
suggested by the fact that species with both spinose and rounded propodea occur also in the
ethologically very homogeneous Group 8. That the forage plant belongs to the family
Leguminosae suggests a closer relationship with Group 2b than with Group 2a, the known
forage plants of which belong to the families Leguminosae (C. clypeatus) and Mesembryan-
themaceae (C. cerceriformis and C. peringueyi) respectively. Certain morphological

201

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 9, MAY 1990

characteristics, in particular those pertaining to the clypeus, indicate that C. micheneri does
not, however, fit satisfactorily in Group 2b. It is therefore suggested that it should be placed
in a additional sub-group, Group 2c.

Clearly the discovery of the nests of C. peringueyi and C. richardsi and the forage plant of
C. richardsi are required before further clarification of Group 2 will be possible.

ACKNOWLEDGEMENTS

The authors wish to thank Mr Klaas van Zyl of the Cape Department of Nature and
Environmental Conservation for permission to work in the Hester Malan Reserve.

Mr David Gess, Mr Harold Gess and Mr Robert Gess are thanked for their enthusiastic
help in the field.

Mrs Estelle Brink of the Albany Museum Herbarium and Mr Ted Oliver of the Stellenbosch
Herbarium are thanked for assistance with the identification of plants.

Gratitude to the C. S. I. R. is expressed by F. W. Gess for running expenses grants (Main
Research Support Programme and National Programme for Ecosystem Research) for field work
during the course of which the present studies were made.

REFERENCES

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

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

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

Gess, F. W. 1968. Further contribution to the South African species of the genus Ceramius Latreille (Hymenoptera:
Masaridae). Novos Taxa ent. 53: 29-33.

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 (6): i09-122.

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

Gess, F. W. and Gess, S. K. 1980. Ethological 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. 1988a. A further contribution to the knowledge of the ethology of the genus Ceramius
Latreille (Hymenoptera: Masaridae) in the southern and western Cape Province of South Africa. Ann. Cape
Prov. Mus. (nat. Hist.) 18 (1): 1-29.

Gess, F. W. and Gess, S. K. 1988b. A contribution to the knowledge of the ethology of the genera Parachilus Giordani
Soika and Paravespa Radoszkowski (Hymenoptera: Eumenidae). Ann. Cape Prov. Mus. (nat. Hist.) 18 (3):
57-81.

Gess, S. K. and Gess, F. W. 1989. Flower visiting by masarid wasps in southern Africa (Hymenoptera: Vespoidea:
Masaridae). Ann. Cape Prov. Mus. (nat. Hist.) 18 (5): 95-134.

Moll, E. J. , Campbell, B. M. , Cowling, R. M. , Bossi, L. , Jarman, M. L. and Boucher, C. 1984. A description of the
major vegetation categories in and adjacent to the fynbos biome. South African National Scientific Programmes
Report 83: i-iv, 1-29.

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

202

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing U or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

BK1515 — CTP Book Printers, Cape

el5

ISSN 0570-1880

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 10 15th March 1991

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

The herpetofauna of the offshore islands of South Africa and Namibia

WILLIAM R. BRANCH

(Port Elizabeth Museum, P.O. Box 13147, Humewood 6013, South Africa)

CONTENTS

Abstract 205

Introduction 206

Herpetofauna of the offshore islands 207

St Croix Island 207

Bird Island 212

Dyer Island 213

Robben Island 213

Dassen Island 216

Marcus Island 218

Malgas Island 218

Jutten Island 218

Meeuw Island 218

Schaapen Island 218

Pomona Island 218

Possession Island 219

Recent geological history of the islands 219

Introduced species 220

Island-species richness relationships 221

Rafting and the effects of floods on reptiles and amphibians 221

Acknowledgements 222

References 222

Appendix 224

ABSTRACT

The herpetofauna of the continental offshore islands of South Africa and Namibia is
reviewed. Historical and recent records are discussed and the origin and affinities of the
herpetofauna are considered. A total of 23 species has been recorded from 11 islands. The
greatest diversity (16 species) occurs on the largest island, Robben Island in Table Bay. The
commonest island species are Phyllodactylus porphyreus (seven islands) and Cordyliis cordyliis
(four islands). All of the islands have rocky coastlines and are unsuitable for nesting sea turtles.

205

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

The islands have been affected to different degrees by sea-level fluctuations associated with
periods of glaciation. All were connected to the mainland during the last Glacial Maximum
(16 000 BP). Human disturbance has occurred on all the islands and many species have been
introduced. A number of islands have relict populations of some species, and these probably
date from the time when the islands were connected to the mainland. No significant correlation
exists between species richness and an island’s size or its distance from the mainland. This is
attributed to the relatively small size of the islands, their depauperate herpetofaunas, and the
habitat requirements of the species. The spotted gecko Pachydactylus maculatus grows to a
much larger size on St Croix Island than on the adjacent mainland but gigantism does not occur
in reptile populations on other islands.

INTRODUCTION

The continental shelf of southern Africa includes a total of 69 small rocky islets and offshore
islands (Skead, 1975). These range in size from small, wave-swept rocks that harbour a few
resting seals or seabirds to the large and well-vegetated Robben Island that has been colonised
for many years, albeit unwillingly, by the many inmates of the old leper colony, insane asylum
and penal settlement (De Villiers, 1971). Almost all the islands are small and uninhabited. Only
two islands exceed 100 hectares in size. Most were periodically visited by early settlers and
seamen who came to collect food, skins and oil from the prolific colonies of seals and seabirds.
Nowadays all of the islands are protected although accumulated guano is still harvested from
some, particularly those along the Namibian coast. The larger islands (Robben, Dassen, Dyer
and Bird islands) have lighthouses, some of which are now automated and require only periodic
visits for maintenance. Human disturbance occurs regularly on most islands and some are
permanently inhabited.

The islands may be conveniently grouped; those of Algoa Bay in the Eastern Cape (St Croix
and Bird islands); those of the southwestern Cape (Robben and Dassen islands, and those
associated with Saldanha Bay, i.e. Marcus, Malgas, Jutten, Meeuw and Schaapen islands); and
those along the Namibian coast near Liideritz Bay (Pomona and Possession islands). More
isolated is Dyer Island, situated off the Bredasdorp coast, southern Cape. All the islands are
situated on the continental shelf and have been affected by sea level changes during periods of
glaciation. During the last Glacial Maximum (16 000 BP) sea levels were very low
(approximately —130 m) and during this period all of the present islands along the southern
African coast would have been joined to the mainland (Tankard, 1976). With the onset of
warmer climates and the subsequent rise in sea level, the islands became progressively cut off
from the mainland. Some very low-lying islands may have been affected by a brief and slight rise
(3m) in sea-levels 2 000 BP during the last climatic optimum or hypsithermal (Flemming, 1977).

Variation in species richness of the major islands of the southwestern Cape has been
analysed by Brooke and Crowe (1982). Alien species on all South African offshore islands have
been reviewed by Brooke and Prins (1986) and Cooper and Brooke (1986). Although these
authors list a number of reptiles from some islands, they do not comprehensively review old
records (Brooke and Crowe, 1982, state explicitly that they have not consulted literature prior
to 1971). The present account is the first detailed review of the herpetofauna of these islands and
the first to consider origins and affinities.

The historical references derive mainly from Skead (1975). Recent literature was reviewed
for references to reptiles and amphibians on the islands. Specimens from the islands housed in

206

BRANCH; HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

the major herpetological collections in South Africa were recorded (see appendix for acronyms
and catalogue numbers); other records, where noted in the literature, are given. Searches for
reptiles and amphibians were made during trips by the author to a number of the major islands:
Dassen Island, 4-10 April 1987 and 4 November 1988; Bird Island (Algoa Bay), 22-26 May
1980; St Croix Island (Algoa Bay), 7 July 1979 and 22 March 1989.

Historical and recent records of island species were assessed and compared to determine
which were likely to represent introductions and which represent naturally occurring species.
The latter may have reached the islands through rafting or may be relict populations from
periods of sea-level fluctuations.

HERPETOFAUNA OF THE OFFSHORE ISLANDS

The geographical situations of the islands discussed are shown in Fig. 1, and the reptiles and
amphibians recorded from the islands are listed in Table 1. Specific details for the islands and
their recorded herpetofauna are given, prefaced by a short note on the marine reptiles found in
the coastal waters of southern Africa.

A number of sea turtles (Hughes, 1974a and b) and a single sea snake occur. The
yellow-bellied sea snake Pelamis platunis is common in Algoa Bay (PEM R 13, 61,
1128,1130-31, 1492, 2360), and there are sporadic records along the southern Cape coast as far
as False Bay (Broadley, 1983). However, the species is excluded from the Atlantic Ocean by the
cold Benguela Current that sweeps up the western Cape coast, bringing cold Antarctic waters
close inshore. Although green sea turtles have recently been reported to bask on desolate
beaches near the Cunene River mouth in northern Namibia (Tarr, 1989), the steep rocky
shoreline of most of the southern African offshore islands precludes sea turtles from beaching.
The shoreline of Bird Island is less steep but still rocky, although sea turtles do forage close
inshore. A large (carapace length approximately 1 m) loggerhead sea turtle, Caretta caretta, was
caught by a fisherman using squid bait from the Bird Island jetty on 24 May 1980, but was
subsequently released (Branch, pers. obs.). None of the islands has sandy beaches suitable for
nesting sea turtles, even though two species, Caretta caretta and Dermochelys coriacea, nest in
northern Natal (Hughes, 1974a, b) and occasionally further south (Branch, 1988a). The only sea
turtle record for any offshore island is a single hawksbill turtle, Eretmochelys imbricata, shell
recorded from the beach of Dyer Island. The specimen was initially catalogued from Dassen
Island but this was later corrected to Dyer Island. This is more likely in view of the cold
Benguela Current on the west Cape coast.

St Croix Island, Algoa Bay (33° 48'S, 25° 46'E; 625 m x 250 m, 2.5 ha, 3.9 km from coast,
59.4 m a.s.l., not manned; Fig. 2.)

Hewitt (1920) noted that the common girdled lizard, Cordylus cordylus, was abundant on
the island and presumed this to be due to the absence of predation by kestrels. He also noted
that they were smaller, had narrower heads, and as adults lacked the bright brick-red colours of
Grahamstown specimens. During a brief (3 hr) visit to the island by the author (21 March 1989),
many Cordylus cordylus were seen basking on rocks or foraging around penguin nests (Fig. 3).
The island is home to a major breeding colony of the Jackass penguin, Spheniscus demersus, and
has a resident population of 13-18 000 individuals. Nine lizards were caught, weighed and
measured. Numerous other specimens were seen, a number of which were slightly larger than
any captured. The largest lizards measured (male, snout-vent (SV) 70 mm, tail 70 mm, weight

207

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

208

BRANCH; HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

Table 1.

Herpetofauna of the offshore islands of South Africa and Namibia

SPECIES

SOUTH AFRICAN AND NAMIBIAN OFFSHORE ISLANDS

TOTAL

Size of island (ha)

507

222

2.5

Distance from coast (km)

6.7

0.5

0.14

1.2

0.8

3.9

0.2

2.7

Human habitation

Amphibians

Strongylopus grayii
Breviceps rosei

Reptiles

Snakes

Pseudaspis cana
Lamprophis inornatus
Lycodonomorphus rufuliis
Aspidelaps lubricus

Lizards

Gekkonidae

Phyllodactylus porphyreus
Phyllodactylus lineatus
Pachydactylus maculatus
Pachydactylus geitje
Chamaeleonidae

Bradypodion pumilum
Agamidae
Agama atra
Scincidae

Mabuya capensis
Mabuya homalocephala
Scelotes bipes
Scelotes gronovii
Acontias meleagris
Cordylidae

Cordylus cordylus
Tetradactylus seps
Chelonians

Eretmochelys imbricata
Chersina angulata
Geochelone pardalis
Pelomedusa subrufa

* *

K?)

3(4?)

4
1

K?)

TOTAL

(excluding doubtful records)

6 3 1

2 0

Southwestern Cape 1, Robben Is; 2, Dassen Is.; 3, Schaapen Is., 4, Meeuw Is., 5, Marcus Is.; 6,
Malgas Is., 7, Jutten Is., 8, Dyer Is.

Algoa Bay 9, St Croix Is.

Namibia 10, Pomona Is.; 11, Possession Is.

209

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

Fig. 2. St Croix Island, Algoa Bay, South Africa.

11.7 g; female SV 72 mm, tail 74 mm, weight 11.5 g) were not significantly smaller than
specimens on the adjacent mainland (Branch, unpubl. obs.). They are a dull dark brown in
coloration. Their relationship to typical Cordylus cordylus and to the more arboreal,
closely-related species Cordylus tasmani, which is endemic to the Valley Bushveld thicket of the
adjacent Algoa Basin, needs to be investigated further. Pending the results of such a study, the
population on St Croix is provisionally referred to typical Cordylus cordylus.

Hewitt (1920) also recorded the spotted gecko, Pachydactylus maculatus, from the island,
noting that it was less abundant than the girdled lizard. During a visit to the island (22 March
1989) the author found six geckos sheltering among stone piles on the barren island. Two were
recent hatchlings. The other four were adult females and were very large and had short,
regenerated tails (snout-vent 55-60 mm, mean 58.00 mm; tail 27-31 mm, mean 29.25 mm;
weight 7.19-7.53 g, mean 7.32 g). Eight preserved specimens, three females and five males, from
the island support these findings. The females all measure over 50 mm SV (50-56 mm) and the
males range from 38-53 mm SV. In a study of the spotted gecko on the adjacent mainland
253 specimens were weighed and measured. Females grew larger than males; largest male, SV
41mm, weight 2.6g; largest female, SV 42 mm, weight 3.1 g. Even the largest mainland female
weighs less than half that of the smallest St Croix female, and the size of males is also
much smaller. These results demonstrate that the geckos on St Croix grow substantially larger
than on the adjacent mainland. Whether this is due to reduced predation or abundant food is
unknown.

Of the six live geckos collected all but the smallest hatchling had regenerated tails. Similarly
all eight preserved geckos have regenerated tails. The incidence of regenerated tails on the
island is therefore very high (91.67%). Tail regeneration frequency on the mainland was found

210

BRANCH: HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

Fig. 3. The common girdled lizard Cordylus cordylus foraging among nesting Jackass penguins on St Croix Island, Algoa

Bay, South Africa.

to be dependent upon age. All geckos over 40 mm SV had regenerated tails. Tail autotomy in
geckos is usually considered to serve as an antipredator device although tail-loss may occur for
other reasons (Arnold, 1988). Some geckos on the mainland were observed to undergo 3-4 tail
regenerations; a principal predator was found to be the Natal hunting spider, Palystes natalius
(Branch, unpublished data). Snakes are also often considered to be major predators of lizards.
The potential efficacy of tail autotomy was noted when a spotted skaapsteker, Psammophylax
rhombeatus was found to contain a complete spotted gecko and the autotomised tail of another
spotted gecko (Branch and Braack, 1987). The cause of the high frequency of tail loss in spotted
geckos from St Croix is unknown as the normal predators (snakes, viverids, large spiders, etc.)

211

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

Fig. 4. Size difference between populations of the spotted thick-toed gecko, Pachydactylus maculatus, from St Croix
Island, Algoa Bay (right), and the adjacent mainland (left).

are absent from the island. Penguins, which were observed to peck at girdled lizards that
came to close to their nests, may be responsible for some tail loss. Whether the nocturnal spot-
ted gecko is treated in the same manner is unknown. Intra-specific fighting may also be
responsible for some tail loss but it is unlikely to be solely responsible for the high percentage
observed.

Bird Island, Algoa Bay (33° 51'S, 26° 17'E; 548 m x 320 m, 19 ha, 8.4 km from coast, 9.1
m a.s.l., manned)

No early visits to the island mention the presence of reptiles on the island (Skead, 1975) and
Hewitt (1937) observed that “On Bird Island there are no reptiles whatever”. However, Rand
(1963) stated “. . . lizards, tortoises and occasionally snakes are found here too”. There are no
preserved voucher specimens in any museum collection to validate this claim. Jordan,
ex-lighthouse keeper on Bird Island, 5 Dec 1951 — 24 Jan 1953, records {in litt. Skead, April
1971) that “There were lizards on the island, also one tortoise, small type. My children collected
a few on my brother-in-law’s farm and took them over for company for him. There were no
snakes . . . There were also a few frogs, mainly brought over with our vegetables . . . my
brother-in-law . . . used to dig out cabbages, carrots . . . [etc.] . . . intact and we used to replant
them on Bird Island”. The identity and the fate of these animals are unknown. The Port
Elizabeth Museum has had an active research program studying the gannet population on the
island for the last ten years. During this period, of almost monthly visits, no reptiles or
amphibians have been discovered living on the island. In 1981, an angulate tortoise was found
freshly dead on the shore of Bird Island (A. Batchelor, pers. comm.) following heavy rain on the

212

BRANCH: HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

mainland. The old records noted by Rand (1963) can either be discounted or they document
temporary colonizations that have subsequently died out.

Dyer Island, Southern Cape (34° 40'S, 19° 25'E, 731 m x 180 m, 20 ha, 3.5 km offshore, 6
m a.s.l., manned)

Brooke and Crowe (1982) record the Cape skink Mabuya capensis and the angulate tortoise
Chersina angulata from the island but give no further details. Symons (1924) visited the islands
to collect bird eggs but records no observations of tortoises, suggesting that these introductions
may have occurred recently, perhaps in a similar manner to that noted above and by Haacke
(1965) for Possession Island (see below).

Robben Island, South-western Cape (33° 49'S, 18° 22'E; 4 km x 2 km, 507 hec., 7 km from
land, 30 m a.s.l, colonised since 1652)

Robben Island is the largest South African offshore island and also the closest to Cape
Town which was one of the first areas on the subcontinent to be settled by Europeans. It has
been continuously inhabited since 1652 and has been substantially altered by more than 300
years of human occupation. In the 19-20th century it was used as a leper colony and insane
asylum (both of which have now been closed). It has been used as a penal colony from the
earliest days (Thunberg, 1793, notes that “criminals [are] exiled there”). At present access to the
island and possibilities for biological surveys there are restricted.

Amphibians

Robben Island is the only offshore island from which an amphibian has been recorded (i.e.
the spotted stream frog, Strongylopus grayii) (Rose, 1950). The species was recently confirmed
to be present at the old quarry (Baard, et al., 1986). This, the only permanent source of
freshwater on the island, dates from the settlement established by Van Riebeeck in 1652. Due
to their intolerence of salt water amphibians do not easily colonise off-shore islands. However,
S. grayii is common in the southern Cape coastal regions and can be found breeding in pools of
freshwater in the dune slacks. It is probable that the species was introduced to Robben Island
by man, possibly in a manner similar to that noted for frogs to Bird Island in Algoa Bay (see
above). That the spotted stream frog can be successfully translocated is shown by its
introducition to St Helena in the middle of the Atlantic Ocean (Frost, 1985).

The absence of standing fresh water need not preclude all amphibians, development in
some terrestrial species being direct. Such a reproductive mode is characteristic of rain frogs
{Breviceps spp.), which are common in the Cape coastal region. Boynton (1964) records Rose’s
rain frog {Breviceps rosei), a species restricted to the sandveld of the western Cape coastal
foreland where it may be locally common, from the island. This record is based on three
specimens in the South African Museum. It should be noted that these specimens and that of the
olive water snake from Robben Island (see below) were donated (collected ?) by Rev. G. Fisk
and accessioned into the SAM collection between 1895 and 1896. Boulenger (1910) does not
record any amphibians from Robben Island and furthermore no recent collections of B. rosei
confirm the presence of the species on the island.

Reptiles

Numerous reptiles have been recorded from the island but it is difficult to determine which
species have been introduced during the period of human settlement. However, early records do
give some idea of the island’s relatively diverse herpetofauna before European settlement. The

213

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

indigenous Khoisan population did not use boats at sea and there is no evidence that they
visited any of the offshore islands. They were thus not responsible for herpetological
introductions.

Snakes

The earliest references note the presence of snakes on Robben Island:

1610 “Upon the Illand be . . . [an] aboundance of great snakes lying upon the ground
against the Sune . . .” — Thomas Best (in Raven-Hart, 1967).

1614 “. . . the illande is very full of Snakes . . .’’—Nicholas Downton (in Raven-Hart,
1967)

1615 . . there are a very great number of snakes in that island . . ., so many of them
venomous worms that a man cannot tread safely in the long grass which grows in it for fear
of them . . .’’ — Edward Terret (in Raven-Hart, 1967)

Such is the universal fear of snakes that soon attempts were made to eradicate them from
the island.

1638’’. . . the blacks while there had destroyed almost all the penguins and the
snakes.” — Arthur Gijsels (in Raven-Hart, 1967)

1654 “. . . there are snakes which we notice do them (i.e. the sheep) no harm, and of these
our men are instructed to exterminate as many as possible” — Van Riebeeck (in Thom,
1952-1958)

1658 “You must make every effort to destroy the numerous snakes during periods of
idleness, so that the island may once and for all be cleared of these poisonous
vermin.” — Van Riebeeck (in Thom, 1952-1958)

That these attempts were unsuccessful is indicated by Le Vaillant’s comment in 1782 quoted
by Skead (1975) that he had

“. . . seen here a great many black serpents four feet (1.21 m) in length but they are not
venomous . . .”

More recently Siegfried {in litt. Skead, November 1965) noted that when the Cape
Department of Nature Conservation wanted mole snakes to restock the Cape Peninsula in order
to control small rodents they “took gangs of men over to Robben Island and found a mole snake
under every pile of rubbish they turned up.”

The first identifications of snake species from the island, based on specimens housed in the
South African Museum, are by Boulenger (1910). He records the brown water snake
Lycodonomorphus rufulus and the olive house snake Lamprophis inornatus. FitzSimons (1962)
repeats these records and also includes the mole snake Pseudaspis cana. It is surprising that
Boulenger did not document this large, non-venomous colubrid as it fits best the early
descriptions of snakes on the island (e.g. ‘great size’, ‘black’). Specimens from the western Cape
populations of this impressive snake are often a uniform black and grow much larger than those
from elsewhere (in excess of 2 m; Broadley, 1983). The mole snake was probably present before
settlement by man. It is possible that the other species were also present but overlooked as both
are small and nocturnal. Small mammals form the usual diet of the olive house snake although
this species will also take small reptiles (Branch, 1988). The brown water snake feeds almost
exclusively on frogs but may occasionally take fish and exceptionally geckos. Its existence on the
island is probably dependent upon the presence of amphibian prey.

214

BRANCH: HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

Lizards

Very early references note the presence of chamaeleons and other lizards on the island:
1620 “There are . . . also some chamaeleons and other lizards” — Augustin de Beaulieu (in
Raven-Hart, 1967)

1773 “Now a resort of chamaeleons and quails.” — Thunberg (in Skead, 1975).

Boulenger (1910) includes Robben Island as a locality for a number of lizards, based on
specimens in the South African Museum, but gives no further information on their origin or
details of their capture. These are:

Ocellated gecko, Pachydactylus geitje (as P. ocellatus)

Cape girdled lizard, Cordylus cordylus cordylus (as Zonurus cordylus)

Redsided skink, Mabuya homalocephala homalocephala
Silvery dwarf burrowing skink, Scelotes bipes
Cape legless skink, Acontias meleagris meleagris
Short-legged seps, Tetradactylus seps

Rose (1926 and 1929) records a number of additional lizard species from the island, as well
as giving notes on the above species:

Marbled leaftoed gecko, Phyllodactylus porphyreus
“. . . very common ... on Robben Island”.

Ocellated gecko, Pachydactylus geitje (as P. ocellatus).

”... may be seen in very large numbers on Robben Island”,

“. . . closely associated with Zonurus cordylus, Phyllodactylus porphyreus and Acontias
meleagris.”

Cape girdled lizard, Cordylus cordylus cordylus

“On Robben Island yellow-brown ones (flavus) are extremely numerous in the stone heaps
. . . but no black ones (niger) were seen.”

Silvery dwarf burrowing skink, Scelotes bipes

“A specimen we obtained from Robben Island produced two active young . . . during . . .
March.”

Cape legless skink, Acontias meleagris

“On Robben Island it is particularly abundant and we once obtained nine specimens in little
over an hour by turning over large stones by the side of the road that skirts the island.”

FitzSimons (1943) records most of these species and adds the Cape dwarf chamaeleon,
Bradypodion pumilum (as Microsaura pumila pumila), based on a specimen in the Albany
Museum.

Baard et al. (1986) in a recent survey of the island confirm the presence of eight of the above
species, including the mole snake, and also document the presence of the Cape skink Mabuya
capensis (two were preserved and nine others observed). As the survey occurred during the dry
summer, the presence of other amphibian species on the island could not be determined.

Bustard (1963) records reproduction in 50 captive ocellated geckos received from Robben
Island. Mouton (1987) discusses the relictual distribution of melanistic girdled lizards, Cordylus
cordylus, in the southwestern Cape and confirms Rose’s record of the typical form on Robben
Island and the presence of the melanistic C. c. niger on Jutten Island and Schaapen Island in the
Saldanha Bay area. Mouton etal. (1987) include Robben Island as a locality for Scelotes gronovii

215

ANN. CAPE PROV, MUS, (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

prior to 1978 but give no further indication of the provenance of this record and exclude it from
their present distribution map. The record is repeated in Baard (1988). It is based on a specimen
in the Transvaal Museum the identity of which has been confirmed (W.Haacke, pers. comm.).
The species is not known from the adjacent mainland and may either be mis-labelled or
represent a northern retraction of the species’ range following separation of the island from the
mainland. During a recent search on the island the only fossorial skinks found were Scelotes
bipes and Acontias meleagris (Baard, pers. comm.). Further searches on the island are required
to resolve the issue.

Chelonians

Tortoises are mentioned in the earliest references to Robben Island:

1503 “On this island they killed many birds, . . . and sea wolves and tortoises, of which

there is great abundance.” — Antonio da Saldanha (in Raven-Hart, 1967).

These have not been discussed by recent authors (Skead, 1975). It is probable that they are
the angulate tortoise. The presence of the angulate tortoise on the island was confirmed by
Baard et al. (1986). As in many other places in the south-western Cape, the leopard tortoise
Geochelone pardalis has been introduced to Robben Island. D'Ewes (1967) noted that in 1959
a police launch had discovered an adult leopard tortoise swimming in the sea “a couple of miles
offshore from Robben Island, and swimming strongly for the mainland.” Skead (1975) has
traced the origin of this specimen to one of 25 tortoises taken to the island from the Paarl Roller
Flour Mills (now SASKO) in 1959. They had been kept as pets by workers at the mill, and were
disposed of to Robben Island (C. W. van der Westhuizen in lift. Skead, 3.3.1971). This species
was not present during the survey of Baard et al. (1986).

Dassen Island, Southwestern Cape (33° 26'S, 18° 05'E; 2.5 km x 1.05 km, 222 ha, 9 km
offshore, 19 m a.s.l, manned)

Records of early visits to the island note the abundance of wildlife, e.g. both Sir Edward
Michelbourne, 1601 (in Raven-Hart, 1967) and Van Riebeeck, 1652 (in Thom, 1952-8)
comment on the “unbelievable multitude of seals, black birds (cormorants) and penguins.”.
They also noted the presence of a ‘rabbit’; in fact the rock hyrax or dassie Procavia capensis. The
island’s early name, i.e. Cony Island, and its present name are based on the presence of this
species on the island. They were later eradicated and were replaced with the European rabbit
Oryctolagus cuniculus which was introduced as food late in the administration of the Cape by
Van Riebeeck (1652-1662). There are no early records of any reptiles or amphibians. Four
reptile species have been reported in recent years although only three are recently confirmed.

Boulenger (1910) includes Dassen Island as a locality for Agama atra and Scelotes gronovii,
based on specimens in the South African Museum but he gives no further information on the
specimens. FitzSimons (1943) repeats the record for the rock agama but with no further
documentation or proof that a viable population existed on the island. These rock-living,
diurnal, and brightly-coloured (at least in males) lizards are highly visible but no evidence of
their presence on the island was found during the author’s recent visits nor has any evidence of
their presence been reported subsequently. Rose (1929) records receiving a specimen of Scelotes
gronovii from Dassen Island but does not mention the rock agama.

During the author’s first trip to the island (4-10 April 1987), 12 Scelotes gronovii were
collected in sandy, gritty soil under cover (granite slabs, old corrugated iron sheets, wooden
logs, etc.), some only a few metres above the high water mark and others among the roots of

216

BRANCH: HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

scrub vegetation. All were adults (N = 11, SV 65-82, mean 73.44 mm), and all but the two
smallest (SV 65-67 mm) had regenerated tails (i.e., 81.82%). Specimens from Elands Bay and
Langebaan on the adjacent mainland are slightly smaller in size (N = 5, SV 61-77 mm, mean
66 mm) but there are too few specimens to determine whether this size difference is statistically
significant. Two young were born whilst the above specimens were retained in a collecting bottle
but it was not possible to determine whether they came from the same or separate mothers. They
measure; 1. 40 + 21 mm (SV + tail); 2. 38 + 12 (tail broken). In adults with original tails the
tail ranges from 74.3-93.8 % (mean 84.5%, n = 5) of the SV length. This contrasts with
52.5-73.6 % (mean 62.0 %, n = 3) in new born specimens, indicating that tails are relatively
smaller in juveniles than adults.

McLachlan (1978) queried the safety of this species on Dassen Island following disturbance
during guano collection. However, little guano collecting now occurs on the island and is
restricted to the northern, barren peninsulas. The dwarf burrowing skink is distributed
throughout the island in areas not disturbed by guano collecting. The species is still included in
the revised South African Red Data Book — Reptiles and Amphibians and, although not
currently threatened, is placed in the Restricted category (Branch, 1988c).

Neither Boulenger (1910), FitzSimons (1943) nor Rose (1929, 1950 and 1962) records the
presence of the marbled gecko Phyllodactylus porphyreus on Dassen Island, even though
FitzSimons (1943) does record the species from Robben and Jutten islands. Its presence on
Dassen Island was first documented by Brooke and Crowe (1982), however, no further details
were given or voucher specimens documented. Numerous specimens of the marbled gecko were
found by the author during a short trip to the island (4-10 April 1987). Twenty one specimens
were found sheltering under a single granite flake (c 35 cm x 40 cm) on granite bed rock, 50 m
NW of the lighthouse on 7 April 1987. Another nine geckos were sheltering under a slightly
larger slab in the same region. Another granite slab, partially embedded in sandy soil in the
centre of the island, sheltered three geckos. An additional four geckos were found individually
sheltering under stones or building debris around the keeper’s house.

The SV and tail length of 29 geckos were measured, and the state of regeneration of their
tails noted. The development of eggs or enlarged endolymphatic sacs (ELS) in the neck region
of females was also recorded. Adult males are easily sexed by the prominent hemipeneal bulge
at the base of the tail. The sex ratio was 0.69 (11:16 M:F). Very few juveniles were found. No
females had obvious well-developed eggs but 5 had well-developed ELSs. Craye (1976) notes
that females produce two eggs in spring and that ELS development occurs in late winter to meet
the stress of rapid calcification of eggs-shells in spring. Two geckos lost their tails during capture.
Of the remaining 27 geckos 14 (51.85%) had regenerated tails. Craye (1976) observes that tail
autotomy is well-developed in the marbled leaftoed gecko but gives no details of frequency. The
number of Dassen Island geckos with regenerated tails, in a situation where many of the normal
predators (e.g. snakes) are absent is relatively high. However, the spotted gecko on St Croix
Island also has a very high tail break frequency, that may be caused by pecks from penguins
defending their nest sites.

The date of introduction of the angulate tortoise, Chersina angulata to Dassen Island is
undocumented. In numerous early references to the island (from 1601 onwards) there is no
mention of tortoises (Skead, 1975). This contradicts the conclusion of Brooke and Prins (1986)
that the species is “probably indigenous”. The first record of tortoises occurs as an appendix in
Rose (1929), who notes that “Testudo angulata is plentiful on Dassen Island, having probably

217

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

been introduced somewhat recently from the mainland.” Sclater (1896) makes no reference to
tortoises on the island during a visit in 1896, which indirectly supports Rose’s comment. It is now
known to be present in high densities on the island.

Studies by Apps (1983) and Berruti (1986) on the impact of feral cats on the the island’s
fauna revealed only minimal predation on the reptiles. Small cats were found to occasionally
take Scelotes gronovii (<1% prey mass), but no tortoise or gecko remains were present in
numerous scats or gut contents studied. The feral cats on the island have recently been
eradicated (Berruti, 1986) to protect the endangered jackass penguin colony.

Marcus Island, Saldanha Bay (32° 02'S, 17° 58'E; 650 m x 200 m, 11 ha, 1.2 km offshore,
7.3 m a.s.l., usually manned)

Brooke and Crowe (1982) record Phyllodactylus porphyreus from the island but give no
further details. The record postdates the connection of the island to the mainland by a causeway
in 1977. It is possible that the gecko colonised the island via this land connection.

Malgas Island, Saldanha Bay (33° 03'S, 17° 55'E, 300 m x 300 m, 9 ha, 800 m offshore, 7
m a.s.l., manned)

The only herpetological record from the island is that of two adult and one juvenile
Phyllodactylus porphyreus collected 25 January 1951 by the Swedish Expedition to southern
Africa (FitzSimons, 1957).

Jutten Island, Saldanha Bay (33° 05'S, 17° 57'30"E, 1550 m x 650 m, 46 ha, 800 m offshore,
60.5 m a.s.l., manned)

FitzSimons (1943) records Phyllodactylus porphreus and Cordylus cordylus from the island
and both species were also collected there by the Swedish Expedition to southern Africa
(FitzSimons, 1957). Brooke and Crowe (1982) also record Acontias meleagris. Mouton (1987)
notes that the population of the girdled lizard on the island is referable to the melanistic form
(niger).

Meeuw Island, Saldanha Bay (33° 05'S, 18° 00'30"E, 500 m x 300 m, 7 ha, 140 m offshore,
9 m a.s.l., not manned)

McLachlan (in litt. to Skead, April 1971) records Phyllodactylus porphyreus and Scelotes
gronovii from the island, whilst Brooke and Crowe (1982) also record Phyllodactylus lineatus.

Schaapen Island, Saldanha Bay (33° 06'S, 18° OTE, 650 m x 600 m, 41 ha, 500 m offshore,
18 m a.s.l., not manned)

Lichtenstein in 1803 noted that “the inhabitants of this as well as the other islands are chiefly
sea-fowl, serpents and lizards” (quoted by Skead, 1975), and later Symons (1926) recorded that
”... Schapen has a very bad reputation for snakes, cobras and puff-adders being said to be very
plentiful”. However, this has not been confirmed recently, and Grindley (in litt., Skead, April
1971) found that “Some years ago 1 did some work on Schaapen island in Saldanha Bay where
I found Cordylus cordylus to be common and the burrowing lizard Acontias meleagris to be
present.” Similarly, Brooke and Crowe (1982) record the lizards Phyllodactylus lineatus,
Scelotes gronovii and Mabuya capensis on the island. The Scelotes gronovii are of typical size (N
= 6, SV 60-74 mm, mean 68.83 mm), not noticeably larger than those on the adjacent mainland.

Pomona Island, Namibia (27° 12'S, 15° 16'E, 91 m x 366 m, 3 ha, 200 m offshore, 2.5 m
a.s.l., not manned)

This small, barren guano island lacks vegetation or permanent water. Nonetheless, Bogert

218

BRANCH: HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

(1940) records a single specimen of the Coral snake Aspidelaps lubricus lubricus from the island.
Mertens (1954 and 1955) transferred this specimen to his newly-described race, A. 1. infuscatus,
subsequently followed by FitzSimons (1962). However, Broadley (1983) does not list this
specimen in his revision although he makes no further comment on its identity or provenance.
It should be noted that Bogert (1940), in his original description of the specimen, records that
“The stomach of this snake contained a small unidentified rodent and nineteen eggs, presumably
of some lizard. The eggs measure approximately 10 mm x 6 mm, with soft shells, typical of
many species of gecko”. No southern African gecko has soft-shelled eggs and no geckos have
been recorded from Pomona Island. The small size and barren nature of the island make it a
most unlikely habitat for this nocturnal, semi-fossorial elapid and it is probable that the
specimen was collected elsewhere.

Loveridge and Williams (1957) record an angulate tortoise from Pomona, based on a
specimen in the AMNH. It is not known whether this specimen came from the adjacent
mainland or the island to which it might have been taken by a guano worker. It is pertinent that
whereas FitzSimons (1950) lists a number of other specimens from Pomona it is obvious from his
text that he refers to the mainland adjacent to Pomona Island in Liideritz District. Specimens of
the angulate tortoise are still found around Liideritz Bay, probably having escaped from
captivity. The nearest natural population occurs 220 km to the south at Oranjemund (Branch,
1989).

Possession Island, Namibia (27° OTS, 15° 12'E; 5.6 km x 0.8 km, 90 ha, 2.7 km offshore,
20 m a.s.l., manned)

This is the largest of the guano islands off the Nambian coast. It has a sparse scrub cover.
Werner (1910) recorded one male and three female marsh terrapins {Pelomedusa subrufa) from
the island but queried the record. Haacke (1965) discounted the record, noting that the small
island contains no standing water and that the original collector (L. Schultze) made no reference
to this unusual discovery in his account of his journey. During a two day visit to the island (May,
1963) Haacke found no evidence of the marsh terrapin but did note the presence of two angulate
tortoises “brought from the Cape as pets”. The permanent resident supervisor also noted that
“the only wild reptiles ever noticed were the odd skink and gecko found amongst timber or
empty bags from Cape Town” (Haacke, 1965).

RECENT GEOLOGICAL HISTORY OF THE ISLANDS

All the offshore islands discussed in this paper occur on the continental shelf and are
relatively close to the South African mainland (maximum distance 9 km offshore). All have
been affected, to a greater or lesser degree, by sea level fluctuations associated with periods of
glaciation. Tankard (1976) has reviewed sea level fluctuations in the region during the Cenozoic.
He notes a rapid fall in sea-level with the advance of the final Wiirm glaciation (17-18 000 BP)
and a minimum sea level 130 m lower than present. At this time all of the South African offshore
islands would have been connected to the mainland. Following the maximum glacial advance,
the rate of retreat of the ice was nearly constant. The present level was reached about 6500 BP.
This infers a sea-level rise of some 1.125 m per 100 year for the period 18 000 — 6500 BP. The
islands are separated from the mainland by water of different depths. Most channels are under
20 m deep. The maximum depth (approx. 40 m) occurs between Ysterfontein and Dassen Island
(which is also the farthest offshore). This island would thus have separated from the mainland

219

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

about 14-13 000 years BP, whereas the islands with the shallowest channels (i.e. those of
Saldanha Bay, and Pomona (9 m) and Possession (12.8 m)) would have separated between
8-7000 BP. The Algoa Basin islands are separated from the mainland by water of depths of
17-23 m and would thus have separated 10-9000 BP.

During the last climatic optimum or hypsithermal, which ended c 2000 years ago, sea levels
were at least 3 m higher than at present (Flemming, 1977). At this time many of the low-lying
islands (e.g. Bird Island, Algoa Bay; Dyer Island; and Pomona Island, Namibia) even if still
exposed, would have been seriously affected by storm swells. They now either lack a
herpetofauna or have one only recently introduced.

INTRODUCED SPECIES

The larger islands have a longer and more intense history of human habitation and there has
therefore been a greater chance for species introductions, be they deliberate or accidental.
Tortoises are the only group for which evidence of deliberate introduction is known: Chersina
angiilata. Dyer Island, after 1924; Bird Island (Algoa Bay), 1951-3; Possession Island, 1960-3;
Geochelone pardalis, Robben Island, 1959. It is highly likely that the colony on Dassen Island
was deliberately introduced between 1896 and 1929.

It is difficult to determine which of the remaining species on the offshore islands represent
accidental introductions. It is possible that many of the species recorded from Robben Island
were present naturally before colonization by European settlers. The early records confirm the
presence of snakes, chamaeleons, lizards and tortoises. Mouton and Oelofsen (1988) present a
model explaining the distribution of melanistic girdled lizards in the region and note the
importance of mountains as refugia for melanistic populations during the current amelioration
of the climate following the last Glacial Maximum (16 000 BP). It is proposed that the rapid
warming of the climate along the coastal lowlands since 14-12 000 BP allowed the typical form
to rapidly expand its range and to colonize Robben Island before it became separated from the
mainland. Whether at this period the species also managed to colonize St Croix Island in Algoa
Bay is not known. The species is common in many coastal regions of the Cape Province
(Burrage, 1974), particularly in the southwestern and Eastern Cape.

The same events probably allowed the burrowing skinks Acontias meleagris (Robben,
Schaapen and Jutten islands), Scelotes gronovii (Robben (?), Dassen, Schaapen, and Meeuw
islands) and S. bipes (Robben Island) and Rose’s rain frog Breviceps rosei (Robben Island) to
colonise a number of the islands on the southwestern Cape coast. It is extremely unlikely that
so many islands could subsequently be colonised by small, fossorial lizards and a terrestrial,
burrowing frog following separation from the mainland. It is equally unlikely that they could
reach the islands by natural rafting or be accidentally introduced by man. They are therefore
likely to represent relict populations, isolated following the rise in sea-level. The comments in
Haacke (1965), concerning the introduction of geckos and skinks to Possession Island with cargo
from Cape Town, offer a probable explanation for the presence of Phyllodactylus porphyreus on
so many islands in the southwestern Cape and for the presence of the Cape skink Mabuya
capensis on Dyer Island. The marbled leaftoed gecko is common throughout the coastal regions
of the southern Cape and is very common in human habitations (Branch, 1988b). The
introduction of geckos to ports in ship cargo is well-established (e.g. Branch, 1987; Loveridge,
1961).

220

BRANCH: HERPETOFAUNA OF OFFSHORE ISFANDS OF SOUTH AFRICA AND NAMIBIA

When discussing Robben Island, it was noted that no permanent freshwater source existed
on the island until the colony was established by Van Riebeeck in 1652. It is thus likely that the
spotted stream frog Strongylopm grayii was introduced during the period of human settlement
and that the brown water snake Lycodonomorphus nifidus which feeds on amphibians was also
subsequently introduced.

The slight rise (3 m) in sea-level during the last hypsithermal (Flemming, 1977) may account
for the current absence of reptiles on Bird Island in Algoa Bay, Dyer Island (where the
herpetofauna is probably all introduced), Marcus and Malgas islands in Saldanha Bay (where the
only record for both islands is the marbled leaftoed gecko which has probably been introduced),
and Pomona Island, Namibia. All are low-lying islands and, although they would still have been
exposed during the hypsithermal, they would have been drastically affected by storm swells.

Among the species that have been introduced to the offshore islands, a number have been
translocated even further distances. The small stream frog, Strongylopus grayii has become
established on St Helena (Frost, 1985). Loveridge (1947) notes a record of the ocellated gecko
Pachydactylus geitje from Ascension Island but does not mention it in subsequent papers on the
introduced herpetofauna of the island (Loveridge, 1959 and 1961). The specimens may have
been mislabelled or the population may subsequently have died out.

ISLAND-SPECIES RICHNESS RELATIONSHIPS

The islands for which a herpetofauna has been identified were studied to determine whether
a relationship between species richness and island size/distance from the mainland exists. Marcus
Island was excluded from the analysis as it was connected by a causeway to the mainland in early
1977. The only record from the island (the gecko Phyllodactylus prophyreus) (Brooke and
Crowe, 1982) occurs after the causeway was constructed. The species could have invaded the
island naturally after connection to the mainland. The islands of Dyer, Pomona and Possession
have also been excluded because in all cases the recorded reptiles have probably either been
deliberately or accidentally introduced, or mistakenly assigned to the island. The angulate
tortoise on Dassen Island and the leopard tortoise on Robben Island are introduced species and
have been excluded from the following anlaysis.

Linear and log-log simple regression models (Sokal and Rohlf, 1969) were computed to test
statistical relations between island herpetofauna richness and island area and distance offshore.
Log/log plots for species/island area and species/distance offshore revealed no significant
correlation. Brooke and Crowe (1982) found a highly significant relation between South African
offshore island area and species richness for the total biota and also for various higher taxonomic
categories (vertebrates, birds and plants). They found no significant relation in these categories
between island species richness and island distance from the mainland. The absence of any
significant correlation for the herpetofauna can be attributed to a number of factors: a) the small
size of the islands (only two exceed 100 ha in extent); b) the depauperate herpetofauna which
on most islands is restricted to 1-3 species; c) the low mobility of many reptiles and amphibians;
d) the low salinity tolerance of most amphibians; and e) the lack of freshwater and thus suitable
breeding sites for most amphibians.

RAFTING AND THE EFFECT OF FLOODS ON REPTILES AND AMPHIBIANS

It is often proposed that species may reach offshore islands by rafting. Arnold (1976) has
discussed the history of fossil reptiles on Aldabra Atoll which has been inundated by rising sea

221

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

levels on at least two occasions. The giant land tortoise Geochelone elephantina has colonised the
atoll on three occasions and it is probable that it originated from the northern areas of
Madagascar. This large tortoise may either enter the sea voluntarily to cool itself or is washed
out to sea by floods. It floats easily in water and could have drifted northwards in sea currents
to wash ashore on Aldabra. This ability is well-illustrated by the discovery of a large leopard
tortoise, G. pardalis, alive and well, drifting in the sea about 4 km offshore between Robben
Island and the mainland (D'Ewes, 1967). The small land tortoises of South Africa, including the
angulate tortoise Chersina angulata and the genera Homopus and Psammobates, are unable to
swim (Branch unpubl. obs.) and can only colonise islands by rafting on flotsam in association
with floods.

The effects of floods on reptiles and amphibians have rarely been documented but some
indication of their impact may be judged by the following record. During the period 22-24
August 1971 the Port Elizabeth region experienced heavy rains and the Gamtoos, Swartkops
and Sundays rivers flooded. Afterwards local newspapers recorded the capture of 93 snakes
sheltering among debris on the beaches at the river mouths. The common slug eater Duberria
lutrix was the most common snake. Other reptiles caught included the puffadder Bins arietans\
sand snakes Psammophis spp.; the skaapsteker Psammophylax rhombeatus; the night adder
Causus rhombeatus-, house snakes Lamprophis spp.; the garter snake Homoroselaps lacteus', the
aurora house snake Lamprophis aurora, the Cape cobra Naja nivea-, and at least 100 unidentified
tortoises (Ross, in lift, to Skead, Sept 1971).

Although no tortoises are currently found on either of the major islands in Algoa Bay,
following heavy rain on the mainland in 1981 an angulate tortoise, Chersina angulata, was found
freshly dead on the shore of Bird Island (A. Batchelor, pers. comm.). It is possible that the two
lizard species on St Croix Island in Algoa Bay could have been introduced to the island by rafting
following floods in the rivers of the adjacent mainland. The spotted gecko Pachydactylus
maculatus regularly shelters under bark on dead logs in the region (Branch, 1989), as does
Tasman’s girdled lizard Cordylus tasmani, a mainly arboreal cordylid endemic to the Algoa
Basin (Branch and Braack, 1987). The Cape girdled lizard Cordylus cordylus is common among
rocks along the Eastern Cape coast (Branch, 1988d).

ACKNOWLEDGEMENTS

The historical base for this study draws heavily on the unpublished studies on the offshore
islands of southern Africa by Jack Skead (doyen of South African historical biogeography); the
author thanks him for permission to use his data. He thanks Ernst Baard (herpetologist,
CDNEC, Jonkershoek Research Station) for allowing him to refer to his results given in an
unpublished report on a recent survey of Robben Island. The Department of Sea Fisheries,
South Africa, provided the author with transport to and from Dassen Island, and gave
permission for him to also visit Bird Island and St Croix in Algoa Bay. The author thanks the
South African Museum (Graham Avery), Transvaal Museum (Wulf Haacke), and CDNEC
(Ernst Baard) for details concerning preserved material in their care.

REFERENCES

Apps. P. J. 1983. Aspects of the ecology of feral cats on Dassen Island, South Africa. S. Afr. J. Zool. 18 (4); 393-399.
Arnold, E. N, 1976. Fossil reptiles from Aldabra Atoll, Indian Ocean. Bull. Br. Mus. nai, //«r. (Zoology) 29(2); 85-116.
Arnold, E. N. 1988. Caudal autotomy as a defence. In: Gans, C. and R. Huey eds. Biology of the reptilia, 16: Ecology
B. New York: Alan R. Liss, pp. 235-274.

