C H6RAL

Bulletin of the

British Museum (Natural History)

Zoology series Vol 37 1979-1980

British Museum (Natural History)
London 1980

Dates of publication of the parts

No 1 29 November 1979

No 2 20 December 1979

No 3 31 January 1980

No 4 .28 February 1980

ISSN 007-1498

Printed in Great Britain by Henry Ling Ltd, at the Dorset Press, Dorchester, Dorset

Contents
Zoology Volume 37

No 1 Miscellanea

Reticulopodia in testate amoebae (Rhizopodea : Protozoa)

Ronald H. Hedley & Colin G. Ogden

Review of the genus Guaranidrilus (Oligochaeta, Enchytraeidae) with

the description of two new species

Brenda Healy ..........

Larval development of British prawns and shrimps (Crustacea :
Decapoda : Natantia). 3. Palaemon (Palaemon) longirostris H. Milne
Edwards, 1837 and the effect of antibiotic on morphogenesis

A. A. Fincham

The larval development of the spider crab Rochinia carpenteri
(Thomson) [Oxyrhyncha : Majidae] with a review of majid subfamilial
larval features

R. W. Ingle

On the spider genus Cynapes (Araneae : Salticidae)

F. R. Wanless

A new species of Phthiracarus (Acari, Cryptostigmata) from Austria

B. W. Parry

No 2 The echinoderms of Aldabra and their habitats

N. A. Sloan, A. M. Clark & J. D. Taylor

No 3 The anatomy, phylogeny and classification of bariliine cyprinid fishes

G. J. Howes

No 4 The Fellodistomidae (Digenea) of fishes from the northeast Atlantic
R. A. Bray & D. I. Gibson

Page

17

47
67
73

81
129
199

LIBRARY

Bulletin of the

British Museum (Natural History)

Miscellanea

Zoology series Vol 37 No 1 29 November 1979

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and
an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers
are works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,

British Museum (Natural History),
Cromwell Road,

London SW7 5BD,
England

World List abbreviation : Bull. Br. Mus. nat. Hist. (Zool.)

Trustees of the British Museum (Natural History), 1979

ISSN 0007-1498 Zoology series

Vol 37 No 1 pp 1-79
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 29 November 1979

L LIBRARY <S

Miscellanea

Contents

Page

Reticulopodia in testate amoebae (Rhizopodea : Protozoa). By Ronald H.
Hedley & Colin G. Ogden 1

Review of the genus Guaranidrilus (Oligochaeta, Enchytraeidae) with the
description of two new species. By Brenda Healy ..... 7

Larval development of British prawns and shrimps (Crustacea : Decapoda :
Natantia). 3. Palaemon (Palaemori) longirostris H. Milne Edwards, 1837 and
the effect of antibiotic on morphogenesis. By A. A. Fincham ... 17

The larval development of the spider crab Rochinia carpenteri (Thomson)
[Oxyrhyncha : Majidae] with a review of majid subfamilial larval features.
ByR.W. Ingle 47

On the spider genus Cynapes (Araneae : Salticideae). By F. R. Wanless . . 67

A new species of Phthiracams (Acari, Cryptostigmata) from Austria. By B. W.
Parry 73

Reticulopodia in testate amoebae (Rhizopodea :
Protozoa)

Ronald H. Hedley & Colin G. Ogden

British Museum (Natural History), Cromwell Road, London SW7 5BD

Summary

The pseudopodial system of Cryptodifflugia oviformis and some other testate amoebae consists of numerous
filopodia when examined at the resolution of the optical microscope. These filopods are seen to be part
of a full reticulum when examined at the ultrastructural level. It is suggested that the general assumption
that testate amoebae produce a filopodial system and foraminifera a reticulopodial system may be
invalid.

Introduction

The basic difference between lobopod, filopod and reticulopod pseudopodia as found in the
class Rhizopodea, Protozoa, is their shape. Lobopodia are blunt extensions of the cytoplasm
such as those found in most of naked amoebae and large testate amoebae; filopodia are usually
single tapering cytoplasmic extensions as, for example, in Euglypha and other small testate
amoebae, whilst reticulopodia are normally the granular anastomosing networks of cytoplasm
produced by foraminfera. These basic differences are recognized in standard classifications of the
rhizopods (e.g. Honigberg et al., 1964; Loeblich & Tappan, 1961, 1964). Most rhizopods produce
pseudopodia which are readily assigned to one of these three basic types. Variants of the three
types are well known: for example, Page (1975, 1976) illustrates some naked amoebae which
produce filose-like pseudopodia and Leidy (1879) in his account of Freshwater Rhizopoda of
North America reports a variety of pseudopodial forms in testate amoebae. There is little
confusion between the two extreme forms - lobose and reticulose - but in some protozoa either
type may merge into the filose type. Greater confusion exists in certain testate amoebae which
possess pseudopodia showing a tendency towards both the lobose and reticulose forms.

