Memoirs

of THE

Queensland Museum

Brisbane
June, 1981

Volume 20
Part 2

Mem. Qd Mus. 20(2): 237-51, pis. 1-2. [1981]

STUDIES IN THE CORONULIDAE (CIRRIPEDIA):

SHELL MORPHOLOGY, GROWTH, AND FUNCTION, AND THEIR BEARING
ON SUBFAMILY CLASSIFICATION

Ronald Monroe

Queensland Museum*

ABSTRACT

A new morphological feature of the shell of Chelonibia caretta is described. A species
incertae sedis is compared with Cylindrolepas darwiniana Pilsbry, 1916. Cylindrolepas may be
a junior synonym of Platylepas on the basis of this comparison. The shell morphology and
probable mode of growth of Coronulidae are described. A new arrangement of genera into
subfamilies is proposed. The subfamily Coronulinae Leach, 1812, is emended to include
Coronula, Cetopirus, Cryptolepas, and Platylepas. The subfamily Xenobalaninae Gruvel, 1 903,
is emended to include Xenobalanus, Tubicinella, Stomatolepas, and Stephanolepas.
Chelonibiinae Pilsbry, 1916, and Emersoniinae Ross, 1967, remain unaltered, and
Platylepadinae (auct.) is rejected. A possible phylogeny for the family is included.

INTRODUCTION

Monroe and Limpus (1979) indentified and
listed the barnacle epifauna on turtles in
Queensland Waters. Subsequent collection has
yielded one other species of uncertain identity, and
subsequent examination of specimens of
Chelonibia caretta has clarified some aspects of
shell morphology. Analysis of growth patterns and
shell structure has thrown new light on the
probable subfamilial relationships.

Specimen numbers prefixed by W refer to the
Queensland Museum.

ADDITIONS TO SPECIES DESCRIPTIONS
OF MONROE AND LIMPUS, 1979.

Chelonibia caretta (Spengler, 1790)
(Plate 1, Figs. 1, 2)

Material Examined

W3654, 15 specimens, North Reef, SE.Q.

Previous authors have remarked on the
roughness of the exterior surface of the shell (e.g.
Pilsbry 1916, Monroe and Limpus 1979).
However, when the outer surface and base of
cleaned specimens are examined, the roughness is
attributable to infoldings of the outer wall (Plate
1, Fig 2, A). These structures are not present in
Chelonibia testudinaria (Plate 1, Figs. 3, 4).

?Cylindrolepas darwiniana Pilsbry, 1916
(Plate 1, Figs. 5-8; Plate 2, Figs. 1, 2)

From host number X9313, Caretta caretta,
Mon Repos, 2.xi.l972, an empty shell was
removed from the area round the base of the tail.
In appearance the shell has obvious affinities with
Platylepas. The midrib structure and the ridges of
beading on the outer walls are very similar to
those of P. decorata.

However the specimen also agrees well with
Pilsbry’s (1916) description of Cylindrolepas
darwiniana, but not as well with his figures. These
latter are not to the usual high standard of
Pilsbry’s drawings. It has not been possible to
borrow the type of Cylindrolepas darwiniana,
through no fault of the Academy of Natural
Science Philadelphia. Pilsbry’s (1916) description
is reproduced below for comparison with the
plates.

‘The barnacle is hexagonal, the carinoros-
tral diameter a little larger than the lateral,
of about equal diameter from base to
summit; whitish, with fine sculpture of
close transverse wrinkles, and on the carina
and carinolateral compartments a few low,
coarse vertical ribs. The compartments

’Present address; ‘Heather Cottage’, Bradda Rd, Pt
Erin, Isle of Man.

238

MEMOIRS OF THE QUEENSLAND MUSEUM

when isolated are square. Their summits
are beveled and polished, apparently by
wear. A median fold or Filled sulcus is
indicated on the polished summit by a
small depression filled with the softer and
dull substance of the outer layer; and on
some compartments a slight, mesial sulcus
is visible externally. The radii are
represented by narrow sulci; their edges are
distinctly septate. The lower edges of the
compartments have about three short,
vertical, blunt teeth on each side of a larger
median tooth, which bends slightly inward,
and is homologous with the prop or midrib
in typical forms of Platylepas. The sheath
is delicately striate transversely, and stops
a little short of the basal edges of the
compartments. The scuta are in contact
with the terga, and together they stretch
from end to end of the orifice. Carinoros-
tral diameter, 4 mm; lateral diameter,
3-8 mm, height, 3 mm. Some individuals
are slightly larger, greater diameter
5-3 mm.’ . . .

‘The specimens of Cylindrolepas were
embedded in a very hard yellowish
substance showing but little structure. Dr
Thomas Barbour, of the Museum of
Comparative Zoology, to whom I applied,
concluded that it is the salt-water cured,
sun-dried skin of either a loggerhead or
green turtle, probably from between the
neck and flippers or around the base of the
tail.’

The main point of difference between Pilsbry’s
(1916) description and the photographs of the
present specimen is the form and presence of the
‘short, vertical, blunt teeth on either side of (the)
larger median tooth.’ One of these may be seen in
Plate 1, Fig. 6, but the ratio of small to large is
quite different to that shown in Pilsbry’s figure.
Nonetheless it is probable that the present
material represents specimens of Cylindrolepas
darwiniana — the first recorded since Pilsbry’s
description. If this is the case (as only comparison
with the types will reveal) then, on shell
characters, Cylindrolepas should be included in
Platylepas.

SHELL MORPHOLOGY AND GROWTH
WITH A DISCUSSION OF PROBABLE
FUNCTION

Shells were cleaned in sodium hypochlorite
solution. Histological specimens were decalcified
in ethyl-diamine-tetra-acetic acid, stained with

Mallory’s collagen stain, Hughesdon’s modifica-
tion (Carleton and Drury, 1957), and mounted in
Canada balsam. The plates illustrating many of
the morphological features discussed in this
section are to be found in Monroe and Limpus
(1979) to which much reference is made.

General Remarks on Shell Structure

Barnacles of the family Coronulidae are diverse
in shape and size. Some of them (subfamily
Chelonibiinae) show relatively ‘normal’ balan-
omorph external morphology while at the other
extreme is Xenobalanus with an external
morphology reminiscent of the Lepadomorpha.

In common with other Balanomorpha the
sheath in coronulids is laid down in zones of
growth (Darwin 1854). The tranverse striae so
produced on the internal surface mark the lines of
junction between successive periods of shell
deposition — i.e. they mark moults. They are
more distinct in some members of the family than
others. Darwin (1854) clearly outlines the process
of formation of the sheath.