222

BRANCH; HERPETOFAUNA OF OFFSHORE ISLANDS OF SOUTH AFRICA AND NAMIBIA

Baard, E. H. W, de ViLLiERS, A. L. and Steyn, M, E. 1986. Verslag oor in herpetologiese opname van Robbeneiland,
Kaapstad. In: Natuurbewaring ondersoek van Robbeneiland. In: Coetzee, D. J., and J. A. van Zyl eds,
unpublished report. Chief Directorate Nature and Environmental Conservation, CPA.

Baard, E. H. W. 1988. Scelotes gronovii: species account. In: Branch, W. R. ed. South African Red Data Book-reptiles
and amphibians, S. Afr. Nat. Sci. Prog. Rpt. 151: 154-155.

Berruti, a. 1986. The predatory impact of feral cats Felis catus and their control on Dassen Island. S. Afr. J. Antarct. Res.
16(3): 123-127.

Bogert, C. M. 1940. Herpetological results of the Vernay Angola Expedition, with notes on African reptiles in other
collections. Part 1. — Snakes, including an arrangement of African Colubridae. Bull. Am. Mus. nat. Hist. 77(1):
1-107.

Boulenger, G. a. 1910. A revised list of the South African reptiles and batrachians, with synoptic tables, special reference
to the specimens in the South African Museum and descriptions of new species. Ann. S. Afr. Mus. 5: 455-538.

Branch, W, R. 1987. Introduced reptiles in the Addo Elephant National Park. Koedoe 30: 165.

Branch, W. R. 1988a. Marine reptiles. In: Lubke, R. A., F. W. Gess and M. N. Bruton eds, A field guide to the Eastern
Cape Coast. Grahamstown: Grahamstown Centre of the Wildlife Society of Southern Africa, pp. 115-119.

Branch, W. R. 1988b. Field guide to the snakes and other reptiles of southern Africa. Cape Town: Struik.

Branch, W. R. (ed.) 1988c. South African Red Data Book-reptiles and amphibians. S. Afr. Nat. Sci. Prog. Rpt. 151: i-vi
+ 1-241.

Branch, W. R. 1988d. Terrestrial reptiles and amphibians. In: Lubke, R. A., F. W. Gess and M. N. Bruton eds, A field
guide to the Eastern Cape Coast. Grahamstown: Grahamstown Centre of the Wildlife Society of Southern
Africa, pp. 251-264.

Branch, W. R. 1989. The Angulate tortoise, Chersina angulata. In: Swingland, I. and M. W. Klemens eds. The
conservation biology of tortoises. Occ. Pap. lUCN Species Survival Commision (SSC) 5: 68-71.

Branch, W. R. and Braack, H. H. 1987. Reptiles and amphibians of the Addo Elephant National Park. Koedoe 30:
61-111.

Broadley, D. G. 1983. FitzSimons’ snakes of southern Africa. Johannesburg: Delta Books.

Brooke, R. K. and Crowe, T. M. 1982. Variation in species richness among the offshore islands of the southwestern Cape.
S. Afr. J. Zool. 17(2): 49-58.

Brooke, R. K. and Prins, A. J. 1986. Review of alien species on South African offshore islands. S. Afr. J. Antarct. Res.
16(3): 102-109.

Burrage, B. R. 1974. Population structure in Agama atra and Cordylus cordylus cordylus in the vicinity of De Kelders,
Cape Province. Ann. S. Afr. Mus. 66(1): 1-23.

Bustard, R. 1963. Notes on the eyed gecko {Pachydactylus geitje) with special reference to incubation. Copeia 1963(2):
433-434.

Cooper, J. and Brooke, R. K. 1986. Alien plants and animals on South African continental and oceanic islands: species
richness, ecological impacts and management. In: MacDonald, I. A. W., F. J. Kruger and A. A. Ferrar eds. The
ecology and management of biological invasions in southern Africa. Cape Town: Oxford University Press, pp.
133-142.

Craye, G. 1976. The ecology of the South African gecko, Phyllodactylus porphyreus (Daudin) in the south-western Cape
Province, with special reference to its calcium metabolism. Hons, thesis. Zoology Department, University of
Cape Town.

De VILLIERS, S. A. 1971. Robben Island: out of reach, out of mind. Cape Town: Struik.

D’Ewes, D 1967. Wayward naturalist. Cape Town: Timmins.

Fitzsimons, V. F. M. 1943. The lizards of South Africa. Transv. Mus. Mem. 1: 1-528.

Fitzsimons, V. F. M. 1950. Notes on a collection of reptiles and amphibians from the west coast of southern Africa. Ann.
Transv. Mus. 21(3); 253-259.

Fitzsimons, V. F. M. 1957. Reptilia. In: Hanstrom, B., P. Brink and G. Rudebeck eds, South African Animal Life 4.
Stockholm: Almquist and Wiksell, pp. 385-405.

Fitzsimons, V. F. M. 1962. Snakes of southern Africa. Cape Town; Purnell.

Flemming, B. W. 1977. Distribution of recent sediments in Saldanha Bay and Langebaan Lagoon. Trans, roy. Soc. S. Afr.
42(3 and 4): 317-340.

Frost, D. R. (ed.) 1985, Amphibian species of the world, Lawrence; Allen Press.

Haacke w. d. 1965. additional notes on the herpetology of South West Africa with descriptions of two new subspecies of
geckos. Cimbebasia no. 11: 1-39.

Hewitt, J. 1920. Notes on the fauna of St Croix Island. S. Afr. J. nat. Hist. 2(1): 98-112.

Hewitt, J. 1937. A guide to the vertebrate fauna of the Eastern Cape Province, South Africa. Part 2. Reptiles, amphibians
and freshwater fishes. Grahamstown: Albany Museum.

Hughes, G. R. 1974a. The sea turtles of south-east Africa. 1 . Status, morphology and distributions. Investl Rep. Oceanogr.
Res. Inst. 35: 1-144.

Hughes, G. R. 1974b. The sea turtles of south-east Africa. 2. The biology of the Tongaland Loggerhead turtle Caretta

223

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 10, MARCH 1991

caretta L. with comments on the Leatherback turtle Dermochelys coriacea L. and the Green turtle Chelonia
mydas L. in the study region. Investl. Rep. Oceanogr. Res. Inst. 36: 1-96.

Loveridge, a. 1947. Revision of the African lizards of the family Gekkonidae. Bull. Mas. comp. Zool. Harv. 98(1):
1-469.

Loveridge, A. 1959. Notes on the present herpetofauna of Ascension Island. Copeia 1959(1): 69-70.

Loveridge, A. 1961. An East African gecko colonizing Ascension Island. J! E. African nat. Hist. Soc. 23: 196-191 .
Loveridge, A. and Williams, E. E. 1957. Revision of the African tortoises and turtles of the suborder Cryptodira. Bull.
Mus. comp. Zool. Harv. 115(6): 161-557.

Mclachlan, G. R. 1978. South African Red Data Book-reptiles and amphibians. S. Afr. Nat. Sci. Prog. Rpt. 23: i-vi,
1-53.

Mertens, R. 1954. Neue Schlangenrassen aus Siidwest- und Sudafrika. Zool. Anz. 152(9-10): 213-219.

Mertens, R. 1955. Die Amphibien und Reptilien Siidwestafrikas. Abh. senckenb. naturforsch. Ges. 490: 1-172.
Mouton, P. le F. N. 1987. Phenotypic variation among populations of Cordylus cordylus (Reptilia: Cordylidae) in the
south-western Cape Province, South Africa. S. Afr. J. Zool. 22(2): 119-129.

Mouton, P. le F. N. and Oelofsen, B. W. 1988. A model explaining patterns of geographic character variation in Cordylus
cordylus (Reptilia : Cordylidae) in the south-western Cape, South Africa. S. Afr. J. Zool. 23(1): 20-31.
Mouton, P. le F. N., Oelofsen, B. W. and Mostert, D. P. 1987. New data on threatened lizard species in the
south-western Cape, South Africa. S. Afr. J. Sci. 83(1): 48-52.

Mullins, A. J. 1984. Leatherback turtle, Dermochelys coriacea. breeding on the Bathurst Coast. The Naturalist 28(2):
14-15.

Poynton j. c. 1964. the amphibia of southern africa; a faunal study, ann. natal mus. 17: 1-334.

Rand, R. W. 1963. The biology of guano-producing birds. Part 5. Composition of colonies on the South West African
islands. Investl Rep. Div. Sea Fish. S. Afr. 46: 1-26.

Raven-hart, R. 1967. Before Van Riebeeck. Cape Town: Struik.

Rose, W. 1926. Some notes on the lizards of the Cape Peninsula. Ann. S. Afr. Mus. 20(6): 491-494.

Rose, W. 1929. Veld and viei: an account of South African frogs, toads, lizards, snakes and tortoises. Wynberg: Speciality
Press of South Africa.

Rose, W. 1950. The reptiles and amphibians of southern Africa. Cape Town: Maskew Miller.

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

ScLATER, W. L. 1896. A visit to Dassen Island, the home of the Jackass penguin. Ibis (7)2: 519-525.

SicEAD, C. J. 1975. The offshore islands of the South African coastline. Vols 1 and 2, unpublished manuscript notes. Port
Elizabeth Museum.

SoKAL, R. R. and Rohlf, F. J. 1969. Biometry. San Francisco: Freeman.

Symons, R. E. 1924. A visit to Dyer's Island. S. Afr. J. nat. Hist. 4(4): 253-259.

Symons, R, E. 1926. Collecting birds' eggs at Saldanha Bay. S. Afr. J. nat. Hist. 6(1): 30-36.

Tankard, A. J. 1976. Cenozoic sea-level changes: a discussion. Ann. S. Afr. Mus. 71: 1-17.

Tarr, P. W. 1987. Non-nesting emergence by Green turtles Chelonia mydas at the Cunene River Mouth, SWA/Namibia.
Madoqua 15(3): 267-8.

Thom, H. B. (ed.) 1952-1958. Journal of Jan van Riebeeck, 1651-1662. 3 vols. Cape Town: Balkema.

Werner, F. 1910. Reptilia et Amphibia. In: Leonhard Schultze, Zoologische und anthropologische Ergebnisse einer
Forschungsreise im westlichen und zentralen Sudafrika. 4. Denkschr. med. naturw. Ges. Jena 16: 279-370.

APPENDIX

Voucher specimens — AMNH, American Museum of Natural History, New York; CDNEC,
Chief Directorate of Nature and Environmental Conservation, Jonkershoek; PEM R, Port Elizabeth
Museum; SAM, South African Museum, Cape Town; TM, Transvaal Museum, Pretoria.

AMPHIBIA.

Breviceps rosei: Robben Island. SAM 2147, 2149, 2154, no date. Rev. G. H. R. Fisk. Strongylopiis
grayii: Robben Island. Poynton (1964) lists specimens from the Albany Museum, Grahamstown but
these are unlocated.

REPTILIA.

Serpentes.

Colubridae. Pseiidaspis cana: Robben Island. SAM 1725, 1882, H. W. Oakley; SAM 1847, no date.
Sir R. Southey. Lamprophis inornatus: Robben Island. SAM 3528, 31 January 1899, D. Denyssen. (?)
Lycodonomorphus nifidiis: Robben Island. SAM 1857, no date. Rev. G. H. R. Fisk.

Elapidae. Aspidelaps liibriciis infuscatus: Pomono Island. AMNH 51836, October 1925, H. Lang.

224

BRANCH: HERPETOFAUNA OF OFFSHORE ISFANDS OF SOUTH AFRICA AND NAMIBIA

Sauria.

Gekkonidae. Pachydactyliis geitje: Robben Island. SAM 157-158, 7 March 1881, H. W. Oakley; SAM
2049, December 1896, W. F. Purnell; SAM 43947-65, 10 June, G. R. McLachlan and U. van der
Heever; TM 3575, 32791. Pachydactylus maculatiis: St Croix Island. PEM R 239, 245, 250-52, 254,
262, 264; J. Spence, 8 October 1963. Phyllodactylus porphyreus: Robben Island. SAM 43979-80, 10
June 1976, G. R. McLachlan and U. van der Heever. Jutten Island. 9431-32, no date. Government
Guano Department; SAM 44827-29, 4 November 1979, R. Brooke. Dassen Island. SAM 44830, 4
November 1979, R. Brooke; SAM 44833-34, 7 November 1979, R. Brooke; PEM R 4607-4623.
Meeuw Island. SAM 44770-71, 27 April 1979, R. Brooke. Marcus Island. SAM 44776-77, 27 April
1979, R. Brooke. Malgas Island. TM 26162, 25 January 1951, Swedish Southern Africa Expedition.
Schaapen Island. PEM R 1572-73, 1575-78, 1580, 1588-89, 18 April 1965, J. Spence and G.
McLachlan. Phyllodactylus lineatus: Schaapen Island. SAM 44765-68, 27 April 1979, R. Brooke.
Meeuw Island. SAM 44772-75, 27 April 1979, R. Brooke.

Chamaeleonidae. Bradypodion puniilunv. Robben Island. FitzSimons (1943) records specimens in the
Albany Museum but these are unlocated.

Agamidae. {l)Agama atra: Dassen Island. SAM 2644, April 1897, R. M. Lightfoot.

Scincidae. Mabuya homalocephala: Robben Island. SAM 1090-91, 1881, H. W. Oakley; SAM
1420-1446, 17 March 1881, H. W. Oakley. Mabuya capensis: Dyer Island. SAM 44836, 16 June 1979,
R. Brooke. Robben Island. CDNEC 5612-13. Scelotes gronovii: Meeuw Island. SAM 44769, 27 April
1979, R. Brooke. Schaapen Island. SAM 43240, 4 May 1961, C. Gow; SAM 44764, 27 April 1979, R.
Brooke; PEM R 554, 556-57, 561-64, 566-68, 18 April 1965, J. Spence and G. McLachlan. Robben
Island. TM 35741, 16 May 1968, Dr J. M. J. Meier. Dassen Island. SAM 44835, 7 November 1979, R.
Brooke; SAM 44831-32, 4 November 1979, R. Brooke; PEM R 4624-4632, 4566-67, TM 65844,
4-10 April 1987, W. R. Branch. PEM R 4633-4634, born in captivity to previous specimens. Scelotes
bipes: Robben Island. SAM 1458-59, 1881, H. W. Oakley; SAM 1524-25, 1882, H. W. Oakley; SAM
1982-88, 21 April 1897, Dr Spencer; SAM 2025, 2027-29, 2031-40, 2042-43, no date, H. W. Oakley;
SAM 2050, December 1896, R. M. Lightfoot. Acontias meleagris meleagris: Schaapen Island. SAM
44538, 2 May 1954, University of Cape Town Ecology Survey; PEM R 2007, 2010-11, 18 April 1965,
J. Spence and G. McLachlan. Robben Island. SAM 1454-57, 1881, H. W. Oakley; SAM 2026, 2030,
2041, 2051, no date, H. W. Oakley; SAM 43937-46, 10 June 1976, G. R. McLachlan and U. van der
Heever.

Cordylidae. Cordylus cordylus: Jutten Island. SAM 9430, no date. Government Guano Department;
SAM 44825-26, 4 November 1979, R. Brooke. Robben Island. SAM 1104-06, 1880, R. Southey;
SAM 2048, December 1896, W. F. Purnell; SAM 43932-36, 10 June 1976, G. R. McLachlan and U.
van der Heever. St Croix Island. PEM R 521, J. Spence, 8 October 1963; R 1358-59, W. R. Branch,
7 July 1979. Tetradactylus seps: Robben Island. SAM 164-165, 1881, H. W. Oakley; SAM 1086-87,
1880, R. Southey; SAM 1187-1193, 7 March 1881, H. W. Oakley; SAM 1396-1419, 17 March 1881,
H. W. Oakley.

Chelonii

Eretmoclielys imbricata: Dyer Island. SAM 9403 (shell), 2 February 1906, H. Jackson, Government
Guano Department.

225

INSTRUCTIONS TO AUTHORS

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

Only generic and specific names should be underlined to indicate italics. All other marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

BK2207 — CTP Book Printers, Cape

./3

ISSN 0570-1880

Annals of the

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 11 20th June 1991

Mr 1 6 1992

J’fBfiAmzs

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS: 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

A revision of the tripterygiid fish genus Norfolkia Fowler, 1953
(Perciformes: Blennioidei)

W. HOLLEMAN

(Albany Museum, Grahamstown)

CONTENTS

Page

Abstract 227

Introduction 227

Taxonomy 228

Genus Norfolkia Fowler 228

Norfolkia brachylepis (Schultz) 233

N. sqiiamiceps (McCulloch and Waite) 236

N. thomasi Whitley 238

Acknowledgements 239

Material Examined 239

References 242

ABSTRACT

The genus Norfolkia is revised and three species are recognised: N. brachylepis (Schultz)
(= N. springeri Clarke) which is widespread throughout the tropical Indo-west Pacific,
N. sqiiamiceps (McCulloch and Waite) from the southern Barrier Reef, Norfolk and Lord Howe
islands and N. thomasi Whitley from the Barrier Reef and south-central Pacific.

INTRODUCTION

Much tripterygiid taxonomy has advanced in a rather haphazard fashion, work tending to
be regional. The first major generic revision was that of Helcogramma by Hansen (1986). Other
genera now revised include Tripterygion by Wirtz (1980) and Gilloblennius , Notochinops and
Forsterygion by Hardy (1986, 1987a, 1989). One of the problems that has beset tripterygiid
taxonomy is the recognition of characters that adequately define genera (see Hardy, 1987a, b).

Most of the 14 genera listed by Hansen (1986) and including Cremnochorites (Holleman,
1982), Karalepis (Hardy, 1984), Ceratobregma (Holleman, 1987), Bellapiscis and Cryptichthys
(Hardy, 1987b) can be defined by a combination of “stable” characters viz. the number of first
dorsal fin and and anal fin spines, the number of pelvic fin rays and the configuration of the

227

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT, 11, JUNE 1991

lateral line. Combinations of these four characters do not hold stable for all genera and in
Forsterygion and Obliquichthys (Hardy, 1987c) number of first dorsal fin spines, and in
Gilloblenniiis and Lepidoblennius number of anal fin spines vary.

Members of the genus Norfolkia are medium-sized (max. 60 mm SL) tripterygiid fishes,
widespread though not abundant throughout the Indian and western Pacific Oceans. One species
is also known from the islands of the south-central Pacific. They are known only from tropical
and subtropical waters occurring as far south as Cockburn Sound on the Australian west coast,
where the mean winter (July to October) sea surface temperature is 18 °C. All species seem to
be associated with coral or rocky reefs.

The colour pattern in preservative of the three species in the genus is rather similar with
head and fins variously speckled with melanophores and the body carrying a number of irregular
dark vertical bars. Colour photographs taken by R. Winterbottom and J. R. Allen and colour
illustrations done by Patricia Hansen show marked differences in live colours.

Kuiter (1986) included in Norfolkia three species with three first dorsal fin spines, whereas
Norfolkia was originally described with four first dorsal fin spines. There are, however, other
osteological characters shared by the “four-spined” species not shared by Kuiter’s “three-
spined” species that preclude their inclusion in the genus. The purpose of this paper is to clarify
which species belong in Norfolkia as well as to provide an adequate definition of the genus.

All measurements were taken as in Hubbs and Lagler (1958). Fin element counts follow
Rosenblatt (1960) and Springer (1968). Snout profile (= snout angle) was measured as described
in Holleman (1982). The method of counting vertebrae follows Hardy (1986, 1987a, b, c) who
considers the first vertebra with a well developed, unforked haemal spine as the first caudal
vertebra. Comparative morphometric data for the species are shown in Fig.l and Table 1.

The following acronyms are used;

AMS - Australian Museum, Sydney

BPBM - Bernice P Bishop Museum, Honolulu

CAS - California Academy of Sciences, San Francisco

LACM - Natural History Museum of Los Angeles County

ROM - Royal Ontario Museum, Toronto

RUSI - JLB Smith Institute of Ichthyology, Grahamstown

USNM - United States National Museum, Smithsonian Institution

WAM - Western Australian Museum.

TAXONOMY

Genus Norfolkia Fowler, 1953

Norfolkia Fowler 1953:262. Type species N. lairdi Fowler 1953 = Gillias squamiceps McCulloch
and Waite 1916:449, by original designation.

Diagnosis

First dorsal fin with four spines; anal fin with two spines; pelvic fin with two long rays united
by a membrane for less than a quarter of their length or not at all; lateral line discontinuous,
anterior series of pored scales ending below second dorsal fin and posterior series of notched
scales from below end of anterior series to base of caudal fin. Orbital and anterior nasal tentacles
present.

228

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

staedard deviation
standard error

Fig.l. Selected meristic and morphometric data for the three species of Norfolkia.

Body and head behind and below eyes heavily scaled; scales ctenoid except for a few cycloid
scales on pelvic fin bases, abdomen and around vent.

Margins of lateral ethmoids and frontals “crenulate”. Septal bone (sensu Springer and
Freihofer, 1976) unossified to partially ossified; vomer with single row of teeth; palatines
edentate. First dorsal fin proximal pterygiophores in front of first vertebra. One or no free
pterygiophore between second and third dorsal fins; caudal skeleton with two epurals and a
substantial hypural 5; short broad neural spine on second pre-ural (penultimate) vertebra.

Description

Small fishes with fusiform bodies and pointed snouts; maximum recorded standard length
57 mm. First dorsal fin with 4 spines, second with 12-15 spines, third with 8-11 undivided
segmented rays, except the last which is divided to its base. Anal fin with 2 short spines, the first
supported by its own pterygiophore, the second sharing a pterygiophore with first of 17-21
undivided segmented rays; the last ray also always divided to its base. Last dorsal and anal fin
rays supported by a pterygiophore and stay (Fig. 2). Pelvic fins with one short hidden spine and
two undivided segmented rays connected by a membrane for part of their length. Pectoral fins
with 14-16 rays, the upper 2 or 3 undivided, the lower 7 undivided and thickened, remainder
bifurcate. Caudal fin with 7 + 6 principal rays, uppermost and lowermost one or two undivided,
remainder divided once; 6-7 dorsal and 4-6 ventral undivided procurrent rays.

Body, head behind eyes, cheeks to lower margin of eye and pectoral fin bases covered with
ctenoid scales. Ctenii large and even-sized, posterior margins of scales slightly raised. Scales on
abdomen and around vent smaller and cycloid.

229

Table 1. Fin and lateral line counts of Norfolkia species

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 11, JUNE 1991

230

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

Fig. 2. Last dorsal and anal fin elements of N. squamiceps. p - distal ptcrygiophore; s - stay; r - divided ray; c - cartilage.

Anterior lateral line series with 12-24 pored scales extending as far as the junction of the
second and third dorsal fins; posterior series with 14-27 notched scales running from three scale
rows below end of anterior series to base of caudal fin. Orbital and anterior nasal tentacles
present. Teeth in jaws and on vomer conical and fixed; palatines edentate. Septal bone
unossified to partially ossified. Margins of lateral ethmoids and frontals “crenulate” to greater
or lesser degree (Fig. 3). Posterodorsal margin of post-temporal bone serrate. One or two spines
lost from second dorsal fin, resulting in no or one free pterygiophore between second and third
dorsal fins. Precaudal vertebrae 10 or 11 (rarely 12), caudal vertebrae 24-27. Caudal skeleton
with two epurals, a large hypural 5 and a short neural spine on the pre-ural centrum.

Discussion

There are only two tripterygiid genera with four first dorsal fin spines, two anal fin spines
and a discontinuous lateral line: Norfolkia and Cremnochorites . These two genera were shown
to be clearly distinct by Holleman (1982). Norfolkia spp. lack the denticle-like head scales,
palatine teeth and completely ossified septal bone of Cremnochorites. In Norfolkia the septal is
usually unossified but some cleared and stained specimens of N. brachylepis and N. squamiceps
show partial ossification, starting from the posterior margin. The cephalic lateralis canals of
Cremnochorites are closed (the bone forms a ‘tube’ with pores), whereas those of Norfolkia are
open (the bone is an open channel).

The crenulate margins of the lateral ethmoids are curious, giving the impression that the
bone has been eroded away, leaving irregular remnants of bone embedded in the tissue (Fig. 3a).

231

ANN. CAPE PROV, MUS. (NAT, HIST.) VOL, 18, PT. 11, JUNE 1991

Fig. 3. a: Lateral view of anterior of cranium N. brachylepis showing "crenulate” margin of lateral ethmoid and partially
ossified septal, b: enlarged portion of lateral ethmoid margin shown in 3a. c: portion of lateral ethmoid and
frontal of N. squamiceps. f-frontal; 1-lateral ethmoid; s-septal.

232

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

Sometimes the remnants resemble thin spines, parallel sided and pointed, the tips protruding
through the skin (Fig. 3b), The orbital margin of the frontals has similar “crenulations” or
“spines”.

This character is autapomorphic for Norfolkia, it not having been observed in any other
tripterygiid species, or in any other blennioid fish. Norfolkia shares with Cremnochorites,
Forsterygion and Apopterygion Kuiter a serrated posterodorsal margin on the post-temporal
(Hardy, 1989). This serrated margin is also present in '''Norfolkia'’ clarkei (= Tripterygion
striaticeps, see Kuiter, 1986). Hardy (1989) suggested — and this author concurs — that the
phylogenic significance of this character is unclear and it may well have been independently
derived by these four genera (and whatever genus "N". clarkei is referable to).

Only three species are here recognised: Norfolkia brachylepis (Schultz), N. sqiiamiceps
(McCulloch and Waite) and N. thomasi Whitley. The three species included by Kuiter (1986) in
Norfolkia have only three spines in the first dorsal fin and lack the distinctive “crenulate”
margins of the lateral ethmoids and frontal bones. They are consequently not referable to the
genus.

The distribution of the three recognised species of Norfolkia is shown in Fig. 4.

Key to the species of Norfolkia

1 (a) Second dorsal fin spines 15 (rarely 14); anal fin rays 21 2

(b) Second dorsal fin spines 12-15, usually 13; anal fin rays 18-20 N. brachylepis

2 (a) Third dorsal fin rays 11; pored scales in anterior lateral line series 21-24 N. squamiceps

(b) Third dorsal fin rays 10, rarely 9 or 1 1 ; pored scales in anterior lateral line series 12-15 . . . . N. thomasi

Norfolkia brachylepis (Schultz) (Fig. 5)

Tripterygion brachylepis Schultz 1960:291, type locality, Bikini Atoll, Marshall Islands.
Norfolkia springeri Clark 1979:95, type locality Gulf of Elat, Red Sea.

Description

Dorsal fins IV + XII-XV -I- 9-11, usually IV + XIII + 10 (except for specimens from
southern western Australia and New South Wales — see below); first dorsal fin lower than
second. Anal fin II + 18-20, usually II -H 19 rays; pectoral fins 15 or 16, usually 16 rays, upper
2 simple, lower 7 simple and thickened, remainder divided once. Caudal fin with 5-7 dorsal, 5
ventral procurrent rays. Pelvic fin rays not united by a membrane. Lateral line anterior series
14-18, usually 16-17 pored scales, ending below last 3 spines of second dorsal fin; posterior
series of 18-23 notched scales from third scale below end of anterior series, overlapping it by 1-2
scales, to base of caudal fin. Transverse scales 3/8; total lateral scales 31-34, usually 32 or 33.

Precaudal vertebrae 11 (rarely 12), caudal vertebrae 24 rarely 23 (as a result of fusion of
PU2 and PU3) or 25. Pleural ribs 7, epipleural ribs 11; no intramuscular bones. One or two
dorsal fin spines lost, leaving no or one free pterygiophore between second and third dorsal fins.

Head profile slightly rounded, interorbital area concave. Nasal bone not bridged or
narrowly bridged. Orbital and nasal tentacles finely serrated. Scales on head extend to below
lower margin of eye. Pectoral fin bases heavily scaled; abdominal scales cylcoid, extending to
base of pelvic fins. Mandibular pores 4+1 + 4 (Fig. 6). Head 3, 0-3, 7 in SL; eye 2, 9-3, 8 in head;
upper jaw 2, 1-2,8 in head; snout 3, 3-4,7 in head; snout profile 58°-70°.

233

ANN, CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 11, JUNE 1991

◄

't'

234

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

Fig. 5. Norfolkia hrachylepis, female, 33,0mm SL, from AMS 120770-012.

Colour

In preserved specimens there are six dark bars running diagonally from the dorsum to the
midline below which they divide to form a series of 10 or more squarish blocks. Anal fin is
distinctly barred, the bars running diagonally forward from the base of the fin. Caudal fin with
broad dark bars, leaving narrow clear bands between the bars. Pectoral fins spotted, dorsal fins
dusky. Head unevenly covered with melanophores with a relatively distinct bar running from
below mid-eye to behind corner of the mouth.

Patricia Hansen made a colour painting of freshly dead material from Japan in 1978, which
shows the second and third dorsal fins suffused with deep red, the body a light brown. Colour
photographs of N. hrachylepis published by Allen (1985:2447) show a mottled orange head, a
light brown flecked body with chocolate-brown bars with red flecks and reddish brown bars on
the second and third dorsal fins. (Allen's Plates 415 and 416 are possibly male and female of the
same species.)

Comparisons

Norfolkia hrachylepis can easily be distinguished from the other two Norfolkia species by
generally lower counts of second and third dorsal and anal fins (see Fig.l and Table 1) and by
the distinctly banded anal fin. Some specimens of both N. thornasi and N. squamiceps do show
light and irregular banding on the anal fin.

Distribution

In a revision of the South African Tripterygiidae (Hollernan, 1978) Norfolkia specimens
from Natal were found to be conspecific with N. springeri described from the Red Sea by Clark
(1979). This was confirmed when the author saw the types of N. springeri at the National
Museum of Natural History in 1979. This study shows that the Natal material is conspecific with
the material from the eastern Indian and Pacific Oceans. N. springeri is thus the junior synonym
of N. hrachylepis.

The distribution of N. hrachylepis is far more extensive than that of either N. squamiceps
or N. thornasi (Fig. 4). Its habitat ranges from coral reefs in tropical waters to rocky reefs in warm
temperate waters around Perth, Western Australia.

235

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 11, JUNE 1991

N. brachylepis also exhibits a greater variation in the number of dorsal and anal fin elements
than do the other two species. Where there are 13 second dorsal fin rays there is one free
pterygiophore between second and third dorsal fins; where there are 14 there is no free
pterygiophore. The most common combination is 13 second dorsal fin spines, 10 third dorsal fin
rays and 19 anal fin rays. The last dorsal ray is always associated with the 8th last vertebra and
the last anal fin ray with the 7th last vertebra.

Norfolkia squamiceps (McCulloch and Waite) (Fig. 7)

Gillias squamiceps McCulloch and Waite 1916: 449 (original description, based on three
specimens from Lord Howe Island).

Norfolkia lairdi Fowler 1953: 264 (type species of the genus, based on one (?) specimen from
Norfolk Island).

Description

Dorsal fins IV + XIV-XV + 10-11, usually IV + XV + 11; first dorsal fin lower than
second. Anal fin 11 + 20-21 rays, usually 21; pectoral fin 16 rays, upper 2-3 simple, lower 7

Fig. 6. Mandibular pore patterns of the three species of Norfolkia.

236

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

Fig. 7. Norfolkia sqtiamiceps, female, 57,5mm SL, from AMS If7368-0f9.

simple and thickened, remainder bifurcate. Caudal fin with 6 or 7 dorsal and 5 or 6 ventral
procurrent rays; pelvic fin rays united by membrane for less than one quarter of the length of the
short ray; lateral line anterior series 21-24, usually 22, pored scales, ending below the junction
of the second and third dorsal fins; posterior series 14-19, usually 16, notched scales, third scale
row below anterior series, from end of anterior series which it may overlap by as many as 4
scales, to base of caudal fin. Transverse scales 4/8; lateral scales 32-35, usually 34.

Precaudal vertebrae 11, caudal vertebrae 24-26, usually 25. Pleural ribs 9, epipleural ribs
14, 3 intramuscular bones. One dorsal fin spine lost, no free pterygiophore between second and
third dorsal fins.

Interorbital concave. Nasal bones long and narrow and narrowly bridged. Palmate nasal
and orbital tentacles present; margins of orbits with many small spines and margins of lateral
ethmoid and frontal bones heavily “crenulate”. Opercles and pectoral fin bases heavily scaled
with ctenoid scales; abdomen with cycloid scales extending to base of pelvic fins. Mandibular
pores 5-6 + 1 + 5-6 (Fig. 6). Head 2, 9-3, 6 in SL; eye 3, 0-3, 9 in head, upper jaw 2, 3-3,0 in
head, snout 3, 3-4, 9 in head. Snout profile 54°-62°, mean 58°.

Colour

No live or freshly caught specimens have been seen. In preservative all Norfolkia species
show remarkably similar pattern of body and fin markings and bars below the eyes.

Light coloured specimens of N. squamiceps show five irregular dark bars along back which
fade toward the midline and disappear below it. All fins except the pelvics have narrow irregular
bars (these are often absent on the anal fin); first dorsal fin usually darker than others, may lack
the barring except on first spine. There is a short, dark bar below and one behind the eye on the
preopercle. The orbital tentacles are dark. In small specimens pigmentation is more intense and
patterning clearer. In dark specimens body bars become obliterated by an even scatter of
melanophores over the body and head. The dorsal fins are very dark; anal fin very dark, without
any barring, and with the margin unpigmented; caudal and pectoral fins retain light, irregular
bars; pelvic fins remain unpigmented. Head is darker but retains bars below and behind eyes.

237

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 11, JUNE 1991

There appears to be no sexual dichromatism but it is reasonable to presume that the males,
like those of many other tripterygiids. will show darker breeding coloration.

Comparisons

N. squamiceps can be distinguished from N. thomasi by its longer anterior lateral-line series
(22 vs 13 pored scales) and concomitantly shorter posterior series of notched scales (16 vs 25),
scales on head which extend to below the eyes and cover the opercula, and a blunter snout (58°
vs 51°). The species can be distinguished from N. hrachylepis by higher dorsal and anal fin
counts.

Distribution

N. squamiceps appears to have a limited distribution range (Fig. 4). It was described from
specimens from Lord Howe and Norfolk Islands. Subsequent material has been collected at
these two islands, the Swain Reefs and One Tree and Heron islands, southern Queensland.

Norfolkia thomasi Whitley (Fig. 8)

Norfolkia thomasi Whitley 1964: 192.

Description

Dorsal fins IV -t- XIV-XV + 9-11, usually IV -t- XV + 10; first dorsal fin lower than second.
Anal fin II + 20-22, usually 21; pectoral fin 14-16, usually 15 rays; upper 1-2 simple, lower 6-7
simple and thickened, remainder bifurcate. Caudal fin with 6 dorsal and 5 ventral procurrent
rays; pelvic fin rays not united by membrane. Lateral line anterior series 12-15, usually 14,
pored scales, ending below spines 9-11 of second dorsal fin; posterior series 24-27, usually 25
notched scales, from third scale row below anterior series, overlapping it by 0-1 scales, to base
of caudal fin; transverse scales Ys lateral scales 34-36, usually 35. Precaudal vertebrae 11, caudal
vertebrae 26. Pleural ribs 4, epipleural ribs 7, intramuscular bones 3. Two dorsal fin spines lost
leaving one free pterygiophore between second and third dorsal fins.

Fig. 8. Norfolkia thomasi, male, 30,Umm SL, from USNM 238854.

238

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

Head flattened, snout long and sharp. Fine “toothing” on margin of frontals and lateral
ethmoids. Interorbitai flat to slightly concave, with ascending processes of premaxillae extending
well beyond midpoint of upper margin of eyes. Large lobate orbital and anterior nasal tentacles
present. Scales on head extend only to lower margin of eye; pectoral hn bases scaled; abdominal
scales cycloid, extending to base of pelvic hns. Mandibular pores 3 + 2 + 3 (Fig. 6). Head 2,9-3, 2
in SL; eye 3, 5-4, 9 in head, upper jaw 2, 1-2,5 in head, snout 3, 0-4, 2 in head. Snout prohle
48°-57°, mean 51°.

Colour

No live or freshly dead specimens have been seen. There are six more or less distinct dark
bars running diagonally back from the dorsum on to the midline. Below the midline they tend
to coalesce and fade. In small specimens markings are more distinct and below midline a number
of irregular, narrow bands are formed. A series of dark spots at base of anal hn extend as light
banding on hn. Other hns irregularly banded except for first dorsal hn which is dark. Head
evenly dusted with melanophores except for single darker band extending from lower margin of
eye diagonally backwards over preopercle. Orbital tentacle dark.

No sexual dichromatism apparent.

Comparisons

See under N. squamiceps.

Dislribiition

N. thomasi is known from the southern Oueensland (Heron and One Tree islands). Lizard
Island in the north of the Barrier Reef (and thus presumably in between) from the Fiji Islands,
and from the Cook (Rarotongla), Society (Tahiti) and Rapa (Rapa Iti) islands (Fig. 4). In the
light of Springer’s ( 1982) work on shore hsh distribution, this species’ distribution across the
western margin of the Pacihc plate to the Cook, Society and Rapa islands is unexpected. There
is, however, no reason to consider the specimens from these islands to be of a different species.
It is not known whether N. thomasi occurs in New Caledonia or the New Hebrides, respectively
marginally off and marginally on the Pacihc Plate.

ACKNOWLEDGEMENTS

The author is grateful to the institutions listed in the introduction for the loan of material
for a very protracted time, particularly to Doug Hoese, the Australian Museum, Jerry Allen,
Western Australian Museum, Rick W'interbottom, Royal Ontario Museum and Jack Randall, B
P Bishop Museum. Particular thanks go to Phil Heemstra for useful comments on the
manuscript.

MATERIAL EXAMINED

Norfolkia brachylepis

Western Indian Ocean

BPBM XXXXX (7, 27,0-43,0); Sodwana Bay, Zululand, Natal.
RUSI 8703 (7, 30,4-40,0); Sodwana Bay, Zululand, Natal.
RUSI 8704 (2, 31,4 & 40,3); Sodwana Bay, Zululand, Natal.
RUSI 8705 (2, 38,8 & 44,5); Boteler Point, Zululand, Natal.

239

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 11, JUNE 1991

RUSI 8706 (5, 28,9-32,7); Sodwana Bay, Zululand, Natal,

RUSl 8707 (5, 27,9-41,0); Sodwana Bay, Zululand, Natal,

RUSI 8708 (16,8); Sodwana Bay, Zululand,

RUSI 8709 (2, 27,6 & 31,1); Sodwana Bay, Zululand, Natal,

RUSI 8710 (6, 19,7-40,8); Hulley Point, Zululand, Natal,

RUSI XXXXX (32,6); Inhaca Island, Mozambique,

RUSI 7386 (34,0); Shimoni, Kenya,

USNM 231788 (2, 33,0); Anjouan Island, Comoro Islands,

Central Indian Ocean

USNM 231789 (37,8); Trincomalee, Ceylon,

USNM 231787 (22,9); Fort Frederick, Trincomalee, Ceylon,

USNM 231783 (23,0); Kaddukulam Pattu, Trincomalee, Ceylon,

USNM 231765 (3,20,7-40,0); Fort Frederick, Trincomalee, Ceylon,

USNM 231758 (2,18,5 & 21,5); Trincomalee, Ceylon,

Eastern Indian Ocean

WAM P26610-016 (2, 29,8 & 30,3); Smith Point, Christmas Island,

WAM P26122-012 (3, 27,9-31,2); Nortwest Point, Christmas Island,

WAM P26656-001 (43,7); Shark Bay, Western Australia,

WAM P26657-016 (2, 36,3 & 38,5); Shark Bay, Western Australia, WAM P26664-009
(42,6); Shark Bay, Western Australia.

WAM P26670-014 (6, 17,0-46,8); Dirk Hartog Island, Western Australia.

WAM P26662-014 (4, 37,6-47,0); Shark Bay, Western Australia.

WAM P276 16-032 (2, 33,1 & 37,6); Rottnest Island Western Australia.

WAM P25374-012 (14, 27,4-46,3); nr. Tantabiddi Creek, North west Cape.

WAM PXXXXX-XXX (6, 33,3-42,2); Garden Island, nr Perth, Western Australia.
WAM P26671-016 (3, 37,2-47,2); Shark Bay, Western Australia.

AMS 120229-038 (38,5); Carnac Island, Cockburn Sound, Western Australia.

Western Pacific Ocean

USNM 227496 (29,5); nr. Ch’uan-fan-shih, Taiwan.

USNM 231769 (8, 20,0-35,5); nr. Ch’uan-fan-shih, Taiwan.

USNM 231763 (11, 17,3-42,3);' Chin-chia-wan, Taiwan.

USNM 231773 (4, 27,2-32,5); Mao-Pi T’ou., Taiwan.

USNM 231762 (30,8); Ch’nan-fan-shih, Taiwan.

USNM 231770 (31,4); Ch’uan-fan-shih, Taiwan.

USNM 231781 (4, 18,2-37,5); Sha Toa, Taiwan.

USNM 231778 (31,5); nr Ch’uan-fan-shih, Taiwan.

USNM 231775 (12, 25,0-41,2); nr. Ch’uan-fan-shih, Taiwan.

USNM 231759 (2, 30,0 & 30,5); nr. Ch’uan-fan-shih, Taiwan.

USNM 231767 (13,2); Chin-chia-wan, Taiwan.

USNM 231760 (34,5); nr. southern tip of Taiwan.

USNM 231772 (3, 18,2-31,4); nr. southern tip of Taiwan.

USNM 231764 (4, 26,9-33,3); Ch’uan-fan-shih, Taiwan.

USNM 231790 (4, 19,9-34,2); Taganayan Island, Palawan, Philippine Islands.

240

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

USNM 231780 (34,0); nr. Maloh, Negros Oriental, Philippine Islands.

USNM 231792 (30,0); Cuyo Island, Palawan, Philippine Islands.

USNM 231791 (4, 24,2-30,5); Cuyo Island, Palawan, Philippine Islands.

USNM 231774 (44,5); Miyake Jima, Izu Islands, Japan.

USNM 231777 (19,2); Miyake Jima, Izu Islands, Japan.

USNM 231784 (24,3); Paeowai Island, Madang Harbour, Papua — New Guinea.

USNM 231768 (28,7); Mare, Loyalty Islands.

AMS 120756-010 (25,0); Great Detached Reef, Raine Island, Queensland.

AMS 120770-012 (3, 25,8-34,3); South Island, Sir Charles Hardy Island, Queensland.
ROM 38215 (32,5); Queensland.

WAM P27065-016 (6, 23,5-31,2); Bagara, Kelly’s Beach, Queensland.

WAM P27075-018 (3, 34,5-42,8); Julian Rocks, Byron Bay, New South Wales.

WAM P27082-014 (3, 35,0-41,0); Solitary Island, New South Wales.

WAM P27076-020 (5, 27,9-41,3); Julian Rocks, Byron Bay, New South Wales.

Norfolkia squamiceps

AMS IA937 (38,5 mm SL); Lord Howe Island.

AMS IB3544 (40,7 mm SL); Heron Island, Queensland.

AMS IB6161 (41,8 mm SL); Gillett Cay, Swain Reefs, Queensland.

AMS 117368-019 (32. 18,9-57,0 mm SL); Lord Howe Island.

AMS 120205-053 (32,2 mm SL); One Tree Island, Queensland.

AMS 120270-014 (9, 25,6-43,0 mm SL); Philip Island, Norfolk Island.

AMS 120271-035 (3, 40,7-45,0 mm SL); Bumbora Beach, Sydney Bay, Norfolk Island.
AMS I 21451-003 (33,8 mm SL); One Tree Island, Queensland.

LACM 39986-37 (4, 19,5-35,0 mm SL); Heron Island, Queensland.

USNM 231782 (9, 19,4-39,0 mm SL); One Tree Island, Queensland.

USNM 231761 (3, 23,0-27,3 mm SL); Heron Island, Queensland.

USNM 231776 (3, 35,0-40,9 mm SL); Heron Island, Queensland.

USNM 231779 (5, 17,3-35,0 mm SL); One Tree Island, Queensland.

Norfolkia thomasi

Holotype — AMS IB 4020 (37,4 mm); Gillett Cay, Swain Reefs, Queensland. (Fowler
(1953:193) gives the number as IB 4040)

Non-types; Queensland

AMS I 19108-105 (5:17,2-28,9 mm); Bird Islet, Lizard Island, Queensland.

AMS I 19338-026 (4:17,6-40,3 mm); One Tree Island, Queensland.

AMS I 20201-019 (5:14,0-35,0 mm); Heron Island, Queensland.

AMS I 20557-005 (2:18,0-212,2 mm); One Tree Island, Queensland.

AMS I 20774-013 (3:22,0-33,3 mm); Cape Melville, Queensland.

AMS I 21422-125 (2:35,1 & 35,5 mm); Lizard Island, Queensland.

AMS I 22634-011 (33,9 mm); Escape Reef North, Queensland.

LACM 39985-58 (4:34,1-37,4); Heron Island, Queensland.

USNM 231766 (36,2 mm); One Tree Island, Queensland.

USNM 231771 (40,5 mm); One Tree Island, Queensland.

USNM 231786 (42,0 mm); One Tree Island, Queensland.

241

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 11, JUNE 1991

WAM P 24735 (33,7 mm); Lizard Island, Queensland.

Western Pacific

ROM 56791 1 (26,5 mm); Dravuni, Astrolabe Reef, Fiji Islands.

ROM 56792 1 (27,0 mm); Dravuni, Astrolabe Reef, Fiji Islands.

ROM 56793 1 (25,2 mm); Vanuakula, Kadavu, Fiji Islands.

ROM 56794 1 (24,8 mm); Astrolabe Reef, Fiji Islands.

USNM 238852 (26,6 mm); Totoya Island, Fiji Islands.

USNM 238853 (33,7 mm); Navutu Ira Island.

LfSNM 238854 (2:24,5 & 30,1 mm); Malamala Island.

USNM 238855 (2:24,9 & 29,5 mm); Totoya Island, Fiji Islands.

South Central Pacific

BPBM 17196 (3:30,5-31,7 mm); Haurei Bay, Rapa Iti.

BPBM XXXXX (35,0 mm); Raratonga, Cook Islands.

CAS 38838 (32,3 mm); Tahiti.

ROM 38215 (32,3 mm); Rarotonga, Cook Islands.

REFERENCES

Allen, G. R. 1985. Fishes of Western Australia. Pacific Marine Fishes Book 9, eds. W. E. Burgess and H. R. Axelrod.
TFH. Neplune City, NJ.

Allen, G. R., D. F. Hoese, I R. Paxton, J. E. Randall, B. C. Russell, W. A. Starck II, F. H. Talbot and G. P.

Whitley. 1976. Annotated check list of the fishes of Lord Howe Island. Records of the Australian Museum 30
(15): 365-454.

Clark, E. 1979. Red Sea Fishes of the Family Tripterygiidae with descriptions of Eight New Species. Israel Journal of
Zoology 28:65-113.

Fowler. H.W. 1953. On a Collection of Fishes made by Dr Marshall Laird at Norfolk Island. Transactions of the Royal
Society of New Zealand 81 (2): 257-267.

Hansen, P. E. Hadley. 1986. Revision of the Tripterygiid Fish Genus Helcogramma . Including Descriptions of Four New
Species. Bulletin of Marine Science 38 (2): 313-354.

Hardy, G. S. 1984. A New Genus and Species of Triplefin (Pisces: Family Tripterygiidae) from New Zealand. National
Museum of New Zealand Records 2 (16): 175-180.

Hardy, G. S. 1986. Redcscription of Gillohlennius Whitley and Phillipps, 1939 (Pisces: Tripterygiidae), and description
of a new genus and two new species from New Zealand. Journal of the Royal Society of New Zealand 16 (2):
145-168.

Hardy, G. S. 1987a. Revision of Aotoc/ino/w Whitley, 1930 (Pisces: Tripterygiidae), and description of a new species from
New Zealand. Journal of the Royal Society of New Zealand 17 (2): 165-176.