This account is concerned with observations at the optical and ultrastructural level of the
pseudopodial system in Cryptodifflugia oviformis, a cosmopolitan soil and moss inhabiting
testacean whose general biology, taxonomy and fine structure have been reported previously
by Hedley et al. (1977).

Materials and methods

A clonal culture of Cryptodifflugia oviformis was isolated from the moss, Eurhync hium praelongum,
by Hedley et al. (1977), and formed the working cultures for this study. It is now deposited at
the Culture Centre for Algae and Protozoa, The Natural Environment Research Council,
Cambridge, England (Reg. No. 1514/2). Live animals from these cultures were examined by
both phase-contrast and bright field illumination. For transmission electron microscopy, speci-
mens were fixed for 12 min. in 1 % glutaraldehyde in 0-025 M cacodylic acid buffer, followed by
7 min. in 3 % glutaraldehyde in the same buffer. After several rinses in the buffer solution, they
were post-fixed with 1 % osmium tetroxide in distilled water. The material was dehydrated and
embedded in Epon 812. Sections cut with a diamond knife on a Porter Blum ultramicrotome
were stained with a saturated solution of alcoholic uranyl acetate and Reynold's lead citrate, and
examined in an A.E.I. 6B electron microscope operating at 60 kV. The results were recorded on
Ilford EM 6 plates.

Bull. Br. Mus. mit. Hist. (Geol.) 37 (1): 1-6 Issued 29 November 1979

R. H. HEDLEY AND C. G. OGDEN

Fig. 1 Phase-contrast micrograph of Cryptodifflugia oviformis to show the typical arrangement

offilopods, x520.

Results

The pseudopodial system of actively moving specimens of Cryptodifflugia oviformis usually
consists of several tapering cytoplasmic extensions which may have small branches, and appear
to represent a typical filopodial system (Fig. 1). At the ultrastructural level (Fig. 2) the pseudo-
podial system is seen to consist of the main filopods and many additional cytoplasmic strands,
ranging in diameter from 45 nm to 3 urn, which are reminiscent of those seen in foraminifera.
If one identifies in such a micrograph as Fig. 2 those pseudopodia which would normally be
seen with the optical microscope, that is those whose diameter is greater than 1 um (Fig. 3),
some explanation is needed for the remainder of the cytoplasmic strands in the micrograph.
At a higher magnification (Fig. 4) these connecting cytoplasmic strands are similar to those
found in foraminifera with a reticulose network (see Figs. 8 and 9).

A model of the type of pseudopodial arrangement which would be consistent with the micro-
graph (Figs. 2 and 3) is presented in Figs 5 and 6. This represents only one of many possible
arrangements of the pseudopodia which would correspond to such a micrograph. The contrac-
tion of the filopods as depicted in Fig. 5 is what one would expect, and does observe when an
animal with an extruded filopodial system is immersed and subsequently fixed in glutaraldehyde.

Discussion

In describing the biology and ultrastructure of Cryptodifflugia oviformis, Hedley et al. (1977)
stated that previous studies, among which de Saedeleer (1932) may be mentioned, had drawn
attention to the fact that the pseudopodia appeared to be intermediate in form between lobose

Fig. 2 Section of a specimen of C. oviformis to illustrate the range of pseudopodial structures

visible at the ultrastructural level, x4300.
Fig. 3 Section as in Fig. 2 showing those pseudopodia with a diameter greater than 1 nm masked

in black, /4300.
Fig. 4 Section of pseudopodia (p) and pseudopodial strands (ps) near the aperture of C. oviformis,

X21 000.
Fig. 5 A model of a specimen of C. oviformis to correspond with the micrographs shown in Figs

2 and 3.
Fig. 6 Diagram to show model cut in the same plane as the sections in Figs 2 and 3.

RETICULOPODIA IN TESTATE AMOEBAE

'2 ^*s.

^. i ' v

4

R. H. MEDLEY AND C. G. OGDEN

Fig. 7 Diagram showing a proposed pseudopodial network for C. oviformis, based on the examina-
tion of several micrographs.