In Chelonibia testudinaria, which has a thick
outer wall and occupies a conventional balan-
omorph position on the surface of the host/
substrate, the transverse striae are only apparent
on the upper one third of the sheath (Fig. 1). In
embedded species such as Stomatolepas sp. or
Stephanolepas muricata with frail outer shells the
transverse striae are apparent for the full depth of
the sheath (Monroe and Limpus, 1979, pi. 2, 3,
6,). The striae may be closely placed as in
Platylepas decorata (Monroe and Limpus, 1979,

Fig. 1 : A portion of the sheath of Chelonibia testudinaria
showing transverse striae in upper portion.

MONROE: STUDIES IN THE CORONULIDAE

239

pi. 4, fig. 8) or more widely as in Stephanolepas
muricata (Monroe and Limpus, 1979, pi. 2, fig.
3 ).

It can be inferred from the appearance of
tranverse striae at different depths on the sheath
that moulting in some superficially situated
species (such as C. testudinaria ) becomes less
frequent with advancing age. However, in
embedded species moulting continues at nearly
constant frequency (as judged by the near regular
spacing of the striae) throughout life. A corollary
of this interpretation of striae spacing is that
‘superficial’ species remain on the host and
continue to deposit shell for some time after their
final moult whereas embedded species do not. This
point is discussed later.

Pilsbry (1916) interpreted the morphology of
coronulid shells as an adaptation to an impact
prone situation. The massive shell of Chelonibia
and the internal buttresses of Platylepas and the
Coronulinae were seen as strengthened of the
shell. This view was shared by Ross and Newman
(1967).

Darwin (1854) and Newman, Zullo, and
Wain wright (1965) recognised that the complex
wall folding of Coronula and Cryptolepas and the
enclosed papillae of whale skin are a means of
improving attachment of the barnacle to the
whale. Darwin described and figured the
ontogenetic development of this mechanism
(Darwin, 1854, p. 15, figs. 4, 10). The latter
authors also interpreted the six basal teeth of
Platylepas as structures which ‘lock the barnacle
to the surface of the host to which it attaches’
(Newman, Zullo, and Wainwright, 1965, p. 172).

Darwin (1854) recognised for Tubicinella
major the role of the annular ridges as ‘necessary
to prevent too easy protrusion’ of the shell.

It is here proposed that there are two
mechanisms operating in the Coronulidae to
facilitate attachment to the host — mechanisms
adapted to the requirements of attaching to a
growing substrate. Both mechanisms are deriva-
tions of the growth processes of the barnacle.

On the one hand there is the convoluted wall
morphology of Coronula and Cryptolepas, and on
the other hand there is the embedding process of
Tubicinella spp., Stomatolepas spp., and
Stephanolepas muricata with the concommitant
requirement for regular moulting throughout life.

Platylepas spp. show elements of both
processes, it being contended here that it is not the
basal teeth but the whole shell wall which anchors
the barnacle to its host.

One of the requirements for embedding in the

surface layer of an animal host is a means of
overcoming the mechanical reaction of the host
skin to the downward force of the growing
barnacle. The holdfast structures of the
established shell provide the necessary anchorage
to permit downward growth as well as prevent
dislodgement. How the fully embedded species
initially gain a purchase to commence their
burrowing is not clear from the adult shells
available.

Chelonibia species

Chelonibia testudinaria and C. caretta are
confined to hard areas of the skin of the host —
carapace, plastron, and head. C. testudinaria
attaches superficially and is not overlain by any
host tissue. The large number of radial septae with
their spiny lower margins (PI. 1, Fig. 4) do not
puncture the host scute but serve to increase the
area of contact between the shell and substrate.
That is, approximately two thirds (by inspection)
of the base becomes effectively calcified and hence
more firmly cemented to the substrate.

The forces most likely to dislodge a superficially
attached barnacle are lateral — for example from
water currents. The form and structure of the shell
in C. testudinaria are well suited to reduce
turbulence and improve adhesion.

When a large specimen of C. testudinaria is
removed from the host it usually brings with it a
portion of the keratinized layer of the scute
showing that adhesion of the barnacle to the scute
is greater than that of scute to host. However in
the case of large (i.e. older) specimens of
barnacles the host is probably nearer to moulting
and the scutes could be expected to be loose at this
time. Being a surface adherent C. testudinaria
must be shed when host moulting occurs —
approximately once a year (Limpus, in Lavery,
1978).

C. caretta has a shell structure and position on
the host similar to those of C. testudinaria.
However there are two differences notable in the
context of this discussion: there is a layer of host
keratin adherant to the outside of the shell wall
which ascends to the level of erosion around the
orifice (Monroe and Limpus, 1979, pi. 1, fig. 4),
and the outer wall is roughened by longitudinal
grooves and ridges (PI. 1, Fig. 1).

These two features are allied. The longitudinal
grooves when seen from their basal ends, appear
as infoldings of the outer wall similar in structure
to the median suture of Platylepas spp. (cf. PI.
l,Figs. 1, 2 with Monroe and Limpus, 1979, pl. 3,
fig. 8; pl. 4, fig. 1) (see below). Examination of the

240

MEMOIRS OF THE QUEENSLAND MUSEUM

host scute after removal of C. caretta shows that
the keratinized layer has been split in two. The
lower of these is the layer below the barnacle and
is constantly being added to from below by the
host epidermis. The upper layer is in keratin which
invests the outer wall of" the barnacle.

This upper layer is carried up the outer wall by
being ‘trapped’ in the infoldings of the outer wall.
C. caretta is more firmly attached to the host by
viture of this partial embedding. However since C.
caretta does not penetrate the scute it will be shed
at each host moult.

Platylepas species and iCylind role pas

DARWINIAN A

Platylepas spp. maintain a more secure hold on
the host than do Chelonibia spp. This is done by
penetrating the scutes and soft epidermis of the
host and anchoring the shell wall in the dermis
and so resisting dislodgement.

The most prominent feature of Platylepas
hexastylos is the median ‘fold’, ‘sulcus’, or
‘midrib’ or each compartment (Monroe and
Limpus, 1979, pi. 4, fig. 3). During growth this
structure traps the host scute at its outer end, and
aided by the surface sculpture of the shell, carries
the layer of scute up and outside the shell (Monroe
and Limpus, 1979, pi. 3, fig. 7) in the manner of
Chelonibia caretta (Monroe and Limpus, 1979, pi.
1, fig. 4). At the same time host scute is carried in
between the walls of the fold and trapped. Along
the lower edge of the fold, host fibrous connective
tissue arising in the dermis is caught. There is no
layer of scute below the barnacle. The same
trapping of scute and dermis occurs at the
compartmental sutures.