Hardy, G. S. 1937b. Revision of some Triplefins (Pisces: Tripterygiidae) from New Zealand and Australia, with
description of two new genera and two new species. Journal of the Royal Society of New Zealand 17 (3):
253-274.

Hardy, G. S. 1987c. Descriptions of a new genus and two new species of tripterygiid fishes from New Zealand. National
Museum of New Zealand Records 3 (5): 47-58.

Hardy, G. S. 1989. The genus Forsterygion Whitley & Phillipps, 1939 Pisces: Tripterygiidae) in New Zealand and
Australia, with descriptions of two new species. Journal of Natural History 23: 491-512.

Holleman, W. 1978. The Taxonomy and Osteology of Fishes of the Family Tripterygiidae (Perciformes: Blennioidei) of
South Africa. Llnpublished Masters Dissertation, Rhodes University, Grahamstown.

Holleman, W. 1982. Three new species and a new genus of tripterygiid fishes (Blennioidei) from the Indo-West Pacific
Ocean. Annals of the Cape Provincial Museum (natural History! 14 (4): 109-137.

HotLEMAN, W. 1987. Description of a new genus and species of tripterygiid fish (Perciformes: Blennioidei) from the
Indo-Pacilic, and the reallocation of Vauclusella acanthops Whitley, 1965. Cyhium 11 (2): 173-181.

Hubbs, C. L. and K. F. Lagler. 1958. Fishes of the Great Lakes Region. Cranford Institute of Science Bulletin 26: 1-213.

Kuiter, R. H. 1986. A new genus and three new species of Tripterygiid fishes of Australia’s South coast. Revue franqaise
d’Aquariologie 12 (3): 39-96.

McCulloch, A. R. and E.R. Waite. 1916. Additions to the Fish Fauna of Lord Howe Island. Transactions of the Royal
Society of South Australia 40: 437-451.

242

A REVISION OF THE TRIPTERYGIID FISH GENUS NORFOLKIA FOWLER

Rosenblatt. R. H. 1960. A Revisionary Study of the hlennioid Fish Family Tripterygiidae. Unpublished Doctoral
Dissertation. University of California, Los Angeles.

Schultz, L.P. 1960. Fishes of the Marshall and Marianas Islands. Vol. 2. Smithsonian Institution United Slates National
Museum Bulletin 202:281-.^00.

Springer. V.G, 1968. Osteology and classification of the Fishes of the Family Blenniidae. Bulletin of the United States
National Museum 284:1-80.

Springer, V. G. and W. C. Freihofer. 1976. Study of the monotypic hsh family Pholidichthyidae (Perciformes).
Smithsonian Contributions to Zoology No. 216:1-43.

WHirLEY, G.P. 1964. Fishes from the Coral Sea and the Swain Reefs. Records of the Australian Museum 26 (5): 145-195.
WiRTZ, P. 1980. A Revision of the eastern Atlantic Tripterygiidae (Pisces: Blennioidei) and Notes on Some West African
Blennioid Fish. Cyhium 3 serie (11): 83-101.

243

SMITHSONfAN INSTITUTION LIBRARIES

3 9088 01429 5588

INSTRUCTIONS TO AUTHORS

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

Only generic and specific names should be underlined to indicate italics. All other marking
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
nomenclature or the International code of botanical nomenclature, as applicable.

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 captions). 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 submitted on a separate sheet.

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

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.

Jlibb, R. a. 1967. Freshwater fishes of southern Africa. Cape Town: Balkema.

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

BK2300 — CTP Book Printers, Cape

ISSN 0570-1880

Annals of th^

Cape Provincial Museums

Natural History

Ann. Cape Prov. Mus. (nat. Hist.)
Volume 18 Part 12 27th June 1991

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.

Parts are obtainable from the Librarian of the Albany Museum. Correspondence in connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS; 1978-

Editorial Assistant
Mrs S. K. GESS: 1980-

bk2301

Some aspects of the ethology of five species of Eumenidae (Hymenoptera)

in southern Africa.

F. W. GESS and S. K. GESS*

(Albany Museum, Grahamstown)

*(The order of names is alphabetical and joint authorship should be understood. The same
applies to previous papers on the ethology of Hymenoptera by the same authors.)

CONTENTS

Page

Abstract 245

Introduction 245

Ethological accounts 246

Allepipona erythrospila (Cameron) 246

Antepipona scutellaris Giordani Soika 249

Euodynerus euryspilus (Cameron) 253

Rhynchium marginellum sabulosum (Saussure) 256

Tricarinodynems guerinii (Saussure) 260

Discussion 267

Acknowledgements 269

References 269

ABSTRACT

Ethological accounts are given for hve eumenids, two excavating nests in horizontal ground,
Allepipona erythrospila (Cameron), Antepipona scutellaris Giordani Soika, and three nesting in
pre-existing cavities above ground, Euodynerus euryspilus (Cameron), Rhynchium marginellum
sabulosum (Saussure), and Tricarinodynems guerinii (Saussure). Associated insects, megachilid
bees, chrysidids, bombyliids and meloids are noted.

INTRODUCTION

Gess (1981) published some aspects of an ethological study of the aculeate wasps and the bees
of a karroid area in the vicinity of Grahamstown. This paper was based on a survey of the
aculeate wasps and bees of a farm Hilton, 18 kilometres WNW of Grahamstown (33° 19'S,

245

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

26° 32'E) in the Eastern Cape Province of South Africa. The location, topography, geology,
climate and vegetation of the study area were outlined. An annotated list of 241 species arranged
on the basis of their ethology was given. There followed 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.

Those species the ethology of which added significantly to the knowledge of the ethology of
the genera or even families to which they belong were selected as the subjects for a series of
papers published in the Annals of the Cape Provincial Museums (nat. Hist.) (1974-1988). There
remains a considerable volume of heldnotes on species the ethology of which is unknown or
poorly known in published accounts but which, though adding to the knowledge of the genera
and families to which they belong, offer nothing of particular note. The accounts of some aspects
of the ethology of five eumenids, Allepipona erythrospila (Cameron), Antepipona scutellaris
Giordani Soika, Euodynerus euryspilus (Cameron), Rhynchium marginellum sabulosum
(Saussure), and Tricarinodynerus guerinii (Saussure), presented here are in this category. The
justification for their publication is that this basic information is required as background to a
project on nest parasites currently being undertaken by A. J. S. Weaving of the Albany
Museum.

The order in which the species accounts are presented is alphabetical and does not imply
relationships.

ETHOLOGICAL ACCOUNTS

Allepipona erythrospila (Cameron)

Geographic distribution

Allepipona erythrospila (Cameron) has been recorded from the Cape Province, the Orange
Free State and Lesotho (Giordani Soika, 1987).

The farms Brak Kloof, one of the two type localities of A. erythrospila, and Hilton, the
locality of the present study, adjoin one another.

Description of nesting sites

Two nests of A. erythrospila were investigated at Hilton. Both were situated in level areas
of denuded clayey soil, one above and one below a furrow (Gess, 1981: Fig. 6), near (137 cm in
one instance) the edge of shallow puddles resulting from recent rain.

Flight period

The flight period of A. erythrospila at Hilton, as deduced from twenty-three specimens
collected, starts in October and continues into March. The two instances of observed nesting
were on 20. xi. 1973 and 27.iii.1974.

246

bk2301

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

Flower visiting

The authors are aware of only one record of A. erythrospila visiting flowers: one female on
the yellow flowers of Berkheya heterophylla (Thunb.) O. Hoffm. (Asteraceae), Strowan,
Grahamstown, 12.x. 1972, F. W. Gess.

Identification of prey and location of sting sites

The single prey obtained was a caterpillar of the family Pyralididae. It was olive-green with
longitudinal paler stripes and pale setal warts, and was 18,3 mm in length.

Sting sites were present in the region of the neck and on the second abdominal segment.

Description of nest

One of the two nests of A. erythrospila consisted of a short burrow surmounted by a collar
constructed from mud pellets and of the same inner diameter as the shaft. The collar was similar
in appearance to that of Parachilus insignis (Saussure) though smaller in diameter.

1 he second nest, when discovered, lacked a collar. It appeared that there had probably
been a collar which had been destroyed by the trampling of sheep. The burrow of this nest
consisted of a short curved shaft slightly dilated at its lower end to form a cell. Provisioning was
still in progress and so it is not known what the nature of the seals would have been.

Method of construction of nest, oviposition and provisioning

Water is required for burrow excavation. It is collected from a nearby pool and carried to
the nest in the crop. When filling her crop, an activity commonly observed, a wasp stands on the
mud at the edge of the water.

Observation of the second nest began at 1 1 .30 am by which time a short burrow had already
been excavated. Consequently initiation of the nest and construction of the turret were not
observed but will almost certainly have followed a similar pattern to that described for
Parachilus insignis (Gess and Gess, 1976).

Further pellets extracted from the burrow were discarded in a distinct pellet dropping area.
This was situated on the far bank of the puddle at a distance of 190 cm from the nest. The wasp
always left the nest in a fixed direction and returned from a fixed direction, the flight path to the
pellet dropping area and back to the nest having been more or less constant and in the form of
an ellipse.

In only one instance was the number of pellets formed with the aid of a single crop-full of
water recorded. In that instance the number was nine.

During burrow excavation the wasp always entered head first (Fig. la and b) and exited
backwards, bearing a pellet (Fig. Ic). At about 12.30 pm, an hour after the start of observations,
the burrow and the cell at its end had been completed and the last pellet was carried away and
dropped.

On returning from dropping this pellet the wasp alighted next to the nest entrance as usual
but then turned around and backed down the burrow (Fig. Id) in order to oviposit in the empty
cell. The egg, 2,2 mm long and 0,68 mm wide at mid-length, was suspended from the ceiling of
the cell by a filament about 0,6 mm long.

Oviposition having taken place provisioning commenced. The wasp emerged from the
burrow (Fig. le) and flew off, returning after ten minutes with a long caterpillar held beneath her
body by her mandibles and legs, the prey’s head being directed forwards. On alighting next to

247

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

Fig. 1. Allepipona erythrospila (Cameron) , Hilton, 27.iii.1974: (a) about to enter nest under construction; (b) entering
nest under construction; (c) About to fly off with mud pellet held between jaws; (d) Entering nest backwards in order to
oviposit in cell; emerging from nest after ovipositing; (f) Entering nest with prey, (x c 2,5)

248

bk2301

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

the nest entrance the wasp retained her hold of the prey with her mandibles and, with her legs
straddling the prey, entered the burrow (Fig. If).

The wasp was captured upon re-emerging from the nest which was then investigated. It is
therefore not known whether any additional caterpillars would have been used to provision this
cell or whether any further cells would have been constructed.

Antepipona scutellaris Giordani Soika

Geographic distribution

Antepipona scutellaris Giordani Soika has to date been recorded only from the Cape
Province (Giordani Soika, 1985).

The farm Hilton, the locality of the present study, is one of the type localities of
A. scutellaris.

Description of nesting sites

At Hilton A. scutellaris has year after year been found nesting in aggregations in the floor
of a sandpit (Fig. 2). Within the sandpit it selects sites where the complete removal of the
overlying sandy alluvial soil has exposed the underlying subsoil which, although itself of a sandy
nature, is very fine and compact and contains a considerable clay factor making it malleable
when mixed with water. The selected sites in addition are bare and weathered, not freshly
exposed.

Two further nests were found at Hilton on the earthen bank of a furrow at some little
distance from the sandpit in an area of clayey soil.

Flight period

The flight period in the Grahamstown area, as ascertained from Malaise trap catches made
at Hilton (1970-71 and 1975-76), at Table Farm (1970-71) and at Belmont Valley and
Howison’s Poort (both 1971-72), is from the beginning of November until the end of March.
Nesting is well advanced by the second week in November and continues throughout the flight
period.

Flowers and young plant growth visited

A. scutellaris was captured on the flowers of Lasiospermum bipinnatum (Thunb.) Druce
(Asteraceae) growing on the bank of a furrow which passes above the sandpit (3.xi.l977, 1 male;
10. xi. 1977, 2 females and 1 male; 15. xi. 1977, 1 female, all F.W.Gess) and on the flowers of
Selago corymbosa L. (Selaginaceae) growing in the sandpit (1 female, 2.xii.l977, R.W.Gess).

Identification of prey and location of sting sites

A. scutellaris at Hilton was found to prey exclusively upon larval Pyralididae. Forty-one
prey were recovered from nests. Two species were taken: one, with longitudinal reddish stripes
and five pairs of prolegs, was represented by thirty-eight specimens; the other, without definite
markings and with only two pairs of prolegs (on the sixth and on the last abdominal segements),
was represented by three specimens. Neither species could be identified beyond the family.
Other than for three prey of the more common species, which were partly devoured and were
represented by shrunken remains, all the caterpillars were alive and responded to tactile
stimulation. In length the thirty-eight measurable prey varied between 3,0 and 10,0 mm
(average 6,3 mm; 76% between 4,1 and 7,0 mm).

249

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12. JUNE IWl

Fig. 2. Sandpit, Hilton, 16. xi. 1975; figures at site of nesting aggregation of Antepipona sciitellaris Giordani Soika.

Eighty sting sites, indicated by dark lesions, were located ventrally on thirty-seven prey and
were found to be distributed as follows: thirty-three in the region of the neck, two on the
mesothorax posterior to the legs, five on the metathorax anterior to the legs, one on the
metathorax between the legs, thirty-three on the metathorax posterior to the legs, two on the
first abdominal segment and four on the second abdominal segment. The favoured target areas
for stinging are clearly the neck and the metathorax posterior to the legs. Stings in other but
nearby areas probably represent cases in which the target areas have been missed.

Description of nest

The nest of A. sciitellaris consists of a subterranean burrow surmounted by a mud turret
which is a vertical flared or funnel-shaped tube (Figs 3 and 4). The burrow consists of a shaft of
the same diameter as the base of the turret terminating in a single cell or branching to terminate
in two cells (Fig. 4 a-d). No nest was found with more than two cells. However, as none of the
nests was sealed it is possible that further cells might have been constructed had the wasps’
activities not been terminated by the destruction of their nests.

Method of construction of nest, oviposition and provisioning

Water, which is required for nest excavation, is fetched in the crop.

A nest site having been chosen, nest construction is initiated by the moistening of the
ground with water to form mud which is then excavated in the form of pellets. Pellets are laid
down around the shaft initial in such a way that the base of the turret will have the same inner

250

bk2301

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

diameter as that of the shaft, that is 4-4,5 mm (average 4,1, sample of 14). Additional pellets
are laid down in such a way that the diameter of the turret increases resulting in its being flared.
Final turret height is 5-8 mm (average 6,2 mm, sample of 13) and the distal diameter is
5-7,5 mm (average 6,3 mm, sample of 13). The outer surface of the turret is left rough but the
inner surface is smoothed.

After the completion of turret construction, mud pellets extracted in excavation are
discarded in a distinct pellet dropping area 60-90 cm from the nest. When leaving the nest with
a pellet, the female flies an elliptical path and always returns to the nest on the same side. She
will be diverted from her regular path by an intruder as she will try to chase it off. For example
a female wasp was observed making successive short downwards flights at a tiger beetle,
Cicindela brevicollis Wied. (Cicindelidae), which was 25 cm from her nest.

The main shaft which is constant in diameter descends vertically and terminates at a depth
of 53-89 mm (average 70 mm, sample of 14) in a cell 10-20 mm long, somewhat ovoid, having
a diameter greater than that of the shaft (Fig. 4a).

A cell having been excavated, the wasp enters the nest backwards and oviposition takes
place. An egg, pale straw yellow, slightly curved, 2, 9-3, 3 mm (sample of 5) and 0,8-0, 9 mm at
mid-length, is suspended from the ceiling of the cell on a short filament, 0,8 mm long.

After oviposition into the empty cell provisioning begins. In the sample up to thirteen prey
were found to have been provided per cell. Provisioning having been completed the cell is sealed
with a mud plate which may be followed by earth and a second mud plate cutting it off from the
shaft at the point where curvature began (Fig. 4c). The shaft is then continued and terminates
in a second similar cell at a depth somewhat greater than that of the first (Figs 4c and d). After
this cell has been supplied with an egg and provision it is sealed and again the seal may be
followed with earth and a second seal (Fig. 4d).

Seventeen nests were investigated, of these three nests had no cells, twelve nests each had

Fig. 3. Female Antepipona scutellaris Giordani Soika, standing on the mud turret surmounting her burrow, Hilton,

ll.xii.I975. (X c 2,6)

251

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 12, JUNE 1991

Fig. 4. Vertical plans of four nests oi Antepipona scutellaris Giordani Soika, Hilton, 12-23.ix.l975. ( x I ) Numbers indicate

sequence of construction of cells.

a single cell and two nests each had two cells. As none of the 17 nests excavated had been finally
sealed it is not known whether more than two cells may be excavated.

Sleeping and sheltering behaviour of adult wasps

Adult female wasps shelter in their nests after work ceases in the afternoon and during the
day, when weather conditions are unsuited to normal nesting activities. A nest opened up after
a period of rain was found to contain a sheltering female below a plug of water-borne earth. A
sheltering or sleeping female positions itself in the nest shaft, facing the entrance. In order to
achieve this position the female enters the nest backwards, as when entering to oviposit. As a
temporary closure is not made at the end of the working day, the presence of the wasp in the nest
serves as a protection to an egg or young larva and provision within an open cell, particularly as
it is common for there to be an open cell at the end of the working day.

Associated insects

Nest parasites were found in the second cell of both of the two-celled nests of A. scutellaris
investigated. All other cells were free of parasites.

One of the affected cells contained a second instar (caraboid) larva of the family Meloidae
(Coleoptera). Rearing the larva was attempted but unfortunately failed. The presence of the

252

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

meloid is of interest as nests of another eumenid, Parachilus insignis (Saussure), investigated at
Hilton (Gess and Gess, 1976) were found to be parasitized by a meloid, Lytta enona Peringuey,
which in its larval stage fed upon the caterpillars laid in as provision by its host for its own young.

The other affected cell contained two small foreign larvae the identity of which could not be
determined. In both cells the Antepipona larva was still alive. Rearing the larvae w'as attempted
but unfortunately failed.

Three instances of usurpation of A. scutellaris nests by a megachilid bee, Megachile
(Eutricharaea) meadewaldoi Brauns were recorded. These bees constructed cells from petals of
a violet flowered Wahlenbergia species (Campanulaceae) growing in the sandpit. This bee has
also been recorded from Hilton nesting in old or abandoned burrows of Bembecinus oxydorcus
(Handlirsch) (Gess and Gess, 1975) and Parachilus insignis (Saussure) (Gess and Gess, 1976).

Euodynerus euryspilus (Cameron)

Geographic distribution

Euodynerus euryspilus (Cameron) is widespread in southern Africa.

Description of nesting sites

No natural nests of E. euryspilus have been located. However, four nests in trap-nests were
obtained from Hilton. All the trap-nests utilized were situated at heights of 25-100 cm above the
ground and were attached to branches of the woody shrub or small tree Acacia karroo Hayne
(Leguminosae: Mimosoideae) growing in the day areas above and below the furrow and on the
flats below the furrow. Natural nesting sites will therefore almost certainly prove to be
pre-existing cavities such as abandoned beetle borings in the branches of bushes of thorn scrub
and small trees along the river banks.

Flight period

E. euryspilus flies at Hilton from October to March.

Flower visiting

There are no records of flower visiting for this species from Hilton. Both females and males
have, however, been recorded visiting the flowers of Acacia karroo Hayne (Leguminosae;
Mimosoideae) at two other localities: 2 females, 1 male, Colesberg, 16-17.1.1985, D. W. Gess;
and 1 male, Oudtshoorn, 9-12. xii. 1986, F. W. Gess.

Identification of prey and location of sting sites

The cells of the nests examined were provisioned with caterpillars belonging to the family
Pyralididae (Lepidoptera). In two nests provisioned during November 1973 the caterpillars were
identified as those of Loxostege frustalis Zeller, the Karoo Caterpillar, a serious pest of Pentzia
incana (Thunb.) Kuntze (Asteraceae). Prey caterpillars ranged in length from 12-16 mm and
numbered from 7-13 per cell, depending upon their size. They were partially paralysed and were
very tightly packed into the cells. Fourteen caterpillars were examined for sting sites. All bore
at least one sting clot on the prothorax anterior to the legs — that is in the neck region, two bore
sting clots on the mesothorax posterior to the legs, ten bore sting clots on the metathorax
(generally posterior to the legs) and five bore sting dots on the first abdominal segment. All the
sting clots were on the ventral surface.

253

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

Description of nest

The four nests of E. euryspilus in trap-nests consisted of linearly arranged cells, in three
instances preceded by a preliminary plug and in the only completed nest succeeded by a
vestibular cell and a final seal. The preliminary plugs, cell closures and the final closure were all
of mud; the cell walls were unlined.

Method of construction of nest, oviposition and provisioning

A large number of trap-nests with borings of 4,8 mm, 6,4 mm, 9,5 mm and 12,7 mm were
offered in mixed bundles in a wide variety of situations at Hilton. Four were accepted for nesting
by E. euryspilus-, three were 6,4 mm borings and the fourth was a 9,5 mm boring.

The building material introduced into the nest was in all cases a reddish clayey mud.

In three of the four trap-nests a preliminary plug had been constructed at or near the blind
end of the boring. In two of these nests this plug consisted of a thin layer of mud filling in the
corners and rounding off the end to form a smooth concave surface; in the third an empty space
10 mm long was left between the end of the boring and the 3 mm thick preliminary plug which
was separated from the first provisioned cell by a 13 mm long empty cell.

The cell walls were not lined.

A cell having been prepared for oviposition, a pale yellowish-white egg is suspended by a
thin filament from the roof of the cell at the inner end of the latter. Egg length ranged from
2, 7-3, 3 mm (average 3 mm, sample of 4) and egg width from 0,84-0,92 mm (average 0,88 mm,
sample of 4). The filament was 0,56-0,60 mm long.

Provisioning takes place after oviposition. The prey caterpillars are tightly packed into the
cell. Completed cells contained 7-13 caterpillars, depending upon size.

After provisioning has been completed the cell is sealed with a mud plate, 1-2 mm thick.
In no nest were intercalary cells present. In all nests the inner end of a succeeding cell was the
seal of the previous cell.

Only one of the nests had been completed. It contained two sealed provisioned cells and a
118 mm long vestibular cell between the outer provisioned cell and the 3 mm thick closing plug
which had been constructed just within the trap-nest opening. The two provisioned cells filled
only 22,5% of the total length of the boring. The other three nests were incomplete and
contained respectively one sealed provisioned cell and one still open and partially provisioned
cell; two sealed provisioned cells; and four sealed provisioned cells.

The measurements of the sealed cells irrespective of the sex of the wasp for which they were
constructed are summarised in Table 1.

Table 1.

Measurements of sealed cells of Euodynerus euryspilus (Cameron) in trap-nests of different sizes
irrespective of the sex of the wasp for which they were constructed.

Boring diameter (mm)

No. of cells

Range in length (mm)

mean length (mm)

6,4

15-29

23,0

9,5

10,5-15

13,6

254

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

Life history

Cell construction and cell provisioning in the four trap-nests utilized by E. euryspilus took
place during the second half of November, 1973 (2 nests), during the second half of February,
1975 (1 nest) and during the second half of February, 1977 (1 nest).

Due to parasitism and other causes no E. euryspilus were reared from the nine cells and
developmental details for the species are thus meagre.

In all cases the egg hatched about three days after oviposition and the larva began feeding
on the prey lying immediately below it at the inner end of the cell.

Sheltering behaviour

During inclement weather females were found sheltering in their nests, facing outwards.
Doubtless they also spend the night in their nests.

Associated insects

Chrysis hoplites Mocsary (as Octochrysis hoplites in Gess, 1981) (Chrysididae) was reared
from six of the nine cells of E. euryspilus.

Some observations were made on three of the cells parasitized by C. hoplites. All the cells
were provisioned during the second half of November 1973. In each cell the larva of
E. euryspilus hatched successfully and commenced feeding upon the stored provision. Within a
day or two of hatching the chrysidid larvae of about the same size as those of E. euryspilus were
seen in the cells, one each in two of the cells and two in the third cell. By the third day after their
hatching the host larvae had been killed and eaten. In the cell in which there were two C. hoplites
larvae, the killing and eating of the host larva was followed a day later by one of the chrysidid
larvae eating the other. When only one chrysidid larva remained in each cell the stored provision
was eaten.

Cocoon spinning by the three C. hoplites larvae, which took about two days, was
completed 15-17 days after the provisioning of the host’s cell and thus 12-14 days after
the hatching of the host’s larva from the egg. Two of these C. hoplites emerged as adults
during the same summer, 41 and 42 days after the date of cocoon spinning, the third
underwent diapause and emerged as an adult the following summer, 364 days after cocoon
spinning.

Toxophora australis Hesse (Bombyliidae) was reared from a single cell provisioned during
the second half of November, 1973. There was no indication of its presence until 53 days after
cocoon spinning by its host was completed when its motile pupa, aided by its strong cephalic
spines broke through the mud cell wall and progressed along the length of the nest to its opening.
Within two minutes of the pupa’s escaping from the trap-nest into the open the pupal skin was
seen to split down the back and the adult fly, a male, emerged. The wings expanded within a few
minutes. It is not clear whether the host was E. euryspilus or C. hoplites but it seems more likely
to have been the latter.

Chrysis species and Toxophora species are recorded by Krombein (1967) as parasites in the
nests of Euodynerus species in North America.

255

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

Rhynchium marginellum sabulosum (Saussure)

Geographic distribution

Rhynchium marginellum sabulosum (Saussure) is widespread and common in the
Afrotropical Region.

Description of nesting sites

R. m. sabulosum nests in pre-existing cavities above ground level in close proximity to a
supply of water and clayey soil.

At Hilton four nests were obtained, each constructed in a 9,5 mm bore trap-nest. These
trap-nests were constituents of a bundle tied together and positioned horizontally in a space in
a north facing river bank. This bank is of recently mineralized fine sand in which there are
numerous pre-existing galleries (Gess 1981, Fig. 27) which are most probably natural nesting
sites of this wasp.

In order to base the description of the nest on a larger sample a further thirty nests of
R. m. sabulosum obtained from a garden in Grahamstown are included. Eighteen were in
trap-nests of bore 6,4 mm, 9,5 mm and 12,7 mm placed horizontally 174 cm above the ground.
Eleven were in cut reeds, the culms of Arundo donax L. (Gramineae), of bore 12-19 mm. Of
these seven were 5-30 cm above ground and were constituents of a horizontally placed bundle
and four were part of a bean frame and varied in angle from 0-70° with the horizontal and in
height from 117-190 cm above the ground. As the reeds had been cut at different distances
along the internodes the available nesting cavities varied in length. Successful nests were made
in cavities from 7-25 cm in length. The thirtieth nest was in a fence post, in a 12,7 mm boring
76 cm above the ground.

Flight period

R. m. sabulosum, in Grahamstown, is actively nesting by the first week in December and
continues into April.

Flower visiting

The authors are aware of only one record of R. m. sabulosum visiting flowers: one male on
flowers of Sarcostemma viminale (L.) R. Br. (Asclepiadaceae), Kommadagga, 14. i. 1986,
R. W. Gess.

Identification of prey and location of sting sites

Over four hundred prey caterpillars were examined and all belonged to the Pyralididae.
Thirteen of these prey caterpillars were from Hilton and were unstriped and bright green. The
remainder were from the Grahamstown sample and were of two forms — one longitudinally
striped and brown and one unstriped, bright green and identical with those from Hilton. These
probably represent two species.

The number of caterpillars constituting the provision for a single cell varied according to the
size of the individual caterpillars. At the beginning of the season the provision consisted of four
or five large brown caterpillars, 14-25 mm long. As the season advanced there was a drop in the
size of this caterpillar prey and the number per cell rose to up to thirteen. From
13.i. 1975-2. ii. 1975 both the brown and the green caterpillars were present in any one cell after
which only the green caterpillars, smaller than the smaller brown caterpillars, were present and
sixteen to thirty-one were supplied per cell.

256

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

Sting sites were located in a total of thirty prey. All bore at least one sting on the prothorax
in front of the legs (that is in the neck region), most bore stings on the metathorax behind the
legs and/or on the first abdominal segment, and a few in addition bore stings on the mesothorax
behind the legs. All sting sites were situated on the ventral surface.

All prey found in the cells were alive, continued to defaecate and responded to tactile
stimulation.

Description of the nest

The nest of R. m. sabulosum is constructed in a pre-existing cavity and consists of a number
of serially arranged cells each sealed with a mud plug, frequently separated by intercalary cells
and succeeded by a vestibular cell which may be subdivided and is closed by a thick usually
layered mud plug which seals the nest entrance (Fig. 5). When the inner end of the boring is
uneven or when the boring is not circular in cross section mud is used to make modifications (Fig.
5c and d). In order that the cells should be of adequate volume their lengths vary according to
the diameter of the boring used. Furthermore cells which will cradle female wasps are larger than
those which will cradle male wasps (Table 2). The diameter of the boring in the lower ranges
becomes a limiting factor, only males being produced in 6,4 mm borings.

Table 2.

Measurements of sealed cells of Rhynchiiim marginelliim sabulosum (Saussure) in trap-nests of
different sizes, showing the relationships between cell length, 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 diam.
(mm)

No. of cells

Range in
length (mm)

Mean length
(mm)

No. of cells

Range in
length (mm)

Mean length
(mm)

6,4

25-42

34,7

9,5

16-27

21,4

22-32

27,1

12,5

15-25

19,7

20-33

25,8

Method of construction of nest, oviposition and provisioning

Nest construction is initiated by the wasp’s selecting a suitable pre-existing cavity and
cleaning out any small pieces of debris and spider spinnings. If, as in A. donax culms, the inner
end of the cavity is firm, smooth and concave no further preparation is required before
oviposition (Fig. 5a and b). However, if the inner end of the cavity is uneven, angular or rough,
the wasp first constructs a preliminary plug with mud applied directly to it (Fig. 5c) (16/22 in the
sample of trap-nests). Furthermore, if some foreign object such as extensive spider spinnings
obstructs the inner end of the boring, a preliminary plug is constructed to seal it off (Fig. 5d)
(6/22 in the sample of trap-nests).

Mud for the construction of the preliminary plug and subsequent plugs is obtained by the
wasp from a dry quarry site, the water for mixing the mud being carried there in her crop from
a selected water source. The horizontal distance from the nests to the quarry sites in the three

257

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

cases where these were located were 12 m, 7,5 m and 7,7 m respectively. In two of these cases
the water sources were also identified and were 7 m and 3,7 m from the quarry sites and 4 m
from the nests. The nest, the quarry, and the water source in each case formed the corners of
a triangle. The wasp having filled her crop with water flies with it to the quarry where she
regurgitates it and kneads it into the earth to form mud which she gathers together with her
mandibles to form a pellet against her front legs which support it from behind and below.
Supporting the pellet in this way and holding it with her mandibles she flies with it to the nest.
On alighting at the nest the pellet is held by the mandibles only. Two large pellets are made for
each crop-full of water and each is approximately 6 mm in diameter. The route taken between
the nest and the quarry is eliptical, the wasp following one long side of the curve on the outward
journey and returning along the other.

The preliminary plug, when present, having been completed the wasp oviposits. The egg is
yellow, gently curved, in average 3,6 mm in length and 1,1 mm in diameter at mid length
(sample of 6) and is suspended from the roof or side of the cell so that it hangs down on a short
filament. The egg is always positioned towards the inner end of the cell.

Oviposition having taken place hunting commences. The wasp captures a prey caterpillar
and subdues it by stinging it several times. She then transports it to her nest in flight clasped
beneath her body with her legs and held by her mandibles near its head end. The caterpillar is
orientated with its head facing the direction of travel and with its venter uppermost. Retaining
the prey in this position the wasp drags it into the nest. Successive prey are added to the cell until
it is fully provisioned. Within the cell the prey are not arranged in any set pattern, that is some
face the inner end of the cell and some the outer end. Provisioning of a cell having been
completed it is sealed with a mud plug 2, 5-5, 5 mm thick at its edges and 1-2 mm thick towards
its centre. Before proceeding with oviposition into and provisioning of another cell the wasp may
construct an empty intercalary cell.

One to three intercalary cells were present in each of 15 nests of the 25 nests having two or
more provisioned cells. The intercalary cells had been constructed singly between pairs of
provisioned cells in all but one instance in which a pair was present (Fig. 5c). The intercalary
cells ranged in length from 3-20,6 mm.

After the full number of cells has been completed the nest is sealed by the construction of
a thick, layered mud plug which is usually positioned at the entrance to the cavity leaving a
vestibular cell which may be subdivided by partitions (Fig. 5). In the sample of 25 nests the
length of the vestibular cells ranged from 5-130 mm and the plug from 5-32 mm.

Female sleeping and sheltering habits during nesting

It was found that the female of R. m. sabulosum regularly spends each night within her nest,
her head facing the open end, a position which she also takes up when weather conditions are
unsuited for normal nesting activities.

As the wasp at no time constructs a temporary closure to the nest and as the end of her
working day is determined by the temperature or light intensity or both and not by the
completion of any particular phase of nesting (such as the sealing of a cell), the retirement of the
wasp to the nest during the night and during unfavourable weather conditions not only serves to
give it shelter but equally importantly ensures the safety of the egg and of any prey caterpillars
which may be present in an unsealed cell. It was in fact common to find a wasp sleeping or
sheltering in a nest with an open cell containing an egg or an egg and a number of prey. In one
instance a wasp spent the night in very cramped quarters, straddling the prey caterpillars, in a

258

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

T crc 2

^2( r rxTfl

100 mm

Fig. 5. Vertical plans of four nests of Rhynchium marginellum sahulosum (Saussure): (a and b) nests in cut culms of
Arundo donax L. (Gramineae), Grahamstown i/ii.l975; and (c and d) nests in trap-nests, “sandstone” bank, Hilton,
21. ii. 1975, Numbers indicate sequence of construction of cells; e indicates presence of and position of egg; and i indicates

interstitial cell.

fully provisioned but unsealed cell situated so close to the nest opening that the wasp was barely
under cover. Cell closure and nest sealing were completed early the following morning. It thus
seems that this behaviour could as much be guarding behaviour as sheltering behaviour.

Life history

In the present study it was found that the time taken from oviposition to the hatching of the
egg took two to three days.

For a sample of 19 larvae developing under laboratory conditions the time taken from the
hatching of the egg to the attainment of larval maturity was five to six days, occasionally seven
days, and from larval maturity to the commencement of spinning one or two days, occasionally
three.

Pupation took place after a further eight to sixteen days and emergence 18 to 25 days later.
Males in individual nests emerged 1-5 days before females ensuring that the males in the outer
cells emerge before the females in the inner cells.

259

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

When a mature larva commences spinning it lines the cell with a layer of yellow
cellophane-like material composed of “varnished” spinnings with a characteristic aromatic
scent.

The nests from Hilton yielded four females and two males which emerged from the third
week in March to the first week in April. Those from Grahamstown yielded thirty-one females
and thirty-five males which emerged from the first week in February to the first week in April.

It seems likely that adults emerging before the end of the nesting season must nest and
produce a second generation of wasps which would overwinter in the pre-pupal stage as it is
highly unlikely that R. m. sabulosum would overwinter in the adult stage and furthermore no
adults have been recorded in the Grahamstown area after April and before December. It seems
highly likely therefore that R. m. sabulosum is bivoltine.

Sequence of sexes in nests

Certain species of aculeates which construct serially arranged cells within pre-existing
cavities commonly show a fixed sequence of sexes in nests in which both males and females
develop. This has been shown by Krombein (1967: 29) for species of Vespidae and some
Megachilidae and by Gess and Gess (1982: 165-166) for a species of Sphecidae, Isodontia
pelopoeiformis (Dahlbom). To establish whether this is also true of R. m. sabulosum an analysis
was made of 19 completed nests of at least two cells, 14 in 12,7 mm bore trap-nests and 5 in
reeds.

Of the 19 nests analysed 2 two-celled nests and 1 three-celled nest produced all males and
three yielded no information concerning the correlation between the position of the cell within
the nest and the sex of the wasp produced within that cell. In the remaining 15 nests, 5
two-celled, 5 three-celled, 4 four-celled and 1 eight-celled, females had without exception
developed within cells constructed and provisioned before those in which the males developed.
Female producing cells are therefore sited towards the inner or blind end of the nesting gallery
and male producing cells towards the outer or open end of the gallery. It follows that, if the first
cell in a nest is male producing, all subsequent cells constructed within that nest will also be male
producing.

Associated insects

No parasites of R. m. sabulosum were recorded. However, some cells in which the wasp’s
young failed to develop yielded instead foreign insects. Thus a single individual of a small species
of Braconidae was reared from each of two cells of one nest and a single individual of a small
species of Tachinidae was reared from one cell of another nest, both nests being from
Grahamstown. It seems probable that the presence of these foreign insects in the cells resulted
from the inclusion amongst the caterpillars introduced into the cells as provision of individuals
which had already been parasitized by Braconidae and Tachinidae.

Tricarinodynerus guerinii (Saussure)

Geographic distribution

Tricarinodynerus guerinii (Saussure) is widespread in its distribution in tropical and
southern Africa and appears to be able to exist in a wide range of climatic conditions. It is
represented in the collection of the Albany Museum by specimens from all four provinces of
South Africa as well as from Lesotho. The species shows considerable variation in its coloration

260

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

and this has lead some authors to split it into a number of so-called subspecies which do not,
however, appear to have any real validity. The wasps concerned in the present study are of the
colour pattern associated with the “subspecies” ruhens.

Description of the nesting sites

T. guerinii nests in suitable pre-existing cavities above ground level in close proximity to a
supply of water and clayey soil. At Hilton three nesting sites were located along the course of
the New Year’s River at heights reached by water only during times of flood. Each of the nesting
sites is different with respect to the geological nature of the material in which the nests occur and
with respect to aspect.

The most frequented site consists of a west facing bank of firmly compacted very fine sand.
Erosion of the bank has exposed the roots of various shrubs and trees growing above the water
course and has cut through a number of subterranean cavities of uncertain origin (Fig. 6).

The second most frequented site is a bank of north facing crumbly sandstone-like material
probably formed by recent mineralization of a fine sand similar to that at the first site and
referred to as the “sandstone” bank (Gess, 1981, Fig. 8).

The least favoured site is a south east facing cliff cut across the roughly horizontal bedding
of a shale formation (Fig. 7).

In the sand bank and the “sandstone” bank the pre-existing cavities utilized are burrows
made in previous years by nest excavating Hymenoptera (Fig. 9). A preference is shown for
holes in sheltered positions such as the upper parts and sides of cavities in the sandbank (Fig. 8)
and in a gap between the “sandstone” bank and a “sandstone pillar” separated from the bank
by weathering. On the shale cliff nesting is in crevices occurring within and between layers of
shale.

Nesting sites of T. guerinii have also been located by the authors at Tierberg, Prince Albert
where they were situated in sheltered positions in a recently mineralized eroded river bank and
in mud wails inside a ruined mud walled building close to the river. In both instances the gallery
initiators were solitary bees.

Empty cells of aerial mud nests of Sceliphron (Sphecidae) and Synagris (Eumenidae) are
used for nesting at False Bay, Lake St Fucia, Natal (Weaving, 1990, pers. comm.).

At Hilton bundles of trap-nests were lodged in cavities in the sandbank, in the gap in the
“sandstone” bank and in cracks in close proximity to natural nests in the shale cliff. Only one
trap-nest of 12,7 mm bore was accepted. It was sited in the shale cliff.

This low incidence of acceptance of trap-nests is not attributable to the very different nature
of the substrate as Weaving (pers. comm., 1990) found T. guerinii nesting extremely abundantly
in cut reeds used in buildings at False Bay.

Flight period

T. guerinii flies at Hilton from early November, in which month a pair in copula was
observed resting on the face of the “sandstone” bank. Nesting was observed in January and
February but undoubtedly starts earlier in the summer.

Nesting at Tierberg was in progress in late November/early December.

T. guerinii appears to be univoltine at Hilton.

Flower visiting

At Hilton both sexes of T. guerinii have been collected on flowers: Celastraceae: Maytenus

261

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

Fig. 6. Sand bank, eroded bank of very fine compacted sand, NewYear’s River, Hilton, 15.ii.f974.

Fig. 7. Cliff cut across roughly horizontal bedding of shale formation by New Year’s River, Hilton, 1974.

262

GESS & GESS; ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

Fig. 8. Close view of cavity in sand bank showing eight Tricarinodvnerus guerinii (Saussure) turrets. Hilton, 15. ii. 1974.

(X 0,25)

Fig. 9. Mud turret and nesting gallery (cut through vertically) of Tricarinodvnerus guerinii (Saussure), “sandstone" bank,

Hilton, 27. i. 1975. (x 1,3)

263

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

linearis (L. /. ) Marais, 1 female and 1 male, ll.xii.l969, F.W.Gess; and Leguminosae:
Mimosoideae: Acacia karroo Hayne, 1 male, 15. ii. 1974, F.W.Gess. Both plants were growing
near the sandbank nesting site.

Other flower visiting records from specimens in the Albany Museum collection are:

Leguminosae: Mimosoideae: Acacia caffra (Thunb.) Willd., 2 males, Oudtshoorn,
9-12. xii. 1986, F.W. Gess; 1 male, Oudtshoorn, 9-12. xii. 1986, R.W. Gess. Acacia karroo
Hayne, 1 male, Colesberg, 17. i. 1985, D.W. Gess.

Leguminosae: Papilionoideae: Calpurnia glahrata Brummitt, 1 male, Mamathes, Lesotho,
2.xi.l952, C.F. Jacot Guillarmod.

Umbelliferae: Foeniculum vulgare A.W. Hill, 1 male, Alexandria/Salem, H.W. Gess.

Rhamnaceae: Ziziphus mucronata Willd., 2 females, 1 male, Adelaide, C.F. Jacot
Guillarmod.

Identification of prey and location of sting sites

From Hilton seven specimens of prey were obtained from three open cells of nests in the
“sandstone” bank, all were small reddish caterpillars, probably Tortricidae, and ranged in
length from 8 — 11 mm. All the caterpillars were alive and responded to tactile stimulation.
Sting sites were recorded for six of the prey. All were on the ventral surface. All prey had been
stung in the region of the neck and on the metathorax posterior to the legs. In addition one had
been stung on the mesothorax anterior to the legs and another on the mesothorax posterior to
the legs.

Description of nest

The nest of T. guerinii is characterized by a downwardly curved ribbed mud entrance turret
most commonly slightly flared at the distal opening (Figs 8-13).

Within (Fig. 13) are several cells arranged, depending upon the shape of the pre-existing
cavity which forms the nesting gallery, either in a linear series (Fig. 13g) or radiating from the
inner end of an entrance burrow (Fig. 13f).

The degree of modification of the cavity is determined by its suitability for cell construction.
The cavities in the sand and in the “sandstone” are old burrows and are therefore circular in
cross section. In these very little mud is used for smoothing the walls whereas in the irregular
shaped crevices in the shale a greater amount of mud is required to round off unevennesses and
to create cells and an entrance passage.

The size of the nest is not necessarily determined by the available space as a preliminary
plug may be constructed reducing the depth of a deep cavity (Fig. 13h). Each cell is sealed by
a thin mud plug and the last cell is separated from the closing plug by a vestibular cell (Fig. 13c,
e and g). In addition to the plug closing the gallery the entrance turret is sealed at its distal
opening (Figs 12 and 13c-g) and may be divided along its length by additional mud partition
(Fig. 13g).

Method of construction of nest, oviposition and provisioning

Nest construction is initiated by the selection of a suitable cavity for use as a nesting gallery.
The wasp flies off and returns with a load of mud which she lays down either around the rim of
the entrance to the gallery or a short distance inside the entrance to the gallery. With successive
loads of mud she first constructs a foundation ring. If the entrance to the cavity is of suitable
diameter, this will be reflected in the diameter of the turret. If, however, the diameter is too

264

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

Fig. 10. Three mud turrets of Tricarinodynems guerinii (Saussure) to show curvature and surface texture, (x f ,28)

Fig. ff. Adult Tricarinodynems guerinii (Saussure) (male above, female below) and mud turret showing outer closure.

(X 1, 35)

Fig. 12. Trap-nest furnished with a mud turret of Tricarinodynems guerinii (Saussure), Hilton, 11. i. 1975. (x 0,95)

265

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

Fig. 13. Vertical plans of eight nests of Tricarinodynerus giterinii (Saussure): (a-g) showing sequence of cell construction
and provisioning of seven nests in sand bank. New Year’s River, Hilton, 15. ii. 1974 and (h) showing preliminary gallery
modifications of a 12,7 mm bore trap-nest, Hilton, i.l974.

266

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

large as was the case in the 12,7 mm bore trap-nest used for nesting, the opening will be partly
closed to reduce it to a suitable diameter, 5,5-8 mm, before turret construction proceeds (Fig.
13h). As turret construction advances the wasp carefully smoothes the inner surface of the walls
but leaves the external surface rough-cast giving the turret its characteristic ribbed and knobbly
appearance Figs 11-12). No interstices are left. Either immediately after the first ring of the
turret has been laid down or after a short horizontal section of tube has been constructed, the
structure is added to unevenly; the layering away from the bank being wider than that towards
the bank so that the tube curves over and downwards. If the tube reaches the vertical plane,
curvature ceases but further additions may be made to extend its length which in the sample was
up to 31 mm. The final section of the tube may be laid down in such a way that the rings are of
increasing diameter causing the tube to be Hared towards its distal opening.

The wasp prepares the first cell using mud to make the neccessary modifications to the
gallery. The first cell having been prepared a yellow, very slightly curved egg is laid suspended
from the ceiling on a short filament (Fig. 13b). Two of the four eggs found were measured, one
was 2,8 mm long and 0,9 mm wide at mid-length and the other was 3,1 mm long. Each was
suspended on a filament 0,2 mm long.

Hunting then commences. The wasp captures a prey caterpillar, stings it and flies to her nest
with it held by her mandibles, supported beneath by her legs and facing the direction of travel.
On reaching the nest she enters rapidly.

As the cells examined were either still open and only partially provisioned or were sealed
but contained fully grown larvae and no prey, no information is available on the number of or
weight of prey with which each cell is provisioned.

The cell having been fully provisioned it is sealed with a mud plug of an even thickness of
approximately 1 mm. Work then starts on the second cell. After the full number of cells has been
completed the nest entrance is sealed with a mud plug leaving a vestibular cell. The distal
opening of the turret is also sealed and the length of the turret may be subdivided by the
construction of partitions. The final closure and the closure of the turret are very little thicker
than the cell closures.

Life history

Egg hatch to full grown larva was observed in one instance only. The time taken was 9 days.

When a larva has consumed all its provision, it lines the cell with a parchment-like substance
composed of “varnished” spinnings so firmly attached to the cell walls and the sealing plug that
it cannot be removed without a coating of earth.

Female sleeping and sheltering habits during nesting

One instance of sheltering was observed on an overcast day. A female facing outwards was
sheltering in a turret which was furnished with a seal at its inner end.

Associated insects

Four Chrysis laminata Mocsary (as Octochrysis laminata in Gess, 1981) (Chrysididae) and
one male Anthrax ?tetraspilus (Hesse) (Bombyliidae) were reared from cells of T. giierinii from
the sandbank at Hilton.

DISCUSSION

In order to place the present ethological studies in context it seems useful to give a brief
review of the published ethological accounts for the five genera here represented.

267

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18. PT. 12, JUNE 1991

Nothing appears previously to have been published concerning the nesting of those species
consigned by Giordani Soika (1987) to his new genus Allepipona. The present study of the
ethology of /I. erythrospila does not indicate any unique features but rather shows it to be of the
general pattern exhibited by many eumenids excavating their nests in horizontally presented
clayey soils.

The only previously published accounts of the nesting of Antepipona species are those of
Roubaud (1916) and of Bonelli (1973a). Roubaud’s study concerned the nesting of a species
identified by him as Odynems tropicalis Saussure (= Antepipona tropicalis (Saussure)) in
Dahomey (Benin according to Giordani Soika (1985)). Whereas Giordani Soika (1985: 120)
accepts Roubaud’s species as belonging to Antepipona he points out that tropicalis does not
occur in that part of west Africa and suggests that the species is more likely to have been
rnucronata (Saussure) ox fervida (Saussure). Bonelli’s studies concern the nesting in Ethiopia of
A. silaos (Saussure) and of its subspecies A. s. quartinae (Gribodo) (as A. asmarensis
(Schulthess)) (for synonymy see Giordani Soika, 1985: 91).