C

.y
t-

?

L

* -V A

-8

Fig. 8 Section of the pseudopodial network of Allogromia laticollaris (Foraminifera), x 30 000.
Fig. 9 Section showing the size range of pseudopodial structures in Shepheardella taeniformis

(Foraminifera), x 15 000.

RETICULOPODIA IN TESTATE AMOEBAE 5

and filose structures. We did not comment on them at the time, as there was insufficient informa-
tion on the ultrastructural organization in other testate amoebae. It is now suggested that in
C. oviformis the pseudopodial system is an anastomosing network. This appears to be so in
other forms possessing filose pseudopodia, for example, Euglypha and Trinema, and also for
some with lobose pseudopodia, for example, Arcella, Centropyxis and Nebela (unpublished
observations). If this proves to be generally true of testate amoebae then the distinction between
a filopodial system in such rhizopods and a reticulopodial system in foraminifera ceases to be
valid. The difference remaining between the two groups would then be whether or not the net-
work in testate amoebae is granular, as it clearly is in the foraminifera when viewed with the
optical microscope (Hedley, 1964). At present the only report of granular filose pseudopodia
in testate amoebae is that of Berrend (1966) who described the bi-directional flow of granules
in Cyphoderia ampulla,

The ultrastructure of the pseudopodia of C. oviformis (Fig. 4) compares well with our previous
studies of the fine structure of foraminifera, such as Shepheardella taeniformis (Fig. 9), Allogromia
laticollaris (Fig. 8), India diaphana and Boderia turneri, and with descriptions of foraminiferal
pseudopodia by other workers (Wohlfarth-Botterman, 1961; Lengsfeld, 1969; Marsalek, 1969;
Schwab, 1969; Febvre-Chevalier, 1971; Anderson & Be, 1976, 1978).

The range in diameters of the pseudopodia in C. oviformis is 45 nm-3 urn and this compares
favourably with those we have observed in foraminifera - Allogromia laticollaris 30 nm-2 urn,
Boderia turneri 40 nm-1-5 um, and India diaphana 40 nm-2 um - and those reported previously
for the benthic foraminifer Shepheardella taeniformis 30nm-l um (Hedley et al., 1967) and
the planktonic foraminifer Globigerinoides sacculifera 80 nm-2 um (Anderson & Be, 1978).

In conclusion it is suggested that the basic architecture of pseudopodial fine structure in
testate amoebae supports the view that these forms produce a reticulopodium. In many charac-
teristics they are similar to the reticulopodia found in foraminifera except that in the foraminifera
the pseudopodia are invariably granular and exhibit bi-directional streaming when observed with
the optical microscope.

References

Anderson, O. R. & Be, A. W. H. 1976. The ultrastructure of a planktonic foraminifer, Globigerinoides

sacculifer (Brady), and its symbiotic dinoflagellates. J.foramin. Res. 6 : 1-21.
1978. Recent advances in foraminifera fine structure research. In Foraminifera, Vol. 3, R. H.

Hedley & C. G. Adams (eds.). London, Academic Press, pp. 121-201.
Berrend, R. E. 1964. Filopodial movement in Cyphoderia ampulla (Ehr.). in Primitive motile systems in

cell biology, R. D. Allen & N. Kamiya (eds). London, Academic Press, pp. 433-443.
Febvre-Chevalier, C. 1971. Constitution ultrastructurale de Globigerina bulloides d'Orbigny, 1826

(Rhizopoda - Foraminifera). Protistologica 7 : 311-324.

Hedley, R. H. 1964. The Biology of Foraminifera. In The International Review of General and Experi-
mental Zoology, Vol. 1, W. J. L. Felts & R. J. Harrison (eds). London, Academic Press, pp. 1-45.
Ogden, C. G. & Mordan, N. J. 1977. Biology and fine structure of Cryptodijflugia oviformis (Rhizo-

podea : Protozoa). Bull. Br. Mus. Nat. Hist. (Zoo/.), 30 : 311-328.
Parry, D. M. & Wakefield, J. St. J. 1967. Fine structure of Shepheardella taeniformis (Foraminifera :

Protozoa). //. R. microsoc. Soc., 87 : 445-456.
Honigberg, B. M., Balamuth, W., Bovee, E. C., Corliss, J. O., Gojdicis, M., Hall, R. P., Kudo, R. R.,