Along with the membraneous basis the folds
hold back the host tissues in a bowl-shaped cavity
that allows the body of the barnacle to be partly
below the host surface level and the shell to be of
lower profile (Monroe, 1979, pi. 6, fig. 9 and pi. 2,
fig. 3). The downward force to maintain the cavity
may be exerted against the grip on the scutes.

Newman, Zullo, and Wainwright (1965, p. 172)
have described the mode of growth of the
buttresses of Platylepas. ‘These are developed by
allometric marginal growth increments. Lateral
growth is suppressed at the location of each
buttress so that it grows only downwards as the
rest of the wall grows outwards and downwards’.
That is the shell grows in diameter by the twelve
lobes of the periphery pushing out into the host
and trapping host tissue between themselves as
they go. It is clear that the whole wall, and not just

the basal teeth, holds the barnacle so firmly to the
host.

When the host moults its scutes, P. hexastylos
is able to adhere because of its grip on the
underlying dermal connective tissue.

Platylepas coriacea is very similar in shell
structure to P. hexastylos. There is no host tissue
on the outer surface of the shell as in P.
hexastylos and all the anchorage appears to be
provided by the nearly horizontal lower edges of
the midrib folds. The leathery nature of the skin of
the host — the leatherback turtle — may account
for the failure of (or lack of necessity for) this
species to embed.

P. decorata has midrib folds and props similar
to those of the foregoing species but has more
elaborate, and efficient, surface sculpture on the
parietes to facilitate anchorage (Monroe and
Limpus, 1979, pi. 4, fig. 7; pi. 6, fig. 8). Scute
tissue becomes engaged in the rows of ‘double
beads’ (cf. Monroe and Limpus, 1979, pi. 4, fig. 7,
with pi. 2, fig. 4) and the median fold, while the
props penetrate deeply allowing the barnacle to
embed almost to its total depth.

?Cylindrolepis darwiniana is very like
Platylepas and would appear, on the basis of the
external shell morphology, to attach in a similar
manner to P. decorata.

Stephanolepas muricata

In this species the simplest development of
non-mural holdfast structures is found — blunt
‘spines’ (Monroe and Limpus, 1979, pi. 2, fig. 1,
2). Their position and structure are intimately
associated with the pattern of growth of the shell
wall. Once the barnacle has embedded the spines
offer a secure mechanical grip on the host.

As outlined above, the outer wall and sheath are
deposited in layers (Monroe and Limpus, 1979, pi.
2, fig. 3). In every second layer at both edges of
each compartment an outgrowth is formed which
curves upwards and reflects slightly back across
the outer lamina of the compartment. On the
radial side of the compartment these outgrowths
are confluent with ridges that cross the radii and
engage with the spines of the alar side of the
contiguous compartment. These outgrowths
become embedded in host connective tissue.

Stomatolepas species

Stamatolepas spp. are similar to Stephanolepas
muricata: the walls are thin and frail, and the
barnacles are deeply embedded in the host. The
outgrowths described in Stephanolepas are in
Stomatolepas species much larger and completely

MONROE: STUDIES IN THE CORONULIDAE

241

reflected back across the outer face of the
compartments (Monroe and Limpus, 1979, pi. 2,
fig. 8). There is also an outgrowth for each layer
of the sheath. The projections are confluent with
ridges on the radii (Monroe and Limpus, 1979, pi.
2, fig. 7). At the lower (growing) edge of the shell
these reflected outgrowths are quite small
(Monroe and Limpus, 1979, pi. 2, fig. 8), but
successively higher (older) ones are longer with
secondary outgrowths (scales) on their upper
edges. By mid-height the reflected outgrowths
from each side have met in the middle leaving a
slight depression (PI. 2, Fig. 5) which has been
interpreted a residual rudimentary midline fold
(Pilsbry, 1910).

The implications of this for the classification of
the Coronulidae are discussed below. PI. 2, Fig. 6
shows an histological section (radial) of a
decalcified portion of the wall of Stomatolepas
praegustator; PI. 2, Fig. 7 is a similar tangential
section. The space x represents the wall, the spaces
y the reflected projections and the scales. Ct is
host connective tissue and p the periostracum of
the barnacle. The space enclosed between the
reflected projections and the compartment wall is
filled by host connective tissue. Connective tissue
is also entwined around the scales, the whole mass
forming a complex holdfast structure.

Tubicinella cheloniae

In this species the holdfast mechanism consists
of large upward and outward curving blunt spines
whose formation is described in Monroe and
Limpus (1979) and Nilsson-Cantell (1932). This
species, as does Stephanolepas muricata and
Stomatolepas species, becomes completely embed-
ded in the host.

The Whale Barnacles

Darwin (1854) showed clearly the ontogeny of
the holdfast arrangement of the walls of Coronula
diadema, C. regina, and Cetopirus complanata.
The process is similar to that in Platylepas and
Chelonibia caretta in that it involves trapping of
the upper epidermal layer in folds of the shell. It is
probable that Cryptolepas rhachianecti can be
included in this category also.

On the other hand Tubicinella major and
Xenobalanus globicipitis, like T. chelonie,
Stomatolepas spp. and Stephanolepas muricata,
all ‘burrow into’ rather than ‘grab hold of the host
ep'dermis. While acknowledging that Cetopirus,
Cryptolepas and some Platylepas become covered

to a greater or lesser degree by host epidermis it is
not the same process of burrowing as in
Tubincinella, Xenobalanus, Stomatolepas and
Stephanolepas .

Host Reaction

Chelonibia species do not penetrate the host
scutes and so there is no host tissue reaction.
However Tubicinella, Stomatolepas, Stephan-
olepas, and to some extent Platylepas, all
penetrate the basement membrane below the
epidermis and invade the dermis. The host
reaction is to lay down a capsule of fibrous
connective tissue around the barnacle (Monroe
and Limpus, 1978, pi. 2 fig. 1, and PI. 2, Figs.
3-7). This is very thick and tough in the case of
Tubicinella cheloniae. As the barnacle expands
during growth the capsule becomes very tightly
and intimately adherent to the barnacle shell. In
the case of Stomatolepas spp., because of the
action of the reflected projections and scales, the
capsule is difficult to remove without damage to
the shell and, while present, conceals the true
nature of the shell structure (PI. 2, Fig. 5). In
Platylepas hexastylos and P. decorata the host
capsule covers just the basal surface of the
barnacle (PI. 2, Figs. 3-4).