A. silaos is one of the commonest species of Antepipona and is widespread over the whole
of central, eastern and southern Africa (Giordani Soika, 1961: 448 and 1985: 91) and is known
also from the Grahamstown district where it is sympatric with A. scutellaris .

The nesting behaviour of A. silaos appears to be similar to that of A. scutellaris. The form
of the underground workings and of the nest turret figured by Bonelli show no apparent
significant differences except that only single-celled nests are recorded. A. siloas is recorded as
making a final closure of the shaft using pieces of turret and then soil from around the entrance,
water being used to moisten the soil to make a plug. This facet of behaviour was not recorded
in the present study but it is likely that A. scutellaris also constructs a final closure.

The account of the nesting of Roubaud’s A. “tropicalis" as rendered by Bequaert (1919:
185-186) shows that this species differs from A. silaos and A. scutellaris in that it excavates its
burrows in clay walls (that is in a vertically rather than a horizontally presented substrate) though
like the latter two species it furnishes the nest with an entrance turret. This turret is removed
when the nest is closed. Possibly the materials of the turret are used for nest closure as are those
of the turret of A. silaos. The nest galleries are short; as a rule bifurcate, each of the branches
containing one, rarely two cells. A. "tropicalis" differs most markedly in apparently being a
progressive provisioner and moreover in tending to care for more than one nest and more than
one larva at a time.

Whereas the nesting behaviour oi Antepipona species as judged from the studies of A. silaos
and A. scutellaris appears to be both interspecifically uniform and unremarkable, the study of
A. "tropicalis" suggests that considerable variation may occur within the genus as a whole.

The genus Euodynerus has a wide distribution, being represented in both the Old and the
New worlds. Krombein (1979: 1491-1497) has indicated the diversity of the nesting habits
known for the genus in North America and has briefly given the essential details of the nest form
of individual species. A few North American species construct mud nests on rocks and a few
species such as E. annulatus (Say) nest in clayey soils, the excavated vertical burrow being
surmounted by a curved mud turret. The majority of species, however, nest in pre-existing
cavities, principally in plant stems. Eight of these species, nesting in trap-nests, were studied by
Krombein (1967: 56-85). In all these species the nest architecture appears to be very similar.
Whereas cell partitions and closing plugs are always constructed of mud or agglutinated sand and
nesting cavities appear never to be furnished with mud entrance turrets, the presence or absence

268

GESS & GESS: ASPECTS OF ETHOLOGY OF FIVE SPECIES OF EUMENIDAE

of preliminary plugs, intercalary cells and vestibular cells appears to be a variable character even
intraspecifically.

The nesting of E. euryspiliis is therefore typical of that of the Euodynerus species which nest
in pre-existing cavities in plant stems.

The only published accounts of the nesting of Rhynchium seem to be a brief account of the
nesting of R. marginellum in abandoned nesting burrows of Xylocopa (Anthophoridae) in
Ethiopia (Bonelli, 1973b) and three accounts for Rynchium oculatum Spinola, a note on nesting
in rose canes in southern France (Lichtenstein, 1869), a short account of nesting in reeds used
in the construction of a wall in Italy (Grandi, 1961) and a fuller account of nesting in trap-nests
in Egypt (Krombein, 1969). Nest construction in all cases seems to be similar to that of R. m.
sabulosiim. Empty intercalary and vestibular cells seem to be a common feature. Pyralidids are
recorded as being used for provisioning by both species and noctuids in addition by R. oculatum.

As far as the authors can ascertain there is no published account of the nesting of any species
of Tricarinodynerus . A photograph of T. guerinii (as Odynerus sp.) on its entrance turret has
been reproduced in Skaife’s African Insect Life (Revised edition, 1979: Plate 133) but it is not
accompanied by a nesting account. The nests described from Hilton, Grahamstown in the
present paper were all in relatively short cavities which only allowed the construction of a small
number of cells. However, nests from False Bay, St Lucia in cut reeds, used in building
construction, with cut open hollow internodes of up to 150 mm in length were constituted of up
to 23 cells (Weaving, pers. comm.). T. guerinii, making use as it does of cavities in vertical earth
banks, in “sandstone” banks, in shale cliffs, and in plant tissue, and of empty cells of wasps
building aerial mud nests, shows considerable flexibility in its choice of nesting substrate unlike
many nesters in pre-existing cavities which seem to be more restricted. E. euryspilus and R. m.
sabulosum have for example not been found nesting in any substrate other than plant tissue.

ACKNOWLEDGEMENTS

The authors wish to thank Mr T. C. White of the farm Hilton for his much appeciated
kindness over the years in allowing them free access to his land; Mrs Sue Dean and Mr R. Dean
of the Karoo Biome Research Station at Tierberg, near Prince Albert, for their hospitality.

Dr D. J. Greathead is thanked for his identification of the bombyliid reared from a nest of
T ricarinodynerus guerinii (Saussure).

Gratitude to the C.S.I.R. is expressed by F. W. Gess for running expenses grants for
fieldwork.

REFERENCES

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. Mm. nai. Hist. 39: 1-384.

Bonelli, B. 1973a. Osservazioni eto-ecologiche sugli imenotteri acuieati dell'Etiopia 1. An. Acc. Rov. Agiati 10-13:
67-72.

Bonelli, B. 1973b. Osservazioni eto-ecologiche sugli imenotteri acuieati dell'Etiopia II. Att. Acc. Rov. Agiati 10-13:
105-110.

Gess, F. W. 1981. Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the vicinity
of Grahamstown, South Africa. Ann. Cape Prov. Mm. (nai. Hist. ) 14 (1): 1-80.

Gess, F. W. and Gess, S. K. 1975. Ethological studies of Bembecinus cinguliger (Smith) and B. oxydorcm (Handl.)
(Hymenoptera: Sphecidae), two southern African turret-building wasps. Ann. Cape Prov. Mus. (nat. Hist. ) 11
(2): 21-46.

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

269

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 12, JUNE 1991

Gess, F. W. and Gess, S. K. 1982. Ethological studies of hodontia simoni (du Buysson), /. pelopoeiformis (Dahlbom) and
I. stanleyi (Kohl) (Hymenoptera: Sphecidae: Sphecinae) in the Eastern Cape Province of South Africa. Ann.
Cape Prov. Mii.s. not. Hist.) 14 (5): 139-149.

Giordani Soika, a. 1961. Hymenoptera (Aculeata): Vespidae. In: Hanstroin, B. et al. eds. South African Animal Life 8.
Uppsala: Almquist and Wiksell, pp. 440-451.

Giordani Soika, A. 1985. Revisione dclle specie afrotropicali del genere Antepipona Sauss. e generi affini (Hym. Vesp.).
Boll. Mas. civ. St. nat. Venezia 34: 29-162.

Giordani Soika, A. 1987. Nuovo contributo alia conoscenza degli Eumenidi afrotropicali (Hymenoptera). Boll. Mas. civ.
Si. nat. Venezia 36: 117-214.

Grandi. G. 1961 . Studi di un entomologo sugli imenotteri superior). Bolletino dell' Instiiuto di Entomologia dell' Universita
di Bologna 25: i-xv, 1-659.

Krombein, K. V. 1967. Trap-nesting wasps and bees: life histories, nests and associates. Washington, D.C.: Smithsonian
Press.

Krombein, K. V. 1969. Life history notes on some Egyptian solitary wasps and bees and their associates (Hymenoptera:
Aculeata). Smithson. Conir. Zool. 19: 1-18.

Krombein, K. V. 1979. Superfamily Vespoidea. In: Krombein, K. V. el al. eds. Catalog of Hymenoptera in America North
of Mexico. Washington: Smithsonian Institution Press, pp. 1469-1522.

Lichtenstein, J. 1869. Une note sur le Rygchiiim oculatum Spinola. Ann. Soc. Ent. France 9 (4), Bull.: Ixxiii-lxxiv.
Roubaud, E. 1916. Rechcrches biologiques sur les guepes solitaires et sociales d'Afrique. Awi. Sc. nat. Zool. (9) 1: 1-160.
[not seen]

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

270

INSTRUCTIONS TO AUTHORS

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

Only generic and specific names should be underlined to indicate italics. All other marking
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
nomenclature or the Internationcd code of botanical nomenclature, as applicable.

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 captions). 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 he 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 submitted on a separate sheet.

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

REFERENCES should he listed alphabetically at the end of the paper. Abbreviations of
periodical titles should conform with those in the World list of periodicals (4th ed.). Layouts
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.

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.

BK2301 — CTP Book Printers, Cape

,4^

i ■?

'■'"■VT^jr

. '■ /-.'" . ,'"'^'^‘'"W ,.:i

•4. '••.j'v^.fj!; /;•<*#> *.- '. !•»?»<*?:. C’jivv if 5 '■;:

M'*..

,-Jf f

'■.<«. «>y-

# I.

‘iS.

■%8- * \'r'*''<!!r*

ti - i^i

;iF«i(f-*'.-:, CK’' ..qi

■;v

i"v'

,,i, :■•■». ,\'.» :

■<5» •■

«^‘(. .; . •••• •■ ’ ,■ ! ;}■'■ •■% «J

•*■• !f ' *-^.'"> '

<f>,i .».*.•* ‘ -■

- ' , 7 ■ 1 .

.. *!F

,t\;^

■ A

-.yv.; .-m'

I'f* 'iii'frtt jjf- 1
i 'r.. ' ... ;

'iTSir'-

W'--

■ t}K<' '•' '.r''-- ■ '^iCJ’

'/ ,ti ■

,a« '.‘fV-^v'^

-T/fr '.v"^'-- '

»■ - - ‘' i- .'■'•^V^lf'v^

';v-,A.r o;y.- fv
-r .

iv-

/ " ■■ ' 5

...4'

^ 6i3

j\ll+

Annals of the
Cape Provincial

Natural History

Ann. Cape Prov. Mus. (not. Hist.)

' Volume 18 Part 13 20th June 1992

\ -

ISSN 0570-1880

' - i ■-)

-ao

Museums

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

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 connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS

Assistant Editor
Mrs S. K. GESS

Parasites, generalist and specialist predators and their role in limiting the
population size of blackflies and in particular Simiilium chutteri Lewis
(Diptera: Simuliidae) in and along the Vaal River, South Africa.

F. C. DE MOOR

(Albany Museum, Grahamstown)

ABSTRACT

Mennithid and microsporidial parasites found in simuliid larvae were monitored at weekly
intervals during one year of a three and a half year study on the biology and ecology of Simuliidae
in the Vaal River near Warrenton. Predators and their feeding behaviour on all life cycle stages of
Simuliidae were recorded over the entire three and a half year period in and along the river. Several
species of vertebrates and invertebrates were recorded as simuliid predators for the first time.
Opportunistic predatory behaviour was observed in most vertebrate and the majority of invertebrate
species identified in the study. Specialist predatory behaviour was observed in some Trichoptera
species. An assessment is made of the role these various predator species play in limiting the popu-
lation size of Simiilium chutteri Lewis, a bloodsucking livestock pest in the region. The larval stages
of two species of hydropsychid Trichoptera, Cheumatopsyche thomasseti (Ulmer) and Amphipsyche
scottae Kimmins, were found sharing the stones-in-cunent biotope with the pest simuliid larvae and
pupae. Population sizes of the two trichopteran species closely follow population size fluctuations
of S. chutteri. These trichopterans contributed significantly to simuliid reduction at certain times of
the year.

Observations made subsequently in and along other river systems and streams have revealed
further instances of predation on larval and adult stages of Simuliidae. Examination of the gut
contents of Chiloglanis species (mochokid rock catlets) sharing the stones-in-current biotope with
several species of Simuliidae in rivers in the eastern Transvaal revealed possible specialist vertebrate
predators of larval and pupal blackflies in Africa.

INTRODUCTION

Aquatic Mermithidae (Nematoda) that attack simuliid larvae have long been considered as
potential biological control agents (Poinar 1981 ). Crosskey (1990) records 67 species in five genera
of mermithids which have been collected from blackfly hosts. Infection rates of up to 68% have been
recorded in populations of simuliid larvae but a rate of between 1-10% is perhaps more commonly
encountered (see references in Crosskey 1990). The taxonomy and culturing techniques for
developing suitable numbers of desired specific mennithid parasites are not well enough understood
at present to enable isolation and large scale breeding of suitable species (Finney 1981 ). About 30
species of Microsporidia representing six families of Microspora (Protozoa) are so far known to be
parasitic on simuliid larvae. Recorded rates of infection by Microsporidia in nature are usually below
1% but are occasionally above 15% (Crosskey 1990). Microsporidia, although used successfully to
control mosquito colonies (Alger and Undeen 1970), have so far not been used for blackfly control.

271

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

Predators of both the aquatic and non-aquatic stages of Simuliidae have been well documented.
In a recent review Davies ( 1981 ) records 206 species of invertebrates and 96 species of vertebrates
as predators of blackfly larvae, pupae, adults and eggs. Direct observation of predation and the
examination of gut contents of potential predators for prey remains are the most frequently used
techniques. More recently the serological study of smears of gut contents of potential predators to
identify protein remains of Simuliiim damnosum sensu lato Theobald have been used (Service and
Lyle 197.6, Service and Elouard 1980). A note of caution should be added as both predation and
scavenging would be revealed using this last technique.

The influence predators have on the population size of blackflies has been estimated from
minimal to very significant. Davies (1981) states that Speir (1976) estimated that predation
accounted for 82.6% of larval simuliid mortality in western Oregon streams (USA). Some of the most
revealing studies on the importance of predators in controlling blackfly population size were
reported by Hynes and Williams (1962) and Ide (1967) who found that insecticidal control of
simuliids also led to a decline in their natural trichopteran and plecopteran predators. Because of their
slow rate of recovery the numbers of these predators remained low in subsequent years whereas
increased numbers of blackflies were observed in years following insecticidal treatment of streams.

In most instances it has only been possible to determine predators without assessing their
importance as population size regulators of Simuliidae (Crosskey 1990). Some recent studies (Merritt
and Wotton 1988, Wotton and Merritt 1988) attempted to evaluate predation on blackflies in a
quantitative way. A three and a half year study of a community of benthic invertebrates in stones-in-
CLurent biotopes in the Vaal River near Warrenton (de Moor 1982a) and subsequent observations and
discussions with colleagues have provided some insight into the impact of parasitism and predation
on simuliid populations.

METHODS

Between June 1977 and March 198 1 a study on the ecology, biology and population fluctuations
of Simuliidae in and along the Vaal River near WaiTenton was undertaken (de Moor 1982a). This
study included a one year assessment of parasites in simuliid larvae. Predators of Simuliidae were
recorded and the possible impact on limiting the population size and growth rate of S. chutteri were
studied over the whole three and a half year period. Observations of predatory activity were
recorded. Where possible photographic records were made, specimens of predators were collected,
gut contents of suspected predators were examined, and identified predatory invertebrates were sent
to specialists for verification.

In order to estimate population sizes of benthic simuliids and associated fauna monthly samples
of natural substrates were collected from the stones-in-current biotope from rapids in the Vaal River
on the Earm Witrand between July 1978 and March 198 i . In addition, samples of animals drifting in
the flowing water column were collected at regular intervals during this period using Cushing-Mundie
or Paddle-wheel drift samplers (de Moor 1982a, de Moor, Chutter and de Moor 1986). Eor an assess-
ment of parasites weekly samples of simuliid larvae and pupae were collected from artificial
substrates for one year (de Moor 1982a).

As the distribution of benthic animals on substrates at any one date was in most instances found
to be contagious (sensu Elliott 1977) it did not allow for the expression of population estimates using
standard arithmetical mean values. After logg transformation it was found that the majority of data
sets conformed to a normal distribution. Erom Sichel ( 1966) an estimate of the average count, given
by the maximum likelihood estimator for the mean of a lognormal population was determined and
is referred to as the Sichel mean (de Moor 1982a). All data sets presented give population size

272

DE MOOR : PARASITES, GENERALIST AND SPECIALIST PREDATORS OE BLACKFLIES

estimates as Sichel mean counts.

Voucher specimens of the predators observed and collected are housed in the National Museum,
Bulawayo (Odonata); British Museum of Natural History, London (Diptera); Plant Protection
Research Institute, Pretoria (Arachnida); and Albany Museum, Grahamstown (Diptera, Trichoptera,
Plecoptera). Parasites of simuliid larvae were not identified to generic level. They are preserved
together with their hosts in the Albany Museum, Grahamstown.

OBSERVATIONS OF PARASITES IN SIMULIIDAE

Two kinds of parasites, memiithid nematodes and microsporidial protozoans were identified
from Simuliidae collected in the Vaal River. A decrease in the percentage of S. chutteri pupae as a
total of all Simuliidae and an increase in merinithid parasitism found by Chutter (1968) was not
discernible in the data gathered over a year from the Vaal River (Fig. 1 ). The highest percentage of
memiithid parasitized S. chutteri larvae were found between August and October 1979 when
simuliid population levels were increasing (see Figs 6, 7 and 8) and when S. chutteri pupae, expressed
as a percentage of all Simuliidae (to make this data set comparable with Chutter’s (1968) study), were
also increasing (Fig. 1 ). The observed decrease of pupal numbers in September 1979 was because
vandals had removed all the rods with artificial substrates from the river between 3 1 August and 6
September (this is indicated by the arrow in Fig. 1). This disturbance prevented data from being
gathered during this week and it took a further three weeks before the simuliid population on

Fig. !.a) Simuliiim chutteri pupae as a percentage of the total number of Simuliidae. b) The percentage of small simuliid larvae
parasitised by Mennithidae. c) The percentage of 5. chutteri larvae parasitised by Mennithidae. Samples of Simuliidae
examined were collected from artificial substrates at weekly intervals from 10 March 1979 to 20 March 1980.

273

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

artificial substrates again attained a level comparable to the natural population. It should be noted
that there are natural seasonal fluctuations in population sizes of simuliid larvae and pupae (de Moor
1982a) and that, when pupae are expressed as a percentage of the total Simuliidae, the role parasitism
by Mermithidae plays in controlling pupal numbers may be erroneously correlated. The discovery
by Mr G.J. Begemann in September 1977 (pers. comm.) of a live mermithid worm in an adult female
S. chutteri indicated that infestation of S. chutteri larvae by Mermithidae did not necessarily prevent
pupation, confirming Anderson and Dicke’s (1960) observation on North American species.
Reproduction would, however, be prevented because the ovaries of this female were atrophied.

The percentage of parasitism in Simuliidae observed by Carlsson (1967) increased markedly
above certain minimum population densities. The population density of S. chutteri in 1979 and 1980
was apparently low enough to prevent intense parasitism by Mermithidae and observed parasitism
remained below 1% throughout that period (Fig. lb and c). This indicated that parasitism played a
minor role in controlling the population size of S. chutteri.

It should be noted that parasitism by Nematodes was only discernible when large worms were
seen in the gut and body cavity of Simidiwn larvae. Many larvae infested by small mermithids, as
well as those that may have perished due to mermithid parasitism may have been missed in the
survey. By standardizing the technique by only recording observed parasites, results are comparable
throughout the period of the study.

Siuiuliiim adersi Pomeroy was also parasitized by Mermithidae as well as by Microsporidia
(Fig. 2). The percentage of larvae parasitized by Mermithidae was below 2% for the entire period
(Fig. 2b) but was noticeably higher when the population size of S. adersi larvae was higher (Fig. 2a).
Mermithidae thus appear to have played a more significant role in controlling the population size of

Fig. 2. a) Sichel mean numbers of large SimuUiim adersi larvae on artificial substrates, b) Percentage of S. adersi larvae parasitised
by Mennithidae. c) Percentage of S. adersi larvae parasitised by Microsporidia. Artificial substrates collected at weekly
intervals from 10 March 1979 to 20 March 1980.

274

DE MOOR :PARASITES, GENERALIST AND SPECIALIST PREDATORS OF BLACKFLIES

S. adersi than that of S. cimtteri. Parasitism by Microsporidia was recorded in up to 24% of large
larvae {Fig. 2c). High parasitism by Microsporidia may have caused the observed decline of the
larval S. adersi population in June 1979 and may also have caused the low number of S. adersi
observed from November 1979 through to March 1980 (Fig. 2a). Although not entirely temporally
separated infestations by Microsporidia were highest during the summer months whereas
Mermithidae were more prevalent in the cooler autumn through to spring.

Table 1. Predators of Simuliidae collected or observed, and identified in a three and a half year
study along the Vaal River near Wari'enton.

SPECIES

FEEDING ON

IDENTIFIED BY

Osteichthyes

Clarias gariepimis (Burcheil)

Larvae, pupae and adult Simuliidae#*

F C de Moor

Aves

Motacilla flava Lin.

Picking adults off reed.s#

F C & I J de Moor

Zosterops paUidiis Swainson

Picking adults off reeds#

F C & I J de Moor

Cossypha cajfra (Lin.)

Picking adults off reeds#

F C & 1 J de Moor

Motacilla capensis Lin. 1

Picking adults off water

Tunliis oUvaceus Lin. >

surface while standing on

F C & 1 J de Moor

Ploceiis vekitus Vieillot \

emerging stones in water#

Hinindo spilodera (Sundevall)

Catching adults in flight

Riparia paliidicola (Vieillot)

(Hawking)#

F C & I J de Moor

Hirudinoidea

Sail fa perspicax Blanchard

simuliid larvae*

F C de Moor

Crustacea: Decapoda

Potamonautes warren! Caiman

simuliid larvae & pupae#*

F C de Moor

Arachnida

Larinia sp

Adult Simuliidae#

A S Dippenaar

Tetragnatha andonea Lawrence

Adult Simuliidae#

A S Dippenaar

Insecta:

Plecoptera

Neoperla spio s.l. (Newman)

simuliid larvae*

M Picker**

Odonata

Adult Simuliidae#

E C G Pinhey

Diptera

Limnophora bella Pont

Pupae and adult Simuliidae#

(see Table 2)
A Pont

Wiedeniatmia sp.

Adult Simuliidae#

B R Stuckenberg

Trichoptera

Ecnomus thomasseti Mosely

simuliid larvae*

K M F Scott

Cheumatopsyche thomasseti (Ulmer)

simuliid larvae*

K M F Scott

Amphipsyche scottae Kimmins

simuliid larvae and eggs*

K M F Scott

# Animals observed feeding on Simuliidae

* Simuliid remains identified from gut contents

** Identified as two different species from eggs found in numphs

275

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

It was apparent that S. adersi, even though present in the rapids at Witrand in lower numbers
than 5. chiifteri during 1979 and 1980, was more severely infested by parasites, both Memiithidae
and Microsporidia, than was that species. This could be due to S. adersi occupying a biotope not
optimally suited to the average individuals of that species. A broadening of this species’ niche would
then lead to reduced fitness, and hence greater susceptibility to parasitism, in individuals found
occupying marginal regions of their extended biotope (de Moor 1982a).

As parasites in the study on the Vaal River were found not to have a noticeably significant impact
on simuliid populations they will not be further discussed in this paper.

OBSERVATIONS AND EXAMINATION OE PREDATORS OE SIMUEIIDAE

During the course of the three and a half year study a number of vertebrates and invertebrates
were observed feeding, identified as feeding from gut content analysis and associated with feeding
on blackflies because of their activity and abundance around the rapids on the Vaal River where the
study was conducted (Tables 1 and 2). Each group of animals is described and details of observa-
tions are recorded.

Osteichthyes

In September and October 1977, October 1978 and August 1980 when simuliid pupal numbers
were high (see Eig. 6) and adult simuliids were emerging and actively flying around rapids, large (up
to c 800 mm long) sharp toothed catfish Clarias gariepiniis (Burchell) were noted to be conducting
a combination of surface and formation feeding (Bruton 1979) in large pools below rapids. On each
occasion numbers of fish were observed to swim in evenly-spaced regular formation, forming a semi-
circular front line with individuals in the centre further back than those on the extreme ends. More
individuals were staggered behind the leaders. They swam slowly towards the bottom end of rapids,
the origin of the food source, with their bodies nearly parallel to the water surface and their mouths
open with barbels extended along the surface of the water, thus guiding into their mouths material
drifting on or just below the surface. They periodically closed their mouths, dived below the surface
and then surfaced again, continuing the activity. This type of feeding concentrates the food source
and maximizes food gathering effort. Eish behind the leaders scooped up any food items not
gathered (Eig. 3).

Close examination of the water surface confirmed that there were large numbers of adult
Simuliidae caught in patches of scum. The examination of gut contents of C. gariepiniis collected
below the rapids, conducted on several occasions, revealed that mostly adult but also larval and pupal
simuliids comprised numerically the major food source. One fish collected in October 1977 had in
its stomach a crab Potanumautes warreni Caiman, three mayfly nymphs, two Baetis glaucus Agnew
and one Choroterpes ? elegans (Barnard), two chironomid larvae and 57 larvae of Simidium spp.,
predominantly S. cinitteri. In October 1978 another fish stomach contained one P. warreni, two
anthomyid pupae, one tipulid pupa, one hydroptilid trichopteran [Catoxyethira sp.) pupal case, three
larvae, nine pupae and 641 adults of S. chutteri. In addition some adult teiTestrial insects which had
accidentally fallen into the water had been taken and included beetles (five Rutelinae, one
Chrysomelidae and one Carabidae) and bugs (two Lygaeidae and one Aphidae). The gut contents of
a third fish was composed almost entirely of S. chutteri, several thousand pupae and a few adults.
This would suggest that the fish had actually scraped pupae off stones in and below the rapids. Bruton
(1979) records that C. gariepiniis feeds most frequently on abundant and easily accessible prey

276

DE MOOR rPARASlTES, GENERALIST AND SPECIALIST PREDATORS OF BLACKFLIES

animals and that invertebrates are numerically the most important food item in their diet. In times of
shortage of a particular prey animal C. gariepimis readily switches to other more easily obtainable
food sources. The almost exclusive feeding on simuliid pupae by one individual of C. gariepimis
confimis an innovative opportunistic feeding strategy.

Stomach content analysis of Barbus aeneiis (Burchell) during the study on the Vaal River did
not reveal any feeding on Simuliidae. In a study of feeding behaviour of Oreochromis mossamhiciis
(Peters) it was found that the juveniles of this predominantly detritivorous or herbivorous fish species
fed almost exclusively on benthic insect larvae during an early phase of their life (de Moor, Wilkinson
and Herbst 1986). For this reason juveniles of large herbivorous fish species in the Vaal River should
also be considered as potential predators of Simuliidae.

Observations and studies subsequent to the research conducted on the Vaal River have
identified other fish species as predators of Simuliidae. Analysis of the stomach contents of the
mochokid rock catlet, Chiloglaiiis pretoriae van der Horst (undertaken for P. de Villiers by H. M.
Barber) confirmed that this species feeds almost exclusively on aquatic insects found in riffle or
stones-in-current biotopes. Orthoclad chironomid as well as simuliid larvae formed a significant
component of the stomach contents of the mochokid rock catlets investigated. Chiloglaiiis
pretoriae is found in rapids and riffles with water flow ranging between 0,8 - 1,0 ms‘* in tropical
and sub-tropical rivers in southern Africa (de Villiers 1991). Miss S. Pollard (pers. comm.) also
confirms that Chiloglaiiis anoterus Crass feeds on orthoclad chironomid larvae indicating that this
species too confines its feeding activity to swift running water biotopes. Another species Chiloglaiiis
paratus Crass was collected by the author in the swiftest of currents on bare bedrock in the lower
reaches of the Sabie River in the Kruger National Park in October 1990. The correspondingly low
density of simuliid larvae in this and other rivers in the eastern Transvaal at this time of the year.

Fig. 3. A fonnatioii of sharp toollicd catfish Clarias garicphiKS feeding on the water surface in the Vaal River downstream of large
rapids on the famt Witrand.

277

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

when one might expect high densities of simuliid larvae and pupae, could possibly be construed as
being a result of the abundance of these small mochokid fish. These fish would very successfully
remove sedentary simuliid larvae and pupae from substrates in the swiftest of rapids. To verify this
hypothesis a detailed study on the life history of simuliid species and their predators in the rivers of
the Kruger National Park would have to be undertaken.

Aves

Birds were often seen feeding on blackflies when large numbers of adult S. chutteri were observed
during the spring in all three years of the study on the Vaal River (Table 1). Certain species picked
floating adults off the water surface while perched on exposed stones in and below rapids, while
others were observed actively searching among reeds while perched on their stems along the banks
of the river. It was recorded that large numbers, mostly of male adult S. chutteri, settled on reeds and
sedges along the river banks at this time of the year. Swallows and martins were often observed
hawking, catching adult flies in flight, when there were swarms of male S. chutteri flying above and
around rapids (Fig. 4). They also hawked above reeds lining the banks of the river. Each year between
August and December swallows and martins were most active around rapids harbouring large
populations of simuliids. In June 1977 when abundant adult simuliid activity was still observed
African Sand Martins Riparia paludicola (Vieillot) were observed hawking above the rapids in the
Vaal River on the farm Witrand.

Fig. 4. Dense swarm of adult male SimuHim chutteri flying above rapids on the Orange River below Marksdrift Weir.

278

DE MOOR :PARASITES, GENERALIST AND SPECIALIST PREDATORS OF BLACKFLIES

Hirudinoidea

The predatory freshwater leech Salifa perspicax Blanchard was identified as a predator of
Simuliidae by Chutter (1968). Gut content analysis of two specimens collected by the author in
December 1980 revealed respectively eight and 1 6 whole simuliid larvae ranging from fourth to final
instars. This confirms that S. perspicax is a voracious predator of simuliid larvae. This leech was
found co-existing with simuliids in the stones-in-cunent biotope in June 1977, August 1979, October
and December 1 980.

Crustacea

Davies (1981) records species of four families of Crustacea, including two species of Potamonidae
Potamonautes niloticus (H. Milne-Edwards) and P. herardi (Audouin), as predators of Simuliidae.
In the present study, from 30 November to 3 December 1980, when the level of the Vaal River was
low and the flow in the nomially swift rapids was reduced to a trickle, medium sized Potamonautes
warreni Caiman (carapace width 50-70 mm) were observed amongst the stones where sampling was
regularly conducted. On closer examination they were seen to scrape their chelae along the submerged
stone surface and go through feeding motions. A crab was collected and preserved in formalin.
Examination of the gut contents of this crab, although very finely masticated, confirmed that it had
been feeding on simuliid larvae and pupae. Regurgitated, un-masticated gut contents contained 88
large S. chutteri larvae and 78 smaller simuliid and ten orthocladiine chironomid larvae.

Arachnida

Web spinning spiders were frequently observed with simuliid adult remains. Webs were spun
amongst vegetation bordering the river banks, between large boulders and stones, and occasionally
on drift sampling apparatus left near rapids. Spiders were observed to be most active when adult
simuliids were abundant between August and Eebruary. Two species were identified in the present
study (Table 1).

Insecta

Plecoptera

Stoneflies identified as Neopeiia spio sensu lato (Newman) but belonging to a species complex
(Picker 1980) were found to be one of the most voracious aquatic invertebrate predators in the
present study. They were encountered in the same biotope as simuliid larvae and pupae in August
and September 1977, July and November 1978, January, March, November and December 1979,
January, March, July, November and December 1980, January and March 1981. Large nymphs, final
instar nymphal shucks and adults were observed from November through to March each year. Nymphs
of N. spio encountered in July, August and September were at an earlier stage of development. Gut
content analysis of a number of nymphs revealed that they fed on fourth to final instar larvae of
Simuliidae as well as small to large larvae of Cheumatopsyche thomasseti (Trichoptera), and nymphs
of Baetis glaucus and Afronurus sp. (Ephemeroptera).

279

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

Odonata

Crisp ( 1936) found that only 7 percent of adult Odonata collected over a two week period had
taken simuliid adults. They could therefore not be considered to be very important predators. Although
the nymphal stages of Odonata are recognised as voracious predators on all forms of aquatic animals
they were never encountered in the same biotope as the simuliid larvae collected in the present study.
Adult Odonata were often observed hawking amongst swamis of adult simuliids around rapids from
October to April. As all adult Odonata are known predators of a number of flying insects a
collection of these was made (Table 2). It should be noted that dates when Odonata were collected
were not the only times when they were present. Smaller species belonging principally to the
Zygoptera, in particular Ischnwa seiiegalensis (Rambur) which was present throughout the summer
period, are potentially more efficient predators of adult simuliids than some of the larger Libellulidae
belonging to the Anisoptera. Ischmira seiiegalensis was the most abundant dragonfly and would,
because of its smaller size, have to expend less energy to obtain sufficient food from capturing adult
simuliids than would the larger Anisoptera.

Table 2. Adult Odonata collected alongside rapids on the Vaal River on the farm Witrand from
1977 to 1980. Months when collected indicated. Species identified by E C G Pinhey.

SPECIES

MONTH AND YEAR COLLECTED

Platycnemididae

Mesocnemis singiilaris Karsch

Dec 1979

ProtoneLiridae

Elattonewa glaitca (Selys)

Oct, Nov, Dec 1977; Oct 1978; Dec 1980

Chlorocyphidae

Platycypha caligata (Selys)

Jan 1980

Coenagrionidae

Pseudagrion salishiiryense Ris

Oct 1977; Oct, Nov 1978; Jan 1980

P. vaalensc Chiitter

Oct, Nov 1977; Oct 1978

P. massaicwn Sjdstedt

Nov 1977

Ischnwa seiiegalensis (Rambur)

Nov 1977

Gomphidae

Crenigomphus hartmanni (Forster)

Jan 1980

Corduliidae

Macromia picta Selys

Nov 1977

Libellulidae

Orihetnim caffrum (Burmeister)

Oct, Nov 1977

0. ahhntti Calvert

Mar 1979

0. chrysostigma (Buirneister)

Oct 1978; Jan 1980

Syinpetruin fonscolomhel (Selys)

Nov 1977

Trithemis fwva Karsch

Nov, Dec 1977; Apr 1979; Jan 1980

T. kirhyi ardens Gerstaecker

Mar 1979; Jan 1980

Palpopleiira jucunda Rambur

Mar 1978

Crocoihemis servilia erythraea (Brulle)

Oct 1978

C. sangidnolenta (Burmeister)

Jan 1980

Diplacodes lefehvrei (Rambur)

Oct 1978

280

DE MOOR :PARASITES, GENERALIST AND SPECIALIST PREDATORS OF BLACKFLIES

Coleoptera

Crosskey (1990) notes that several families of aquatic beetles have been recorded as feeding on
simuliids. He suggests, because of their close association in the swift water biotope, that adult elmids
regularly prey on Simuliidae and adds that serological tests have verified this. In the present study
both larvae and adults of several species of elmid beetles were collected in the same biotope as
simuliid larvae and pupae. They were encountered in 1 2 of the 32 monthly samples collected. Predation
by Elmidae was not, however, confirmed.

No actual observations of predation on Simuliidae by any Coleoptera were made in the present
study. Larvae of a gyrinid, Aulonogyrus sp., were found coexisting with simuliids in the stones-in-
current biotope during 21 of the 32 months. They were on occasion found running around amongst
dense masses of simuliid larvae and pupae on stones collected from rapids. Gyrinid larvae kill their
prey and inject digestive Juices into them with their hollow sickle-shaped mandibles. They then imbibe
the partially digested contents of their prey leaving the empty skin-cast behind. As serological deter-
mination of prey species of the Aulonogyrus sp. was not carried out in the present study predation on
Simuliidae could not be confirmed.

Diptera

Crosskey ( 1990) states that among the Diptera, Simuliidae must be regarded as their own worst
enemies as predation by larger simuliid larvae on smaller ones may be one of the most important
predatory impacts on larval simuliid populations. Cannibalism is recorded in several species (Burton
1971, Chutter 1972, Disney 1972a) but an examination of the larval gut contents of 100 S. cinitteri
in the present study revealed no identifiable blackfly remains. Cannibalism in S. chutteri would be
low because the various instar stages are found to concentrate in different regions of the river during
their life cycle thus limiting contact between individuals at different developmental stages (de Moor,
Chutter and de Moor 1986).

Adults of the muscid fly identified by Mr A. Pont as “very near or identical to Limnophora hello
Pont” were observed on a number of occasions feeding on larval and adult simuliids. In September
1978 when the flow of the river had suddenly decreased and left many stones with simuliid larvae
stranded in shallow trickles of water and pupae exposed to the air, adult L. hello were observed
flying around pupae and settling on them. They attacked pupae in their cocoons and pulled out
partially emerged adult simuliids and started feeding on them. It was also noted that male S. chutteri
were also hovering around pupae apparently waiting for females to emerge so that they could
commence mating. In November 1978, when metal rods with attached artificial substrates were
removed from the water and left exposed while substrate samples were collected for study
purposes, it was noted that a large number of simuliid larvae and pupae had colonised the metal rods.
Adult L. hello were observed walking along these rods catching and eating larvae. They also settled
on exposed stones just above the splash zone and searched for emerging adult simuliids which they
attacked on the surface of the water. In addition they were observed on stones around rapids in
September 1977, January 1980 and February 1981.

In October 1979 a number of individuals belonging to another species of fly (identified, from a
female collected, as an empidid, Wiedemannio sp.) were observed catching adult simuliids in flight
and devouring them while settled on stones (Fig. 5). In February 1980 individuals of this species
were again observed feeding on adult simuliids.

281

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

Fig. 5. An empidid fly Wiedemannia sp., settled on a boulder, feeding on a newly emerged simuliid fly.

Trichoptera

Although all the above recorded invertebrate predators of Simuliidae played some role in
controlling simuliid population sizes they were all found in low densities compared to the simuliids.
Some of the trichopteran predators discussed below, however, appeared in sufficiently large
numbers to have an effective impact on simuliid numbers at certain times of the year. Hydropsychid
Trichoptera, because they closely share the same substrates with many simuliid species, are
documented as efficient predators of Simuliidae in many countries (see Davies 1981). The impact of
certain species of predatory caddis on simuliid population size has been briefly discussed by the
author before (de Moor in press) and is considered in greater detail here.

The mouthparts of the larvae of Orthotrichia species belonging to the microcaddis family
Hydroptilidae are modified for piercing enabling them to get at the cytoplasmic contents of algal cells
on which they normally feed. An Orthotrichia species was, however, observed feeding on simuliid
pupae by Burton and McCrae (1972) who suggested that large numbers of this hydroptilid could
cause considerable pupal mortality in simuliids. Disney (1972b) recorded the larvae of a species of
Orthotrichia feeding on simuliid pupae and eggs. No direct observation of any Orthotrichia species
larvae feeding on simuliid pupae was made in the present study. Total numbers of hydroptilid larvae
(comprising almost entirely an Orthotrichia species), found on the same substrates as simuliid pupae
in 31 of the 32 monthly samples collected, reached peaks in November each year and in January

282

DE MOOR :PARASITES, GENERALIST AND SPECIALIST PREDATORS OE BLACKELIES

1981. This coincided with a noticeable decrease in numbers of simuliid pupae in subsequent months
(Fig. 6).

Fig. 6. Numbers of a) pupae andb) Hydroptiiidae larvae collected from natural stones-in-current substrates at month-

ly intervals between July 1978 and March 1981 . Counts expressed as Sichel mean numbers per 1000 cm- of stone surface area.

283

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

The net spinning Hydropsychidae capture food by means of a silk spun net suspended in the
cun ent from their retreat structure. They periodically move out of their shelter and remove any food
items which have become entangled in the net. Two species of larval hydropsychids, Amphipsyche
scottae Kimmins and Cheimiatopsyche thomasseti (Ulmer), and one species of ecnomid, Ecnomas
thomasseti Mosely, were identified from stomach content analysis as predators of Simuliidae in the
Vaal River near Warrenton. Predation by the two species of hydropsychid was recorded in the months
of November, December, March and April of several years and that by the ecnomid in May and
January. The examination of the gut contents of a range of different instars of C. thomasseti larvae
revealed that this species became more carnivorous with increasing size (Table 3). This confirmed
the findings of Wallace ( 1975) that small hydropsychid larvae tend to be plant and detritus feeders
whereas larger larvae feed more on animal matter. It is thus clear that large C. thomasseti larvae
would have a bigger impact as predators of simuliid larvae than small ones. Third to fifth instar
simuliid larval remains were most commonly found in the foregut of medium to large C. thomasseti
larvae.

Table 3. The number and percentage of Cheimiatopsyche thomasseti (Ulmer) larvae feeding on
simuliid larvae. Instar stage of caddis larvae determined from head capsule measurements
made from the base of mandibles to posterior margin of postgena. Larvae collected on 12
November 1980.

Average head
capsule size

Instar stage
of larvae

Number of larvae
examined

Larvae with simuliid
remains in gut

Percentage of larvae
feeding on

120

1 1

18.S

245

Larvae of C. thomasseti were found coexisting with simuliids in the stones-in-current biotope
in all the 32 months that samples were collected and larvae of A. scottae were found in 28 months.
Simuliidae (comprising 80% or more S. chutteri in all samples collected) were always the most abun-
dant and hydropsychid larvae were either the second or third most abundant benthic macro-inverte-
brates co-existing in the natural stones-in-current biotope in the Vaal River. As C. thomasseti was
the most numerous hydropsychid the discussion on the impact of predators on the simuliid popula-
tion will concentrate on this species.

A decrease in the numbers of C. thomasseti larvae between July and September in each of the
three years of the study coincided with an exponential increase of simuliid larvae in the stones-in-
current biotope between August and October in 1978 and 1979 (Fig. 7). Samples of adult
Hydropsychidae collected from swarms Hying above large boulders along the banks of the Vaal River
on the farm Witrand during September 1977, 1979 and 1980, consisted entirely of male C.
thomasseti. This indicated that a large synchronous emergence of this species had occurred which
suggests that males form mating swanns in spring. The increase of small hydropsychid larvae in drift
samples from late September onwards (Fig. 8) and the appearance of large numbers of small
hydropsychid larvae in the stones-in-current biotope in October each year and C. thomasseti larvae
during November 1978 and 1979 and October 1980 (Fig. 7) confirmed that mating and recruitment
had occLined. The observed increase in the size of the C. thomasseti population was followed by a
decline in simuliid numbers during November in all three years of the study (Fig. 7). It was also
notable that the size of the C. thomasseti population was considerably higher during years when
population levels of Simuliidae were also high (Fig. 7).

284

DE MOOR iPARASITES, GENERALIST AND SPECIALIST PREDATORS OE BLACKFLIES

Fig. 7. Numbers of Simuliidae and Hydropsychidae larvae collected from natural stones-ln-ciirreiit substrates at monthly intervals
between July 1978 and March 1981. Counts expressed as Sichel mean numbers per 1000 cm’ of stone surface area.

285

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

Fig. 8. The number of small liydropsychid larvae per 1000 litres of drifting river water collected upstream of rapids on the farm
Witrand at weekly intervals between 10 March 1979 and 20 March 1980.

From December 1978 to May 1979 and from January to May 1980 there was an increase of
A. scottae larvae (Fig. 7). Although the numbers of larvae of this species were lower than those of
C. thomasseti in the preceding months the temporal separation of large larvae of these two species
would allow for an extended period of predation on simuliids by hydropsychids.

The expected annual increase of simuliid larvae between August and October did not occur in
1980. This was because a successful waterflow regulation programme effectively kept in check the
population growth of a large winter population of S. chutteri. The water flow regulation programme
was carefully timed that year to ensure that it did not disrupt the breeding cycle and recruitment of
C. thomasseti (de Moor 1982a, 1986).

The small larvae of C. thomasseti initially feed on detritus and algae but as they grow they
require progressively more animal protein. Provided that the simuliid population size is not too large
predation by large C. thomasseti larvae in November should halt any further exponential growth of
the simuliid population. The effect a predator species can have on limiting the population size of a
prey species is determined by the size of both the prey and predator populations. If the size of the
prey population is disproportionately large the impact of the predators on the prey population will be
negligible (Carlsson, Nilsson, Svensson, Ulfstrand and Wotton 1977).

Hymenoptera

Gess ( 1980) records the sphecid wasp Dasyproctus westennanui (Dahlbom) on the fann Hilton
near Grahamstown in the eastern Cape as provisioning its nests with several families of Diptera
including simuliid adults. Examination of 13 of these flies by the author revealed that four males and
seven females were Simuliwn adersi Pomeroy and one female was identified as Simuliian nigritarse

286

DE MOOR :PARASITES, GENERALIST AND SPECIALIST PREDATORS OF BLACKFLIES

Coquillett. Five of the females had swollen abdomens revealing the remains of a blood meal. This
would indicate that these flies were caught after taking a bloodmeal, probably while resting waiting
to digest their bloodmeal and complete egg development. Digestion of blood takes three days or more
and females resting on foliage would be vulnerable to searching predators such as sphecid wasps.
Crosskey (1990) records several instances where wasp blackfly-predators wait around mammalian
hosts, including man, and pick off female blackflies while they are engorging on their hosts.

In December 1985 on a field outing with F. W. Gess to Hilton the author was shown nests of
D. westermanni. It appeared that the majority of flies they used for provisioning these nests were
Simuliidae. Material was not collected and the simuliid species were not identified.

DISCUSSION

In the cooler temperate regions of the northern hemisphere, distinct seasonal breeding of many
species of Simuliidae has been recorded and a succession of emergences of simuliid species occurs
from spring through to autumn each year. In the Vaal River with a much warmer water temperature
(de Moor 1982b) breeding and emergence of adults of various species occurs throughout the year.
There are, however, distinct increases of various species at certain times of the year (de Moor 1986)
and in the dominant species, S. chutteri, this usually occurred around September-October each year.

Crosskey (1990) notes that there are more recorded predators on the aquatic stages of blackflies
than on adults. In the many studies conducted to control simuliid population size no real specialist
predators have been found. It should be realised that predators in general are seldom restricted to one
prey species. The simuliid predators discussed in the present paper can in the author’s opinion, how-
ever, be placed into a continuum ranging from opportunistic to specialist predators.

The more specialist predators are species that for obvious reasons must temporally and
spatially closely share the same biotope as their prey species. They must furthermore show some
identifiable population fluctuations which tie in with fluctuations of their prey species. The hydro-
psychid Trichoptera clearly did show this. They were found on the same stones from which dense
populations of S. chutteri were collected and their population sizes closely followed those of this
species. They were less abundant when the simuliid population size was low during the same season
in consecutive years (Fig. 7).

For controlling the population size of a rapids dwelling simuliid species such as S. chutteri a
predatory caddis species such as C. thomasseti would be most effective. Although hydroptilid larvae
of the genus Orthotrichia have been recorded as voracious predators of simuliid eggs and pupae
(Burton and McCrae 1972, Disney 1972b) their low numbers found in association with the Simuliidae
on stone substrates in the present study would suggest that they probably play a minimal role in their
control.

During the study on the Vaal River water flow was periodically regulated to study the effect this
would have on the simuliid population size (de Moor 1982a). Flow regulation effectively reduced
the population size of both simuliids and trichopterans during 1978 and of simuliids during 1980.
This was particularly noticeable in 1980 when the large population of Simuliidae between June and
July did not develop into the expected large population seen in October of the previous years. The
increase of small hydropsychid and C. thomasseti larvae in October, which was not interrupted by
water flow regulation that year, could effectively contain population growth of S. chutteri within
acceptable levels (de Moor 1982a). During 1979 when the population size of simuliids increased
between August and October the appearance of C. thomasseti larvae was undoubtedly the major fac-
tor causing their decline in November as no water flow regulation was implemented that year.

287

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13. JUNE 1992

Fig. 9. Clustering of siniuliid larvae and pupae observed on stones exposed after a drop in the water level following water flow
regulation. These stones during flow conditions exceeding 3 m’s ‘ are in swift flowing regions (c I ms ') of rapids in the
Vaal River.