Levine, N. D., Loeblich, A. R., Weiser, J. & Wenrich, D. H. 1964. A revised classification of the phylum

Protozoa. /. Protozool. 11 : 7-20.
Leidy, J. 1879. Freshwater Rhizopods of North America. Report of the United States Geological Survey

of the Territories, Washington, pp. 1-324.
Lengsfeld, A. M. 1969. Zum Feinbau der Foraminifere Allogromia laticollaris. II. Mitteilung : Aus-

gestreckte und durch Abreiben isolierte Rhizopodien. Helgolander wiss. Meeresunters 19 : 262-283.
Loeblich, A. R. & Tappan, H. 1961. Suprageneric classification of the Rhizopodea. /. Paleont. 35 :

245-330.
1964. Thecamoebians. In Treatise on Invertebrate Palaeontology, Part C, Protista 2, Vol. 1.

The Geological Society of America, University of Kansas Press, Lawrence, C16-C54.
Marsalek, D. S. 1969. Observations on Iridia diaphana, a marine foraminifer. /. Protozool. 16 : 599-611.

6 R. H. MEDLEY AND C. G. OGDEN

Page, F. C. 1975. A new family of amoebae with fine pseudopodia. Zool. J. Linn. Soc. 56 : 73-89.
1976. An illustrated key to freshwater and soil amoebae with notes on cultivation and ecology.

Freshwater Biological Association Scientific Publication No. 34, 1-155.
Saedeleer de, H. 1932. Reserches sur les pseudopodes des Rhizopodes Testaces. Les comcepts pseudopodes

lobosa, filosa, et granulo-reticulosa. Archs Zool. exp. gen. 74 : 597-626.
Schwab, D. 1969. Electronenmikroskopische Untersuchung an der Foraminifere Myxotheca arenilega

Schaudinn. Z. Zellforsch 96 : 295-324.
Wohlfarth-Bottermann, K. E. 1961. Cytologische Studien VIII. Zum Mechanismus der Cytoplasma-

stromung in dunnen Faden. Protoplasma 54 : 1-26.

Manuscript accepted for publication 7 February 1979

Review of the genus Guaranidrilus (Oligochaeta,
Enchytraeidae) with the description of two new
species

Brenda Healy

Zoological Department, University College, Dublin

Synopsis

Guaranidrilus Cernosvitov, 1937 was erected to contain three new species from Argentina, G. glandulosus,
G. rams and G. fridericoides nom. nud. and a species of doubtful affinity, G. (Henlea) columbianus from
Columbia. Only one other species, G. lamottei from Africa, has since been described. Although G.
fridericoides was never described, re-examination of Cernosvitov's specimens reveals that the species
is distinctive. The genus is emended and two new species are described: G. cernosvitovi sp. nov. (for G.
fridericoides nom. nud.) and G. europeus sp. nov. from France and Spain.

Introduction

The genus Guaranidrilus was erected by Cernosvitov (1937a) to include four species: G. glandu-
losus, G. rarus and G. fridericoides, new species from Argentina; and G. columbianus (Michaelsen,
1913) from Columbia. The new species were not described at the time but subsequently descrip-
tions were provided for two, G. glandulosus and G. rarus, together with a redefinition of the
genus (Cernosvitov, 19376). The third, G. fridericoides, has never been described so the name
does not fulfil the requirements of Article 13a of the International Code of Zoological Nomen-
clature and is considered to be a nomen nudum. Only one other species has been recorded,
G. lamottei Omodeo, from the Ivory Coast (Omodeo, 1958). Recently new material of Guarani-
drilus was collected from France and Spain and led to an investigation of the genus. Re-examina-
tion of Cernosvitov's material of G. fridericoides in the British Museum (Natural History)
showed that the species is distinct and a description is given in this paper together with details
of a new species from S.W. France and the Pyrenees.

The three previously recorded localities of Guaranidrilus lie between 8 N and 27 S, thus
the genus appears to have a mainly tropical distribution. Cernosvitov (19376) expressed the
opinion that the genus is probably widely distributed in both tropical and subtropical parts
of South America, although he presented no evidence to support this view. The discovery of
a new species in south-west France and the Pyrenees is of special interest because it extends the
known geographical range of Guaranidrilus. This new European species displays the essential
characteristics of the genus but departs in two respects from Cernosvitov's diagnosis : a seminal
vesicle is absent and there are no oesophageal glands. Similarly G. lamottei also exhibits characters
not shared with Cernosvitov's species. The inclusion of these Old World species makes it necessary
to emend the genus.