The host tissues are weakest at the interface of
the dermal tissue and the capsule — the presumed
site of growth of the capsule. It is at this interface
that the host tissues yield when barnacles are
removed from the host. This allows comparatively
easy removal of the barnacle without damage to
the host, except in the case of Stephanolepas
muricata, and some Platylepas spp. which are
deeply embedded on harder parts of the host. In
these cases the host epidermis and scutes tend to
grow back over the orifice of the barnacle
enclosing it in a narrow-necked sack.

The growth strategy of embedded species

The remarks of this section are derived from a
consideration of embedded species occurring on
turtles but outline a growth strategy that probably
applies also to embedded species on whales, e.g.
Tublicinella major and Xenobalanus globicepitis .

The nature of the attachment region of the
superficially situated Chelonibia testudinaria is
not especially different from that of any other
barnacle having a membranous basis and
attaching on a hard surface. The same remark
applies to the attachment of Chelonibia patula to
the surfaces of its hosts — molluscs, xiphosurans.

242

MEMOIRS OF THE QUEENSLAND MUSEUM

and Crustacea (and old bones*). Though these
barnacles are sloughed at moulting times and
require a reproductive strategy to cope with this
exigency, while on the host they are able to grow
in the normal balanomorph fasion. Chelonibia
caretta is more intimately attached to the upper
stratum of its host, but there is still no necessity
for special growth strategies.

The interface between these superficial
barnacles and their hosts is, when compared to
that of the species considered below, comparative-
ly inert.

When embedded species are considered, i.e.
those which penetrate into the dermal layer and
are encapsulated by host connective tissue, it
becomes apparent that the growth strategy of the
normal balanomorph does not suffice. The
barnacle would soon be ‘grown off the host by the
connective tissue layer with which it is walled off
by the host. The problem is analogous to a fish
swimming upstream.

The barnacle, having established itself in the
host, has to provide for its own increase in size,
and also to grow a little faster than would
otherwise be required in order to maintain its
position against the sloughing process of the host.

Judging by scars on the host’s surface it is
possible for a barnacle to be completely sloughed

by the connective tissue capsule. In some cases
barnacles may be sloughed because they have died
and ceased to continue growing against the
sloughing process. This is seen in the case of
empty barnacle shells found in varying degrees of
protrusion from the host.

The extra rate of growth required is expressed
as shell height. Consider a barnacle B (Fig. 2a),
with a host connective tissue capsule Ct, in host H.
Assume that the barnacle is not at present
growing but that the host capsule is active in
sloughing the barnacle. After a time the barnacles
will protrude a little above the host surface (Fig.
2b). Utilizing its purchase on the capsule and the
adhesion of that to the host, the barnacle grows
downward into the host, maintaining its position
on the host, but the body is still at the upper end of
the shell, attached by the opercular membrane,
and this is projecting above the host surface (Fig.
2c). The barnacle now moults and the body is
drawn down the shell, the upper portion of which

♦Specimen lot W7354, Chelonibia patula, 20 in-
dividuals, was removed (together with 20 Balanus
trigonus W7355) from a human femur recovered in
20 m from Moreton Bay near Moreton Is. on July 10th
1977. It may be that Chelonibia patula has a substrate
preference involving organised calcium matrices.

Fig. 2: Schematic drawing of mode of embedding of burrowing coronulidae in turtle hosts: H, host; CT, connective
tissue capsule; B, barnacle. For explanation see text.

MONROE: STUDIES IN THE CORONULIDAE

243

flakes off along the weakness of the transverse
striae in the sheath (Fig. 2d). The connective
tissue capsule is degraded round the orifice,
probably by the epidermal bacteria of the host.
The wounds of Tubicinella cheloniae are specially
septic.

It can be seen then that although there is a
requirement in Chelonibia for the orifice to
enlarge by erosion in the early period of the
barnacle’s adult life, facilitated by the weaknesses
of the transverse striae and growth of the radii, at
maturity this is no longer so. However in the
embedded species, continued moulting and growth
are required — radial growth maintaining shell
diameter, and moulting, with loss of the
protruding shell, maintaining the barnacle’s
position on the host.

DISCUSSION

Subfamily Classification

Definitions and component taxa of three of the
four previously accepted subfamilies within the
Coronulidae have been given by Newman and
Ross (1976) and were used by Monroe and
Limpus, 1979. A definition of Platylepadinae was
omitted by Newman and Ross, 1976. Zullo (1963)
had previously proposed such a subfamily using
Pilsbry’s (1916) definition of the platylepad series.
Newman and Ross’s (1976) list of genera within
their Platylepadinae is the same as that for
Pilsbry’s platylepad series and presumably their
subfamily would rest upon the same definition.

Further discussion of the authorship of the
Platylepadinae will not be undertaken here as
morphological information from this study casts
doubt upon its integrity as a grouping. The
previously accepted subfamilies and their
characters are presented in Table 1.

The prime character separating the Platy-
lepadinae from the other subfamilies is the
presence or absence of a median sulcus, midrib, or
a vestige thereof (Pilsbry, 1916; Zullo, 1963). The

presence of this character has been claimed in
Platylepas, Cylindrolepas, Stomatolepas, and
Stephanolepas. It is certainly present in
Platylepas (Monroe and Limpus, 1979, pi. 3, fig.
8, pi. 4) are most likely in Cylindrolepas (see
?Cylindrolepis darwiniana and PI. 1, Figs. 5-6).
However it is clearly absent from Stomatolepas
and Stephanolepas (Monroe and Limpus, 1979,
pi. 2; pi. 3, figs. 1-5). No vestige or precursor of
such a structure can be detected. Pilsbry’s (1910;
1916) interpretation of the mid compartmental
groove in Stomatolepas as a rudimentary sulcus is
due in his misunderstanding of the mode of growth
of the external sculpture. Also he observed shells
clothed in connective tissue (as judged by his
paratypes) which hinders observation (cf. PI. 2,
Fig. 5 with Monroe and Limpus, 1979, pi. 2, fig.
4). Pilsbry (1916) never saw Stephanolepas but
inferred from Fischer’s (1886) description of
smooth mid compartmental areas that it would be
similar to Stomatolepas.

Similarly a median fold or sulcus is said to be
absent in the other subfamilies. However
structures very similar in formation and function
to the median fold of Platylepas can be detected
in Chelonibia caretta (PI. 1, Fig. 2).