Predation by C. thoniasseti was more intense when the population size of simuliids was high.
Predation by hydropsychids on simuliids occurred even when population levels of the latter were low
and this would typify a density independent type of predation. When population levels of simuliids
were high a larger percentage of developing hydropsychid larvae would find sufficient food and this
would then lead to a larger population of predators (Fig. 7). The effectiveness of C. thoniasseti
larvae as predators was mainly due to the fact that the larval life cycle of this species coincided with-
high population levels of the aquatic stages of simuliids in spring. The larval drifting behaviour of
S. chiitteri which made it a successful coloniser of newly inundated regions (Chutter 1968, de Moor.
Chutter and de Moor 1986) also made it susceptible to predation by net-spinning Hydropsychidae.

Other potential specialist predators would be the mochokid rock catlets belonging to the genus
Chiloglanis. Their close co-existence with simuliid larvae and pupae in swift-running-water biotopes
makes them preadapted to being specialised predators of Simuliidae.

For simuliid larvae, escape and avoidance mechanisms from predators work most efficiently in
their natural swift-flowing-water biotope. The larvae usually avoid contact with a predator by attach-
ing a silk life line, releasing hold of their substrate and drifting downstream in the current for a
limited distance before resettling. Because of the swift current most other aquatic invertebrates find
it difficult to maintain a good grip on the substrate while moving about in the running water. For this
reason inching away by using the slow looping locomotion, spinning a patch of silk hooking the

288

DE MOOR iPARASITES, GENERALIST AND SPECIALIST PREDATORS OE BLACKELIES

anterior proleg onto this and releasing the posterior proleg and then attaching this to the spun patch
before unhooking the anterior proleg once more, is also an effective avoidance technique. In a
slower current this advantage is lost and predators become much more efficient in locomotion and
procuring of larval simuliid prey. This may explain the dense clustering and almost exclusive occur-
rence of large numbers of larvae and pupae of S. chiitteri on certain stones collected from rapids (Fig.
9). Such densely colonised stones are found only in the swiftest of water flows where most other
species of invertebrates and vertebrates in the Vaal River would find it difficult to obtain a hold on
the substrate let alone actively prey on blackfly larvae and pupae. When water flow temporarily
decreased, during flow regulation applied to control simuliid densities, a host of opportunistic
predators such as birds, fish, crabs and several adult Diptera species made use of these conditions.
They all concentrated around regions where flow was reduced to a trickle and commenced vigorous
feeding on the simuliids left stranded. When mass emergences of adult S. chutteri occurred in spring
predation on adults was also noted to be intense.

Even though there was an abundance of larvae and pupae of S. chutteri in the Vaal River near
Warrenton their specialised adaptation to exist in the swiftest of water flows provides them with
effective predator avoidance mechanisms. Their rapid breeding also allows for exponential popula-
tion growth during periods when inundation of a previously dry river bed occurs or when larval
populations of C. thomasseti decrease because the majority of individuals are pupating or emerging
as adults. ITydropsychid larvae coexisting with simuliid larvae in the same biotope are able to exploit
them as prey with a minimum of effort. Hydropsychid larvae can thus be considered as stones-in-
current biotope, specialist predators and are therefore also exapted (sensu Gould and Vrba 1982)
simuliid specialist predators. The mochokid rock catlets of the genus Chiloglanis also fit into this
class of predator because of their biotope and feeding preferences. Clarias gariepinus exploiting the
abundance of simuliid adults on the water surface is considered to be a generalist predator. Its
opportunistic behaviour in exploiting the dense concentrations of pupae on stones, possibly when
flow rates are sufficiently reduced to allow them access to the normally swift flowing rapids where
water velocities are in excess of 1 ms ', makes it an opportunistic generalist able to take advantage
of conditions outside its normal feeding behaviour pattern. Most of the other predators discussed
above are random generalist predators. They use their normal feeding behaviour approaches and
switch to concentrated feeding on simuliids when these become abundantly available.

Some of the species such as Salifa perspicax which is a voracious predator on simuliid larvae
were only rarely found coexisting with dense populations of Simuliidae. They could thus not be
considered as biotope specialist predators even though they would apparently be able to success-
fully maintain their position on substrates in swift flowing waters.

If predators are to be used in conjunction with other methods to keep population growth of
simuliids under control, a detailed ecological assessment of the life history of the predators to be used
should be undertaken. It is important to identify factors that will optimize environmental and
developmental conditions for the predator and hence increase the chances of efficient control of the
target prey species.

ACKNOWLEDGEMENTS

Most of the data used in this paper were collected while the author was conducting research for
a Ph.D thesis. The research was funded by the Co-operative Scientific Programmes of the CSIR.
Subsequent research and information gathered has been undertaken while working at the Albany

289

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 13, JUNE 1992

Museum and time to conduct research was granted by the Director. Additional funding from an FRD
grant has made the continuation of this work possible. The author would like to thank his supervi-
sors, Dr F. M. Chutter and Prof H. E. Paterson, for their encouragement and guidance during the early
stages of the study, Mr and Mrs W. Vorster for allowing him to carry out the research on their farm
Witrand, and his wife I. J. de Moor for technical assistance and encouragement throughout the study.
The author would furthermore like to extend his gratitude to the following people for advice on and
identification of various taxa; Drs A. Pont and B. R. Stuckenberg (Diptera), Drs R. W. Crosskey and
D. J. Lewis (Simuliidae), Dr K. M. F. Scott (Trichoptera), Dr E. C. G. Pinhey (Odonata), Mrs A. S.
Dippenaar (Arachnida), Dr R. Phelps (Mermithidae). Miss H. M. Barber assisted with the final
preparation of some of the figures.

REFERENCES

Alger, W. and Undeen, A. 1970. The control of a microsporidian Nosema sp. in an anopheline colony by an egg-
rinsing technique. Journal of Invertebrate Pathology 15: 321-327.

Anderson. J. R. and Dicke R. J. 1960. Ecology of the immature stages of some Wisconsin blackflies (Simuliidae:
Diptera). Annals of the Entomological Society of America 53: 386-404.

Bruton, M. N. 1979. The food and feeding bdnayiour ol Clarias gariepinus (Pisces: Clariidae) in Lake Sibaya South Africa
with emphasis on its role as a predator on cichlids. Transactions of the Royal Society of London 35: 47-1 14.

Burton, G. J. 1971. Cannibalism among Simulium damnosum (Simulidae [sic]) larvae. Mosquito News 31: 602-603.

Burton, G. J. and McCrae, T. M. 1972. Observations on trichopteran predators of aquatic stages of Simulium damnosum and
other Simulium species in Ghana. Journal of Medical Entomology 9: 289-294.

Carlsson, M. 1967. Environmental factors influencing blackfly populations. Bulletin of the World Health Organisation 37:
139-L50.

Carlsson, M., Nilsson. L. M.. Svensson. Bj., Ulfstrand, S. and Wotton, R. S. 1977. Lacustrine seston and other factors
influencing the blackflies (Diptera: Simuliidae) inhabiting lake outlets in Swedish Lapland. Oikos 29:229-238.

Chutter, F. M. 1968. On the ecology of the fauna of stones in the cuiTent in a South African river supporting a very
large Simulium (Diptera) population. /ow na/ of Applied Ecology 5: 531-561.

Chutter, F. M., 1972. Notes on the biology of South African Simuliidae, particularly Simulium (Eusimulium)
nigritarse Coquillet. Limnological Society of southern Africa Newsletter 18: 10-18.

Crisp, G. 1956. An ephemeral fauna of toirents in the northern territories of the Gold Coast, with special reference to
the enemies of Simulium. Annals of Tropical Medicine and Parasitology 50: 260-267.

Crosskey. R. W. 1990. The Natural History of Blackflies. Chichester: John Wiley and Sons.

Davies, D. M. 1981. Predators upon blackflies. In: Laird, M. ed„ Blackflies: the future for biological methods in
integrated control. London: Academic Press, pp. 139-158.

Disney, R. H. L. 1972a. Observation of sampling pre-imaginal populations of blackflies (Dipt., Simuliidae) in West
Cameroon. Bulletin of Entomological Research 61: 485-503.

Disney, R. H. L. 1972b. Larval Hydroptilidae (Trichoptera) that prey on Simuliidae (Diptera) in Cameroon. Entomologist’ s
Monthly Magazine 108: 84-85.

Elliott, J. M, 1977. Some methods for the statistical analysis of samples of benthic invertebrates. Ereshwater Biological
Association Scientific Publication No. 25 (2nd Edition), 160 pp.

Finney, J, R. 1981. Potential of mermithids for control and in vitro culture. In: Laird, M, ed.. Blackflies: the future for
biological methods in integrated control. London: Academic Press, pp, 325-333.

Gess, F. W. 1980. Some aspects of the ethology of Dasyproctus westermanni (Dahibom) (Hymenoptera: Sphecidae:
Crabroninae) in the Eastern Cape Province of South Africa. Annals of the Cape Provincial Museums (Natural
History) 13: 95-106.

Gould, S. J. and Vrba E. S. 1982. Exaptation - a missing term in the science of foim. Paleobiology 8: 4-15.

Hynes. H. B. N. and Williams T. R. 1962. The effect of DDT on the fauna of a central African stream. Annals of
Tropical Medicine and Parasitology 56: 78-9 1 .

290

DE MOOR iPARASITES, GENERALIST AND SPECIALIST PREDATORS OE BLACKFLIES

Ide, F. P. 1967. Effect of forest spraying of DDT on aquatic insects of salmon streams in New Brunswick. Journal of
the Fisheries Research Board Canada 24: 769-805.

Merritt, R. W. and Wotton, R. S. 1988. The life history and behavior of Linmophora riparia (Diptera: Muscidae), a
predator of larval black flies. Journal of the North American Benthological Society 1: 1-12.

Moor, F. C. de. 1982a. A community of Simulium species in the Vaal River near Warrenton. Ph.D. thesis. University
of the Witwatersrand, Johannesburg. 2 vols. 317pp.

Moor, F. C. de, 1982b. Determination of the number of instars and size variation in the larvae and pupae of Simulium
chiitteri Levtis 1965 (Diptera: Simuliidae) and some possible bionomical mplicatiom. Canadian Journal of Zoology
60: 1374-1382.

Moor, F. C. de, 1986. Invertebrates of the Lower Vaal River, with emphasis on the Simuliidae. In: Davies B. R. and
Walker K. F. eds. The ecology of river systems. Dordrecht, The Netherlands: Dr W. Junk Publishers, pp. 135-140.

Moor, F. C. de, (in press). The role of Cheumatopsyche thomasseti Ulmer (Hydropsychidae) in controlling population
growth of Simuliidae in the Vaal River, South Africa. In: Tomaszewski M. ed.. Proceedings of the 6th International
Symposium on Trichoptera. Poznan, Poland: Adam Mickiewicz University Press, pp. 1 19-122.

Moor, F. C. de, Chutter, F. M. and Moor, I. J. de. 1986. Drift behaviour and microhabitat selection in the pre-
imaginal stages of the blackfly Simulium chutteri (Diptera Simuliidae). Hydrohiologia 133: 143-154.

Moor, F. C. de, Wilkinson, R. C. and Herbst, H. M. 1986. Food and feeding habits of Oreochromis mossamhicus
(Peters) in hypertrophic Hartbeespoort Dam, South Africa. South African Journal of Zoology 21: 170-176.

Picker, M. 1980. Neoperla spio (Plecoptera): a species complex? Systematic Entomology 5: 185-198.

PoiNAR, G. O. 1981. Mermithid Nematodes of Blackflies. In: Laird, M. ed., Blackflies: the future for biological
methods in integrated control. London: Academic Press, pp. 1 59- 1 70.

Service, M. W. and Elouard, J. M. 1980. Serological identification of the predators of the complex of Simulium
damnosum Theobald (Diptera: Simuliidae) in the Ivory Coast. Bulletin of entomological Research 70: 657-663.

Service, M. W. and Lyle, P. T. W. 1975. Detection of the predators of Simulium damnosum by the precipitin test
Annals of Tropical Medicine and Parasitology 69: 105-108.

SiCHEL, H. S. 1966. The estimation of means and associated confidence limits for small samples from lognormal
populations. In: Symposium on: Mathematical statistics and computer applications in ore valuation, Johannesburg.
Journal of the South African Institute of Mining and Metallurgy 106-122.

Speir, j. a. 1976. The ecology and production dynamics of four black fly species (Diptera, Simuliidae) in western
Oregon streams. Ph.D. Thesis, Oregon State University.

ViLLiERS, P. de. 1991. The ecology and culture of the rock cadet Chiloglanis pretoriae (Pisces: Mochokidae). M.Sc.
Thesis, Rhodes University, Grahamstown. 148 pp.

Wallace, J. B. 1975. The larval retreat and food of Arctopsyche with phylogenetic notes on the feeding adaptation in
Flydropsychidae larvae. Annals of the Entomological Society of America 68: 167-173.

Wotton, R. S. and Merritt, R. W. 1988. Experiments on predation and substratum choice by larvae of the muscid fly,
Linmophora riparia. Holarctic Ecology 11: 151-159.

291

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing 1/: or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature 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 temied 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 captions). 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 submitted on a separate sheet.

TABLES should be submitted 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 confoirn with those in the World list of periodicals (4th ed.). Layouts
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.AfrJ.Sci. 65 (8): 229-242.

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

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.

CADAR, P.E.-K659

ISSN 0570-1880

^ fci3
N\4-

Annals of the

Cape Provincial Museums

Natural History

\nn. Cape Prov. Mus. (nat. Hist.)

Volume 18 Part 14 30th July 1993

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

ANNALS OF THE CAPE PROVINCIAL MUSEUMS

The primary objective of these Armais 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 connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS

Assistant Editor
Dr S. K. GESS

Three recently erected Trichoptera families from South Africa, the
Hydrosalpingidae, Petrothrincidae and Barbarochthonidae
(Integripalpia: Sericostomatoidea)

K.M.F. SCOTT

(Albany Museum, Grahamstown)
with

A cladistic analysis of character states in the twelve families
here considered as belonging to the Sericostomatoidea

F.C. DE MOOR

(Albany Museum, Grahamstown)

CONTENTS

Abstract 294

Introduction 294

The three recently erected families 296

Family Hydrosalpingidae Scott 296

Genus Hydrosalpinx Barnard 298

Hydrosalpinx sericea Barnard 298

Biology of genus Hydrosalpinx 305

Family Petrothrincidae Scott 307

Genus Petrothrincus Barnard 308

Petrothrincus circularis Barnard 310

Petrothrincus triangularis (Hagen) 318

Petrothrincus demoori sp. nov 322

Biology of genus Petrothrincus 328

Key to species of Petrothrincus, male imagos, and larvae

and their cases 328

Family Barbarochthonidae Scott 329

Genus Barbarochthon Barnard 331

Barbarochthon brunneum Barnard 331

Biology of genus Barbarochthon 343

Discussion 344

Zoogeographical note 346

A cladistic analysis of character states in the twelve families

here considered as belonging to the Sericostomatoidea 347

Acknowledgements 352

References 353

293

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

ABSTRACT

Three southern African families of Trichoptera, the Hydrosalpingidae, Petrothrincidae and
Barbarochthonidae, erected recently (Scott, 1985) to accommodate the genera Hydrosalpinx,
Petrothrincus and Barbarochthon (all Barnard 1934), are fully described, illustrated and discussed.
As far as is known all three genera are endemic to South Africa and appear to be relicts of the
Gondwanan south temperate fauna. They are largely confined to the western, southwestern and
southern coastal folded belt. Petrothrincus and Barbarochthon have also been recorded from the
eastern Cape Province and Barbarochthon from Natal.

Specific descriptions are given for the single species of Hydrosalpinx and Barbarochthon, and
for the three species, one new, of Petrothrincus. A key to the species of Petrothrincus is given.

The three families are placed in the superfamily Sericostomatoidea Stephens 1836, sensu Weaver
1983 of the suborder Integripalpia.

Phylogenetic relationships of the twelve families considered to belong to the Sericostomatoidea
are investigated using cladistic methods.

INTRODUCTION

Several southern African genera of Trichoptera have presented difficulties in classification since
their original description by Barnard (1934). These genera include Rhoizema, Cheimacheramus,
Petroplax and Barbarochthon, all accommodated by him in the family Sericostomatidae Stephens
1836 (emend. McLachlan, 1874), in its old wide sense. Others are Hydrosalpinx and Petrothrincus,
placed by Barnard under Aequipalpia near, but not in, the families Molannidae and Beraeidae. To
Barnard’s group Morse (1974) added a new genns Aclosma which he attributed to the Sericostomatidae
sensu stricto. This genus he erected for his new species A. bispinosa from Natal. He believes that
Petroplax anomala Barnard from the eastern Cape should also be included in his genus Aclosma
(Morse in litt, 15. xi. 1974). The differences between Ac/oiwa and Petroplax are small and Aclosma
may prove to be a junior synonym. All the above genera appear to be endemic to South Africa and
restricted mainly to the mountains of the coastal folded belt in the Cape Province. Aclosma and
Barbarochthon have also been found in Natal.

Over the years most of the older genera have suffered various vicissitudes, taxonomically
speaking. Barnard himself did not indicate to which of the then existing subfamilies he considered
that Rhoizema, Cheimacheramus, Petroplax and Barbarochthon might belong. He did, however,
comment that the family Sericostomatidae was a repository for a number of forms the systematic
position of which was not clear. Shortly afterwards Lestage (1936), terming the Sericostomatidae
sensu lato “this old curiosity shop” as had McLachlan earlier, suggested that Rhoizema,
Cheimacheramus and Petroplax might belong to the Sericostomatinae but remarked that the whole
family diagnosis needed revision with the creation of a special division for Petrothrincus and
Hydrosalpinx. He considered that Barbarochthon was more likely to belong to the Brachycentrinae
than to the Sericostomatinae on account of the structure of the maxillary palps. Ulmer (1955) followed
Lestage in placing Barbarochthon under Brachycentrinae with a query but remarked that the immature
stages did not fit well there. Fischer (1970) allocated it to the Brachycentridae without comment.
Marlier (1962) reunited all four genera {Rhoizema, Petroplax, Cheimacheramus and Barbarochthon)
in the Sericostomatinae as an expedient, though artificial, temporary measure pending revision when
greater knowledge should become available.

After consultation with Dr G.B. Wiggins and comparison with Canadian material kindly sent by

294

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

him, it was decided that Rhoizema, Cheimacheramus, Petroplax and Aclosma could well be
accommodated in the Sericostomatidae sensu stricto as at present understood, together with Aselas
Barnard, another endemic genus, more classically sericostomatid than the other four. Barbarochthon
does not, however, belong to this family.

Although, seen in life, adults of both Petrothrincus and Hydrosalpiivc somewhat resemble
sericostomatids in appearance and posture they do not fit into the Sericostomatidae, primarily because
the males have five-segmented maxillary palps similar to those of the females. Barnard (1934)
contented himself with placing both under Aequipalpia, commenting that they might be included in
the Molarmidae-Beraeidae were it not for the presence of a discoidal cell in the fore wings. He also
commented (p. 323) that “the genus [Petrothrincus] bears a general resemblance to Thremma in the
venation of the female, and the scutiform larval case. There are several peculiarities, however, in the
venation, including the dissimilarity in the sexes. The larval resemblances are evidently due to
similarity of habitat, and do not necessarily indicate relationship.” In that Barnard appears to be
perfectly correct. The two families show such major differences that Petrothrincus carmot possibly
be accommodated in the Uenoidae: Thremmatinae (in which Thremma is now placed). Lestage (1936)
discussed their position in some detail, but left it open, suggesting (as mentioned above) that they be
placed in a special division, a division more primitive than the Molannidae in its retention of the
discoidal cell in the fore wings, and heralding the Beraeidae in larval type. Ulmer (1955) placed
Petrothrincus in the Helicopsychidae with a query. Fischer (1964) put it into the Molannidae, leaving
Hydrosalpinx as incertae sedis, quoting Scott (1967) as tentatively placing it in the Beraeidae. Fischer
was under a misapprehension since in Scott’s preliminary key the family Beraeidae was keyed out
per se, followed by a note that Hydrosalpinx and Petrothrincus did not entirely agree with that family
as diagnosed and had therefore been given separate positions in the key under their generic names.
Marlier (1962) left both genera as incertae sedis, pending a world revision. When the present author
originally started to construct keys to southern African Trichoptera, it was found convenient at both
adult and larval levels to lump the five sericostomatid genera together with Barbarochthon in a
heterogeneous group as “western Cape sericostomatids”. Hydrosalpinx and Petrothrincus were keyed
out separately as genera. More recently, however, when revising the keys to Afrotropical families,
the author had occasion to study all the southern African genera in detail and with more experience,
and tried again to fit Barbarochthon, Hydrosalpinx and Petrothrincus into those families from
elsewhere that appeared to be nearest to them, but again without success. It would only be possible
to do so by altering the family diagnoses to fit, not a very desirable or satisfactory procedure,
particularly for endemic genera from another part of the world.

After much consideration and subsequent consultation with Dr F. Schmid, Professor G.B.
Wiggins and Dr A. Neboiss, the author decided to erect three new families, the Hydrosalpingidae,
Barbarochthonidae and Petrothrincidae, to accommodate them. This brings southern Africa more into
line with the Australian Region where several new families have been erected to accommodate
endemic genera which did not fit into existing families although they had previously been allocated
to one or another of the older families despite the difficulties encountered in so doing.

The original intention had been to erect the three new families in the present paper. As it transpired,
however, they were unintentionally, but validly, erected in the chapter on Trichoptera (Scott, 1985)
for the book Insects of Southern Africa (Scholtz and Holm, 1985). This poses problems to research
workers because the families were not formally erected nor do they appear under the author’s name.
The present paper, although long delayed by the author’s illness, fills the lacunae, providing the
necessary descriptions, illustrations and discussion. In addition cladistic analyses of the phylogenetic
relationships of the twelve families considered to belong to the Sericostomatoidea are presented by

295

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Dr F. C. de Moor.

A diagnosis is given for each of the proposed new families, all of which are monogeneric. In
view of the fact that two of the three genera on which the families are based are also monotypic,
namely Hydrosalpinx and Barbarochthon, it has been decided in those cases to omit a generic
diagnosis as it is covered by the full description of the species. As Wiggins (1984) has pointed out,
the diagnostic characters of a monotypic genus are not objectively separable from those of the species.
In the third genus, Petrothrincus, there are three species, two known and one described in this paper,
making a separate diagnosis for the genus possible. Comments on such biological aspects as are
known for each genus are given.

Fully illustrated descriptions of the type-species of each of the three genera and therefore of the
three families are given. These include descriptions of the male and female imagos, and the larva and
pupa together with their cases. With respect to Petrothrincus differences between the type-species,
the second known species and the new species are noted, the two latter are described briefly and
illustrated, and a key to the species is provided. The possible origins of the three families are discussed.

In this paper southern Africa is regarded as being delimited to the North by the Cunene, Okavango
and Zambezi rivers, including their watersheds but excluding Lake Malawi. The north-flowing rivers
on the other side of the watershed are regarded as central African.

Geographical references are taken from 1:50 000 maps, each of which covers a quarter degree
square. Each degree square is numbered and divided into 16 such maps (15' x 15' squares). As an
example 3318 AB refers to a map between 33° to 33°15' S and 18° 15' to 18“ 30' E (see Leistner and
Morris 1976, introductory 4 pp.). Indication of localities in this way is necessary as older localities
are often not precise and may refer to a range of mountains or a whole river.

Abbreviations and names of collectors cited in the text are as follows: ACH - A.C. Harrison; ADH
- A.D. Harrison; BCW - B.C. Wilmot; DFH - D.F. Houck; FCdM - F.C. de Moor; FMC - F.M. Chutter;
HB - H. Bertrand; HGW - H.G. Wood; HM - H. Malicky; HMB - H.M. Barber; JDA - J.D. Agnew;
JMK - J.M. King; KHB - K.H. Barnard; KMFS - K.M.F. Scott; NK - N. Kohly; RD - R. Dick.

THE THREE RECENTLY ERECTED FAMILIES
Family Hydrosalpingidae Scott 1985

Hydrosalpingidae Scott 1985: 331, 337; Scott 1986: 231, 234 (table 1).

Ty'pe-gtnns Hydrosalpinx Barnard 1934: 321, 323.

The family Hydrosalpingidae was erected to receive the single genus Hydrosalpinx Barnard.
One species of Hydrosalpinx, H. sericea Barnard, is known. It has been recorded from a number of
mountain streams in the western and southwestern Cape Province.

Recognition

Imago medium-sized, hairy, golden-brown. Maxillary palps 5- segmented in male and female,
very long in male. Labial palps 3- segmented, very long in male. Pronotum with 1 pair warts;
mesonotum without warts, usually with a pair of single setae; scutellum with single large wart with
lateral setae.

296

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Family diagnosis
Male imago

Ocelli absent; eyes large, glabrous; vertex with pair of small inter-anteimary setose warts and
pair of large, subtriangular posterior setose warts, also with pair of small tubercles, one on each side
of median sulcus. Antennae slightly longer than fore wings; scape stout, slightly shorter than head;
flagellum tapered. Maxillary palps very long, 5-segmented, with fifth segment longest, not annulate.
Labial palps long, 3-segmented, with third segment longest, not aimulate.

Pronotum with single pair of large warts; mesonotum without warts; scutellum elongate with
single large sub-oval wart bearing lateral setae. Tibial spurs 2, 2, 4; some tibial and tarsal segments
with double row of small black spine-like setae and several similar setae at most tarsal apices. Fore
wings with discoidal cell closed, thyridial cell very long, forks 1, 2, 3 present, fork 5 absent. Hind
wings with discoidal cell open, R2 and R3 fused; fork 2 only present. Wing-coupling macrotrichia
present along part of costal margin of hind wings, linking with inturned anal margin of fore wings.

Genitalia with branched paired claspers (inferior appendages) arising from ninth segment; paired
preanal appendages arising from tenth segment which is long; aedeagus large, simple; parameres
absent.

Female imago

Larger than mate and with more complete wing venation, the discoidal cell being closed in both
fore and hind wings, although in the latter the closure may be unclear. In the fore wings forks 1, 2,
3 and 5 are present, as Cuj is forked. In the hind wings R2 and R3 are separate, so forks 1 and 2 are
present; R2 is complete. Other characters as in male.

Genitalia with simple dorsal plate (presumably those of ninth and tenth segments fused). Sternites
of tenth segment unsclerotized; pair of terminal appendages visible, very small.

Larva

Case-dweller; larva not flattened; head, pro- and mesonota strongly sclerotized; no prosternal
horn. Head rounded; frontoclypeal apotome with one pair of indentations; antennae very small, at
base of mandibles; eyes prominent, set fairly far forward; mouthparts small; ventral apotome a short
triangle, not completely separating the genae, only clearly seen in juveniles. Metanotum membranous
with single pair of small sclerites. Fore legs stout, middle and hind legs slender. First abdominal
segment with flattened dorsal hump, lateral humps each with a small, oval, dorsally pubescent sclerite;
abdominal segments smooth, lacking setae; lateral fringe absent; lateral tubercles present on eighth
segment; gills absent; ninth segment with a dorsal sclerite. Anal prolegs very short, their bases fused
to form an apparent tenth segment; anal claw with one or two dorsal hooks.

Larval case

A somewhat tapered, gold-coloured, silken tube with slightly flared opening and terminal
membrane pierced by a circular aperture.

297

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Male pupa

No male pupae were available to the present author, however, according to Barnard (1934) the
male is similar to the female but smaller and has very long maxillary and labial palps as in the imago.

Female pupa

Antennae slightly longer than body; labrum semicircular with short median lobe; mandibles
strong with pointed apex, inner margin strongly serrated; maxillary palps only reaching end of
metanotum; labial palp shorter. Middle tarsi fringed for swimming. Lateral abdominal fringes present,
conspicuous, with tufted ends; gills absent; wing sheaths reach begirming of seventh abdominal
segment. First abdominal segment without lappets; second to sixth segments with pre-segmental
dorsal plates; fifth segment with postsegmental dorsal plate; all dorsal plates small. Anal appendages
slender, rod- like.

Pupal case

An altered larval case, closed anteriorly and just behind the pupa with new membranes, each
with a slit, with old posterior membrane persisting, and attached anteriorly with a single dorsal
holdfast.

Genus Hydrosalpinx Barnard 1934

Hydrosalpinx Barnard 1934: 321, 323, Figs 16a-o.

Type-species H. sericea Barnard 1934.

Etymology: Generic name feminine, meaning a water trumpet; specific name, referring to silk;
both names being descriptive of the case.

As Hydrosalpinx is a monotypic genus a generic diagnosis is omitted. It is covered by the full descrip-
tion of the species.

Hydrosalpinx sericea Barnard
(Figs 1-32)

Hydrosalpinx sericea Barnard 1934: 321, 323, figs 16a-o (male, female, larval and pupal parts, pupal
case); Scott 1985: 337, figs p (p. 331), v (p. 332); Scott 1986: 232, 236, 243.

Lectotype male here selected and designated from Barnard’s syntypes. South African Museum, Cape
Town.

Type locality: western Cape Province, Bain’s Kloof, Wellington Mts, [3319 CA].

Barnard did not as a rule select types, the specimens in his collection being unmarked apart from
name, locality, collector(s) and date. A Lectotype has been selected from those specimens listed in

298

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs. 1-13. HYDROSALPINGIDAE ; Hydrosalpinx sericea Barnard, male, female (Scale lines = 1 mm unless otherwise indicated).

Material used: KHB

1. male: Head, pro- and mesonota, dorsal. 2. male: face, showing maxillary and labial palps. 3, 4. male: fore and hind wings. 5, 6, 7. male gen-
italia, lateral, dorsal and ventral views (a - aedeagus, cl - clasper, ic/icl - internal branch of clasper, pa - preanal appendage, x - hood formed by
tenth tergum) . 8. male: left middle leg, tibia and tarsus, 8a, tarsal apex with crown of spinelike setae, 8b, fifth tarsal segment with claw, 8c, sin-
gle plumose spinelike seta, all further enlarged. 9. male: entire insect, lateral. 10, 11. female: fore and hind wings. 12, 13. female: genitalia, ven-
tral and lateral views (ixs - ninth stemite, ixt - ninth tergum, vg - vagina, xt - tenth tergum) (13 after Barnard 1934 fig. 16h).

299

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

his 1934 paper and still remaining in the South African Museum collection. It is the best available
male (not perfect but nearly so, right fore leg and labial palp only being missing) and is from Bain’s
Kloof (14.iv.l933). The other specimens are regarded as Paralectotypes. The extant material and
original records are listed after the descriptions of the different stages.

Description ofimagos (Figs 1-13).

Male imago (Figs 1-9)

Medium-sized, hairy; fore wings 9-10 mm in length.

Colour description given by Barnard, probably from freshly pinned specimens - “Head and
thorax fulvous with, pale golden hairs. Legs and antennae ochraceous or fulvous, the latter darker
proximally. Wings thickly pubescent; fore-wings bright golden-brown, the costal area and the apical
cells duller and darker brown, the veins paler and brighter; hind-wings greyish-brown, fringe grey.”

Further description from male imagos, pinned and in spirit, from Barnard’s collection.

Ocelli absent; eyes large, black, glabrous; vertex with a pair of small inter-antennary warts, a
pair of large postero-lateral warts and a pair of bare median protuberances (Barnard’s conical warts);
mid-cranial sulcus complete (Fig. 1). All warts on head and thorax bear long, upstanding golden setae.
Antennae longer than fore wings; scape stout, slightly shorter than head, with long setae; flagellum
basally thick (less so than scape), tapering to slender tip. Maxillary palps (Fig. 2) laterally flattened,
5-segmented, very long (5.0 mm), with first and second segments very short, third about equal to
first and second together, fourth more than double length of third, fifth a little longer, simple, with
second segment bearing several long dorsal setae. Labial palps 3-segmented, long, with first segment
short, second longer, third longest, simple (length 4.0 mm).

Pronotum with single pair of large warts; mesonotum without warts; scutellum elongate,
subtriangular, with single large suboval wart bearing scattered setae. Legs with tibial spurs 2, 2, 4:
fore and middle tibiae and tarsi and hind tarsi with double or treble row of black spinelike setae, and
with a crown of three or four similar setae at tarsal apices on middle and hind legs (Figs 8 and 8a, b).
These spinelike setae are plumose (Fig. 8c). Wings (Figs 3, 4, 10, 11) with venation differing in both
wings in the two sexes. Male fore wings (Fig. 3) with discoidal cell closed, median cell open, thyridial
cell very long; there are forks 1, 2, 3 and a large jugal lobe; M is 3-branched; Cu2 joins Cui, A^ -f- A2
join basally, meeting the hind margin at the arculus. Male hind wings (Fig. 4) with discoidal cell
open; only basal part of present, R2 and R3 fused, M and Cu^ unbranched, fork 2 only present and
jugal lobe rounded and inturned. Corneous points present in both wings in fork 2 and in the thyridial
cell in the fore wings. (Barnard gave the male fore wings forks as 1, 2, 3, (5), remarking that fork 5
was spurious owing to Cui not being forked; it appears preferable to omit it, as the area between Cuj
and M3+4 cannot be mistaken for a true fork 5.) The hind wings bear long macrotrichia (not hamuli)
on the basal part of the costa, evidently linking with the inturned hind margin of the fore wings in
flight. There is a long setal fringe along the anal margin of the hind wings.

The dorsum of the abdomen bears a pair of oval setose warts on each segment except the first.

Genitalia (Figs 5 -7) with ninth and tenth tergites dorsally fused; paired preanal appendages arising
from tenth segment near junction with ninth, long, slender, with blunt apices bearing a few setae, and
small setose basal lobe. Tenth tergite forms a median dorsal apical hood with bifurcate apex,
transparent and difficult to see. Claspers with lower branch broadly lobate, rounded and setose in
lateral view, with upper branch sinuously blade-shaped, strongly curved dorsally, posteriorly and
inwards with a stout triangular process near base and with a few setae. In the two pirmed specimens

300

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

the upper branches cross over one another dorsally. Aedeagus stout, lacking parameres or endothecal
processes.

Female imago (Figs 10-13)

Larger than the male (fore wings 11-12 mm). Palps similar but shorter (maxillary palps 3.0 mm,
labial palps 1.5 mm in length); otherwise very similar apart from wing venation and genitalia, which
are dealt with below.

Wing venation more complete than in male, the discoidal cell being closed in both fore and hind
wings, although in the latter the closure may be unclear. In the fore wings forks 1, 2, 3 and 5 are
present as Cuj is forked. In the hind wings R2 and R3 are separate, so forks 1 and 2 are present; Rj
is complete.

Genitalia (Figs 12, 13) with ninth and tenth tergites apparently fused, simple, rounded, partly
pubescent, with a few longer lateral setae, apex slightly projecting, bilobed, without appendages;
sternal plates of ninth segment unsclerotized or only slightly so, rounded, with numerous striae or
corrugations, vagina strongly sclerotized; sternum of eighth segment thickly setose along posterior
margin.

Description of mature larva (Figs 14-25, 30-32).

The following description was made from material (MISC 306a) (in spirit) collected in 1976
from a mountain stream, tributary of the Vet River, Garcia’s Pass, north of Riversdale in the
southwestern Cape, which was compared with the earlier specimens from the western Cape and
Barnard’s own material, descriptions and drawings.

Case dweller; length 15-16 mm; larva not flattened.

Head (Figs 17-20, 31, 32) slightly longer than broad, very dark brown; cuticle strongly sclerotized,
pitted; muscle spots smooth, shining, colour as head; genae paler; frontoclypeal apotome with single
pair of lateral indentations; ventral apotome short, triangular, darker brown anteriorly, pale posteriorly,
triangular shape only discernible in young larvae (Fig. 20); ventral ecdysial line unclear or absent;
pair of brown pigmented areas present lateral to mid-line. Eyes prominent, situated fairly far forward,
beneath a short, strong ridge, giving a faintly 'beetle-browed’ effect, each eye in a clear area under
cuticular lenses. Antennae inconspicuous, at bases of mandibles. Labrum small with rounded anterior
margin, partially retractile. Labium with clear paired labial palps. Maxillary palps stout, 5-segmented,
strongly setose; galeae also long, stout; stipites large. Mandibles (Figs 24 and 25) small, heavily
sclerotized, hollowed on inner side with strong brushes, two small apical teeth and two setae on outer
side of each.

Thorax (Figs 14, 17) with prosternal horn absent. Pronotum without carina or anterolateral
expansions, mainly very dark brown and thickly set with long, fine setae, posterior part paler with
dark spots and few setae. Pleural sclerites (Fig. 21) light brown; pre-episternum strong, curved,
pointed, not fused with episternum. Mesonotum with large oblong paired plates, dark brown anteriorly,
paler posteriorly, with dark spots; plates with many relatively long, fine, pale, inconspicuous setae,
mainly anterolaterally. Metanotum membranous apart from a pair of very small, transversely oblong,
lightly sclerotized plates, each with a single seta. Sternum membranous. Fore leg (Fig. 21) shortest,
stout; coxa with long, mainly ventrally placed setae; trochanter divided, proximal part bare, distal
part with ventral trochanteral brush; femur widely subtriangular, with many fine setae and long ventral

301

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs. 14-29. HYDROSALPINGIDAE; Hydrosalpinx sericea Barnard, mature larva, pupa and cases
(Scale lines = 1 mm unless otherwise indicated).

Material used: larva MISC 306a; pupa KHB

14. Mature larva, habitus, lateral, a, anal claw, much enlarged. 15. Anal region, further enlarged. 16. Larval case. 17. Head and
thoracic nota, dorsal view (setae shown only on right thoracic nota), muscle spots indicated. 18. Head, dorsal with antenna fur
ther enlarged. 19. Head, ventral. 20. Ventral apotome of young larva. 21. Right foreleg, with plumose and bladelike setae further
enlarged. 22, 23. Right middle and hind legs. 24. Right mandible, dorsal view. 25. Left mandible, ventral view. 26. Pupal case
showing holdfast; a,b,c, membranes closing front aperture, end of part of case occupied by pupa and hind aperture respectively
(after Barnard 1934 fig. 160). 27. Immature pupa, female, habitus, with dorsal plates further enlarged showing hooks. 28. Apex
of abdomen; apices of anal rods (appendages) shown further enlarged, a. of specimen drawn, b. variations (after Barnard 1934
fig. 16m). 29. Left pupal mandible.

302

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

fringe, like that of trochanter including feathered setae and strong blade-like bristles; tibia with smalt
distal ventral protuberance bearing two spinelike setae; tarsus smaller; claw stout, sharp, about same
length as tarsus, with long basal seta. Middle leg (Fig. 22) longer, hind leg (Fig. 23) longest, similar
in form, each with long, curved coxa, short divided trochanter (proximal part small, distal part apically
widened, rounded); femur shorter than tibia, noticeably so in hind leg; tarsi and parts of tibiae finely
pubescent in both; tarsal claw considerably shorter than tarsus with small basal seta part way along
claw.

Abdomen (Figs 1 4, 1 5, 30) smooth, creamy white, with segmental divisions shallow, setae lacking
or minute, no lateral fringe visible (feeble according to Barnard), gills absent; first segment with
flattened dorsal protruberance and lateral protuberances, lateral humps each with an oval faintly
sclerotized area bearing an anterior pubescent patch and a single seta; eighth segment with row of
about 24 lateral tubercles on each side; ninth segment with very pale dorsal sclerite bearing two pairs
of long posterolateral setae; apparent tenth segment (fused bases of paired anal prolegs) short, rounded,
not produced into apical lobes or projections, with few setae; anal prolegs with large curved lateral
sclerites almost meeting ventrally, bearing long apical setae, sclerotization variable with darker and
paler areas. Ventral sole plates with dark dorsal margin; anal claws small, strong, with one or two
stout, curved accessory hooks.

Larval case (Figs 16, 30)

Composed entirely of brighter or darker gold-coloured silk, tubular, widening anteriorly to the
slightly flared aperture; posterior aperture circular, centrally placed on raised membranous base.
Young larvae have a few sand grains incorporated near base of case; very early instars have a small
sand-encrusted basal section; this section is evidently cut off later. Length of case (mature larvae)
17-18 mm.

Note on identification of larvae and cases of Hydrosalpinx-, the golden, silken larval case is
distinctive and easily recognizable. Care, however, must be taken to identify the larva itself, as empty
cases are frequently utilized by larvae of Athripsodes species (Leptoceridae) which are usually found
further down stream below the habitat of Hydrosalpinx. The antermae of the Hydrosalpinx larva are
very small (Fig. 18) whereas those of Athripsodes are very long. Furthermore, Wte Athripsodes larva
usually adds a collar of sand grains to the case.

Description of pupa (Figs 26-29).

The description of the pupa is based on the single available specimen, a damaged immature
female, and on Barnard’s description and drawings.

Male and female pupae

Antennae longer than body. Labrum semicircular with short median lobe. Maxillary and labial
palps very long in male, somewhat shorter in female. Mandibles stout with broad base, strongly
dentate inner margin and two dorsal setae (Fig. 29). Middle tarsi fringed for swimming. Lateral fringes
present on seventh abdominal segment, curving round onto eighth, forming strong ventral tufts as
indicated (Figs 27, 28). Gills absent. Presegmental dorsal plates on second to sixth abdominal segments
(one hook per plate on second, two on third to sixth); postsegmental plates on fifth segment (three
anteriorly directed hooks per plate). Anal appendages rod-like, slender, with scattered setae; dorsally

303

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

curved apices serrated with apical finger and two stout setae (Figs 28a, b).

Pupal case (Fig. 26)

An altered larval case, fastened to a rock or stone by a single anterodorsal attachment disc. Hind
end closed by the larval membrane with a circular aperture. Posterior part of the case closed off from
the section in which the pupa lies by a new membrane with a narrow vertical opening. Anterior end
of case closed by a membrane with a transverse crescent-shaped slit on a central boss (Fig. 26c, b, a).

Distribution

South Africa, mountain streams in the coastal ranges in the western and southwestern Cape
Province.

Material examined
South African Museum material
Material in spirit

Lectotype: western Cape: Bain’s Kloof, east side [3319 CA] (KHB, 14.iv. 1-933, male).
Paralectotypes: Western Cape: Bain’s Kloof, east side [3319 CA] (KHB, 14. iv. 1933, 5 males,
one with head off and genitalia of two in microvials, 1 female abdomen lacking genitalia).

Other material: western Cape: Bain’s Kloof, east side [3319 CA] (KHB, 14. iv. 1933, an immature
pupa); Bain’s Kloof (larvae); Du Toil’s Kloof [3319 CB] (CWT andHGW leg., I.iv.l934, KHB det.,
6 males, one lacking abdomen).

All the specimens are faded and most are more or less damaged.

Pinned material

Paralectotypes: Western Cape: Bain’s Kloof (KHB, l.v.1933, 2 males); Hottentots Holland Mts
[3418 BB] (3000 ft. [= 914 m] KHB, March 1919, 1 female).

The males with wings spread are entire but the palps are much distorted, some curled up and some
broken. The female, the only one still available, had had the wings spread but one pair and the abdomen
had been removed and tipped onto card points by Barnard. The present author removed the fore wing
(which was damaged) and the abdomen from the card points. She mounted the fore wing on a celluloid
strip, removed, cleared and mounted the genitalia in Euparal on another celluloid strip and placed both
on the pin with the specimen. The hind wing was found to be almost entirely missing.

Albany Museum material

Material in spirit

Western Cape: Great Berg River, Fransch Hoek [Franschhoek] Forest Reserve, Assegaibos
stream tributary [3319 CC] (GBG 2k: ADH, 24.V.1950, larva), Assegaibos cold stream waterfall

304

SCOTT; THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

(GBG 132c: ADH, 20. xi. 1950, larva andGBG 752a: ADH, 26.xi.1953, larvae), source of Berg River
at Sneeugat 3319 CC (GBG 372a: ADH, 28. ix. 1951, larva); stream on Helderberg Mt., near Somerset
West, [3418 BB] (MISC 157c: DFH, iv.l963, 4 larvae).

Southern Cape: Vet River, Garcia’s Pass, north of Riversdale, tributary below Tolhuis [3421
AA] (MISC 306a: KMFS, 9.ii.l976, 23 larvae).

Material cited in literature

Barnard (1934: 323): Hottentots Holland Mts [3418 BB](KHB, 12 March 1919, 1 male, 1 female);
Wolwenhoek Kloof, French Hoek [Franschhoek] [3319CC] (KHB, April 1931, 1 male); Jonkershoek,
Stellenbosch [3318 DD] (HGW, February and April 1931, male and female pupae); Bosch Kloof
Keeromberg, Worcester [3319 DA] (KHB, January 1930, larvae); River Zonder End
[Riviersonderend] Mts [3419 BA] (HGW, December 1931, larva); Cedar Mts [Cedarbergen],
Clanwilliam [3219 AA] (KHB, January 1930, larvae); Du Toil’s Kloof, Rawsonville [3319 CB]
(KHB, March 1932, cases); Gt. Winterhoek Mts, Tulbagh [3319 CB] (KHB and HGW, November
1932, larvae); Bain’s Kloof, Wellington Mts, [3319 CA] (KHB, 14th April 1933, males; KHB and
HGW, 1st May 1933, males, females). A1 were deposited in the South African Museum, Cape Town.
The extant material is listed above under material examined.

Jacquemart (1963: 347): Kogelbaai 10 miles south of the Strand, [western Cape] [3418 BD],
shaded stream rushing down mountain slope, at road forming a waterfall, 19.xii.l950, larvae. The
identity of these larvae is given by Jacquemart in his text as Hydrosalpinx (sericea)l and in the caption
to his figure as Hydrosalpinx sericea. From a consideration of Jacquemart’s drawings the present
author concludes that they could be Hydrosalpinx, probably sericea. Unfortunately the cases were
neither drawn nor described.

Biology of the genus Hydrosalpinx

Hydrosalpinx sericea larvae inhabit high-lying, cold, acid montane streams in which they are
found in waterfalls, stony runs and riffles. Common in Barnard’s day, this species is now rarely found,
though it may still be locally common in remote areas. Its disappearance from some of the recorded
localities is more likely to be due to alteration of river beds by the building of reservoirs and constant
bush fires than to pollution, as this does not occur in them. As Barnard was a great mountain climber
many of his collecting places are inaccessible to collectors who are not mountaineers.

The adults, at rest and in flight, can easily be confused with sericostomatids such as Rhoizema
spp. being little smaller than the smaller species of Rhoizema and having somewhat similar colouring.
They have been collected in March, April and May.

The larvae may be seen grazing on the stones or rocks of the substratum. The gut contents and
mouthparts, particularly the mandibles with their small, blunt apical teeth and stout inner brushes,
indicate that the food comprises algae and detritus together with any animalcules and bacteria present.

305

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs. 30-32. HYDROSALPINGIDAE: Hydrosalpinx sericeu Barnard, mature larvae.

30. Photograph of 2 mature larvae in cases and one ex case. 31, 32. Photographs of head of mature larva further enlarged.
Photographs by J.C. Hodges, Jr. (Ex. MISC 306a).

306

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Family Petrothrincidae Scott 1985

Petrothrincidae Scott 1985: 331, 337-338; Scott 1986: 231, 234 (table 1).

Type-genus Petrothrincus Barnard 1934: 323, 325.

The family Petrothrincidae was erected to receive the single genus Petrothrincus Barnard. Three
species of Petrothrincus are known. Two occur in high mountain streams in the western Cape
Province, having been collected together by Barnard at 4 000ft [= 1 219 m] in the Hottentots Holland
Mountains. The distribution of circularis Barnard, however, extends further downstream than that
of triangularis Barnard. The third species was found in similar streams in the southern Cape Province.
It has been described in this paper as P. demoori.

Recognition

Imago small, dusky; fore wings may have light patches or appear plain grey. Maxillary palps 5-
segmented in male and female. Pronotum with two pairs of warts; mesonotum with median patch of
setae; scutellum with pair of anterolateral warts.

Family diagnosis

Male imago

Ocelli absent; eyes large, glabrous; vertex with pair of small interantennary setose warts and pair
of large posterior warts, without tubercles. Antennae stout, somewhat longer than fore wings; scape
about as long as head; flagellum tapered. Maxillary palps 5- segmented, fifth not annulate. Labial
palps 3-segmented, third simple, not annulate. Palps of normal length.

Pronotum with two pairs of warts; mesonotum without warts but with median field of recumbent
setae; scutellum quadrangular with pair of anterolateral warts. Tibial spurs 2, 2, 4. No black spinelike
setae on legs but colourless spinelike setae are present. Fore wings with discoidal cell closed, thyridial
cell long, Cuj simple; forks 1, 2, 3 present, Cu2 complete. Hind wings with costal area proximally
broad. Sc and Rj close together, sinuous, base of R2+3 and discoidal cell absent, forming median
discal area; M simple, joining Cu^ about midway; only base of Cu2 present. Wing coupling
macrotrichia present on basal part of costa in hind wings.

Genitalia with ninth tergite narrow, fused to tenth, which is somewhat produced; preanal
appendages prominent; paired spatulate claspers with internal branches arising from ninth; aedeagus
accompanied by paired parameres; ninth sternite produced.

Female imago

Similar to male but slightly larger than male; antennae somewhat shorter. Fore wings similar to
those of male, except for slight differences in base of anal veins. Hind wings also similar to those of
male in regard to curvature of Sc and R^ and broad basal costal area, however, base of R2+3 is almost
entirely absent; M is 2-branched; M and Cuj join basally; Cu2 is complete, as are lA and 2A.

Genitalia with tenth tergite with apical projection, partly covered by a lightly sclerotized dorsal
hood; a bilobed supragenital plate and simple or paired vulvar scales flanked by lateral striations, probably

307

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

modified pleural folds but at least partly sclerotized; appendages absent. Vagina small, sclerotized.
Larva

Case dweller; dorsoventrally flattened, widest at metanotum; head, pro- and mesonotum lightly
sclerotized; prosternal horn absent. Head round, frontoclypeal apotome with one pair of indentations;
antermae small, at base of mandibles; eyes small, fairly far forward; mouthparts stout, prominent;
ventral apotome sub-oval, completely separating genae. Metanotum membranous (or with very weak
median plate). Legs fairly long, with sparse long setae. Abdomen smooth, deeply indented ventrally;
first segment with very large dorsal hump, lateral humps lacking sclerotized areas or setae; lateral
fringe and lateral tubercles absent; gills present, small, branched or simple, varying according to
species. Anal prolegs longer than usual in cruciform larvae, not fused basally; anal claws small with
comb of small teeth.

Lar\>al case

Larval case of fine sand grains, more or less limpet-shaped, circular to suboval or triangular,
with ventral plastron supporting larva, and with posterodorsal terminal aperture.

Male pupa

Antennae much longer than body; labrum semi-circular; mandibles strong, triangular, with apex
blunt and inner margin feebly serrated; maxillary palps 5-jointed, palps reaching end of metathorax.
Fore and middle tarsi fringed for swimming; lateral abdominal fringes and gills absent; wing sheaths
reach end of seventh abdominal segment. First abdominal segment with small pair of posterolateral
lappets; third to sixth segments with presegmental dorsal plates; fifth segment with post segmental
dorsal plate; all dorsal plates small and weak. Anal appendages slender, the apical portion set with
blade-like bristles. No special pocket for male genitalia.

Female pupa

Similar to male but slightly larger; antennae much shorter than body, curled round at seventh or
eigth segment meeting ventrally.

Pupal case

An altered larval case, with a transparent ventral membrane sealing off the anterior end from the
plastron forwards, and sealed down around the edges onto the substratum (rock or stone).

Genus Petrothrincus Barnard 1934

Petrothrincus Barnard 1934: 323, 325-327, figs 17a-m, 18a-q.

Type-species P. circularis Barnard 1934, selected by Barnard, 1934.

Etymology: Generic name masculine, meaning coping-stones on a wall.

308

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Generic diagnosis
Male imago

Small in size, densely hairy, with long fringe on posterior margin of hind wings.

Ocelli absent; eyes large, glabrous; vertex with pair of small inter-antennary warts and pair of
larger rounded posterolateral warts; mid-cranial sulcus complete. Antennae stout, considerably longer
than fore wings; scape stout, nearly length of head; flagellum tapered. Maxillary palps of medium
length, 5-segmented, strongly pubescent, carried upwards over face. Labial palps 3- segmented,
pubescent. In neither maxillary nor labial palps is the last segment annulate.

Pronotum with two pairs of warts, median pair long, outer pair very small; mesonotum without
warts but with median area bearing short recumbent setae; scutellum quadrangular with large well
defined pair of anterolateral warts. Legs with tibial spurs 2, 2, 4; middle and hind tibiae with a few
colourless spine-like setae. Fore wings with discoidal cell closed, median cell open, thyridial cell
long; Sc and running parallel, meeting costa separately; forks 1, 2, 3 present; Cu^ and Cu2 meeting
margin separately; 1 A and 2A joining before meeting margin; jugal lobe, small inturned. Hind wings
with a basally wide costal area due to sinuous course of parallel Sc and R^; discoidal cell absent,
discal area vein-free due to absence of base of R2+3; M unbranched, stem fused with M-Cuj cross-
vein; Cu2 reduced.

Genitalia with tent h tergum fused with ninth, forming a blunt median process; preanal appendages
long, prominent; aedeagus simple with a pair of strong spines (parameres: Barnard’s titillators) lying
just above it; clasper stout, spatulate, with subapical indentation, long internal branch and small
uncinate spine (Barnard’s uncinate titillator); ninth sternum produced.

Female imago

Like male but antennae shorter than body and more slender.

Fore wings with slight differences in bases of anal veins. Hind wings differ in that M is 2-
branched, stem present, normal, as is Cu2 which reaches wing margin.

Genitalia with terminal tergum with apical projection; supragenital plate and vulvar scale(s)
present (hard to distinguish); sternites obliquely corrugated; no appendages.

Larva, pupa and their cases

See under family diagnosis and under individual species.

Distribution

South Africa, Cape Province.

309

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Petrothrincus circularis Barnard

(Figs 33-63)

Petrothrincus circularis Barnard 1934: 325, figs 17a-m, 18a-f, r, (male, female, larval and pupal
parts, larval case); Scott 1985: 331, fig. q.; Scott 1986: 236; Harrison and Elsworth 1958: 181,
table 16, as Molannidae; Harrison 1958a: 260, as Molannidae; Harrison 1958b: 311, table 65.

Lectotype male here selected and designated from Barnard’s syntypes. South African Museum, Cape
Town.

Type locality: western Cape Province, Table Mountain, Echo Valley, Cape Town, [3318 CD] (named
by Barnard as the Type locality, being the only place at which imagos were found.)

In regard to his choice of P. circularis as the type-species rather than P. triangularis (Hagen),
Barnard (1934: 325) comments as follows: “Although by rights triangularis Hagen should be made
the genotype, I have no adults from the type locality (Swellendam), and the true triangularis may
possibly, though not probably, prove to be different from the Great Winterhoek specimens which I
am describing as Hagen’s species. In that case the Great Winterhoek specimens would require a new
name. I therefore make circularis the genotype”. Hagen described his species from larval cases only
(Hagen 1864; 225), as Molanna triangularis.

Barnard treated his specimens of circularis as syntypes. Many are missing. From those that
remain the best available male has been selected and marked Lectotype by the present author. This
specimen is in spirit. It is the only male with complete antennae and has all other parts present except
for one hind leg. The other extant specimens mentioned in Barnard’s paper are regarded as
Paralectotypes. All specimens are faded and damaged. The extant material and original records are
listed after the descriptions of the different stages of circularis.

Description of imagos (Figs 33-46).

Description of male, female imagos as in generic diagnosis, with the following additional notes
made from the imagos, pinned and in spirit, from Barnard’s collection, material in the Albany Museum
collection, and checked against fresh material received from Dr H. Malicky in 1988 (MISC 311e-j),
from which Figs 41, 42 and 45 were drawn. New material (in spirit) appears dusky grey.

Male imago (Figs 33-42)

Small, fore wings 4. 8-6. 5 mm in length.

Colour, probably of fresh material, described by Barnard (1934: 325) as follows: “Head and
thorax fuscous with silvery-white hairs. Legs and palps grey. Antennae dark brown. Wings grey-

brown, fore- wing with silvery-white hairs in patches [as indicated in Fig. 35]. Abdomen orange-

fulvous, the tergites and sternites darker brown.” In old pinned specimens the wings appear faded
golden brown mottled with pale silvery patches.

Ocelli absent; eyes large, glabrous; vertex with warts all bearing very long setae; mid-cranial
sulcus complete; face with a pair of lateral warts bearing long setae and paired mesal patches of short
white setae. Antennae stout about a third longer than fore wings; scape stout with long setae, nearly
length of head; flagellum finely pubescent, tapered. Maxillary palps (Fig. 34) of median length.

310

SCOTT; THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs. 35-46. PETROTHRINCIDAE: Petrothrincus circularis Barnard, male, female (Scale lines = 1 mm unless otherwise indicated).
Material used: KHB (Echo Valley), MISC 311h (Female)

33 male; head and thoracic nota, dorsal. 34. male face, showing maxillary and labial palps. 35. male: entire insect, brownish setae
indicated, spaces on wings filled with white setae. 36, 37. male: fore and hind wings. 38, 39, 40, 41 male; genitalia, dorsal, ven-
tral and lateral views, stylized sketch of caudal view showing relative positions of parts (a - aedeagus, cl - clasper, ic/icl - internal
branch of clasper, p - paramere, pa - preanal appendage, us - uncinate spine, x - hood formed by tenth tergum). 42. male: inner pro
cess of clasper and uncinate spine drawn under compound microscope x 400. 43, 44. female; fore and hind wings (after Barnard
1934, fig. 17a). 45. female: genitalia, end view from new material. 46. female, ventral view (after Barnard 1934, fig. I8f, proba
bly actually a caudal view - dorsal and ventral views do not show vulvar scale etc.).

311

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

densely pubescent, 5-segmented, carried upwards over face; two basal segments short, rest subequal
but third longest and fifth shortest, simple. Labial palps (Fig. 34) 3-segmented, with segments
subequal; third segment simple, finely pubescent.

Pronotum narrow, with median pair of warts long, lateral pair small, round, hard to see unless
long setae are still present (they are easily lost); mesonotum with a quadrate median area bearing
recumbent white setae; scutellum with warts bearing long setae. Legs hairy; middle and hind tibiae
with a few colourless spinelike setae, no black ones. Wings (Figs 35-37) with venation as in generic
description and as illustrated in the figures.

Genitalia (Figs 38-42) with terminal tergum triangular; preanal appendages large, curved, with
median expansion, distally curved outward and upward; aedeagus simple with narrow apex and
subapical membranous area with a pair of long parameres arising just above it; paired strongly
spatulate claspers each with a long inner branch with basal expansion and small simple uncinate spine
(Figs 41 and 42); apex of clasper bifid; ninth sternum apically truncate, slightly indented.

Female imago (Figs 43-46)

Similar to male in coloration and general appearance but slightly larger, fore wings 5. 4-7. 5 mm
long.

Antennae shorter than in male, slender.

Fore wings practically identical to those of male except for basal parts of anal veins; hind wings
show considerable differences. Hind wings with M two-branched, joins Cuj at base’Cuj simple, Cu2
complete, anals separate (Fig. 44 compare with Fig. 37). Hind wings have a large patch of blackish
hairs on the lower surface; these tend to stick to the egg mass when laid.

Genitalia (Figs 45, 46) having terminal terga (ninth + tenth) with dorsal projection the shape of
which depends entirely on angle of viewing; as in Fig. 45, 46 or 72 (of triangularis). Sternites
membranous, showing lightly sclerotized ridges resembling pleural folds; single median vulvar scale,
presumably covering vulvar opening, lateral to which the supragenital plates can be seen. Vagina
clearly visible.

No good adult female of Barnard’s material was available for comparison with his drawings, of
which his Fig. 18 f. has been reproduced here as Fig. 46. There is, however, a good female pupa,
which coincides well with them. The genitalia of one of the two females collected by Dr Malicky
have been cleared and drawn (Fig. 45) for comparison with those of P. triangularis (Figs 68, 72).
See also the SEMs of P. demoori female genitalia, which help considerably in elucidation (Figs 82-
85).

Description of mature larva (Figs 47-57 and 61-63).

The following description and drawings were made from Barnard’s own specimens and the more
recently collected larvae in the Albany Museum (all material in spirit). Comparisons were made with
Barnard’s text.

Case dweller; up to 7 mm in length; dorsoventrally flattened; head, thoracic nota and legs light
brown to yellowish or pallid in colour, with patterning of brown spots or marks, often faint; abdomen
white.

Head (Figs 47-49, 62, 63) hypognathous, rounded; frontoclypeal apotome with single pair of
lateral indentations; ventral apotome suboval, completely separating the genae. Eyes situated well
forward, each under four cuticular lenses. Antennae small, near mandibular bases. Labrum small,

312

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs. 47-60. PETROTHRINCIDAE: Petrothrincus circularis Barnard: 47-57. mature larva; 58-60. male pupa.

(Scale lines = 1 mm unless otherwise indicated).

Material used: MISC 291a (Figs 47-57); KHB (Figs 58-60).

47. Habitus of mature larva, dorsal. 48, 49. Head, dorsal and ventral, antenna much enlarged. 50. Thoracic nota, dorsal. 51. Left
anal proleg, outer view. 52, 53, 54. Right fore, middle and hind legs, fore leg with pleural sclerites, middle and hind legs with
trochanteral setulae further enlarged. 55. Right mandible of newly ecdysed larva, ventral. 56, 57. Right and left mandibles of older
larva, ventral. 58. Habitus, dorsal, dorsal plates further enlarged. 59. Anal appendages, ventral, further enlarged. 60. Left
mandible, dorsal.

313

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

rounded, prominent, with very small, paired indentations. Anteclypeus large, colourless. Labium
thick with very small pair of palps. Maxillary palps apically small with large stout bases. Both
mandibles with inner brush, in recently ecdysed mature larvae long, with slightly serrated ventral and
apical cutting edges and with dorsal edge heavily sclerotized, plain; in older larvae the apical third
may be worn away (compare Fig. 55 with Fig. 56), presumably due to grazing on the rock surfaces
on which they live.

Thorax (Figs 47, 50) with prosternal horn absent. Pronotum sclerotized, without carina or
anterolateral expansions; median suture present, sometimes faint or partial; pattern of muscle spots,
often inconspicuous. Mesonotum with a pair of large pale yellow sclerites which may show brownish
spots indicating Barnard’s “three to four pairs of small chitinous plates medianly”; the median spots
tend to run together forming paired median stripes; mesonotal plates shiny and lightly pitted; median
ecdysial line faint or partial. Metanotum very broad, widest segment, with small paired posterolateral
lobes of unknown function, with possible quadrangular median sclerite, hard to distinguish, shining
and lightly pitted, colourless, with pair of vague brownish spots, segment otherwise membranous
(this possible sclerite is not usually visible apart from the spots), with paired raised diagonal strips
(Fig. 50) apparently overlying paired muscles. Sternum membranous with minute paired ventrolateral
extensions of pronotum. Pleural sclerites largely colourless; pre-episternum long, sclerotized, with
upturned tip, wide in lateral view, narrow in dorsal view (Fig. 52). Fore legs (Fig. 52) short, stout,
with few long setae except on trochanter and femur which have a ventral fringe of long, simple setae;
middle legs (Fig. 53) longer; hind legs (Fig. 54) longest; both middle and hind legs slender in dorsal
view, wide in lateral view, with long setae but no ventral brushes; middle and hind trochanters with
minute setulose marginal spines. Fore claw as long as tarsus, middle and hind claws little longer than
fore claw, each with a single seta part way along claw.

Abdomen (Figs 47, 51) with lateral line fringe and lateral tubercles absent; a very narrow lateral
fold along each side; segments not indented dorsally though deeply so ventrally (presumably to
increase the respiratory surface); several segments with small, stout gills; first segment very broad,
with wide flattened dorsal hump and large lateral humps; second segment with three pairs gills of
which dorsolateral gills simple or bifid, lateral and ventrolateral ones 3-4 branched; third to fifth
segments with a single ventral pair of branched gills; sixth segment with a simple pair; anal prolegs
with colourless lateral sclerites and longer than is usual in eruciform larvae, fused only partially, not
really forming an apparent tenth segment; anal claws very small with neat dorsal comb of teeth (Fig.
51).

Larx’al case (Fig. 61)

Limpet-shaped, subcircular, of minute sand grains cemented and lined with silk secretion, with
ventral shelf (plastron) of sand grains supporting the larva, the spaces lateral to this being filled in
with sand grains loosely held together by secretion, with a small oval posterodorsal aperture. The
species name is derived from the subcircular larval case.

Description of pupa (Figs 58-60).

Male pupa

Antennae much longer than body, at least one and a half times body length, curled round as
shown in Fig. 58. Labrum semicircular. Mandibles strong, triangular, with the apex blunt and with

314

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

E'igs. 61-63. PETRO THRINCIDAE: Pelrolhrincus circularis Barnard, mature larvae.

61. Photograph of mature larva in case, ventral. 62. Photograph of head of larva, frontal view, further enlarged. 63. Photograph
of head and thorax of larva, dorsal, further enlarged. Photographs by J.C. Hodges, Jr. (Ex MISC 291a).

315

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

inner margin feebly serrated. Fore and middle tarsi fringed for swimming. Lateral fringe and gills
absent. Wing sheaths reaching end of segment 7. Segment 1 with a small pair of posterolateral lappets;
segments 3-6 with small presegmental dorsal plates each with one tooth and segment 5 with in addition
postsegmental dorsal plates each with two teeth (Figs 58, 58a). Anal appendages slender, slightly
wrinkled, with a few long setae and many small blade-like bristles, particularly apically. No special
pocket for male genitalia.

Female pupa

Similar to male but tends to be larger; antennae shorter, ends curled, meeting ventrally about
segment 7 or 8.

Pupal case

The larval case is utilized, being sealed down onto a rock or stone round the margin. The larva
spins a transparent ventral membrane sealing off the front end anterior to the plastron. Emergence is
through a semicircular anterodorsal aperture in the case, cut by the pupal mandibles at eclosion.

Distribution

South Africa, high mountain streams in the coastal ranges in the Cape Province, mainly in the
western Cape, but also in the southern Cape. There is also a single isolated record from the vicinity
of Rhodes in the Witteberg Mountains, southern outliers of the Drakensberg, eastern Cape Province
(Jacquemart 1963), but this may prove not to be of P. circularis, as only larvae were collected, and
the larval cases are very similar to those of P. demoori.

Material examined

South African Museum material

Pinned material

Paralectotypes: Echo Valley, Table Mountain, Cape Town [3318 CD] [2 990 ft = 911 m] (KHB,
February-April 1933, 5 males in fair condition, one lacks two wings, one lacks one wing and abdomen,
antennae are broken or missing and most palps are damaged or lacking, 9 females in fairly good to
very poor condition, two lack genitalia).

Material in spirit

Lectotype: Echo Valley, Table Mountain, Cape Town [3318 CD] [2 990 ft = 911 m] (KHB,
February-April 1933, male).

Paralectotypes: Echo Valley, Table Mountain, Cape Town [3318 CD] [2 990 ft = 911 m] (KHB,
February-April 1933, 7 males in good to poor condition, bits of others and a cleared male abdomen,
a single female body with thorax, legs and genitalia, the latter badly damaged).

Other material: Echo Valley, Table Mountain, Cape Town [3318 CD] [2 990 ft = 911 m] (KHB,
February-April 1933,3 immature pupae andlarval and pupal exuviae); Hottentots Holland Mountains,
[3418 BB] 4 000ft [= 1 219 m] East side (KHB, January 1933, larvae and pupae in cases).

316

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Albany Museum material
Material in spirit

Western Cape Province: Great Berg River, French Hoek [Franschhoek] Forest Reserve, main
stream [3319 CC] (MISC 288: KMFS, 30.i.76, larvae); Assegaibos tributary [3319 CC] (GBG 131b:
ADH, 30.x. 1950, larvae); same place (MISC 289: KMFS, 30.1.76, larvae). Riviersonderend, French
Hoek [Franschhoek] Pass [3319 CC] (MISC 290: KMFS, i and ii.l976, larvae); Witte River,
Wolwekloof tributary [3319 CA] (MISC 291: KMFS, i and ii. 1976, larvae); Breede River, Elandspad
tributary [3319 CA] (MISC 292: KMFS, i and ii.l976, larvae); Witte River, Bastiaanskloof tributary
[3319 CA] (MISC 295c: KMFS, i and ii. 1976, larvae); Langrivier, mainstream (MISC 297a: KMFS,
i and ii.l976, larvae) and Assegaibos stream (MISC 298d: KMFS, i and ii.l976, larvae), both
tributaries of the Eersterivier [3318 DD]); Witte River, Happy Valley [3319 CA] (MISC 304a: KMFS,
i and ii.l976, larvae) ; Langrivier [3318 DD] (JMK, iii.75, larvae and pupa, vi.83, larva, x.83, 2 larvae,
xi.83, 2 larvae); small feeder streams of the Dutoitsrivier in the Mount Rochelle Nature Reserve,
Franschhoek Pass [3319 CC], at 900-1 000 m altitude (MISC 311 e-g, and h and], HM, 20.iii.88, 3
male imagos and 2 female imagos, at light).

Material cited in literature

Barnard 1934: 325. Type locality: Table Mt. Cape Town [3318 CD] (KHB February - April,
males, females). Groot Drakenstein(ACH); Jonkershoek, Stellenbosch [3318 DD] (KHB and HGW);
Witte River, Wellington Mts [3319 CA] (KHB, November 1922); River Zonder End Mts [3319 AB]
(HGW); Palmiet River [3418 BD] (HGW); Hottentots Holland Mts [3418 BB] (KHB); Great
Winterhoek Mts (KHB and HGW); French Hoek [Franschhoek] Mts [3319 CC] (KHB and HGW);
Elands Kloof, Citrusdal [3219 CA] (HGW).

These localities are all in the western Cape coastal folded belt. Contrary to his usual custom,
Barnard did not give dates under each entry; these do appear on the labels with the specimens, but
where specimens are missing there is now no possibility of dating them. The extant material is listed
above under material examined.

All the records, apart from the Table Mountain and Hottentots Holland specimens were evidently
of larvae only. The Hottentots Holland material also included many pupae, unfortunately immature.

Jacquemart (1963): western Cape Province: near Hermanns [3419 AC], twin waterfalls on
Maanschijnkop (21.xii.l950, larvae); at Hermanns [3419 AC], stream (22.xii.1950, larvae); near
Grabouw [3419 AA], Viljoenspas (ll.ii.l951, larvae). Eastern Cape Province: 5 miles E.N.E. Rhodes
[3027 DC], (on Witteberg Mountain, an outlier of southern Drakensberg Mountains), high mountain
stream (lO.iii. 1951, larvae). The larvae from the eastern Cape may prove not to be of circularis, when
male imagos are available. A recent search of the area by Dr F.C. de Moor and Miss H.M. Barber
failed to discover any, but a further search will be made, as it is important to know if it is indeed
circularis or demoori or another species.

317

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Petrothrincm triangularis (Hagen)

(Figs 64-73)

Molanna triangularis Hagen 1864: 225 (larval case only).

Petrothrincus triangularis (Hagen), Barnard 1934; 325, 327, figs 18g-q (male, female, case); Barnard
1940: 643 (cases); Harrison 1958a: 260.

Neotype male here selected and designated from Malicky’s specimens, Albany Museum,
Grahamstown.

Type locality: Swellendam, western Cape Province [3420 BB].

Neotype locality: Dutoitsrivier in Mount Rochelle Nature Reserve, Franschhoek Pass, western Cape.

The name, Molanna triangularis Hagen 1869, was validly published in spite of the fact that the
only material available to Hagen was an empty larval case because to satisfy the provisions of the
International Code of Zoological Nomenclature (1985) a name published before 1931 need only be
based on an “indication”, in this instance “the description of the work of an extant animal” (Criteria
of Availability, Article 12). The Type specimen, an empty larval case, is presumed lost.

As Hagen had no imagos Barnard described them from his own material but, as was his wont,
did not designate a type. Barnard’s extant material and original records are listed after the descriptions
of the different stages. The extant male lacks its genitalia and so cannot be named a Neotype. Two
males and two females were received recently from Dr H. Malicky. These were collected at light,
together with circularis imagos, from small feeder streams of the Dutoitsrivier in the Mount Rochelle
Nature Reserve, Franschhoek Pass, western Cape [3319 CC], at 900-1 000 m altitude. The best male
(MISC 311a) is here designated the Neotype. It is lodged in the Albany Museum.

Description of imagos

Imagos (Figs 64-68, 71-73)

According to Barnard (1934) the coloration of the imagos and the wing venation are as in
circularis, excepting that the fore wings appear to have more patches of silvery white setae, and the
setal fringes are very long, particularly on the anal margin of the hind wings. (Setae fringing wings
in circularis may be equally long but may have been lost or abraded in the extant material.) The recent
light trap specimens have, unfortunately, lost most of their setae, including wing fringes. The length
of the male fore wings is 5. 5-6.8 mm and that of the female is 6. 0-7. 9 mm. The descriptions are based
on Barnard’s descriptions and drawings and remaining female imagos and upon the material collected
by Malicky. There are very minor differences in wing venation between triangularis and
circularis, not sufficient to affect the generic diagnosis or even to differentiate between the species
in the absence of genitalia.

The male genitalia (Figs 64-66, 71) are compared with those of circularis (Figs 38 - 42). They
have the ninth tergum ovate not triangular; preanal appendages longer and wider, curved in lateral
view, lacking the median expansion, but with a small basal expansion; aedeagus with apex indented,
and two small sclerotized ventral processes (Fig. 65a), the parameres are shorter and stouter with

318

SCOTT; THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs. 64-73. PETROTHRINCIDAE: Petrothrincus triangularis (Hagen), male, female, larva
(Scale lines = 1 mm unless otherwise indicated).

Material used: MISC 311b (male), MISC 311 c and d, and KHB (female), MISC 290b and KHB (larva), King 31 (case).

64-66. Male: genitalia, dorsal, ventral and lateral views. 65a; Ventral view of aedeagus. b, c; variations in shape of spine (titillator) and ninth stern-
ite. 67, 68. female: genitalia, lateral and ventral views. 69. Mature larva: head, dorsal. 70, 70a. Larval case: ventral and lateral views. 70b, pos-
terodorsal end of case to show membrane in which distal opening lies. (Figs 64-68 after Barnard 1934, figs 18 g-n.) 71-73. Drawings made from new
material off. triangularis. 71. Male: genitalia, lateral (compare with Fig. 66); 71a. Uncinate process (x 400); 71b. Ninth sternite. 72,73. Female: gen-
italia, caudal and skewed lateral views; 72a. Variation in shape of tergite. 72b. Supra- genital plate (not to sam.e scale), (a - aedeagus, cl - clasper,
ic/ici - internal branch of clasper, p - paramere, pa - preanal appendage, sp -supragenital plate, v - vulva, vg - vagina, vs- vulvar scale, x - hood formed
by tenth tergum xt - tenth lergum)

319

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

straight apices and wide bases. The claspers are spatulate but longer and basally narrower in lateral
view, their internal processes longer, the uncinate spines larger, with basal part apparently coiled like
a spring (Fig. 71a). The ninth sternum is widely bifurcate with the arms varying in size (compare
Figs 65, 65c, 71, 71b).

The female genitalia (Figs 67, 68, 72, 73) are compared with those of circularis (Figs 45, 46).
They have the terminal dorsal projection stronger, sub-ovate to subtriangular, not apically indented;
paired lateral vulvar scales, each bifid; large median vulvar opening, clearly visible. The supra geni-
tal plate is transparent, narrowly bifurcate (Fig. 72b). The vagina is small, sternites apparently showing
lightly sclerotized pleural folds.

Setae under hind wings of female are pale, not blackish as in circularis.

In comparing Fig. 72 {triangularis) with Fig. 45 (circularis) note that the scale is the same, but
one is a very large individual, the other very small. Although triangularis is the larger species, the
difference in size range is not great.

Description of larva (Figs 69-70)

The larvae fit tightly into their cases and are therefore difficult to extract and as a result their
abdomens are frequently not well preserved. The best available larvae (collected from the Langrivier
(W Cape) by Dr J.M. King) are those which were preserved in Kahle’s fluid immediately after
collection but even in these the metanotum and some sterna are poorly preserved and their structure
unclear, as are the gills.

The largest available larvae are 6-7 mm in length.

The larva (Fig. 69) is as in circularis but the sclerotized parts are more strongly coloured and
easier to see, the colour pattern is clearly distinguishable, somewhat different from that in circularis
(compare Fig. 69 with Fig. 48). In Barnard’s material the head and pronotum are brown with faint
patterning. In fresh specimens the head shows strong markings similar in type to those in circularis
but the spots lateral to the frontoclypeal apotome are confluent and the central spots larger and bolder.
The mandibles, as in circularis, are long in newly ecdysed specimens and worn down in older ones
(compare with Figs 55-57). The thoracic nota also show minor differences in patterning. The pronotum
has a clear median division and the mesonotal plates are also clearer, as is the metanotum. The larvae
do not show a possible metanotal sclerite. No branched abdominal gills were found, but a single pair
of simple ventral gills is present on segments 2-6.

Larval case

The larval case (Figs 70, 70a, 70b) is similar in type to that of circularis, but is narrowly triangular
in shape, giving rise to the specific name. The sides are nearly straight, the tube wide, round, and the
posterodorsal aperture has a central opening in the terminal membrane (Fig. 70b), in some instances
on a small chimney. Barnard’s drawings of the case must have been made from an unusual specimen,
as an incorrect impression is given of the lateral flanges. The case has therefore been re-drawn after
comparison with Barnard’s original material and the Albany Museum specimens of which there are
many.

320

SCOTT; THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Description of pupa

The pupae like the larvae fit tightly into their cases and are therefore difficult to extract and as
a result their abdomens are frequently not well preserved.

The pupa is similar to that of circularis. TTe case is similarly attached to the substratum and
widely opened anterodorsally by the pupal mandibles when the pharate imago emerges.

Distribution

South Africa, western and southwestern Cape.

Material examined
South African Museum material
Pinned material

Western Cape: Great Winterhoek Mountains, 4 000 ft. [= 1 219 m] [3319 AC] (KHB and HGW,
iii.1933, 1 male without genitalia, 2 females).

Material in spirit

Western Cape: Great Winterhoek Mountains, 4 000 ft. [= 1 219 m] [3319 AC] (KHB and
HGW, i. 1933, 10 larvae in cases, 47 immature pupae, most in cases).

Albany Museum material

Material in spirit

Neotype: western Cape: feeder streams of the Dutoitsrivier, Franschhoek Pass (3319 CC) (MISC
311a: HM, 20.iii.l988, male, to light)

Other material: western Cape: Great Berg River, Assegaibos main stream [3319 CC] (GBG 40c:
ADH, 2.viii.l950, larvae); Great Berg River, source, 3 500 ft [= 1 067 m] [3319 CC] (GBG 134e:
ADH, 21. xi. 1950, larvae); same, 3 000 ft [= 914 m] (GBG 134e: ADH, 21. xi. 1950, larvae); Upper
Langrivier, tributary of Eerste Rivier [3318 DD] (JMK, iv.l975, 1 pupa; iv-xi.l984, many larvae,
and in iv and v. 1975, empty pupal cases); Riviersonderend, French Hoek [Franschhoek] Pass [3319
CC] (MISC 290b: KMFS, 30.i.l976, larvae); feeder streams of the Dutoitsrivier, Franschhoek Pass
(3319 CC) (MISC 311b-d: HM, 20.iii.l988, 1 male and 2 females, to light). South western Cape:
Vetrivier tributary high on Garcia’s Pass, [3321 CC] (MISC 306K: KMFS, 9.ii.l976, larvae).

Material cited in literature

Hagen 1864: 225. Swellendam [3420 BB] (larval case) (as Molanna triangularis Hagen).
Barnard 1934: 325. Swellendam (Hagen, case only); Great Winterhoek Mts., Tulbagh (KHB and
HGW, November 1932, larvae, males and females bred out March 1933); Jonkershoek, Stellenbosch
[3318 DD] (HGW); Witte River, Wellington Mts [3319 CA] (KHB, 1922); Bain’s Kloof, Wellington

321

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Mts (KHB and HGW, 1 May 1933, females); Hottentots Holland Mts (KHB and HGW, 1932); Du
Toit’s Kloof, Rawsonville (KHB, 1932); Mostert’s Hoek and Waai Hoek Mts [3319 AD] (KHB,
April 1933); Bosch Kloof, Keeromberg, [33 19 DA], Worcester (KHB, 1930); Cedar Mts, Clanwilliam
[3219 AC] (KHB, 1930); French Hoek [Franschhoek] Mts [3319 CC] (KHB and HGW, 1932).

Barnard 1940: 643, Valsch Gat stream on Ceres side of Matroosberg, Hex River Mts, western
Cape [3319 BD] (KHB, November 1933, empty cases).

The extant material is listed above under material examined. The only adults Barnard had were
males and females bred out from larvae and pupae collected on the Great Winterhoek Mountains at
4 000 ft. [= 1 219 m] [3319 AC]. Of those only two females remain. One is complete and in good
condition, with an egg mass adhering to and distorting the apex of the abdomen. One antenna is
unbroken. It is almost as long as in the newly collected males. The other female has lost two wings
and the abdomen.

Petrothrinciis demoori sp. nov.

(Figs 74-85)

Holotype male (SCR 4A) here selected and designated, from material in the Albany Museum.

Type locality: southern Cape, Plaat River, tributary of the Karatara River, at Klein Plaat se Brug
(33°53'20"S, 22°50'45"E, altitude c 280 m.

Ten male and six female imagos collected using a light trap at the type locality and putative
pupae (SCR 2A and 3C) and putative larvae (SCR 3B) collected from Jubilee Creek, a tributary of
the Homtini River (all by Dr F.C. de Moor and Miss H.M. Barber) were at first thought to be
P. circularis on account of their subcircular larval cases. Both streams are in the same general area
in the southern Cape nowhere near the habitats of the two hitherto known species of Petrothrincus.
On examination the males proved to be a new species of Petrothrincus. The larval cases are subovate
rather than subcircular (Fig. 81) but there is certainly an overlap between the cases of the two species.
The species is named after Dr de Moor.

After study of the specimens, a Holotype male (SCR 4A) was selected and designated by the
author. The remaining nine males (SCR 4B, C, D and 4E with 6 males) were designated Paratypes,
as were the six females (SCR 4F, 4G - used for S.E.M. micrographs - and 4H with 4 females). The
Holotype and the Paratype males and four of the Paratype females are all lodged in the Albany
Museum. The remaining two Paratype females were sent to Dr A. Neboiss, Museum of Victoria,
Melbourne, Australia.

Description of imagos (Figs 74-78, 82-85)

The adults are somewhat teneral and in spirit. They are very similar in general appearance to
circularis and triangularis: small, dusky grey insects, length of fore wings 5. 4-6.0 mm in males, 5.8-
6.8 mm in females. The wings show little trace of lighter and darker patches but this may be due to
the teneral nature of the material. The wing venation is very like that in the other two species with a
few very minor differences that do not affect the generic diagnosis. In the fore wings, there are,
however, long hyaline streaks along the R3+M cross-vein, the base of the R4^5 fork and the adjacent
part of the lower margin of the discoidal cell, also along M for almost the whole length of the thyridial
cell. Such hyaline streaks were not seen in the other two species. Some cross-veins in the hind wings

322

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs. 74-81. PETROTHRINCIDAE: Petrothrincus demoori sp. nov., male, female, larva, pupal case
(Scale lines = 1 mm unless otherwise indicated).

Material used: SCR 4A, 4E (male); SCR 4H (female); SCR 3B (probable larvae); SCR 3C (probable pupal case).

74-76. Male genitalia. 74. Dorsal view, omitting aedeagus. 75. Ventral view, winged stemites and claspers with internal processes. 75a,
Same; stemites and aedeagus (broken lines), parameres and claspers. 76 Lateral view: preanal appendage, clasper and internal process,
uncinate spine, parameres, aedeagus and left side of sternite, 76a, Internal process and uncinate spine further enlarged 77,78. Female
genitalia. 77. Caudal view, showing sclerotized ring, supragenital plates, vaginal apparatus, etc. 77a. Vaginal process raised up; also
setulose membranous covering, shown as seen. 78. Sketch of genitalia as seen in end view; note sclerotized pleural folds resembling a
pile of saucers. 79, 80. Larva. 79. Head, dorsal. 80. Thorax, dorsal. 81. Pupal case showing opening left by pharate imago after
eclosion 81a. Lateral view of chimney-like posterior aperture of case, (a - aedeagus, cl - clasper, ic/ici - internal branch of clasper, ixs
- ninth sternite, ixt - ninth tergum, m - setose membrane, p - paramere, pa - preanal appendage, pi - pleura, sp - supragenital plate, us -
uncinate plate, v - vulva, vg - vagina, x - hood formed by tenth tergum xt - tenth tergum)

323

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

are faint, but both these facts could be due to the teneral nature of the material.

Male genitalia (Figs 74-76) are in general typically petrothrindd. They differ from those of the
other species as indicated below. Preanal appendages lack both the median expansion of circularis
and the basal one of triangularis, and are not strongly angled; they are minutely setulate. Parameres
are basally broad, not narrow, in lateral view (Fig. 76). Claspers are more broadly quadrangular than
in either of the other species, with the apical excision small; left and right claspers are very similar,
whereas in circularis one is larger than the other. The internal process of the clasper is much like that
of triangularis but the uncinate spine resembles that of circularis excepting that it has an additional
basal process tipped with a small seta (compare Figs 42 and 71a with Fig. 76a). The aedeagus is very
long with a lengthy narrowed apical section bearing a membranous expansion (Fig. 76). The lateral
processes of the ninth sternum are strongly winged unlike those of the other two species (Figs 75,
75a).

Female genitalia (Figs 77, 78, 82-85), as might be expected, are very similar in general type to
those of circularis and triangularis. As in those species, they are best seen in caudal view.

The dorsal process and vulva resemble those of triangularis, but the lateral striated areas are
wider and there appear to be paired vulvar scales (Fig. 84). In caudal view (Fig. 78) these striated
areas resemble a pile of saucers. The dorsal process is lightly sclerotized dorsally and this sclerotization
extends laterally, embracing the sides of the genitalia (Fig. 77) as in triangularis. There is a broad,
apically bifid supragenital plate and the ninth tergum and sternum together form a sclerotized ring
(Fig. 77). The vagina is hour-glass shaped.

The species are not easily distinguished from one another unless cleared and checked against
the figures (compare Figs 45, 72 and 77).

The female supragenital plate was first clearly seen in this species (Fig. 77) (its appearance differs
when seen at different angles) and was then found in triangularis (Fig. 72b) and in circularis (Fig.
45). Scanning electron micrographs of the female genitalia (Figs 82-85) shed further light on the
structures. There is an apparently lightly sclerotized, bilobed setose membrane, shown in Fig. 77a,
the function of which is unclear. As seen, it lies above everything else, being very clear in Figs 82-
85. Also very clear are the ridges on the supragenital plate, and what appear to be the paired vulvar
scales (Fig. 84) between the arms of the supragenital plate and the membranous lobe, which is actually
ventral to them.

Description of putative larva (based on the three available larvae in cases, SCR 3B) (Figs 79, 80,).

Length of mature larva 6.0 - 6.6 mm.

The head pattern (Fig. 79) is of the same type as that of circularis, though the frontoclypeal
pattern is nearer that of triangularis (compare Figs 48, 69 and 79). The pattern on the pronotum shows
many more spots than in circularis; the mesonotum is very pale, showing a few very small spots, the
mid-line is clearly visible; the metanotum is entirely membranous. Abdominal gills resemble those
of circularis but have longer branches. There are three pairs on the second abdominal segment (2-4-
branched), two pairs on the third (2-branched) and one pair each on the fourth and the fifth (2-
branched).

Larval case

Length of case 6.5 - 7.5 mm; sub-ovate, similar to that of circularis but on average narrower
relative to length. There is a definite overlap in size and shape between the two species. The larval

324

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs. 82-83. PETROTHRINCIDAE: Pelrolhrincus tlemuori sp. nov., female.

Material used: SCR 4G (subsequently lost).

82. Female genitalia, caudal view. Scale line 100 pm. Seen in situ with sternites, terga and pleura surrounding it. 83. Same. Scale
line 100 pm. Note apex of tergum X on right, supra genital plates, and between them, vulvar scales; above them setose
membrane.

325

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

case is exactly like the pupal case except for the absence of the neat escape aperture (Fig. 81).
Description of putative pupa (SCR 3C, 17 pupae).

The available pupae, all apparently still immature, resemble those of the other two species.
Pupal case (Fig. 81)

An empty pupal case is figured (Fig. 81) to indicate the size and position of the escape aperture
which is much neater than in triangularis in which the openings are evidently torn, not cut, leaving
a ragged hole with upstanding lateral flaps. The posterior opening is also figured (Fig. 81a) showing
the turret. In pupal cases the turret is filled with sand grains (not shown), whereas in the larval case
it is open. A turret is not always present.

Distribution

South Africa, mountain streams in the coastal range of the southern Cape. The original estuaries
and lower reaches of the rivers in this region have been drowned so that the present estuaries are
preceded by what was originally the middle or upper reaches of the rivers. Consequently the present
altiitudes of the tributaries are lower than they will originally have been.

Material examined

Albany Museum material

Material in spirit

Holotype: Southern Cape Province: Plaat River, tributary of the Karatara River, at Klein Plaat
se Brug (33°53'20"S, 22°50'45"E), altitude c 280m (SCR 4A: FCdM and HMB, 7-8.iii.1989, male).

Paratypes: Southern Cape Province: Plaat River, tributary of the Karatara River, at Klein Plaat
se Brug (33°53'20"S, 22°50'45"E), altitude c 280m (SCR 4B-H: FCdM and HMB, 7-8.iii.1989, 9
male and 4 female imagos, to light).

Two male Paratypes and one female Paratype will be presented to the South African Museum.

The two of the Paratype females not in the Albany Museum collection were sent to Dr A. Neboiss,
Museum of Victoria, Melbourne, Australia, for examination of their tentorial systems. He later wrote
(20.i.91) that the tentorium of demoori females looks very much like that in the Australian genus
Caloca (Calocidae), although there are also differences.

Other material: Southern Cape Province: Jubilee Creek (Station Jub. 1.), tributary oftheHomtini
River, 33°53'20'’S 22°58T5"E (SCR 2A and 3C: FCdM and HMB, 7. hi. 1989, 17 putative pupae and
15 empty putative cases; SCR3B: FCdM and HMB, 7.iii.l989, putative larvae); Jubilee Creek (station
at picnic spot) (FCdM and HMB, 7.iii.l989, 1 dead putative pupa and 1 empty putative pupal case).

Note: several Petrothrincus larvae (MISC 310c) were collected by NK from the Blaauwkrantz
(Bloukrans) River, below Staircase Falls [3323 DC], southern Cape. These larvae, however, differ
in colour from those of demoori, being plain pale brownish, with a few small, vague muscle spots,
very hard to distinguish, on the head, none on pronotum (plain brownish) and a few on mesonotum.

326

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Figs 84-85 PETROTHRINCIDAE: Pelroihrincus demoori sp, nov., female.

Material used: SCR 4H (subsequently lost).

84. Female genitalia, scale line 10pm Note lateral ridges on setose membrane, indicating partial sclerotization, and folds on vulvar
scales. 85. Same, scale line lOpm. Note ridges on supragenital plate, also pleural folds.

327

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

It is unclear whether these represent a colour variation or another species. The former could well be
the case as there is a brown colour variation in some of the triangularis specimens from the Langrivier
(Jonkershoek) material, the larval cases and abdominal gills are similar to those of circularis and
demoori but not of triangularis. Male imagos are needed for clarification.

Biology of the genus Petrothrincus.

The larvae of all three species are found only in clean, undisturbed streams. Larvae of circularis
and triangularis live in stony runs in high, cold, acid mountain streams. P. triangularis appears to
be restricted to the higher streams, particularly over 3 (XlOft [= 914 m] altitude, but circularis is also
found further down, as far as the upper foothills [300-1 000 ft = 91-304 m]. Excellent detailed accounts
of the Great Berg River and its tributaries (geology, water chemistry, collecting stations, with details,
vegetation and aquatic fauna) can be found in Harrison and Elsworth (1958) and Harrison (1958 a
and b). The southern Cape streams are being progressively disturbed by the planting of exotics, timber
extraction and bridge building.

P. demoori occurs in places at comparatively low altitudes. De Moor (pers.comm.) noted that
demoori appears to be restricted to small tributaries as it was not found in the main streams.

Petrothrincus larvae are algal grazers, also ingesting detritus and concomitant animalcules. They
are present throughout the year. Adults of P. circularis have been collected in February, March and
April, of P. triangularis in March and May, and of P. demoori in March. Oviposition has not been
observed. Barnard captured several circularis females with an egg mass, a round ball covered with
the long hairs from the underside of the female hind wings, still attached to the tip of the abdomen.

Barnard (1934) observed the duration of the pupal stage in circularis and triangularis finding
that it lasted from three to four weeks in February and March.

Key to the species of the genus Petrothrincus

Male Imagos

1. Preanal appendages strongly angulate, claspers wider basally than apically, internal branch of

clasper with basal lobe, uncinate spine simple; ninth sternum blunt-ended, slightly indented
apically (Figs 38-42). P. circularis

Preanal appendages not angulate, claspers evenly wide, internal branch of clasper lacking basal
lobe, uncinate spine not simple; ninth sternum bifid (Figs 64-66, 71) 2

2. Preanal appendages with basal lobe, without setulae; claspers narrow; uncinate spine with coiled

base; ninth sternum narrow with narrow branches of variable length; aedeagus short, blunt (Figs
74-76). P. triangularis

Preanal appendages without basal lobe, with setulae; claspers broadly quadrangular; uncinate
spine without coiled base but with setate basal process; ninth sternum with winged lobes;
aedeagus very long, apically narrow in lateral view, broad in dorsal view, with membranous
upper part. P. demoori

328

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Larvae and their cases

1. Cases subcircular, with wide lateral flanges, larvae with pattern of small spots on head;

abdominal gills branched. (Figs 47, 48, 61, 81.). 2

Cases narrowly triangular, lacking wide lateral flanges; larvae with spots on head confluent;
abdominal gills simple (on ventral side only) (Figs 69, 70). P. triangularis

2. Pattern of spots on head usually forming paired rows on each side of head; spots on pronotum

and mesonotum tending to be confluent (Figs 48, 50, 62, 63). P. circularis

Pattern of spots on head only partly paired, usually forming larger spots in a single row; pattern
of small separate spots on pronotum and a few on mesonotum (Figs 79, 80).

. . . putative larva of P. demoori

Note: in identifying the imagos it is usually necessary to clear the genitalia in KOH, and to study
them under a research microscope with a magnification x 100, for some details x 400. The genitalia
are very small. For the larvae a stereo microscope with a good magnification is adequate but it is
necessary to prop up the larvae on a sand bed to see the top of the head unless the head is taken off.
The figures should be used in conjunction with the keys.

Family Barbarochthonidae Scott 1985

Barbarochthonidae Scott 1985: 331, 332 (fig. v), 338; Scott 1986: 231, 234 (table 1), 236.

Type-genus Barbarochthon Barnard 1934: 319, 320

The family Barbarochthonidae was erected to accomodate the single genus Barbarochthon
Barnard. Only one species B. brunneum Barnard, is known. It is common in mountain streams in the
western Cape Province, is also present, sometimes commonly, in the southern Cape, and has been
recorded from Natal.

Recognition

Imago small, mid to dark brown. Vertex with paired interantermary warts and a large
posterolateral pair. Male with 3-segmented maxillary palps carried up over face. Pronotum apparently
with a single long wart; mesonotum without paired warts but with a median patch of setae which may
be present or absent; scutellum suboval with scattered setae. The most useful identification mark is
the conspicuous cream-coloured pronotum without visible separate warts.

329

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Family diagnosis
Imagos

Ocelli absent; eyes large, glabrous; vertex with pair of small interantennary warts and a pair of
large posterolateral warts. Antermae about as long as fore wings in male, shorter in female; scape
stout, as long as head; flagellum tapered. Maxillary palps in male 3-segmented, little longer than
labial palps, usually carried upturned over face, terminal segment not armulate either in maxillary or
labial palps. Maxillary palps in female 5-segmented, longer than labial palps.