The Coronulinae are united by the presence of
an ‘oral hood’ (Darwin, 1854, pi. 17, fig. 4a) and
the possession of a ‘single row of wall tubes
formed by infoldings of (the) outer lamina against
the sheath’ (Newman and Ross, 1976, p. 37).

Much confusion has been brought about by the
terms ‘infolding’ and ‘outfolding’. Setting aside
the Chelonibiinae and Emersoniinae, which have
more recently been included in the Coronulidae,
the main division in the Coronulidae, between the
Platylepadinae and Coronulinae, has been based
chiefly on the type of wall folding. The following
paragraphs show that ‘in’ and ‘out’ are relative
terms that have obscured the similarity between
some genera of the Coronulidae.

In Tubicinella, tubes (interlaminate pores) are
formed by the union of lateral outgrowths from

Table 1: The subfamilies of the Coronulidae and their distinguishing characters {after Newman and

Ross, 1976, and Zullo, 1963).

Subfamily

6 or 8
wall plates

opercular plates
weakly articulated

oral hood
+ /-

tergum

reduced

rows of
wall tubes

mid-parietal

sulcus

Chelonibiinae

6/8

yes

+

1

.

Emersoniinae

6

?

9

7

several

-

Platylepadinae

6

yes

-

+

not

specified

+

Coronulinae

6

yes

+

+ + +

1 (wall folds)

-

244

MEMOIRS OF THE QUEENSLAND MUSEUM

the outer edges of the interlaminate septa (Darwin
1854). There are no ‘infoldings’ of the wall in the
sense of Newman and Ross (1976). Xenobalanus
has incomplete interlaminate pores (PI. 2, Fig. 8)
but no ‘infoldings’ of the outer layer. In Coronula
and Cetopirus such interlaminate pores exist
(Darwin 1854) as well as a ‘single row of tubes
formed by infoldings of the outer lamina against
the sheaths’ (Newman and Ross, 1976, p. 37). In
Cryptolepas (which has interlaminate pores
(Pilsbry, 1916)) these ‘infoldings’ produce an
incomplete outer wall analogous with that of
Coronula (Pilsbry, 1916, p. 280, pi. 66).

As shown above Stomatolepas, Stephanolepas,
Tubicinella and Xenobalanus have no wall
folding. What is more, they are all united by other
characters: 1, possession of structures arising from
the lateral edges of the compartments and forming
holdfasts (this is not strictly so in Tubicinella
major as the annular ridges are formed across the
whole of the face of the compartment. The ridges
are, however, produced in synchrony with the
layers of sheath as are the holdfasts on the other
genera included here); 2, the shell is completely
embedded in the host, is thin-walled and frail, and
grows continually and rapidly (as judged by the
relatively wide layers of the sheath) to maintain its
position on the host; and 3, the shell is conical;
straight sided and only slightly tapering in
Tubicinella, ‘glubulo-conic’ (Fischer, 1886) and
more strongly tapering in Stomatolepas and
Stephanolepas, Since only three or four sheath
layers of Xenobalanus are retained at any one
time (PI. 2, Fig. 8) it is difficult to determine if or
how the shell tapers, but in general shape it is in
conformity with those of the other genera.

Platylepas is said to have an ‘infolded’ buttress
(Pilsbry 1916; Newman, Zullo, and Wainwright,
1965) and this is a convenient way to describe the
buttress — especially that of P. hexastylos when
seen from the exterior, e.g. Monroe and Limpus,
1978, pi. 4, fig. 3. However when seen in basal
view, Monroe and Limpus, 1979, pi. 4, fig. 1 the
inner edge of the buttress (except in the carina)
aligns well with the plane of the intercompartmen-
tal sutures (roughly a cylinder), and the lower
projection of the buttress — curving in under the
lower edge of the sheath — is strongly reminiscent
of the structure of Cetopirus complanata of which
(as Coronula balaenaris) Darwin (1854, p. 416)
says ‘the inner ends of the folded walls . . . descend
some little way beneath the basal edge of the
sheath, as low, or lower, than the circumference of
the shell’ (Darwin, 1854, pi. 16, fig. 3). Platylepas
could be interpreted as a Coronula with two

outfoldings in each compartment. Whereas in
Coronula the space between the two layers of the
outfolded outer lamina is narrow and filled by
calcareous material, in Platylepas, they are wide,
though still solidly filled. The ‘tubes’ in Coronula,
filled with whale epidermis, are homologous with
the narrow space between the walls of the buttress
in Platylepas in which turtle host tissue is held.

Newman, Zullo and Wainwright (1965, p. 172)
recognise that ‘in Platylepas the regions occupied
by the buttresses . . . correspond to the “canals” in
the whale barnacles’ and that, ‘in Coronula and
Cryptolepas, the intervening growth areas
(between the canals) . . . could be called
buttresses, but they are homologous with the wall
proper, rather than the buttresses of Platylepas ’,
but they fail to realise the implication of this
homology for the classification of the Coronulidae.
There has been so much attention focused on the
‘infolding’ of buttresses in Platylepas and the
‘outfolding’ to form ‘canals’ in Coronula and
Cryptolepas that, coupled with a lack of
understanding of the function of these structures
in Platylepas, it has led to the obscuring of the
close relationship between these genera. The walls
of Platylepas, Coronula, Cetopirus, and
Cryptolepas are homologous in both structure and
function.

On the basis of wall structure then, a second
group of genera can be recognised composed of
Coronula, Cetopirus, Cryptolepas, Platylepas and
Cylindrolepas.

The remaining genera, Chelonibia and
Emersonius, are at present accommodated in their
own subfamilies. Emersonius cybosyrinx is a fossil
form and is not considered further here. The
inclusion of the Chelonibiinae in the Coronulidae
is not questioned, and is supported by the presence
of involutions in the outher lamina of Chelonibia
caretta which this study has revealed (PI. 1, Fig.
1 ).

The arrangement of subfamilies in the
Coronulidae proposed as a result of this study
follows.

CORONULINAE LEACH 1812, EMEND.

Wall of 6 plates, exterior lamina of wall folded
to produce holdfast structures by trapping host
epidermal tissues, attached to surface of host or
only shallowly embedded. Coronula, Cetopirus,
Cryptolepas, Platylepas, Cylindrolepas.

Xenobalaninae Gruvel, 1903, EMEND.

Walls of 6 plates, holdfast structures produced
by outgrowths from edges of compartments in

MONROE: STUDIES IN THE CORONULIDAE

245

synchrony with layers of sheath, shell deeply
embedded in host. Xenobalanus, Tubicinella,
Stomatolepas, Stephanolepas.

Chelonibiinae Pilsbry, 1916

Chelonibia.