Pronotum with a pair of long warts fused medially, appearing single; mesonotum without warts
but with small median field of setae and a pair of very small anterior setate spots (all are invisible
unless the long setae are still present, those are very easily lost, and in any case seem to vary between
specimens, being present or absent or partially present); scutellum suboval, sparsely covered with
setae. Middle legs with 6 pairs of colourless spines on first to fourth segments of tarsus. Hind legs
with 6 pairs on basitarsus, 2 at end of each segment from second to fourth. Fore wings similar in male
and female, discoidal cell closed, median cell open; thyridial cell present; anastomosis very clear,
forks 1, 2, 3 present. Hind wings considerably smaller than fore wings, male differing considerably
from female; in both sexes, discoidal and median cells absent and the venation much reduced,
particularly in the male; in male Sc and fused, stems of fork 1 and of M and Cu2 barely indicated,
sometimes M completely absent, fork 1 and sometimes 2 present; a patch of androconia between
bases of RS and Cuj, covered with long setae; in female Sc and Rj largely separate, forks 1 and 2
present, stalked; base of M absent. Wing coupling by macrotrichia, and in male in addition by thickly
placed marginal setae.

Male genitalia with short two-branched pre-anal appendages; tenth segment bifid; claspers strong,
unbranched, with basal lobes; aedeagus stout, divided apically into a dorsal bifid lobe and a ventral
lobe with apex scoop-shaped.

Female genitalia with short overhanging paired lobes of tenth tergite. Sternites of tenth segment
unsclerotized, without appendages.

Lan’a

Case dweller; larva rounded; head and pronotum strongly sclerotized; prosternal horn absent.
Head dorsally flattened; frontoclypeal apotome with single pair of indentations; antennae very small,
at base of mandibles; eyes medium-sized; mandibles large, strongly sclerotized, other mouthparts
small, very hairy; ventral apotome rhomboidal, completely separating genae. Mesonotum less strongly
sclerotized than pronotum, particularly posteriorly; metanotum membranous with anterior transverse
band of setae. Fore leg stout, middle and hind legs long, with very long claws. Abdomen smooth,
lacking lateral fringes and lateral tubercles; first abdominal segment with small dorsal hump, lateral
humps each with small setate sclerite, ventrally a tough “lip”; gills absent; ninth segment without
dorsal plate. Anal prolegs short, with bases fused, strong sclerites present. Anal claw with a long
comb of teeth.

Larval case

A long, sometimes very long, tapered tube of dark coloured silk, ornamented with very small
sand grains; terminal membrane with circular aperture.

330

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Male pupa

Antennae slightly longer than pupa; labrum transversely ovate; mandibles slender, falcate, inner
margin faintly serrulate; maxillary palps 3-jointed, with the labial palps reaching just beyond end of
metathorax. Middle tarsi fringed; lateral fringes feebly present from end of sixth segment to eighth,
with small tufted ends; wing sheaths reach to ninth segment. First abdominal segment without lappets;
second to sixth segments with presegmental dorsal plates, fifth segment with postsegmental dorsal
plate, all dorsal plates small. Apical appendages slender, straight, rod-like. Genitalia obscured by
brown, somewhat sclerotized integument of ninth segment.

Female pupa

Similar to male but somewhat larger.

Pupal case

LFnaltered or shortened larval case closed anteriorly by a membrane with a central boss and
transverse slit and having a membrane with a dorsal slit just beyond the pupa; anchored by one or
two anterior holdfasts.

Genus Barbarochthon Barnard 1934

Barabarochthon Barnard 1934: 319, 321, figs la and 15 a-p.

Type species: B. brunneum Barnard 1934 (the only species).

Etymology: Generic name neuter, referring to the early Dutch name for the range of mountains where
these caddis were discovered. These mountains were termed the “Holland” or Home of the
Hottentots (Barbarians).

As Barbarochthon is a monotypic genus a generic diagnosis is omitted. It is covered by the full
description of the species.

Barbarochthon brunneum Barnard
(Figs 86 - 124)

Barbarochthon brunneum Barnard 1934: 319, 321, figs la and 15 a-p (male, female, larval and pupal
parts, pupal case); Barnard 1940: 643; Harrison and Elsworth 1958: tables 16, 24, 25, 26 (as
Sericostomatidae), 207; Harrison 1958a: 260 (as Sericostomatidae); Scott 1985: 338; Scott 1986:
234, 236.

Lectotype male here selected and designated from Barnard’s syntypes. South African Museum, Cape
Town.

Type locality: western Cape Province, Hottentots Holland Mountains, East side, 4 000ft, [= 1 218 m].

331

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs. 86-97. BARBAROCHTHONIDAE: Barbarochthon brunneum Barnard, male, female
(Scale lines = 1 mm unless otherwise indicated).

Material used: KHB

86. Male: head and thoracic nota, dorsal. 87. Male: head and palps, lateral view. 88. Male: face. 89. Male (and female) forewing.
90. Male hind wing. 91. Female hind wing. 92. Male: entire insect, lateral view. 93, 94, 95. Male genitalia, dorsal, lateral and
ventral, showing branches of 10th tergite, pre-anal appendages, claspers and aedeagus. 96. Female genitalia, caudal view. 97a and
b. Female genitalia (from Barnard 1934, figs 15g and h, p. 322), a. lateral, b. ventral, (a - aedeagus, cl - clasper, ixs - ninth
sternite, ixt - ninth tergum, pa - preanal appendage, pi - pleura, v - vulva, vg - vagina, xt - tenth tergum)

332

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Barnard did not as rule select types, the specimens in his collection being unmarked apart from
name, locality, collector(s) and date. The best male from those specimens listed in his 1934 paper
and still remaining in the South African Museum collection has been selected as the Lectotype. The
other specimens of the same locality and date are regarded as Paralectotype males and females. The
1916 male has not been designated a Paralectotype as it was glued to its mount in small bits. The
extant material and the original records are listed after the descriptions of the various stages.

Description of imagos (Figs 86-99)

Male imago (Figs 86-90, 92-95)

Fore wings 5-6 mm in length.

Colour description given by Barnard (1934) as follows: “Head and thorax dark sepia-brown with
paler hairs. Legs and antennae fuscous. Wings brown with pale hairs. Membrane with a clear patch
on upper margin of thyridial cell and the cross-vein between R5 and M, and a spot on the connecting
vein [= cross-vein] between Cu^ andCu2.. .”. Barnard (1940) added to this a note that fresh specimens,
particularly from the Hottentots Holland Mountains, had a pale or white band along the side of the
abdomen.

Ocelli absent; eyes large, black, glabrous; vertex with small, paired, contiguous, inter-antennary
warts and with posterolateral warts somewhat crescent -shaped; median sulcus present with a small
tubercle posteriorly on each side; face with one pair large setose median warts and one pair small
lateral warts at antennal bases, these pairs are separated by paired glassy yellow strips. Maxillary
palps longer than labial palps; first and second segments together about as long as third, moderately
stout, thickly covered with setae. Labial palps 3-segmented; second and third segments subequal,
together slightly longer than first. Third segment not annulate in maxillary or labial palps.

Pronotum with a single long wart, evidently representing a fused pair; mesonotum with a me-
dian patch of setae and a pair of very small anterior patches, variable in size and presence or absence,
practically impossible to see unless the long setae covering them are still present (they are, however,
very easily lost and in any case seem to vary between specimens, being present, partially present or
absent, as is the case with the Australian family Antipodoeciidae); scutellum with scattered long
setae, also difficult to make out unless the setae are still present. Fore wings (Fig. 89) with discoidal
cell closed, median open, thyridial present, anastomosis very clear; forks 1, 2, 3 present; Rj separated
from Sc; M3^ fused; Cuj present, simple; Cu2 connected to Cu^ by a cross-vein; A2 joining Aj near
base, meeting margin basal to Cu2 at arculus. Hind wings (Fig. 90) with venation considerably reduced,
variable; R^ probably fused with Sc (compare with female in which they are separate. Fig. 91), ending
before margin and bearing a row of macrotrichia near base; RS complete, continuing as R4+5; usually
only fork 1 present;R2 and R3 present, but only tenuously connected with RS; occasionally R4 and
R5, which are normally fused, may be separated apically to form a very small fork 2; M undivided,
may be completely absent but, if present, most of its stem is missing or very faint, as are the stems
of fork 1 and CU2; Cu^ complete, simple; anals short; a cluster of scent scales (androconia) always
present between bases of RS and Cuj , covered by a dense tuft of setae. Only one cross-vein, sometimes
very faint, between Sc -1- Rj and faint base of fork 1. Wing-coupling by many strong setae along basal
half of hind wings, also by the macrotrichia on SC Rj. Jugal lobes large.

Genitalia (Figs 93-95) with short two-branched preanal appendages, the inner branch pointed,
the outer with expanded setose tip; tenth segment bifid, forming 2 long slender processes; strong
single-jointed claspers each with broad inturned basal lobe (Fig. 95); aedeagus divided towards

333

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs. 98, 99. BARBAROCHTHONIDAE: Barbarochlhon bnmneum Barnard, female
(Scale lines = 1 mm unless otherwise indicated).

Material used; KHB

98. Female genitalia, caudal view. Scale line 100 |.im. Note 10th tergites on right, and very simple genitalia.

99. Female genitalia, same. Note tergites and apparent curling in of pleura-like folds, into vulva.

334

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

middle into dorsal apically bifid lobe and ventral lobe with scoop-shaped apex and membranous in-
fill.

Female imago (Figs 91, 96, 97, 98, 99)

Female larger than male; fore wings 6.5-7.0 mm in length.

Vertex with setose warts differing slightly from those of the male in position and shape, the
interantennary warts being slightly separate and the posterolateral warts being subovate. Antennae
shorter and more slender than in male. Maxillary palps 5-segmented, longer than labial palps; both
palps smaller than in male.

Fore wing venation as in male; hind wings (Fig. 91) with forks 1 and 2; Rj largely separated
from Sc and, as in male, cormected to stem of fork 1 by a cross-vein; RS present, complete; apex of
M present, simple, lacking basal part of stem; stem ends just beyond cross-vein; Cu^, Cu2 and A all
present, entire. Wing coupling by macrotrichia. Jugal lobes large.

Genitalia with short overhanging paired lobes of tenth tergites; ninth sternites evidently soft,
unsclerotized, not rugose as in Hydrosalpingidae. Genitalia much simpler than in Petrothrincidae.
Position of vagina shown in Fig. 96. There are no appendages.

Figs 98, 99 show finely rugose pleura-like folds, evidently within the vulva. In another female
the vulvar opening appears to be plugged.

Description OF MATURE LARVA (Figs 100-112; 116-124).

Description from material from the Homtini River, southern Cape, from stones in current (SCR
5A: FCdM and HMB, 8.iii.89; MISC 251a: KMFS and BCW, 27.V.70) compared with Barnard’s
specimens and drawings.

Length of larva up to 8 mm; larva rounded.

Head dorsally flattened (used to plug case), strongly sclerotized, blackish-brown to rich chestnut
in colour, with pigment arranged in an apparent honeycomb pattern (see Figs 100, 102 and 119-121);
ventral apotome narrowly quadrangular (Fig. 102), completely separating genae in young larvae,
difficult to distinguish posteriorly in older larvae in which only the dark brown anterior bar may be
clearly visible. Eyes fairly small, under small clear lenses in white area. Antermae very small (Figs
119-121), on side of head near base of mandibles. Labrum (Fig. 102) small, rounded, partially
retractile. Maxillae and labium very tightly packed into a small space behind mandibles; maxillary
palps clear, thick, very hairy; stipes and cardo fused forming a single stout structure; labium and
labial palps stout, both sets of palps on large bases; a dense brush of setae anterior to labium. Mandibles
(Figs 111, 112) very heavy, blackened apically due to strong sclerotization; left mandible with two
irmer brushes of setae set deep in a hollow; right mandible with a single brush; brushes apparently
variable, in some instances right one appears to be missing; both mandibles appear to have a cutting
edge and two very long, strong basal ligaments (shown cut short in Figs 111 and 112); setae very
strong but not feathered; apex long with three or four teeth when unworn (as in the Petrothrincidae
the mandibles may show heavy wear as in Figs 111, 112).

Pronotum strongly sclerotized with carina ending in strong anterolateral points, with strongly
marked dark sepia pattern behind carina (obvious on the pale greenish-yellow background), and with
small posterolateral divisions seen in lateral view (Fig. 103). Mesonotum with sepia muscle spots on
greenish-orange background, less strongly sclerotized particularly on posterior half. Metanotum
membranous with anterior band of setae, in colour pale greenish mixed with orange, with small sclerites

335

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs. 100-11.5. BARBAROCHTHONIDAE; Barbarochthon brunneum Barnard, iarva and case, pupa
(Scale lines = 1 mm unless otherwise indicated).

100, 101. Mature larva, head and thorax, dorsal. 102. Younger larva, head, ventral. 102a. Same. Maxillary and labial palps.
103. Pronotum, lateral. 104. Mature larva, habitus, lateral. 105. Larval case. 106. Fore leg and pre-episternum. 107, 108. Middle
and hind legs. 109. Anal proleg and claw, lateral view. 110. Anal prolegs, dorsal view. 111,112. Right and left mandibles,
ventral view. 113. Male pupa, habitus. 114. Same, anal appendages further enlarged. 115. Pupal mandible.

336

SCOTT; THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

as shown in Barnard’s drawing of metanotum (Barnard 1934 Fig. 15i) not visible. Pre- episternum
large with apex bluntly rounded (Fig. 106). Pleural sclerites with black marks. Legs (Figs 106-108)
whitish to pale brown in life with light brown to sepia markings. Fore leg stout; coxa large; trochanter
2 and femur with setal brush along ventral margin but without distoventral process; ventral brush with
plumose setae and a few peg-like ones; tibia with small clusters of minute setulae and two blade-like
setae on disto-ventral angle, one larger than the other; claw stout, same length as tarsus. Middle leg
longer, more slender; trochanter 2 and femur with partly plumose setal fringe; tibia with many minute
clusters of setulae; tarsus set with smalt setae; claw same length as tarsus. Hind leg much longer; coxa
slender, curved; femur with a few plumose setae; tibia with a few setulae; tarsus without small setae;
claw nearly as long as tarsus. All three claws stout with sharp apex and stout seta near base.

Abdomen (Figs 104, 109, 1 10 and 122-124) orange-coloured anteriorly, fading and mixing with
pale green posteriorly in life, white or cream in spirit; first segment with low dorsal hump, with lateral
humps each with small pubescent sclerite, and with ventrally an apparent toughened area or “lip”;
ninth segment without dorsal sclerite; anal prolegs short; anal claw very small, with dorsal comb;
tenth segment with two small dorsal sclerites of variable length (between anal prolegs), with large,
dark, lateral sclerites and ventral sole plates, lacking tufts of setae (Figs 109, 110). Anal claw as seen
in the SEM micrographs (Figs 122-124) shows a few teeth of comb just above anal claw, continued
round to back as a number of sharp teeth, about 8-10 in all.

Larval case (Fig. 105)

A long, often very long, slender, curved, tapering, dark- coloured, silken tube, ornamented
particularly towards the posterior end with rows of minute sand grains; terminal membrane with
circular aperture.

The tube looks more brightly coloured when it contains a live larva as the orange-green colouring
of the abdomen and parts of the thorax show through.

Description OF PUPA (Figs 113-115).

Male pupa

Antennae slightly longer than body. Labrum transversely ovate. Maxillary and labial palps extend
to first abdominal segment. Mandibles slender, falcate, inner margin very slightly serrated. Middle
tarsi fringed. Lateral fringe present from sixth segment, ending in a tuft on eighth segment. Gills
absent. Fore wing sheaths reaching to begirming of ninth segment. Presegmental dorsal plates on
second to sixth segments (2 teeth each); postsegmental dorsal plates on fifth segment (4 teeth); ninth
segment apparently lightly scelerotized, brown in colour. Anal appendages slender, rod-like, each
with a long stout apical seta, and two or three long slender ones.

Pupal case

At pupation the long posterior end of the larval case is usually cut off and the terminal membrane
replaced by a new one with an oval vertical slit; the anterior end is closed by a convex membrane with
narrow horizontal slit. Should the end of the larval case not be cut off at pupation, the terminal membrane
of the larval case remains and the pupal membrane is formed within the case below the end of the
pupa. The pupal case is anchored underneath a stone or on vegetation by one or two anterior holdfasts.

337

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs 116-118. BARBAROCHTHONIDAE: Barbarochthon briinneum Barnard, larvae.

Material used: MISC 270c.

116. Photograph of two larvae in cases, one larva ex. case, and three empty cases. 117. Larva showing dorsal view of head.
118. Lateral view of head and thor^ of larva in case. Photographs by J.C. Hodges, Jr.

338

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Distribution

South Africa: the western and southern Cape Province, in mountain streams in the coastal ranges;
and Natal where it has been recorded from Karkloof, near Howick.

Material examined

South African Museum material

Firmed material

Lectotype: Southwestern Cape: Hottentots Holland Mountains [3418 BB], 3 500-4 000 ft
[= 1 066-1 218 m] (KHB and HGW, i.l933, male).

Paralectotypes: Southwestern Cape: Hottentots Holland Mountains [3418 BB], 3 500-4 000 ft
[= 1 066-1 218 m] (KHB and HGW, i.l933, 8 males).

Other material: Southwestern Cape: Hottentots Holland Mountains [3418 BB], 3 500-4 000 ft
[= 1 066-1 218 m] (KHB, i.l916, male, glued to its mount in small bits).

Fig. 119. BARBAROCHTHONIDAE: Bflffcaroc/u/io/i brunneum Barnard, larva.

Material used: MISC 251a.

Scale line 100 (.im. 119. Head of larva, to show sculpturing of cuticle, tightly packed mouthparts, and position of very small
antenna (indicated by arrow).

339

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Figs. 120, 121. BARBAROCHTHONIDAE: Barbarochllwn hrunneum Barnard, larva.

Material used: MISC25la.

120. Scale line 10 f.im. Part of head of latv'a showing antenna. 121. Scale line 10 fjm, Antenna.

340

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Material in spirit

Paralectotypes: Southwestern Cape: Hottentots Holland Mountains, East side, 3 500-4 000 ft,
[= 1 066-1 218 m] [3418 BB] (KHB and HGW, i.l933, 21 males, 12 females).

Other material: Southwestern Cape: Hottentots Holland Mountains, East side, 3 500-4 000 ft,
[= 1 066-1 218 m] [3418 BB] (KHB and HGW, i.l933, larvae and few pupae); Wellington Mountains
[3318 DB], 3 000 ft, Witte River, Bains Kloof (KHB, ix.l922, 2 males); Cape Peninsula, Table
Mountain, Orange Kloof [3318 CD] (KHB, l.iii.l933, pupal cases, largely empty but with a few
pupae). [From the Table Mountain material Barnard (1940) obviously had had imagos, since lost,
which he used in establishing its identity.]

Albany Museum material

Material in spirit

Western Cape: Great Berg River, Driefontein [3319 CC], (MISC 89: KMFS, 12.xii.54, 1 male
with pupal pelt and case); same (GBG 142B: ADH, 22.xi.50, 1 pupa); Great Berg River, Railway
Bridge [3319 CC] (GBG 780, 783: KMFS, x.53, 1 male, 1 female); Great Berg River, Sneeugat [3319
CC] (Source; GBG 372A(1): ADH, 28.ix.51, larvae); Great Berg River, Driefontein, (MISC 139:
KMFS, 8.x. 59, larvae); Great Berg River, Assegaibos [3319 CC] (MISC 288d: KMFS, 30.i.76, 2
larvae); Smalblaar River, tributary of the Breede River, Du Toit’s Kloof [3319 CA] (MISC 217: RD,

Fig. 122. BARBAROCHniONIDAF: Barbaruchthon bnmneiim Barnard, larva.

Material used: MISC 2,‘ila.

Scale line 100 |.rm. Posterior segments of larval abdomen showing anal claws (ex. MISC 2.51a).

341

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

igs. 123, 124. BARBAROCHTHONIDAE: ftruwic’um Barnard, larva.

Material used: MISC 251a.

123. Larva: Right anal claw seen trom outer side, showing long claw with smaller hook and smaller teeth curling round to the
inner side. Scale line 100 |.im, 124. Left anal claw seen from inner side, showing row of teeth curling right round to the inside,
making in all a comh with about 8-10 teeth. Scale line 100 |.im.

342

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

i.l962, 1 larva); Upper Witte, Happy Valley [3319 CA], (MISC 269d: KMFS, 3.X.74, 1 larva); Upper
Witte, Happy Valley (MISC 304e: KMFS, 6.ii.76, 2 larvae, 1 pupa); Upper Witte, Happy Valley,
higher up, (MISC 270c: KMFS, 3.X.74, 8 larvae); Witte River, Wolwekloof (MISC 291c: KMFS,
30.xi.76, 2 larvae); Witte River, main stream [3319 CA] (MISC 296b: KMFS, 2.ii.76, 23 larvae);
Witte River, Leeuklip Kloof tributary [3319 CA] (MISC 267a: KMFS, 3.x. 74, 4 larvae); Palmiet
River near Elgin [33 1 9 AA] (MISC 48: ADH, 1 7.xii .52, 1 male); Table Mountain, Disa Gorge, Orange
Kloof [3318 CD] (MISC 285g: KMFS and BCW, 27.xi.76, 4 larvae) and (MISC 286d: KMFS and
BCW, 27.1.76, 1 prepupa in case); Langrivier, tributary of Eersterivier [3318 DD], Jonkershoek Nature
Reserve, Stellenbosch (JMK, different times of the year, larvae); Upper Langrivier (MISC 271b:
KMFS, 4.x. 74, 2 larvae).

Southwestern Cape: Vetrivier, Garcia’s Pass, above Riversdale, tributary near top of pass, below
Tolhuis [3421 AA] (MISC 306L: KMFS, 9.ii.76, 1 larva).

Southern Cape: Homtini River, Homtini Pass at road bridge [3322 DD] (MISC 251a: KMFS and
FMC, 27.vii.70, 102 larvae); Homtini River, Phantom Pass [3322 DD] (MISC 278e: KMFS, 22.i.76,
10 larvae); Karatara River Jubilee Creek tributary (SCR 2A and 3C and SCR 3B, 6 and 7: FCdM and
HMB, 7. hi. 1989, putative pupae and larvae); Blaauwkrantz River, Tsitsikama, [3323 DC], 180 m
(SU 67: HB, 25.i.59, 4 larvae); Kruis River, [3319 AB] (FRW 164H, 166J, 174D, 183F: ADH and
JDA, 8.iii.60, all larvae); Storms River [3423 BB] (FRW 159Z, 186C, 8.iii.60, ADH and JDA, larvae);
Kaaiman’s River [3322 DC] (FRW 180F: ADH and JDA, 10.iii.60, larvae).

Material cited in literature

Barnard (1934): Hottentots Holland Mts., 3 500-4 000 ft (KHB, January 1916, 1 male and KHB
and HGW, January 1933, males, females, larvae and pupae); Wellington Mts, 3 000 ft (KHB,
November 1922, 1 male); Table Mt, Cape Peninsula, 1 000 ft (KHB, 1st March 1933, 1 pupa, empty
cases); Montagu Pass, north of George, Outeniqua Range (HWG, April 1933, larvae) [appear to have
been lost].

Barnard (1940): Upper Olifants River, north of Ceres,(KHB and CWT, October, 1937, males
and females) [appear to have been lost].

Jacquemart (1963): Bainskloof [=Bain’s Kloof] about 10 miles E.N.E. Wellington [the Witte
River] [3319 CA]. In the eastern part of the kloof: fast-running stony mountain stream. 12.2.1951. 1
male.

Additional material

Morse (in litt 16.xii.74), 1 male from Karkloof near Howick, Natal [2930 AC], 27.x. 70, H. and
M. Townes leg.

Biology of genus Barbarochthon

Much background information regarding the localities on the Great Berg River can be found in
Harrison and Elsworth (1958), as was indicated for Petrothr incus. A brief faunistic summary for the
Berg River (for Trichoptera see pp. 260-263) is given in Harrison (1958a). In these studies
Barbarochthon brunneiim is usually referred to by name and placed under Sericostomatidae or
sometimes Beraeidae. Barbarochthon larvae were commonly found from the uppermost zone of the
Berg (sponges and cliff waterfalls, altitude from 4 000-5 000 ft [= 1 219-1 514 m]; cliff waterfalls

343

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

down to 2 500 ft [= 762 m]); through the Mountain Torrent Zone at 1 000-4 000 ft, [304-1 219 m];
Assegaaibos main stream in the Franschhoek Forest Reserve - Stn 1 at 1 000 ft [= 304 m]; to the
Foothill Stony Run Zone at 300-1 000 ft [= 91-304 m], subdivided into two subzones, an upper one
which included stations 3 and 5 at Driefontein and the Groot Drakenstein Railway Bridge respectively
and a lower one which included at its upper end Station 9 at Simondium (Harrison and Elsworth,
1958, collecting stations shown in Plates X-XII). From the tables of significant animals from stones-
in-current, from Scirpus, and from marginal vegetation (Harrison and Elsworth 1958: Tables 16, 24,
25 and 26) it seems that Barbarochthon larvae are far more abundant in the marginal vegetation than
in either the Scirpus or the stones-in-current biotopes. In the marginal vegetation they were notably
present in considerably higher percentages of the fauna in summer and autumn in the Mountain
Torrent Zone (Stn 1) and in spring, summer and autumn in the upper foothills (Stn 3) but always
present in lower numbers at both these stations during other seasons. Barbarochthon is present in
much smaller numbers and only at certain times of the year at Stations 5 and 9, the lowest stations
at which it occurred at all.

Barbarochthon larvae have also been recorded (see material, above), but not usually commonly,
from streams in the southern Cape.

The larvae live mainly amongst marginal vegetation and on submerged or partially submerged
clumps of Scirpus digitatus. They can be found in both fast current and backwaters. They feed on
leaves, crawling about actively at all seasons of the year. They are also present, though much less
commonly, on stones in current and are evidently capable of feeding there. They do not, however,
use leaves in making their cases which are always of silk and are ornamented with sand grains. The
structure of the mandibles bears out the method of feeding, the inner margin having a cutting edge
and the sturdy brushes of setae presumably being used to help hold the leaves.

Adults have been collected in the Cape Province in October, November, December and January.
A single male was collected at Karkloof near Howick in Natal in October.

DISCUSSION

Over the years the author has drawn up comparative tables for adults, larvae and pupae in order
to compare the genera Hydrosalpinx, Petrothrincus and Barbarochthon with a number of families:
both those into which earlier authors attempted to place them, and certain of the Australian families
which seemed comparable at least to some degree. The tables were derived from the author’s notes
and drawings for the three South African genera and from the best sources that could be found for
the families considered. Literature consulted in building up the tables and in the preparation of the
present paper included Cowley (1976 and 1978), Harrison and Elsworth (1958), Harrison (1958a and
b, 1965 and 1978), Lepneva (1966 (1971)), Mosely (1939), Mosely and Kimmins (1953), Neboiss
(1977, 1981, 1983, 1984, 1986, 1988 and 1991a and b), Ross (1967 and 1978), Scott (1955), Ulmer
(1951 and 1955).

Based on the author’s analysis of the literature she selected for comparison the Australian fami-
lies Helicophidae, Calocidae and Antipodoeciidae as possibly being nearest to the three genera
Hydrosalpinx, Petrothrincus and Barbarochthon. To these were added the Bereidae and the
Sericostomatidae, primarily because of their repeated linkage with the three in the literature, and the
South American family Anomalopsychidae as a Neotropical comparison. It should be mentioned that
not all the Australian families had been studied and that a number of other families, for example the
Molannidae, Brachycentridae and Thremmatidae, had been included in earlier analyses.

Having placed the threti genera in three families, Hydrosalpingidae, Petrothrincidae and

344

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Barbarochthonidae, it is necessary to set the families in their place in the classification of the
Trichoptera. All clearly belong to the Sub-Order Integripalpia Martynov (1924) (see Schmid, 1980,
pp. 14-17). All have adults with maxillary palps either 3 or 5-segmented in the male, 5-segmented
in the female, and with the last segment simple. The fore wings have the discoidal cell closed, rather
long, the median cell open and fork 1 sessile. In the male genitalia the ninth tergum is narrowed, the
tenth tergum is somewhat hood-like or present but elongated and forked, the phallic apparatus is
provided with an aedeagus, and in the Petrothrincidae with paired parameres. In the female the
genitalia are without cerci but may have small apical dorsal appendages, evidently modifications of
the tenth tergites. There are definitely separate anal and vaginal apertures and, in the Petrothrincidae,
a clear supragenital plate and vulvar scales. The larvae are eruciform and live in portable cases in
which they are anchored by their short anal prolegs which in the Petrothrincidae are rather longer
than is usual though not as long as in the Annulipalpia.

The division of Integripalpia to which all three appear to belong is the Superfamily Leptoceroidea
Schmid (1980). The Leptoceroidea were characterized by Flint (1981, quoting Ross, 1967) as having
tost their ocelli, the male fore wings retaining M4 in primitive families, the supratentorium reduced
and larval pronotum without a crease. The last appears to suggest a posterior suture of the pronotum
as opposed to an anterior suture in the Limnephiloidea. Flint (1981), however, could find no con-
sistent pattern in this, with which the present author agrees. In fact some taxa within one group may
have both, others neither. Flint further characterized the leptocerid branch (the Leptoceroidea Schmid)
as having larvae which lack a prosternal horn, have bifid tubercles on the eighth abdominal segment
only, and lack a strongly sclerotized dorsal sclerite on the ninth abdominal segment. None of the three
South African families has a prosternal horn. In the Hydrosalpingidae there are bifid tubercles on the
eighth segment only and a weak dorsal sclerite on the ninth segment. Neither of the other two families
has either bifid (lateral) tubercles or a dorsal sclerite on the ninth segment. It should be noted that, in
southern Africa at least, certain genera of Leptoceridae (for example, Setodes, Trichosetodes and
Leptecho) do have a dorsal sclerite on the ninth abdominal segment, in some cases strongly sclerotized.
Judging from the illustration in Wiggins (1977), the North American Setodes incertus (Walker) also
does. On the other hand, many leptocerids do not, so that it would not seem to be a good distinguishing
character for the Leptoceridae, and therefore for the Leptoceroidea.

Weaver (1983, 1984), in his proposed classification of the Trichoptera, splits Schmid’s
Leptoceroidea into Sericostomatoidea Stephens (1836) and Leptoceroidea, both included in his new
Infraorder Brevitentoria. Schmid (1980) indicated this division in his phyletic tree by its branching.
In such a split the Hydrosalpingidae, Petrothrincidae and Barbarochthonidae would all fall into the
Sericostomatoidea rather than the Leptoceroidea. This is indicated by the apparent relationship of all
three to the Australian families Helicophidae, Calocidae and Antipodoeciidae, and possibly of the
Hydrosalpingidae and Barbarochthonidae also to the Beraeidae. The characters given by Weaver,
however, are so few as to render comparisons difficult. Certainly all three families share with his
Sericostomatoidea adults with tibial spurs 2, 2, 4 (in common with many southern hemisphere families
and certain others) and lacking ocelli, and larvae which are phytophagous. In the larva the pre-
episternum (Weaver’s trochantin), however, is not small but large in both Petrothrincidae and
Barbarochthonidae and is normal in Hydrosalpingidae, and the tergite on the ninth abdominal segment
is present, although weak, in Hydrosalpingidae but is absent from the other two families.

Incidentally, Weaver gives the larvae of his Sericostomatoidea as phytophagous and those of
Leptoceroidea as predatory. In fact most southern African Leptoceridae show the whole gamut of
feeding habits from purely phytophagous to mainly carnivorous, the latter being rare. The related
Calamoceratidae are certainly plant-eaters.

345

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

AJthough the Hydrosalpingidae, Petrothrincidae and Barbarochthonidae were individually
compared with the Brachycentridae and the Thremmatidae there seems to be no point in drawing
comparisons between them, since the latter families fall into the Limnephiloidea not the Leptoceroidea
(sensu Schmid). Molarmidae falls into the leptocerid branch rather than the sericostomatid branch of
the Leptoceroidea and so no detailed comparison between it and the three families is necessary.

In the present paper Dr F. C. de Moor, using cladistic analyses, has compared the
Hydrosalpingidae, Petrothrincidae and Barbarochthonidae with the families in Weaver's
Sericostomatoidea and with the Antipodoeciidae.

The original decision to treat the three geneia Hydrosalpinx, Petrothrincus and Bar abarochthon
as each belonging to a new family is supported by the cladistic analyses. The three families do appear
to fit into the Sericostomatoidea, as may be seen in both analyses (Figs 125, 126). It appears from
the cladogram based on 44 characters (Fig. 126) that the closest relatives of the Hydrosalpingidae
are all Australian or South American and those of Barbarochthonidae and Petrothrincidae are the
Beraeidae and Antipodoeciidae. These are of course preliminary results, however, it is of interest that
this cladogram separates out into two branches, one almost exclusively Australian, the other almost
exclusively non-Australian, with the South African families divided between the two.

ZOOGEOGRAPHICAL NOTE

Ross (1967), when discussing the origin of the fauna of Eurasia- Africa-Madagascar, referred to
two of these endemic South African genera, Petrothrincus and Hydrosalpinx. He queried whether
they, together with Paulianodes (Philopotamidae) and the Pisuliidae, were not survivors of the
Cretaceous or more recent immigrants that had become extinct in their original, unknown home, or
were not even older lineages arising from ancestors that had reached Africa or Madagascar perhaps
as early as the Jurassic.

Petrothrincus, Hydrosalpinx and Barbarochthon of the presently defined families
Petrothrincidae, Hydrosalpingidae and Barbarochthonidae are all endemic to South Africa and indeed,
almost entirely to the Cape Province, where they are found in the acid mountain streams of the western
Cape coastal folded belt and in its extension into the southern Cape. In this area there are many relicts
of the cool-adapted Gondwanaland fauna to which the present three families most probably belong
(see also Scott, 1986). This is borne out by the close linkage (Fig. 126) of the Hydrosalpingidae with
the Helicophidae, a family also found in South America and Australia.

In early Jurassic times, over two hundred million years ago, Africa lay well to the South of where
it lies now (Harrison, 1978 and W. J. de Klerk, pers. com.), still linked with Antarctica, Australia and
South America to the South. What is now our south-west coast would have been much cooler and
wetter. It would have formed a southern part of the very large continent, Gondwanaland, of which
the northerly parts were becoming subtropical and then tropical as they slowly drifted northwards.
Gondwanaland finally broke up in the late Jurassic and the continents eventually reached something
like their present positions.

Many cool-adapted caddisflies must have become extinct as the climate warmed up, but quite a
number did survive: in addition to the Hydrosalpingidae, the Petrothrincidae and the
Barbarochthonidae, certainly also the southern African Sericostomatidae (6 genera), many species
of Athripsodes (Leptoceridae) and certain genera, for example, Leptecho (also Leptoceridae), and the
hydropsychid genus Sciadorus.

Ross (1967) was therefore correct in tentatively placing Petrothrincus and Hydrosalpinx as
survivors of the Cretaceous rather than as more recent immigrants. They might indeed have an older

346

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

lineage, perhaps from the Jurassic or even earlier as was also suggested by Ross. Ross could quite
justifiably have incXudud Barbarochthon with them. Wiggins (1984) suggested that the origins of the
Trichoptera could have been in the early Mesozoic or even the Permian (in the Palaeozoic). Whenever
they originated, it must have been prior to the final break-up of the southern continents in order for
there to be links between some of the southern families found on the present day continents of South
America, South Africa and Australia.

A CLADISTIC ANALYSIS OF CHARACTER STATES IN THE TWELVE FAMILIES HERE
CONSIDERED AS BELONGING TO THE SERICOSTOMATOIDEA (by F. C. de Moor)

A cladistic analysis of the states of a number of characters of the three recently erected endemic
South African families of Scott (1985), the Antipodoeciidae and the 8 families recognised as belonging
to the Sericostomatoidea Stephens (1836) sensu Weaver (1983) was carried out. For the analysis 59
characters (from the larvae, pupae and adults) each with two to several possible states, were chosen
(Table 1). Characters were selected from morphologically useful identification features within these
twelve families that appear to belong to the Sericostomatoidea. They were chosen irrespective of
whether they were shared by members of the three new South African families or were considered
to contribute to a phylogenetic evaluation of the families. Characters considered as plesiomorphic
were given a state of 0, and numerical values for the other characters assumed that they were derived.
Where the primitive or derived status of characters could not be decided upon enumeration of character
states commenced with one. The family Anomalopsychidae was chosen as the outgroup as it was the
only family in the Sericostomatoidea with ocelli and showed the highest number of character states
considered as being plesiomorphic. The full selection of characters is not included for this preliminary
analysis as it is still being refined and will be presented when an analysis of the genera of these
families can be conducted.

In a number of instances characters showing several states were found within one family (see
Table 1 for definition of characters). For the Sericostomatidae wing venation characters were shared
with the three endemic South African families and Antipodoeciidae (forks 1, 2, 3 present in male
fore wings), or with the Calocidae, Conoesucidae and Anomalopsychidae (forks 1, 2, 3, 5), or only
with the Chathamiidae (forks 1, 2). Likewise for the forks in the male hind wings alternative character
states existed in the Sericostomatidae (states 1 or 2), Barbarochthonidae (states 3 or 4), Beraeidae
(states 5 or 7), Helicophidae (states 6 or 7), Calocidae (states 1 or 4), Anomalopsychidae (states 1 or
2), Conoesucidae (states 5 or 7) and Helicopsychidae (states 1 or 4). The development of the male
maxillary palps also showed variation within one family and states 0, 1 or 2 were found in the
Sericostomatidae, states 0 and 1 in both the Calocidae and Anomalopsychidae and states 1 and 2 in
both the Conoesucidae and Helicopsychidae. In the cladistic analysis it was assumed that if the most
primitive state of a character (e.g. the highest number of segments in the maxillary palps and the
fullest complement of cells in the wings) occurs in genera of a particular family, it is the general state
of that character for that family.

An analysis on this set of 59 characters using Hennig86 version 1.5 (Farris, 1988) determined
shared characters in the twelve families without prior selection for any criteria. This produced two
trees indicating parsimonious relatedness amongst the terminal taxa (Fig. 125). The Petrothrincidae,
Barbarochthonidae and Hydrosalpingidae group closely together in both trees. These three families
appear most closely related to the Helicophidae and Conoesucidae in the first tree and the whole
cluster of families then shows close relationship to the Beraeidae. The second tree shows the three
endemic South African families are apparently related to a cluster of families made up of the Beraeidae,

347

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Table 1. The description of the status of 59 selected characters of larvae, pupae and adults from the
eight presently described families of the Sericostomatoidea plus the three new families
and the Antipodoeciidae. A status of 0 represents the primitive plesiomorphic state for a
character. States such as 1,2 etc. rraresent derived apomorphies, usually but not always in
a hierarchical phylogentic order. Where the primitive or derived state of a character could
not be reasonably estimated valuation of character states commence with one.

LARVAE

1. Head; shape round 0, shape oval 1

2. Head; without carina 0, with Carina 1

3. Antennae; near base of mandibles 0, midway between eyes and base of mandibles 1, close to eye 2

4. Shape of ventral apotome; an equilateral triangle 1, an elongated triangle 2, quadrate or oval 3

5. Ventral apotome; entirely separates genae 0, separates genae anteriorly only 1

6. Pronotum; one single plate may be partially divided I, one pair of large plates 2

7. Pronotum; without a carina 0, with a carina 1

8. Pronotum; with a posterior division f , without a posterior division 2

9. Pre-episternum; large and prominent f, small and inconspicuous 2

fO, Pre-episternum; with upturned pointed apex f, with blunt apex 2

f 1. Mesonotum; a single plate 0, a pair of plaies f, more than two plates or sclerites 2

12. Metanotum; one pair of small plates f, membranous with setae 2, with more than two plates .3

13. Forelegs; long and slender 0, short and stout 1

14. Foreleg tarsal claw; shorter than tarsus 0, approximately as long as tarsus 1, longer than tarsus 2

15. Midlegs; long and slender 0, short and stout 1

16. Midleg tarsal claw; shorter than tarsus 0, approximately as long as tarsus 1

17. Hindlegs; long 0, short 1

18. Hindleg tarsal claw; shorter than tarsus 0, approximately as long as tarsus 1

19. Abdomen with; small dorsal hump 0, large dorsal hump 1, lateral humps only 2

20. Lateral abdominal humps with; setose sclerites 1, no sclerites or setae 2

21. Ninth abdominal segment; with sclerite 1, with no sclerite 2

22. Abdomen with; simple and branched gills 0, simple gills 1, no gills 2

23. Eighth abdominal segment; with lateral tubercles 1, with no lateral tubercles 2

24. Lateral abdormnal fringe of setae; present 1, absent 2

25. Anal prolegs; forming an apparent 10th segment 1, separated 2

26. Anal claw with; 2-3 dorsal hooks 1, a dorsal comb 2

27. Larval case; tubular and tusk shaped 0, limpet or shell shaped 1

28. Larval case constructed of; pure silk 0, silk with some embeded sand grains 1, fine sand grains 2, a mixture of
sand and plant matter 3, coralline or other algae 4

PUPAE

29. Antennae in male; shorter than body 0, as long as body 1, longer than body 2

30. Antennae in female; shorter than body 0, as long as body 1, longer than body 2

31. Inner margin of mandibles; feebly serrate 1, dentate 2

32. Wing sheaths reach end of abdominal segment; five 1, six 2, seven 3, eight 4,

3.3. No swimming fringes on tarsi 0, mid tarsi only fringed 1, fore and mid tarsi fringed 2
.34. Presegmental plates on abdominal segments; two to six 1, three to six 2, no plates present

35. Post segmental plates on abdominal segment five with; one or two hooks 1, three hooks 2, two to four hooks 3

36. Lateral fringe of abdominal setae on segments; six to eight 1, seven and eight 2, no lateral fringe of setae 3

37. Pupal case anchored; around margin of case 1, by one or more anterior holdfasts 2

Table 1 continued on page 349

348

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Table 1 continued from page 348

ADULTS

38. Antennae; shorter than length of forewing 0, same length as forewing 1, longer than length of forewing 2

39. Maxillary palps in male; five segmented 0, three segmented 1, one or two segmented 2

40. Ocelli; present 0, absent 1

41. Facial warts; one pair at antennal bases 1, two pairs at antennal ba.ses 2, one pair and median patch of setae at
antennal bases 3

42. Pronotum with; two pairs of warts 0, one pair of warts 1, single large wart (fused) 2

43. Mesonotum with; one pair of warts 1, median field of setae 2, two lateral strips of setae 3, one or two setae only 4

44. Mesoscutellum with; one pair of warts 0, a single large wart 1

45. Leg spines; colourless 1, some black 2

46. Thyridial cell in forewing; present 0, absent 1

47. Androconia; absent 0, present on head 1, present on forewing 2, present on hindwing 3, present on fore and
hindwing 4

48. Di.scoidal cell in forewing; clo.sed 0, open 1, absent 2

49. Median cell in forewing; open 1, absent 2

50. Sc and R1 in forewing; separate 0, joined 1

51. In male forewing presence of forks: one to five 0; one, two, three and five 1; one and two 2; two and five 3;
one, three and five 4; two or two and four 5; one, two and three 6; one and three 7

52. In female forewing presence of forks: one, two, three and four, or one, two and five 1; one, two, three and five 2;

one, two and three 3; one and two 4; two and five 5; two, four and five 6; one and three 7

53. Jugal lobes in hindwing; large 0, small 1, absent 2

54. In male hindwing presence of forks: one, two and five 1; two and five 2; one and two 3; one 4; two 5; five 6;
none 7

55. Wings; with no apparent coupling device 0, coupled by macrotrichia 1, coupled by hamuli 2

56. Parameres on male genitalia; absent 0, present 1

57. Male pre-anal appendages; with basal branch 1, unbranched 2, absent 3

58. Female genitalia with vagina; sclerotised 1, not sclerotised 2

59. Ninth abdominal sternite; separate from tenth 0; fused to tenth 1

Helicophidae and Conoesucidae. In both trees this grouping of families appears closest to the
Antipodoeciidae.

It appears that certain synapomorphies link the three South African families (Table 2 characters
14, 29, 5 land 53).

Unfortunately the pupae of Antipodoeciidae are at present unknown which results in many mis-
sing characters being introduced into the analysis leading to a lack of resolution (Platnick, Griswold
and Coddington, 1991). To overcome this problem all the pupal characters and a number of other
characters for which character states in several of the families were missing were excluded and a
cladistic analysis on only 44 characters was carried out (Fig. 126).

The most parsimonious single tree produced (Fig. 126) indicates a set of relationships rather
different from that in Fig. 125. The Hydrosalpingidae are most closely related to the Helicophidae
and these two families are then most closely related to the Conoesucidae. This cluster of families is
related to the Chathamiidae and Calocidae. The family Petrothrincidae is closely related to the
Barbarochthonidae which, however, appears most closely related to the Beraeidae which in turn are
most closely related to the Antipodoeciidae. This cluster of families is most closely related to the
Helicopsychidae and Sericostomatidae.

This second analysis is interesting because the Helicophidae are found in both Australasia and
the Neotropical Region. It suggests that the Hydrosalpingidae are more closely related to the
Australasian and South American families than to the other two South African families. This would
strengthen the hypothesis of a temperate Gondwanaland ancestral origin of these families. A more

349

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993
mhennig length 239 ci 42 ri 35 trees 2

1=21

j=8 Anomalopsychidae

[p=0 Sericostomatidae

11 Helicopsychidae
;:{| [j=7 Calocidae

lt=i 6^1=9 Chathamiidae
5 Antipodoeciidae
n| [j=4 Beraeidae
tl7^ Conoesucidae

14=!!=6 Helicophidae

C5 Ani

L-c;

Hydrosalpingidae

i=j| ip=2 Petrothrincidae
L=12Jt=3 Barbarochthonidae

Fig. 12.S.

^0 Sericostomatidae
1=21= =8 Anomalopsychidae

|j^9 Chathamiidae
18^=7 Calocidae

[f=ll Helicopsychidae

lj=5 Antipodoeciidae

[7=3 Barbarochthonidae
|7=12=!1=i Hydrosalpingidae
14^=2 Petrothrincidae
4 Beraeidae

L-c:

6 Helicophidae
13J!=i0 Conoesucidae

Cladograms showing two possible relationships of the presently considered twelve families of Sericostomatoidea.
Analysis based on 59 selected characters (Table 2).

mhennig length 188 ci 41 ri 37 trees 1

[f=8 Anomalopsychidae
1=21= |{=9 Chathamiidae

17JL7 Calocidae

Cj./s=/ caiociaae

|j=10 Conoesucidae
18l [7=1 Hydrosalpingidae

li=15Jt=6 Helicophidae

CO Sericostomatidae

[7=11 Helicopsychidae
lej [7=2 Petrothrincic

1!=14| p=3 Barbarochthonidae

Petrothrincidae

ii=i3=r [j=4 Beraeidae
1!=i2=!!=

122!=5 Antipodoeciidae

Fig. 126. Cladogram showing the possible relationships of the presently considered twelve families of Sericostomatoidea.

Analysis based on 44 selected characters. Character slates 18, 20, 29-37, 41 45, 58 and 59 (Table 2) were lett out for this analysis.

350

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Table 2. The status of 59 selected characters in the larvae, pupae and adults of the eight presently
described families of the Sericostomatoidea plus the three new families and the
Antipodoeciidae. Where there are variations for a character state within one family the
most plesiomorphic state for that character in the family is chosen. (See Table 1 for
explanation of character states). Abbreviations in the table refer to: Serico =
Sericostomatidae, Hydro = Hydrosalpingidae, Petro = Petrothrincidae, Barba =
Barbarochthonidae, Berae = Beraeidae, Antip = Antipodoeciidae, Helico =
Helicophidae, Caloc = Calocidae, Anoma = Anomalopsychidae, Chat = Chathamiidae,
Cono = Conoesucidae, Heli = Helicopsychidae. .S:6

Characters

Serico

Hydro

Petro

Barba

Berae

Antip

Helico

Caloc

Anoma

Chat

Cono

Heli

LARVAE

1 Head shape

2 Head carina

3 Antenna pos

4 Vent apo shp

5 Genae sep

6 Pronot pis

7 Pron Carina

8 Pron div

9 Pre epi size

10 Pre epi shp

11 Meso plates

12 Meta plates

13 Foreleg size

14 Claw size

15 Midleg size

16 Claw size

17 Hindleg size

18 Claw size

19 Abd hmps

20 Setal scler

21 9th abd scler

22 Abdom gills

23 Lat tuberc 8

24 Lat frnge seta

25 10 abd proleg

26 Anal claws

27 Case shape

28 Case struct

PUPAE

29 M pup antenna

30 F pup antenna

31 Pup mandibles

. .