Emersoniinae Ross, 1967
Emersonius.

Evolution

Ross and Newman (1967) proposed a
phylogeny for the Coronulidae in which the more
generalised ‘primitive’ chelonibiines exploited the
phylogenetically older hosts (Crustacea and
molluscs) while the more advanced coronulines
were confined to more recent hosts — cetaceans.

The scheme suggests coevolution of the
Coronulidae and the major groups of marine
vertebrates, the ‘higher’ subfamilies of barnacles
supposedly arising as new host forms became
available.

In the light of the above proposed altered
subfamily classification it is appropriate to
re-examine the probable phylogeny of this family.

Under the proposed classification the Coron-
ulinae are no longer epizooans only of cetaceans
but are to be found on phylogenetically ‘older’
hosts such as turtles. Similarly the Xenobalaninae
(which in part takes the place of the Platylepadin-
ae) are no longer confined to the ‘middle’ range of

Fig. 3: The relationship between hosts and subfamilies
proposed as a result of this study.

hosts but now include species which occur on hosts
from turtles to whales (fig. 3).

Ross and Newman (1967, p. 16) maintain that
‘the most generalised members of (the Coron-
ulidae) are Platylepas and Chelonibia patula’. If
epizooites are host conservative then Chelonibia
patula would be a candidate for the position of
most generalised coronulid. Platylepas however
shows all the attributes of the (generally accepted)
more advanced Coronula (see above).

It is possible to see a relationship between the
Chelonibiinae and Coronulinae in the involutions
of the exterior wall of Chelonibia caretta and the
convoluted exterior wall of the coronulines. It is
also possible to see a relationship between the
Chelonibiinae and Emersoniinae in the well
developed radial septae of both groups. However it
is not possible to recognise any general
morphological affinity of the Xenobalaninae with
the above three subfamilies except in the
possession of porose walls by some of the
xenobalanines and an epizooic habitat.

Figure 4 shows an inferred phylogeny of the
Coronulidae based on wall structure and function.
The ancestral form is postulated as being similar
to Chelonibia patula, having porose walls, and
radial septae projecting into the cavity of shell. C.
testudinaria is similar in construction to C. patula
but has further developed radial septae and the
pores in the walls are filled with calcareous matter
in the upper parts.

The Emersoniinae are separated from the
chelonibiid stock by the formation of intercalary
and transverse septae. The Coronulinae arise from
the chelonibiid stock somewhat later, there being a
strong similarity in the structure and function of
the involutions in the outer wall of C. caretta and
the convoluted wall of the Coronulines. C. manati
is supposed, on the basis of Pilsbry’s 1916
description, to have similar involutions.

Whence the Xenobalaniinae arise it is difficult
to say. Since they are all burrowing forms they are
unlikely to be associated with the ‘older’
exoskeletal hosts. It is possible that they do not
belong in the Coronulidae, but considerations of
the anatomy of the soft parts make that
improbable.

Darwin (1854) makes note of the ‘beaded’
nature of the sculpture on the exterior of Coronula
and Platylepas , These beads are synchronous with
the layers of the sheath and can be quite elaborate
— e.g. in Cryptolepas rachianecti (Pilsbry, 1916,
pi. 66) and? Cylindrolepas darwiniana (PI. 1, Figs.
5, 6). It may be that the Xenobalaninae arose
from a stock in which the radial septae were

246

MEMOIRS OF THE QUEENSLAND MUSEUM

suppressed and there was no development of wall
convolutions, the development of the holdfast
structures taking place by the elaboration of the
‘bead’ structure at the edges of each compartment.
In Tubicinella major the annular ridges may be
an expression of the growth ridges as seen across
the exterior of the compartments in other species
— e.g. Platylepas hexastylos and ?Cylindrolepas
darwiniana.

The presence of an oral hood in Tubicinella
major and Xenobalanus in the Xenobalaninae on
the one hand and in Coronula, Cetopirus and
Cryptolepas in the Coronulinae on the other is an
objection to the proposed arrangement of the
genera. This structure though, may be developed
only in those species occurring on whales as an
adaptation to the higher swimming speeds of these
hosts.

Concluding Remarks

The Coronulidae are a very specialized group of
barnacles. In the main, they exhibit a high degree
of host and site specificity. The principal hosts,
turtles and whales, are migratory animals, in some

cases traversing many degrees of latitude in their
migrations. The coronulids occurring on these
animals can hardly be considered as sessile
animals in the sense of the ‘normal’ barnacle.
While it is true that Lepas species are to be found
only on floating substrates, it can at least for them
be pointed out that the substrate moves with
nearly the same vagaries as the general body of
water around them. Only Conchoderma virgatum
(Spengler, 1790), an epizooite of epizooites, and
Analesma squalicola (Love, 1844), share with the
coronulids ‘active’ motion through the water as
being normal for the entire population. This fact
prompts questions about the breeding biology of
the Coronulidae. When are larvae released?
Perhaps in synchrony with host aggregations. How
long is their pelagic life? The only published
information available is for Chelonibia patula
(Crisp and Costlow, 1963) and that species is
found attached to relatively localised hosts —
crabs and molluscs.

It was observed during the summer at Mon
Repos that newly settled and small Chelonibia
testudinaria were comparatively abundant (C.

Fig. 4: A possible phylogeny of the barnacles of the family Coronulidae.

MONROE: STUDIES IN THE CORONULIDAE

247

Limpus, pers. comm.) which lends some support to
the idea that larvae are released during
aggregation of the hosts on the breeding grounds.

Some species of coronulid are comparatively
rare — e.g. Stephanolepas muricata and
Tubicinella cheloniae. What mechanisms operate
to ensure sufficient settlement of larvae is not
known. These two species are embedded in firm
tissues underlying hard epidermal structures. T.
cheloniae penetrates the dermal bone of the
carapace. Tomlin (1967, not seen, reported in
Mitchell and Kozicki, 1975) cites L. Mathews to
the effect that an enzyme is secreted by
Tubicinella major which dissolves skin proteins
‘which they then probably consume’. If such a
mechanism operates in T. major to facilitate
embedding (though there is no evidence of
reduced feeding ability, as judged by the cirri) a
similar one may be present in other
Xenobalaninae.

Barnacle cyprids respond positively to specific
proteins, arthropodins, but these active substances
are not confined to arthropods alone (Crisp 1974).
What substance(s) provides the stimulus to
settlement of coronulids is not known. The
mechanism of site selection on the host is not
known though it is possible that turbulence may
play some part in defining areas on the host
suitable for particular species of barnacles.

ACKNOWLEDGMENTS

Thanks are due to the following Queensland
Museum staff who provided assistance with this
study: Mr. N. Hail who cut the histological
sections; Messrs. A. J. Easton and R. Burgess who
printed the photographs; and Mr. B. M. Campbell
and Dr. L. J. Cannon for valuable advice,
criticism, and encouragement. Dr. J. G.
Greenwood, Zoology Department, University of
Queensland, is especially thanked for advice and
help in the course of this study.

LITERATURE CITED

Carleton, H.M. and Drury, R.A.B., 1975.
‘Histological Technique’. Oxford Universi-
ty Press: London. 343 p.

Crisp, D.J., 1974. ‘Factors influencing the
settlement of invertebrate larvae’, p.
177-265 in Grant, P.T., and Mackie,
A.M., (Eds) ‘Chemoreception in Marine
Organisms’. Academic Press: London. 295
P-

Darwin, C., 1854. *A Monograph of the subclass
Cirripedia with figures of all the species,
the Balanidae’. The Ray Society: London,
viii + 684 p.

Fischer, P., 1886. Description d’un nouveau
genre de Cirrhi pedes {Stephanolepas)
Parasite des tortues marines. Actes Soctite
Linneenne Bordeaux. (4)10: 193-6.

Mitchell, E. and Kozicki, V.M., 1975. Autumn
strandings of the northern bottlenose whale
{Hyperoodon ampullatus ) in the Bay of
Fundy, Nova Soctia. J. Fish. Res. Board
Can. 32(7): 1019-1040.

Monroe, R.W., 1979. Studies in the Coronulidae:
Barnacles on Turtles in Queensland
Waters. Unpbl. M.Sc. Thesis, Zool. Dept.
Univ. Qd.

Monroe, R.W. and Limpus, C.J., 1979.
Barnacles on turtles in Queensland waters,
Mem. QdMus. 19(3): 197-223, pis. 1-6.

Newman, W.A. and Ross, A., 1976. A Revision
of the balanomorph barnacles; including a
catalogue of the species. Mem. San Diego
Soc. Nat Hist. 9: 1-108.

Newman, W.A., Zullo, V.A., and
Wainwright, S.A., 1965. A critique on
recent concepts of growth in the Balan-
omorpha (Cirripedia, Thoracica).
Crustaceana 12: 167-178.

Nilsson-Cantell, C.A., 1932. The Barnacles
Stephanolepas and Chelonibia from the
turtle Eretmochelys imbricata. Ceylon
Journal Science, Sect. B. {Spolia Zeylan).
16(3): 257-64, 3 pis, 2 figs.

Pilsbry, H.A., 1910. Stomatolepas, a barnacle
commensal in the throat of the Loggerhead
turtle. Am. Nat. 44: 304-6.

1916. The sessile barnacles (Cirripedia)
contained in the collections of the U.S.
National Museum, including a monograph
of the American species. Bull. U.S. Nat.
Mus. 93: 1-366, pis. 1-76.

Ross, A. and Newman, W.A., 1967. Eocene
Balanidae of Florida, including a new
genus and species, with a unique plan of
‘turtle- barnacle’ organisation. Am. Mus.
Nov. 2288: 1 21.

Zullo, V.A., 1963. Classification and Phylogeny
of the Balanomorpha (Cirripedia). Ph.D.
thesis, University of California, 372 p.
University Microfilms: (Ann Arbor and
London).

MEMOIRS OF THE QUEENSLAND MUSEUM

Plate I

Figs. 1-2: Chelonibia caretta (Spengler), W3654; 1, basal edge of
exterior of shell showing grooves; 2, peripheral edge of base
showing infoldings of outer layer (A).

FiGS. 3-4: Chelonibia testudinaria (Linnaeus), W7845; 3, basal edge
of exterior of shell; 4, peripheral edge of base.

Figs. 5-8: ?Cylindrolepas danviniana, W7848; 5—6, exterior of
compartments; 7, basal view; 8, radial edge of compartment.

MONROE: STUDIES IN THE CORONULIDAE

Plate 1

MEMOIRS OF THE QUEENSLAND MUSEUM

Plate 2

Figs. 1-2: ?CyIindrolepas darwiniana W7848; 1, opercular view of
compartment; 2, radial edge of compartment.

Fig. 3: Platylepas hexastylos (Fabricius), W7814; section of
decalcified animal showing a, cavity of shell, ct, connective tissue
capsule of host, k, keratin layer of host epidermis, p, periostracum
of shell.

Fig. 4: Platylepas decorata Darwin, W7811; approximately median
radial section of one wall illustrating intimate association of
keratin of host, k, with the fine ‘beaded’ sculpturing on the
exterior of the parietes, s, the basal membrane, b, is covered by
host connective tissue, ct.

Fig. 5: Stomatolepas praegustator Pilsbry, exterior view of
compartment with investing layer of host connective tissue.

Fig. 6: Stomatolepas praegustator Pilsbry, W7815; radial section of
decalcified wall, x, reflected projection, y, host connective tissue,
ct, and barnacle periostracum, m.

Fig. 7: Stomatolepas praegustator Pilsbry, W7813; tangential section
of decalcified wall, symbols as in fig. 6.

Fig. 8: X enobalanus globicipitis Steenstrup, W7384; exterior view of a
compartment showing incomplete fusion of outer lamina.

MONROE: STUDIES IN THE CORONULIDAE

Plate 2

Mem. Qd Mus. 20(2): 253-77, pis. 1-3. [1981]

THE BIOLOGY OF CEPHALODESMIUS, A GENUS OF DUNG BEETLES
WHICH SYNTHESIZES “DUNG” FROM PLANT MATERIAL
(COLEOPTERA:SCARABAEIDAE:SCARABAEINAE).

G.B. Monteith,

Queensland Museum
and

R.L Storey,

Department of Primary Industries, Mareeba.

ABSTRACT

The complex and aberrant biology of the endemic, Australian dung beetle genus,
Cephalodesmius, is described from a detailed study of C. armiger and supporting observations
on the other two species of the genus, C. laticollis and C. quadridens. Permanently bonded pairs
occupy fixed, subterranean nests in the rainforest floor during their year-long lifespan. Males
forage outside the nest for leaves and other plant materials which are processed and cultured by
the female into a malleable, dung-like material which is used to construct small brood balls
containing larvae. During larval growth the parent beetles cooperate in progressive provisioning
of the larval brood balls, a complexity unknown elsewhere in the dung beetles and probably
unique in the Coleoptera. Audible stridulation is recorded for the first time in larval
Scarabaeidae, and a new mechanism of sound production described in which the tip of the
abdomen is rubbed against the underside of the head. Stridulation is interpreted as
communication between larva and adult.