32 Wing sheaths

33 Leg fringes

34 Preseg plates

35 Postseg pis

36 Ab setal fnge

37 Pup case anch

ADULTS

38 Anten length

39 M maxil palps

40 Ocelli

41 Facial warts

42 Pronot warts

43 Mesonotal wts

44 Mesoscutel wts

45 Leg spines col

46 Thyrid cell

47 Androconia

48 Discoid cell

49 Median cell

50 Sc R1 joined

51 M Forks FW

351

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

52 F Forks FW

53 Wing lobes

54 M Forks HW

55 Wing coupling

56 Parameres

57 M prean app

58 Vagina scler

59 9 slern fused

■

careful determination of the relationships of the pupae would provide a clearer picture of the placement
of the Antipodoeciidae.

The variation of possible character states within certain families of the Sericostomatoidea
indicates either a polyphyletic origin for these families or else a convergence of character states. It
confirms Wiggins’ (1982) findings, that the relationships within the sericostomatoid families have
not been sufficiently synthesised and fully resolved. A more detailed analysis of character states
within world genera would possibly improve the determination of relationships within this interesting
and diverse array of families which are so well represented in the southern hemisphere. More detailed
morphological information on larvae, pupae and adults as well as behavioural and life history
information are required before a comprehensive analysis of phylogenetic relationships within the
Sericostomatoidea can be undertaken. A more rigorous analysis to test character consistency among
all genera of the sericostomatoids is necessary before unbiased hypothesis testing of the monophyletic
origin of the sericostomatoid family grouping can be performed.

ACKNOWLEDGEMENTS

Grateful thanks go to Professor G. B. Wiggins, Dr A. Neboiss and Dr F. Schmid, for their kindly
encouragement and advice in the preparation of this paper; also to Dr Neboiss for investigation of
the tentorium of Petrothrincus demoori females. They are not, however, responsible for any mistakes
of the author. Dr K. M. F. Scott wishes to offer particular thanks to Dr F. C. de Moor for his cladistic
analyses, also to both him and Miss H. M. Barber for much assistance in the preparation of this work,
and for the new species Petrothincus demoori, discovered by them in the southern Cape.

Dr H. Malicky (Lunz, Austria) kindly presented to the Albany Museum, on exchange, images of
Petrothrincus circularis andP. triangularis, the latter providing a Neotype male. Special thanks are also
extended to Dr J. M. King for special collections of specimens from the Jonkershoek Nature Reserve
and to Mr J. C. Hodges Jr for the gift of his beautiful photographs of larvae and their cases belonging to
the three new families. Warm thanks are also due to Mr R. Cross and Mrs S. C. Pinchuck from the Rhodes
University Electron Microscopy Unit for assistance with the preparation of material for the scanning
electron micrographs, and to Mrs V. Scott of the Albany Museum for word processing assistance.

Appreciation is expressed to the Entomology Department of the South African Museum for the
long loan of Barnard’s material of these genera, which made this work possible.

Thanks are also due to Mr J. L. Jackson of the Classics Department, Rhodes University and to
Professor K. O. Matier, currently of the University of Durban-Westville, for the provision of the
correct family names for the genera.

Dr F. C. de Moor wishes to thank Dr E. Anderson, J. L. B. Smith Institute of Ichthyology, for
reading through and commenting on the cladistic analysis.

Acknowledgement must also be made to the Department of Environmental Affairs for a permit
to collect in their area in the southern Cape.

Dr F. W. Gess, Editor of the Annals of the Cape Provincial Museums, and Dr S. K. Gess, Assistant

352

SCOTT: THREE RECENTLY ERECTED TRICHOPTERA FAMILIES

Editor, are also warmly thanked for encouragement and much editorial work. Mr B. C. Wilmot,
Director of the Albany Museum, in which this work was carried out, is thanked for accommodation
and assistance.

REFERENCES

Barnard, K. H. 19.34. South African Caddis-flies (Trichoptera). Transactions of the Royal Society of South Africa 21 (4):
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).
Annals of the South African Museum 32 (6); 609-661.

Cowley, D. R. 1976, Family characteristics of the pupae of New Zealand Trichoptera. New Zealand J ourna! of Zoology
3: 99- 109.

Cowley, D. R. 1978. Studies on the larvae of New Zealand Trichoptera. New Zealand Journal of Zoology 5: 639-750.

Farris, J. S. 1988. Hennig 86, version 1.5. Computer program and documentation. Port Jefferson Station, New York.

Fischer, F, C. J. 1964. Trichopterorum Catalogus Vol. V. Phryganeidae, Limnocentropodidae, Molannidae. Amsterdam:
Nedcrlandsche Entomologische Vereeniging.

Fischer, F. C. J. 1970. Trichopterorum Catalogus Vol. XI. Philanisidae, Lepidostomatidae, Brachycentridae, Beraeidae,
Sericostomatidae, Thremmatidae, Helicopsychidae. Amsterdam: Nedcrlandsche Entomologishe Vereeniging.

Flint, O. S. 1981. Studies of Neotropical caddisflies, XXVII: Anomalopsychidae, a new family of Trichoptera. In: Moretti,
G.P. ed, Proceedings of the Third International Symposium on Trichoptera. University of Perugia (Italy)
28 July - 2 August 1980. (Series Entomologica Vol. 20). The Hague: Junk, pp. 75-85.

Hagen, H. A. 1864. Ueber Phryganiden-Gehause. Stettiner entomologische Zeitung 25: 221-263.

Harrison, A. D. 1958a. Hydrobiological studies on the Great Berg River, Western Cape Province. Part 2. Quantitative
studies on sandy bottoms, notes on tributaries and further information on the fauna, arranged
systematically. Transactions of the Royal Society of South Africa 35 (3): 227-276.

Harrison, A. D. 1958b. Hydrobiological studies on the Great Berg River, Western Cape Province. Part 4. The effects of
organic pollution on the fauna of parts of the Great Berg River System and of the Krom River, Stellenbosch.
Transactions of the Royal Society of South Africa 35 (4): 299-329.

Harrison, A. D. 1965. Geographical distribution of riverine invertebrates in southern A'rica. Archiv fiir Hydrobiologie
61 (3): 387-.394.

Harrison, A. D. 1978. Freshwater Invertebrates (except Mollusca). In: Werger, M. S. A. and A C. van Bruggen eds.
Biogeography and ecology of southern Africa. The Hague: Junk, pp. 1141 - 1152.

Harrison, A. D. and 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. Transactions of the Royal
Society of South Africa 35 (3): 125-226.

International Code Of Zoologicai. Nomenci.ature adopted by the XX General Assembly of the International Union of
Biological Sciences. 1985, International Trust for Zoological Nomenclature.

Jacquemart, S. 1963. Trichoptera. In: Hanstrdm, B. et a!., eds, South African animal life. Vol. 9. Uppsala: Amquist and
WiLsels, pp. 337-415.

Leistner, O. a. and Morris, J. W. 1976. South Arican place names. Annals of the Cape Provincial Museums 12: 1-565.

Lepneva, S. G. 1966. [Fauna of the U.S.S.R. Trichoptera. Larvae and Pupae ofintegripalpia Vol. 2 (2). Moscow: Zoological
Institute of the Academy of Sciences of the USSR.) Translated from Russian and published in 1971 by the Israel
program for Scientific Translations, Jerusalem.

Lest AGE, J. A. 1936. Notes trichopterologiques. XIV. Les aimposantes dc la fauna sud-africaine et la dispersion transafricaine de
quelques especes. Bidletin et Annales de la Societe entomologiqiie de Belgique 76: 165-192.

Marlier, G. 1962. Genera des Trichopteres de FArique. Annales du Musee royal de TAfrique centrale. Serie in 8vo.
(Sciences zoologiques) 109: 1-263.

Martynov, A. V. 1924. Preliminary revision of the family Phryganeidae, its classification and evolution. Annals and
Magazine of Natural History (9) 14 (80): 209-224.

McLachlan, R. 1874-1880-1884. a Monographic Revision and Synopsis of the Trichoptera of the European Fauna. 2 vols.
(With Supplement and First Additional Supplement). London: John van Voorst.

Morse, J. C. 1974. New Caddisflies (Trichoptera) from southern Arica. Journal of the Kansas Entomological Society 47
(3): 328-344.

Mosely, M. E. 1939. The British Caddis Flies (Trichoptera): A Collector’s Handbook. London: Routledge.

353

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 14 30 JULY 1993

Mosely, M. E. and Kimmins, D. E. 1953. The Trichoplera (Caddis-flies) of Australia and New Zealand.

London: Trustees of the British Museum (Natural History).

Neboiss, a. 1977. A Taxonomic and 21oogeographic study of Tasmanian Caddis-flies (Insecta: Trichoptera).
Memoirs of the National Museum of Victoria 38: 1 - 208.

Neboiss, A. 1981. Tasmanian Caddis-flies. (Fauna of Tasmania Handbook no. 4). Hobart: University of Tasmania.

Neboiss, A. 1983. Checklist and bibliography of the Australian Caddis-flies (Trichoptera). Australian Society for
Limnology. Special Publication 5: 1 - 132.

Neboiss, A. 1984. Calocidae of North Queensland. In: Morse, J. C. ed. Proceedings of the 4th International Symposium on
Trichoptera. The Hague: Junk, pp. 267 - 276.

Neboiss, A. 1986. Adas of Trichoptera of the 51T Pacific - Australian Region. (Series Entomologica Vol. 37).
Dordrecht: Junk.

Neboiss, A. 1988. Trichoptera. In: Walton, D. W. and W. W. K. Houston eds, Zoological Catalogue of Australia Vol. 6
(Ephemeroptera, Megaloptera, Odonata, Plecoptera, Trichoptera). Canberra: Australian Government Publishing
Service, pp. 177-280.

Neboiss, A. 1991a. Illustrated key to Australian Trichoptera Families and Genera. Trichoptera Taxonomy Workshop,
Murray- Darling Freshwater Research Centre, Albury, Melbourne. Internal Publication.

Neboiss, A. with contnbution by J, C. Dean. 1991b. Trichoptera. In: Division of Entomology, CSIRO ed. The Insects of
Australia. Carlton: Melbourne University Press, pp. 787 - 816.

Platnick, N. I., Griswold, C. E. and Coddington, J. A. 1991. On missing entries in cladistic analysis,
Cladistics 7: 337- 343.

Ross, H. H. 1967, The evolution and past dispersal of the Trichoptera. Annual Review of Entomology 12: 169-206.

Ross, H. H. 1978. The present disposition of components of the Sericostomatidae s. lat. (Trichoptera). In: Crichton, M. I.

ed, Proceedings of the second International Symposium on Trichoptera. University of Reading, England, 25-29
July 1977. The Hague: Junk, pp. 1-30.

Schmid, F. 1980. Les Insectes et Arachnides du Canada. Partie 7. Genera des Trichopteres du Canada et des etats
adjacents. Hull, Canada: Agriculture Canada.

ScHOLTZ, C. H. and Holm, E. (eds) 1985. Insects of Southern Africa. Durban: Butterworths.

Scott, K. M. F. 1955. Some new caddis flies (Trichoptera) from the Western Cape Province - 1. Annals of the South
African Museum 41 (6): 367-380.

Scott, K. M. F. 1967. Trichoptera: provisional key to families (imagos) from Africa south of the Sahara. News Letter of
the Limnological Society of southern Africa 8: 21-34.

Scott, K. M. F. 1985. Order Trichoptera. In Scholtz, C. H. and Holm, E. eds, Insects of Southern Africa. Durban:
Butterworths, pp. 327-340, figs 24.1-24.32.

Scott, K. M. F. 1986. A brief conspectus of the Trichoptera (Caddisflies) of the Afrotropical Region. Journal of the
Entomological Society of Southern Africa 49 (2): 231-238.

Stephens, J. F. 1836. Illustrations of British Entomology. Vol. 6. London (?): Baldwin and Cradock.

Ulmer, G. 1951 , Kocherfliegen (Trichopteren) von den Sunda-Inseln (Teil 1). Archiv fur Hydrobiologie Supplement- Band
19: 1-528.

Ulmer, G, 1955, Kocherfliegen (Trichopteren) von den Sunda-Inseln (Teil II) : Larven und Puppen der Integripalpia, unter
Berucksichtigung verwandter Formen und deren Literatur aus anderen Faunengebieten. Archiv fur
Hydrobiologie Supplement-Band 21 (3/4): 408-608,

Weaver, J. S, 1983. The Evolution and Classification of Trichoptera, with a revision of the Lepidostomatidae and the
North American synopsis of this family. PhD thesis, Clemson University, South Carolina.

Weaver, J. S. 1984. The evolution and classification of Trichoptera. In: Morse, J. C. ed. Proceedings of the Fourth
International Symposium on Trichoptera. Clemson, South Carolina, 11 - 16 July 1983. The Hague: Junk, pp.
413-419.

Wiggins, G. B. 1977. Larvae of the North American Caddisfly Genera (Trichoptera). Toronto: University of Toronto Press.

Wiggins, G. B. 1982. Order Trichoptera. In: Parker, S. P. ed. Synopsis and Classification of Living Organisms.
New York: McGraw-Hill, pp. 599-612.

Wiggins, G. B. 1984. Trichoptera, some concepts and questions. In: Morse, J. C, ed, Proceedings of the fourth International
Symposium on Trichoptera, Clemson, South Carolina. 11-16 July 1983. The Hague: Junk, pp. 1 - 12.

354

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing lYi or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature 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.

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

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

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

Alexander, J. 1969. The indirect evidence for dornestication. 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.AfrJ.Sci. 65 (8): 229-242.

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

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.

CADAR. P.E.-K887

. ■llh

' -rn:

I . .

! y,

■' I

- I

ISSN 0570-1880

,:6I3

MI+

Annals of the

Cape Provincial Museums

Natural History

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

Volume 18 Part 15

15th September 1993

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 m parts at
irregular intervals as material becomes available.

The primary objective of these Annals is to disseminate the results of research work undertakeii 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 connection
with these Annals should be addressed to the Editor, Albany Museum, Grahamstown 6140.

Editor

Dr F. W. GESS

Assistant Editor
Dr S. K. GESS

Dicrotendipes pilosimanus Kieffer: a description of all life stages, and features
which distinguish it from Dicrotendipes septemmaculatus

(Becker)

(Insecta, Diptera: Chironomidae)

A. D. HARRISON

(University of Waterloo, Ontario N2L 3G1, Canada)

ABSTRACT

Dicrotendipes pilosimanus Kieffer andD. septemmaculatus (Becker), both widespread in Africa
and the Palaearctic, have been synonymized under septemmaculatus by some authors. This paper
shows clearly that they are in fact both good species and can be separated in all life stages.

INTRODUCTION

Kieffer (1914) described Dicrotendipes pilosimanus, without illustrations, from a male and a
female collected near Cape Town, South Africa. He described the wing markings, the hypopygium
and the bearded tarsomeres of the male forelegs, the main distinguishing features of this species.

Freeman (1957) gave a fuller description of the male and a short description of the female
D. pilosimanus. He also reduced the status of quatuordecimpunctatus Goetghebuer to a subspecies
of pilosimanus. D. quatuordecimpunctatus is found in central and east Africa and is ‘very similar to
the typical subspecies in colour and general structure including hypopygium but differing in the
absence of the tarsal beard’. With the limited data he had available, he considered that pilosimanus
and quatuordecimpunctatus were the South African and east and central African representatives of
one species.

His opinion was strengthened by the finding near Johannesburg, Transvaal (reported in Harrison,
Keller and Dimovic, 1960; collected by Harrison) of what he regarded as apparent intermediate stages.
However, the specimens mentioned in the paper were obtained from a small, swampy lake,
Olifantsvlei, which was polluted by acid sulphates from the gold mines, and were not intermediates
but were stunted specimens of pilosimanus from a very unsuitable, acid habitat.

These Olifantsvlei specimens have resulted in some taxonomic confusion and were one of the
factors that led Epler (1988) to synonymize pilosimanus Kieffer with septemmaculatus (Becker) and
quatuordecimpunctatus Goetghebuer. Cranston and Armitage (1988) also synonymized pilosimanus
and septemmaculatus.

D. septemmaculatus is a widespread species found mostly in the Palaearctic, with the type locality,
Tenerife, Canary Islands; Epler (1988) considers quatuordecimpunctatus to be a synonym but
Contreras-Lichtenburg (1986, 1988) considers that these names apply to two distinct species, basing
her opinion on larval and pupal characters. The present author has identified specimens found in and
around Ethiopian Rift Valley lakes as septemmaculatus, as the larvae are similar to those described
by Contreras-Lichtenberg (1988) for this species.

The aim of this paper is to define clearly in detail all stages of pilosimanus, using specimens
collected near Cape Town, as well as some from Ethiopia and Zimbabwe, and to show that this species
can be distinguished easily from septemmaculatus, using specimens from Ethiopia and the published
descriptions.

357

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 1993

METHODS

Adults were caught by sweeping vegetation near lakeshores; many were caught at lights. Some
Ethiopian specimens were bred out in the laboratory in Addis Ababa and some South African
specimens in the author’s home near Cape Town. Larvae and pupae were caught by hand net in rivers
and lakes.

All specimens were preserved in 70% alcohol and mounted in Canada Balsam dissolved in
Cellosolve, or in Euparal. Specimens were not macerated in 10% KOH, as this may damage fine
structures and may upset the arrangement of spermathecal capsules and ducts in females. Canada
Balsam is the better mountant, as muscles are more transparent in it.

Measurements were made with an eyepiece micrometer and all drawings with a drawing tube
on the microscope.

Generic definitions of males follow Epler (1988) and Cranston et al. (1989); of females Epler
(1988) and Saether (1977); of pupae Finder and Reiss (1986); and of larvae Finder and Reiss (1983).
Morphological terminology is from Saether (1980).

Abbreviations used in this paper are:

AR = antennal ratio. Ratio of length of apical flagellomere to combined length of basal
flagellomeres.

LR = leg ratio. Ratio of length of tarsomere 1 to length of tibia.

SV = ‘Schenkel-Schiener-Verhiiltnis.’ Ratio of femur plus tibia to tarsomere 1.

BV = ‘Beinverhaltnisse.’ Combined length of femur, tibia and tarsomere 1 divided by combined
length of tarsomeres 2 to 5.

DESCRIFTIONS

Dicrotendipes pilosimanus Kieffer 1914

Chironomus (Dicrotendipes) pilosimanus, Freeman 1957.

Dicrotendipes pilosimanus, Freeman and Cranston, 1980.

Dicrotendipes pilosimanus, Contreras-Lichtenberg 1986.

Dicrotendipes septemmaculatus (part), Epler 1988.

The type series of this species in the South African Museum, Cape Town consists of two
specimens glued to a card on a pin; the male cotype has wings, with characteristic pattern of spots
but the abdomen and forelegs are missing; the female cotype also has no abdomen but the wings are
entire. The specimens were collected in the vicinity of Cape Town by Dr L. Feringuey.

The following descriptions are based on fresh material collected by the present author in the
vicinity of Cape Town, and in the Ethiopian Highlands. All stages have been associated by rearing
larvae.

Adult male (N = 14 mounted)

Close to generic definition in Cranston et al. (1989).

Wing length. 2.2 - 3.3 mm; mean 2.7 mm.

Colour (in alcohol). Head creamy yellow, antennae brown, mouthparts light brown. Thorax

358

HARRISON: DICROTENDIPES PILOSIMANUS KIEFFER: A DESCRIPTION OF ALL LIFE STAGES

background creamy yellow, mesonotal stripes brown with darker edges, central stripes fused,
postnotum dark brown, median anepisternum 11, epimeron 11 and preepisternum dark brown. Legs:
all coxae creamy yellow; foreleg - femur light with dark ring at tip, knees light, tibia and tarsomeres
1 and 2 light with dark tips, tarsomere 3 light at base, darkening to tip, tarsomeres 4 and 5 dark; mid
and hindlegs - femora light, tibiae light, darkened at tip (specially hind), tarsohieres 1 to 3 light,
darkened at tip, 4 and 5 dark. Wings: light elongated spot at tip of r4^5, dark spots - one central in
r4+5, one at base of r4^_5 near cross vein continuing into m, one in mi^.2, one in m3^, 2 in anal cell (see
Freeman’s photograph). Abdomen background light, large darker spots in centre of tergites I - Vlll,
and lateral edges of tergites darkened, tergite IX and hypopygium darker.

Head. AR 2.3 - 3.1; mean 2.6. Frontal tubercles about twice as long as wide. Palp segments:
small specimen, 31, 47, 146, 149, 205 pm; large specimen, 58, 62, 183, 171, 233 pm; large Ethiopian
specimen, 78, 78, 183, 189, 280 pm. Five to six subapical sensilla on segment 3.

Thorax. Scutal tubercle present. Setation: Lateral antepronotals absent, dorsocentrals 10-15
(mean = 12) uniserial, number not obviously related to size, posterior prealars seven, scutellars seven
to eight uniserial.

Wings. Setation: Brachiolum two, R 20-22, Rj 15-19, R4^5 17-19, squama in large specimens
22, small specimens 16.

Legs. LR fore 1.35-1.5, midO.5-0.6, hind 0.6; SV fore 1.5-1. 6; BV fore 1.8-1. 9. Sensilla chaetica
on tarsomere 1: midleg 10-13, hindleg absent. Beard on tarsomeres 1, 2 and 3 of foreleg. Ratio of
beard setal length to width of tarsomere : tarsomere 1 about 3.8, tarsomere 2 in South African specimens
5. 5-6. 5, in Ethiopian specimens about 5. Tibia and tarsomeres of hind legs also have very long setae.

Hypopygium (Figs 1 and 3). Bands on anal tergite usually short and indistinct but clearer in some
specimens and meeting centrally to form central band (Fig.l); about 16 or 17 apical anal tergite setae;
apex of anal tergite with lateral hyaline lobes that show best on pinned specimens (Freeman) but
usually become tucked under the tergite in specimens preserved in alcohol (see Fig. 1 where they are
not visible); anal point down-turned. Superior volsella long and curved with two terminal setae;
inferior volsella bowed dorsoventrally, bifid usually with five apical and one subterminal spinose
setae, dorsally (Fig. 1), some with only the five apical ones; the Ethiopian specimens are like this but
one has also a small subterminal seta on one side. Gonostylus evenly curved. Apodemes as in Fig.
3, the phallapodemes are very long and cross one another, in some specimens they are moved nearer
to the anal point; the tips are usually blunt.

Adult FEMALE (N = 8 mounted)

Close to generic definition.

Wing length. South African specimens 2.4 - 2.6 mm (mean 2.5 mm). Ethiopian specimens 3.1 -
3.6 mm (mean 3.3 mm).

Colour (in alcohol). Similar to male. Spots on abdominal tergites lighter than in male and edges
of tergites or|ly slightly darkened; tergite IX light. Genitalia: Sternite Vlll brown and chitinized
portions of IX and X brown.

Head. AR 0.45 - 0.50. Frontal tubercles similar to male. Palp segments in South African
specimens: larger, 62, 47, 140, 140, 217 pm; smaller, 53, 53, 124, 142, 217 pm; in large Ethiopian
specimen: 93, 78, 177, 186, 298 pm. Five to six subapical sensilla on segment 3.

Thorax. Scutal tubercle present. Setation: Lateral antepronotals absent, dorsocentrals 21-22,
posterior prealars six to seven, scutellars 11-12 irregularly biserial.

359

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 199.'!

Figs 1-5. Dicrotendipes pilosimanus. Adult: 1. male hypopygium; 2. female genitalia, ventral; 3. male phallapodemes; 4.
female genitalia, lateral; 5. seminal capsules and ducts, dorsal.

360

HARRISON: DICROTENDIPES PILOSIMANUS KIEFFER: A DESCRIPTION OF ALL LIFE STAGES

Wings. Setation: Brachiolum 1-2, R 24-25, Rj 18-23, R4^_5 34-36, squama in South African
specimens 20 - 25, in Ethiopian specimen about 30 partly biserial.

Legs. LR fore 1.48, mid 0.46 - 0.53, hind 0.59 - 0.62. Sensilla chaetica on tarsomere 1: midleg
about 50 in South African specimens, about 94 in Ethiopian specimen, biserial to triserial; hindleg
absent.

Genitalia (Figs 2, 4 and 5). Sternite VllI lightly chitinized, ventral ridge of VIII narrow;
dorsomesal and ventrolateral lobes large, ventrolateral lobe is usually seen in side view (Fig. 2, right)
but appears much larger when specimen is compressed; apodeme lobe small and narrow with
microtrichia (Fig. 2, left); gonocoxapodemes light (stippled in figure), joined. Coxosternapodemes
dark and curved (black in figures); gonocoxite IX small with three setae and small dorsal process
(Figs 2 and 4). Segment X with chitinized anterior edge (Fig. 4) and with 8 - 13 setae per side (most
specimens with 8), postgenital plate pointed. Labia without microtrichia. Seminal capsules (Fig. 5)
with short, narrow neck, 127 - 158 pm long, ducts almost straight joining to common opening, tending
to run parallel to each other a short distance before joining, central wider portion glandular.

Pupa (N = 25 mounted)

Similar to generic definition.

Colour. Exuviae are a fairly uniform yellowish brown, shagreen on abdomen darker.

Cephalothorax. Cephalic tubercles (Fig. 6) high, base granulose, frontal setae long; anterior of
dorsum and humeral callus (see Epler, 1988) covered with short, blunt spines, rest of dorsum pebbled;
dorsocentral setae 1 and 2 close together.

Abdomen (Fig. 7). Tergites I, and VI - VIII with weak reticulate pattern, not shown in figure.
Tergite I bare, the rest with shagreen of posteriorly pointing spines, which become larger on posterior
part of tergites II - VI, forming paired patches; on tergites V and VI these posterior spines are larger
than on the other tergites, notably on VI. No shagreen in conjunctives. Posterior hook row on tergite
II. No ventral spinal rows on sternites. Caudolateral spurs (Fig. 8) 1, 2 or 3 in the following pairs: 1
and 1-1 specimen; 1 and 2 - 10; 1 and 3 - 1; 2 and 2 - 6; 2 and 3 - 2; 3 and 3-5. Pedes spurii B are
on segment II, and A on sternite IV.

Larva (N = 16 mounted)

Similar to generic definition.

Colour. Head capsule light brown to brown, darker on ‘cheeks’, posterior rim black. Body green
in life, yellowish preserved, anterior claws light brown, posterior claws brown.

Head capsule. Length 585 - 780 pm; mean 702 pm.

Dorsal surface of head (Fig. 9). Frontal apotome with small and narrow frontal process (Epler’s
term) or mark, frontal margin smooth, not crenate. Anterolateral projections short.

Antenna (Fig. 10). AR 0.63, blade slightly shorter than flagellum, segment 4 about six times as
long as wide.

Labrum. S I with nine points. Pecten epipharynx (Fig. 11) with five teeth, chaetulae laterales
comparatively small (Fig. 11), premandible with two teeth and brush (Fig. 12).

Mandible (Fig. 13). Dorsal tooth pale, four other teeth very dark. Pecten mandibularis (dorsal)
with about 13 setae, mostly long and curved. Seta subdentalis pointed.

Mentum (Figs 14, 15). Unworn central tooth pointed with lateral notches, six lateral teeth, the

361

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 1993

Figs 6-15. Dicrotenclipes pilosimanus. Pupa; 6. cephalic tubercle; 7. abdomen, dorsal; 8. caudolateral spurs. Larva: 9. dorsal
surface of head; 10. antenna; 11. pecten epipharynx; 12. premandible; 13. mandible; 14. mentum, unworn; 15.
mentum, worn, and ventromental plate.

362

HARRISON: DICROTENDIPES PILOSIMANUS KIEFFER: A DESCRIPTION OF ALL LIFE STAGES

second forming a notch on the first. Most specimens were worn (Fig. 15). Ventromental plates have
40 - 46 strial ridges.

Maxilla (Fig. 16). Similar to Epler’s figure showing generic structure. Numerous pointed
chaetulae of palpiger on dorsal surface, not shown in Fig. 16.

Body. Claws of anterior parapod mostly long and curved, specially at tip, surface appears to be
scaly at xlOOO magnification, shorter claws pectinate towards tip, longer claws simple. Posterior
claws simple; anal tubules short and blunt.

Specimens examined. Adults: Numerous males and females netted on the shore of the small
Noordhoek Lake, near Cape Town, ii.l992; 2 males and 2 females caught at lights, Addis Ababa,
xi.l982; 2 males and 2 females caught at lights, Addis Ababa, ix.l983; 1 male and 1 female bred out
from samples from stony run in polluted Kebena River, Addis Ababa, ix.l985; 1 male netted near
sewage maturation ponds, Marendera (=Marendellas), Zimbabwe 18.x. 61, Coll. A.D.H.. Pupae: 25
exuviae from Noordhoek Lake, ii.l992. Larvae: numerous larvae netted in weed beds in Noordhoek
Lake, ii.l992; numerous larvae netted in stony runs in Kebena River, Addis Ababa, 1983-84.

Comments. Kieffer’s cotypes both have the wing markings characteristic of the species and the
female has more than 20 squamal setae. Kieffer states that the male has long setae on tarsomeres 2-
4 of the foreleg about five to six times as tong as the thickness of the joints; this is similar to the
specimens described here.

Ecology. Harrison (1958) found this species in alkaline lakes in the Western Cape Province, South
Africa, but not in those with a pH in the acid range. The specimens used in this paper came from
weed beds of Potomogeton sp. in a small alkaline lake near Cape Town; pilosimanus is also found
in slow-flowing alkaline to neutral streams in the same region.

In Zimbabwe it was breeding in the sewage maturation ponds at Marendera (=Marendellas),
about 50 km from Harare.

In Ethiopia it was found in ponds, one above 4500 m in the Afro-alpine region of the Bale
mountains, but none was found in the lakes; most larvae came from the torrential Kebena River (at
about 2600 m) running through Addis Ababa. This stream received considerable amounts of organic
pollution from the city, which encouraged this species as well as Chironomus alluaudi (Tesfaye Berhe
era/., 1989).

D. pilosimanus prefers slow-flowing streams and small eutrophic lakes and ponds. It is able to
take advantage of polluted torrential streams in the same manner as Chironomus spp.

Distribution. Found over most of southern Africa (Freeman, 1957); it is replaced over most of
central and east Africa by D. quatuordecimpunctatus but occurs again in the Ethiopian Highlands;
Contreras- Lichtenberg (1986) reports it from north Africa and Israel and the Balkans, as well as from
the Oriental and Australasian regions.

Dicrotendipes septemmaculatus (Becker 1908)

Chironomus septemmaculatus Becker, 1908.

Dicrotendipes pilosimanus subsp. quatuordecimpunctatus (part?) Freeman 1957.

Dicrotendipes septemmaculatus, Cranston and Armitage 1988.

Dicrotendipes septemmaculatus, Epler 1988.

363

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 1993

364

HARRISON; DICROTENDIPES PILOSIMANUS KIEFFER; A DESCRIPTION OF ALL LIFE STAGES

Dicrotendipes septemmaculatus, Contreras-Lichtenberg 1988.

Becker’s description is based on one female type from the Canary Islands (Cranston and
Armitage). Freeman does not illustrate the male of quatuordecimpunctatus but says that it is similar
to that of pilosimanus s.str. except that it lacks the beard on the tarsomeres of the foreleg. Epler and
Contreras-Lichtenberg describe the immatures.

Epler considers that septemmaculatus and quatuordecimpunctatus are conspecific but Contreras-
Lichtenberg considers that they are distinct species; she separates them mainly on the structure of
the dorsal sclerites of the larval head capsule. The larvae of the Ethiopian populations dealt with here
are similar in some respects to her septemmaculatus.

Because of the controversy regarding the identity of septemmaculatus, quatuordecimpunctatus
and pilosimanus, these Ethiopian specimens, including the female, are described here in detail.

Adult male (N = 5 mounted)

Close to generic definition.

Wing length. 2.15 - 2.60 mm., mean 2.40 mm.

Colour. Unmounted specimens had been in alcohol for more than seven years by the time they
were examined and had lost almost all their colour. Nevertheless, judging from other Ethiopian
specimens mounted while they were still fresh, the colour pattern is very similar to that of pilosimanus,
including the legs.

Head. AR 3.1. Frontal tubercles twice as long as wide. Palp segments: 71, 78, 177, 174, 267 pm.
Five subapical sensilla on segment 3.

Thorax. Scutal tubercle present. Setation: Lateral antepronotals absent, dorsocentrals nine
uniserial, posterior prealars five, scutellars six uniserial.

Wings. Dark markings as in Cranston and Armitage (1988), but elongated spot at tip of cell r^
faint. Setation: Brachiolum 1, R 19 - 23, R, 15-18, R4^5 16 - 20, number does not seem to be correlated
to size of wing, squama 10.

Legs. LR fore 1.5, mid 0.5, hind 0.6; SV fore 1.4- 1.5; BV fore 1.8. Colour: foreleg with dark
tips to femur, tibia and tarsomeres 1 and 2, rest of tarsomeres wholly dark; midleg with femur plain,
dark tips to tibia and tarsomeres 1 -4, 5 dark; hindleg with femur plain, dark tips to tibia and tarsomeres
1-3, 4 and 5 dark. Sensilla chaetica on tarsomere 1 : midleg 10, hindleg absent.

Hypopygium (Figs 17, 18). Bands on anal tergite short and indistinct (not in figure), median anal
tergite setae absent, about 10 apical anal tergite setae; apex of anal tergite with lateral hyaline lobes;
anal point down-turned. Superior volsella long and curved with two to three terminal spinose setae
(Table 1); inferior volsella bowed dorsoventrally, bifid, with strong apical spinose setae, the variable
number of which appears to be related to size of specimen (Table 1). Gonostylus evenly curved.
Apodemes as in Fig. 18.

Adult female (N = 5 mounted)

Close to generic definition.

Colour. Similar to female pilosimanus, including the legs.

Wing length. 2.3 mm.

Head. AR 0.4. Frontal tubercles similar to male. Palp segments: 47, 47, 99, 124, 189 pm. About
six subapical sensilla on segment 3.

365

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 1993

Table 1. Dicrotendipes septemmaculatus. Number of terminal spinose setae on volsella related to
wing length. (Left/right)

Wing length mm

2.60

2.40

2.35

2.15

Sup. volsella

3/3

2/2

2/3

2/2

Inf. volsella:

main

7/8

7/7

5/6

5/5

branch

4/4

5/5

4/4

3/3

Thorax. Scutal tubercle present. Setation: Lateral antepronotals absent, dorsocentrals 16uniserial,
posterior prealars five, scutellars four or five.

Wings. Markings as in male. Setation: Brachiolum two, R 26, Rj 20, R4^_5 35, squama six.

Legs. LR fore 1.5, mid 0.5, hind 0.6. Markings as in male. Sensilla chaetica on tarsomere 1:
midleg 34 - 46 partly biserial; on hindleg absent or one. Three of the specimens had one on tarsomere
2 of midleg.

Genitalia (Figs 19, 20 and 21). Sternite VIII lightly chitinized; ventral ridge of VIII narrow;
dorsomesal and ventrolateral lobes large (Fig. 19, right), apodeme lobe small and narrow, with
microtrichia (Fig. 19, left); gonocoxapodemes light, do not appear to be joined. Coxosternapodemes
dark and curved; gonocoxite IX small with two setae and small, dorsal chitinized process. Segment
X with six setae per side, postgenital plate pointed. Labia without microtrichia. Seminal capsules
(Fig. 21) oval with short neck, 108 pm long, ducts almost straight joining at an angle of about 45o
or more to common opening, central part glandular.

Pupa

None found, but pupae are described by Contreras-Lichtenberg and Epler.

Larva (N = 7 mounted)

Similar to generic definition. Drawings were made from a newly moulted 4th instar with
mouthparts unworn.

Head capsule. 520 - 585 pm long; mean 553 pm.

Dorsal surface of head (Fig. 22). Frontal apotome with distinct frontal process or mark, frontal
margin crenate on either side of this. Anterolateral projections small and short.

Antenna. Very similar to those of D. pilosimanus. AR 0.73. Antennal blade longer than flagellum,
segment 4 about six times as long as wide.

Labrum. Similar to pilosimanus. S 1 with nine points. Pecten epipharynx (Fig. 23) with five, six
or seven teeth. Premandible (Fig. 24) with two teeth and brush.

366

HARRISON: DICROTENDIPES PILOSIMANUS KIEFFER: A DESCRIPTION OF ALL LIFE STAGES

Figs 22-26. DicrotemUpes sepieiwnaculatus. Larva: 22. dorsal surface of head; 23. pecten epipharynx; 24. premandible;
25. mandible; 26. mentum and ventromental plale.

Mandible (Fig. 25). Dorsal tooth pale, the four other teeth very dark, seta subdentalis simple and
pointed.

Mentum (Fig. 26). Central tooth rounded and notched, six lateral teeth, the second forming a
notch on the first. The ventromental plates have 26 - 30 strial ridges.

Maxilla. Similar to that of pilosimanus.

Body. Claws of anterior parapods mostly long and thin, the larger with ends hooked and minutely
serrate towards tip, the smaller not hooked or serrate. Claws of post parapods not serrate, anal tubules
slightly tapered with rounded tips, much shorter than parapods.

Specimens examined. Adults: numerous males and 5 females, caught at lights. Lake Awasa, Rift
Valley, 15.iii.l981; 1 male and 1 female, netted at Lake Awasa, 25.viii.1984; Larvae: 4 netted in
weed beds. Lake Awasa, 1983-84; 1 from shallow bottom. Lake Ziway, 5.vii.l984; 1 from Akaki
River, downstream from Debre Zeit, 16.ii.l984; 1 from Bulbulla Shet flowing out of Lake Langano,
Rift Valley, 24.ii.1984.

Comments. The female holotype of septemmaculatus from the Canary Islands has been mounted
and redescribed by Cranston and Armitage. The arrangement of the spermathecal ducts and the number
of squamal setae (about 10) in the Canary Island specimen seem to indicate that the Ethiopian
specimens are correctly placed in this species. Also the association here between the females, males
and larvae indicates that the specimens described by Contreras-Lichtenberg and Epler also are
correctly placed in this species.

In addition, Cranston and Armitage redescribed the male lectotype of Stictochironomus
sexonolaliis Goetghebuer, from southern Spain, and consider it to be a junior synonym of D.
septemmaculatus. The number of terminal spinose setae on the superior volsella (three) and inferior

367

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 1993

volsella (six) fit in with septemmaculatus and not pilosimanus. (See Table 2).

Ecology. The larvae were found in weed beds, on the bottom of shallow lakes or in slow-flowing
rivers.

Distribution. Epler lists material from the Afrotropical Region, the southern Palaearctic, Australian,
and Oriental Regions. In Africa this species or its (?) allied species quatuordecimpunctatus, has not
yet been found south of Zimbabwe although one of them probably exists in the subtropical regions
of South Africa.

Table 2. Dicrotendipes pdosimanus and D. septemmaculatus contrasted.

D. pilosimanus

D. septemmaculatus

Adult male:

wing length

2.20 - 3.30 mm

2.15 - 2.60 mm

dorsocentrals

10-15

squamal setae

16-22

about 10

beard on tarsomeres of foreleg

present, longest on 2 and 3

absent

hypopygium: terminal spinose

setae on inferior volsella

5, in all specimens

5-8, depending on size

apical anal tergite setae

about 16-17

about 10

Adult female:

dorsocentrals

21 -22

squamal setae

S. Africa: 20-25
Ethiopia: about 30

genitalia: seminal ducts

running almost parallel
to each other just
before joining

joining at an angle of
about 45^ or more

Pupa:

cephalic tubercle

large

not specially large

shagreen on conjunctives

absent

ivA', vm, VI/VII

caudolateral spurs of VIII

1,2 or 3

one

Larva:

head capsule length

585-780 mm.

520-585 mm.

frontal apotome

frontal margin smooth

frontal margin crenate

ventromental plates

40-46 strial ridges

26-30 strial ridges

368

HARRISON: DICROTENDIPES PILOSIMANUS KIEFFER: A DESCRIPTION OF ALL LIFE STAGES

DISCUSSION

The two species, D. pilosimanus and D. septemmaculatus, are very similar and obviously fall
into the same species group. However, it is apparent from Table 2 that it is fairly easy to distinguish
all life stages.

The males can be separated by means of the tarsal beard and large number of squamal setae of
pilosimanus, and the females by the large number of squamal setae of pilosimanus and its special
and consistent arrangement of the seminal ducts which run almost parallel to each other before joining.

The pupae can be separated by the multiple caudolateral spines of pilosimanus. Also, the shagreen
on tergites II and V is more extensive than that of septemmaculatus and quatuordecimpunctatus, as
illustrated by Contreras-Lichtenburg.

The larvae can be separated by the smooth frontal margin of their frontal apotome and the larger
number of striae (40-46) on the ventromental plate of pilosimanus.

The two species differ also in their ecological requirements. D. pilosimanus is found in productive
lakes and ponds, mainly in weed beds, but not where the pH is in the acid range (Harrison, 1958). It
is also found in slow-flowing streams. In Ethiopia it was found in a torrental stream strongly polluted
with organic matter from the city of Addis Ababa; here it formed part of the usual 'pollution
community’ oiTubifex, Chironomus, Psychodidae and resistant Baetidae(TesfayeBerhe era/., 1989).
It was also found in a pond in the Afro-alpine region of the Bale Mountains. In Zimbabwe it was
found in sewage oxidation ponds but nowhere else, although the region had been extensively collected
by the author during 1961-63 (Harrison: unpublished data); other small ponds and impoundments
there contained quatuordecimpunctatus.

In Ethiopia D. septemmaculatus was found in Rift Valley lakes, mainly in weed beds but also
on shallow bottoms (Tilahun Kibret and Harrison, 1 988, listed as quatuordecimpunctatus; Tudorancea
et ai, 1988, under Dicrotendipes spp.). It is yet to be found in South Africa. Dejoux (1983) does not
report it from Lake Chad although he did have some unidentified Dicrotendipes larvae.

It is too early to attempt to resolve the septemmaculatus vs. quatuordecimpunctatus controversy.
If they are separate species they resemble one another more closely than either resemble.? pilosimanus.
More material of all life stages from the type localities is required, preferably correlated by rearing.

ACKNOWLEDGEMENTS

This study forms part of a programme of cooperative research on fisheries and limnology,
developed between Addis Ababa University, Ethiopia, and the University of Waterloo, Canada, and
aided by the Canadian International Development Agency. The author wishes to thank Dr J. Rankin
for preparing the ink drawings from his drawing tube outlines.

REFERENCES

Becker, T. 1908. Dipteren der Kanarischen Inseln. Mill. zool. Mus. Berlin. 4: 1-180.

Contreras-Lichtenberg, R. 1986. Revision der in der Westpalaarktis verbreiteten Arten des Genus Dicroiemlipes
Kieffer, 1913 (Diptera, Nematocera, Chironomidae). Ann. Naturhist. Mus. Wien 88/89: 663-726.
Contreras-Lichtenberg, R. 1988. Zur Kenntnis der Jugendstadien von Dicroiemlipes sepiemmaculaius (Becker,
1908) (Diptera, Nematocera, Chironomidae). Zeii. Arheiisgemeinschafi dsierr. Eniontologen 40: 45-48.
Cranston. P. S. and Armitage, P. D. 1988. The Canary Islands Chironomidae described by T. Becker and by Santos
Abreu. Disch. eni. Z., N. F. 35: 341-354.

369

ANN. CAPE PROV. MUS. (NAT. HIST.) VOL. 18, PT. 15 15 SEPTEMBER 199.1

Cranston, P. S., Dillon, M. E., Finder, L. C. V. and Reiss, F. D, 1989. The adult males of Chironomidae (Diptera:
Chironomidae) of the Holarctic region - keys and diagnoses. In: Wiederholm, T. ed., Chironomidae of the Holarctic
Region. Part 3. Adult males. Ent. Scand. Suppl. 34, pp. 353-502.

Deioux. C. 1983. The fauna associated with the aquatic vegetation. In: Carmouze, J.-P., J.-R. Durand and C. Leveque ed.,
Lake Chad, ecology and productivity of a shallow tropica! ecosystem, Monographiae Biologicae 53, The Hague:
Junk, pp. 273-292.

Epler. J. H. 1988. Biosystematic.s of the genus Dicrotendipes Kieffer, 1913 (Diptera: Chironomidae: Chironominae) of the
world. Mem. Amer. ent. Soc. 36: 1-214.

Freeman, P. 1957. A study of the Chironomidae (Diptera) of Africa south of the Sahara, Part 3. Bull. Brit. Mus. (Nat. Hist)
Ent. 5: 323-426.

Freeman, P. and Cranston, P. S, 1980. Family Chironomidae. In: Crosskey, R. W. ed., Catalogue of the Diptera of the
Afrotropical Region. London: British Museum (Natural History), pp. 175-202.

Harrison, A. D. 1958. Hydrobiological studies on alkaline and acid still waters in the Western Cape Province. Trans, r. Soc.
sth.Afr. 36: 213-244.

Harrison, A. D., Keller, P. and Dimovic, D. 1960. Ecological studies on Olifantsvlei, near Johannesburg. Hydrobiologia
15: 89-134.

Kieffer, J. J. 1914. South African Chironorrudae (Diptera). Ann. S. Afr. Mus. 10: 259-270.

Finder, L. C. V. and Reiss, F. 1983, The larvae of Chironominae (Diptera: Chironomidae) of the Holarctic region - keys and
diagnoses. In: Wiederholm, T. ed., Chironomidae of the Holarctic region. Part 1. Larvae. Ent. Scan. Suppl. 19:
293-435.

Finder, L. C. V. and Reiss, F. 1986. The pupae of Chironominae (Diptera: Chironomidae) of the Holarctic region - keys and
diagnoses. In: Wiederholm, T. ed., Chironomidae of the Holarctic region. Part 2. Pupae. Ent. Scan. Suppl. 28:
299-456.

Saether, O. a. 1977. Female genitalia in Chironomidae and other Nematocera. Bull. Fish. Res. Bd Canada. 197: 1-209.

Saether, O. a. 1980. Glossary of chironomid morphology terminology (Diptera: Chironomidae). Ent. Scan. Suppl. 14: 1-51.

Tilahun Kjbret and Harrison, A. D. 1989. The benthic and weed-bed faunas of Lake Awasa (Rift Valley, Ethiopia).
Hydrobiologia 174: 1-15.

Tesfaye Berhe, Harrison, A. D. and Hynes, H. B. N. 1989. The degradation of a stream crossing the city of Addis Ababa,
Ethiopia, Tropical Freshwater Biology 2: 112-120.

Tudorancea, C, Baxter, R. M. and Fernando, C. H. 1989. A comparative limnological study of zoobenthic associations
in lakes of the Ethiopian Rift Valley. ArcJi. Hydrobiol. Suppl. 83: 121-174.

370

INSTRUCTIONS TO AUTHORS

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

MANUSCRIPTS should be submitted on floppy disk accompanied by three print-outs
(spacing I/2 or 2). In exceptional cases typed manuscripts will be accepted. These 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 marking
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
nomenclature 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.

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

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

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

Aliixander, 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.AfrJ.Sci. 65 (8): 229-242.

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

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.

CADAR, P.E.-K887

:■ ■ -V , i:j:: .

r-rV'""' ■=

■■; ‘ ‘ .i. ■

/■

913..,Y™if

10/03/ IG ^^15750 ^

f-';

■

■i

t',;'i51

1 -

wJ'V''

SMrTHSONIAN INSTITUTION LIBRARIES

III

III I I

III

088 01740 7404

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Annals of the Cape provincial Museums"

See other formats