The brood nests of Cephalodesmius are inhabited by a suite of 8 other insects and mites
which enter into a variety of trophic and phoretic relationships with the beetles and these are
summarized. The evolutionary significance of the behaviour of Cephalodesmius is discussed.

INTRODUCTION

The subfamily Scarabaeinae is a large, diverse
and successful group which has earned the
common name of ‘dung beetles’ through their
almost universal utilization of animal dung as
larval and adult food. Although a few species are
known to feed on other diets such as fruits,
carrion, feathers and fungi it is generally accepted
that dung is the primitive and primary food of the
subfamily. The classic observations of J.H. Fabre
(1897) were the first to reveal that some dung
beetles have complex behaviour patterns by which
they manipulate and transport dung to subterran-
ean nests where they rear their immature stages.
It is now known that there is a remarkable array
of variations on the basic breeding behaviour
described by Fabre, that of some species
approaching the most intricate known in the
Insecta. Halffter and Matthews (1966) give a
comprehensive review and analysis of the overall

biology of the world fauna, while Halffter (1977)
gives a comparative account of nesting behaviour
(nidification) in the subfamily, including possible
evolutionary pathways.

The six tribes of Scarabaeinae show a basic
behavioural dichotomy into (i) those that bury the
dung in preformed burrows at the food source
(Onthophagini, Oniticellini, Onitini and Coprini),
and (ii) those that form the dung into balls which
are transported away and buried at a distance
from the source (Scarabaeini and Eurysternini)
(Halffter, 1977). The latter group are often
termed ‘ball-rollers’. All Australian ball-rollers
belong to the tribe Scarabaeini and have been
comprehensively revised by Matthews (1974) who
placed all 16 genera into the primitive, Gondwana
subtribe, Canthonina. Matthews notes that 1 1
genera of the Australian Canthonina, comprising
what he terms the ‘mentophiline-group’, have not
been observed to make balls and he suggests they
represent an early evolutionary stage before the

254

MEMOIRS OF THE QUEENSLAND MUSEUM

acquisition of this ability in the Scarabaeini. He
draws attention to the paucity of information on
the biology of these ‘mentophilines’ and suggests
such data might throw light on the evolution of
behaviour in dung beetles.

One of these mentophilines is Cephalodesmius
Westwood, an endemic Australian genus with
three flightless species restricted to wetter forests
of SE Queensland and N.S.W. All three species
are common insects where they occur.
Cephalodesmius laticollis Pascoe and C. qua-
dridens Macleay are confined to the northern part
of the generic range, while C. armiger Westwood
is virtually ubiquitous in rainforests and some
adjacent sclerophyll forests throughout the range
of the genus from the Bunya Mountains south to
Wollongong. All three species overlap in the
Macpherson Range complex on the N.S.W. / Qld
border.

Previous observations on the biology of
Cephalodesmius are fragmentary and somewhat
conflicting. Fricke (1967) excavated a burrow
near Sydney to find a pair of adult C. armiger, a
large, oval pellet of what he took to be horse dung,
and 5 smaller pellets. He kept the pellets in a
container with the adults during which time they
consumed part of the ‘horse dung’ and tended the
round pellets. Eventually an adult emerged from
one pellet and the others, when opened, contained
dead larvae and pupae. Matthews (1974) records
his personal observations of armiger digging short
burrows in captivity, and of both armiger and
laticollis shaping rough balls of dung which they
dragged backwards instead of rolling; he also
notes an ‘extraordinary observation’ by G.B.
Monteith who reported C. quadridens ‘in holes
filled with pieces of green leaves under logs’ at the
Bunya Mountains. Finally, a most important early
observation, overlooked by other writers but
brought to our notice by Dr B.P. Moore, is that of
Walker (1905, p. 269) who writes:

‘The curious small Coprid,
Cephalodesmius armiger Westwood, is also
somewhat diurnal in its habits. In the
Illawarra in April, 1903, I found the females
of this beetle busily engaged in filling their
burrows under the logs with fresh minced-up
leaves of clover and other low-growing plants,
presumably as food for the larvae; a habit
which recalled to my recollection the
proceedings of the still more singular Lethrus ,
which I had observed in Turkey many years
ago provisioning its burrow with vine leaves.’

The major inconsistency in these isolated
reports is the contrast between the apparently

normal association with dung seen by Fricke and
Matthews, and the apparently abnormal associa-
tion with leaf material noted by Walker and
Monteith. Accordingly, a detailed field and
laboratory study of the biology of
Cephalodesmius was carried out during the period
1972-74 in an attempt to resolve this anomaly.
The results of this study are presented in this
paper. We know now that all the previous
observations are indeed correct, except for the
certainty that what Fricke saw was not ‘horse
dung’ but plant material mashed by the beetles
themselves. Cephalodesmius adults do, as noted
by Matthews, feed on dung when the opportunity
arises but most of their foraging activity goes into
gathering leaves and other vegetable material
which is mashed and cultured in preformed
underground nests into a dung-like substance
which forms the larval brood material and much
of the adult diet. This remarkable adaptation is
quite unique even in an enormous subfamily
famous for its behavioural eccentricities. Other
striking features of Cephalodesmius biology
which emerge from this study include: permanent
pair-bonding of males and females during a
lifetime of almost 12 months; residence in a single
burrow during the life of the pair; deposition of
eggs in tiny ‘egg-balls’ which are progressively
provisioned with brood material as the larvae
inside feed and grow; and phoretic and trophic
associations with a large suite of other organisms
which share their nests.

Due to the delay in publication of this
comprehensive account, certain aspects of our
findings have been referred to, with our
permission, in other papers dealing with related
matters. These include Richards (1973) on
phoretic flies, Matthews (1974) with footnotes on
use of leaf material, Costa (1975) on phoretic
mites, Hammond (1976) on phoretic staphylinids,
Matthews (1976) with reference to leaf utiliza-
tion, but erroneously citing eucalypt leaves,
Halffter (1977) who comments on the significance
of Cephalodesmius to the evolution of nest
behaviour in dung beetles, and Edmonds and
Halffter (1978) who refer to larval stridulation.

MATERIALS AND METHODS

This study is based on a detailed examination of
the biology of Cephalodesmius armiger. This was
supported by sufficient observations on C.
laticollis and C. quadridens to confirm that all
species of the genus have similar behaviour
patterns. Studies on C. armiger took place in both
field and laboratory.