GUARANIDRILUS Cernosvitov, 1937 (emended)
TYPE SPECIES. G. glandulosus Cernosvitov, 1937 (designated by Brinkhurst & Jamieson, 1971).

DIAGNOSIS. Setae straight, without nodulus, two per bundle. Head pore at the tip of the pros-
tomium; dorsal pores absent. Three pairs of septal glands united dorsally, all with ventral
lobes. Sudden transition between oesophagus and intestine with a pair of bilobed intestinal
diverticula at the oesophago-intestinal transition. Dorsal vessel originating in the clitellar region

Bull. Br. Mas. not. Hist. (Zool.) 37 (1) : 7-15 Issued 29 November 1979

8 B. HEALY

or just anterior to it. Interstitial tissue of the nephridia well developed; anteseptale large, con-
taining coils of the canal, efferent duct terminal. Sperm funnel cylindrical, duct long and thin
and wound into a spiral. Penial bulb compact. Spermatheca simple, not communicating with the
oesophagus.

HABITAT. Terrestrial and freshwater.

SPECIES. G. glandulosus, G. cernosvitovi sp. nov. and G. europeus sp. nov. are distinctive species
which cannot be confused with any of the enchytraeid species presently known and which fall
clearly within the genus as defined. G. columbianus and G. lamottei are both incompletely
described, there is therefore some doubt concerning their taxonomic affinities. G. rarus is close
to G. glandulosus and the validity of the species is doubtful.

REMARKS. The intestinal diverticula have a characteristic bilobed shape which may be unique
among the Enchytraeidae. Although diverticula occur at the oesophago-intestinal transition
in Buchholzia, Enchytronia and some species of Henlea, these taxa differ from Guaranidrilus
in possessing spermathecae which communicate with the oesophagus. Two characters which
Cernosvitov regarded as having generic significance have been omitted from the emended
diagnosis: the presence of oesophageal glands (which he termed 'peptonephridia') and the
presence of a seminal vesicle. Paired, spherical, oesophageal glands are present in VI in all
three Argentinian species but in G. lamottei they originate at IV/V and are long and forwardly
directed. Glands attached to the anterior part of the oesophagus have not been reported in
G. columbianus and are probably absent as in G. europeus. This is seemingly a variable character
as in Marionina and Achaeta. The seminal vesicle is variable in many enchytraeid genera.
In Fridericia, for example, it is present in most species with large individuals but is absent
from many of the species in which the worms are smaller.

SPECIFIC CHARACTERS. The main characters of the species are listed in Table 1. The most useful
diagnostic characters are the presence and form of the oesophageal glands, the position of the
intestinal diverticula, the size of the sperm funnel and seminal vesicle and the shape and size
of the spermatheca. Cernosvitov probably did not have access to live material and for this
reason may have overemphasized the taxonomic importance of the position of the intestinal
swelling and the shape and size of the intestinal diverticula. Examination of living specimens
of G. europeus reveals that the intestinal diverticula are contractile and can vary in shape and
size in an individual (Fig. 3). Also, the intestinal swelling appears to alter position according
to the state of the diverticula, e.g. when the latter are expanded, the swollen region of the gut
is pushed posteriorly. These characters are therefore unreliable for purposes of identification.

Guaranidrilus glandulosus Cernosvitov, 1937
Guaranidrilus glandulosus Cernosvitov, 19376 : 149, figs 1-9.

TYPE MATERIAL. Syntypes: 5 microslides (1 whole mount and 4 vertical longitudinal sections),
British Museum (Natural History) Reg. Nos 1949.3.1.949-954. Collected 17.xi.31 and 18.ix.31;
collector unknown.

Cernosvitov stated that specimens were deposited in the Museo Argentine de Ciencias Naturales.
The material referred to is probably that which is now in the collections of the British Museum
(Natural History) (see G. rarus).

TYPE LOCALITY. Loreto, near Santa Ana, Misiones, Argentina. Soil rich in humus and stream
detritus.

REMARKS. While Cernosvitov's description is adequate, he gives no indication of the number
of specimens examined or the extent of variation observed. The surviving material is, for the
most part, poor, only one of the preparations shows the diagnostic characters clearly. It seems
probable that the description was based on a small number of specimens giving a somewhat
narrow concept of the species. Some variation is to be expected, especially in the size of the
intestinal diverticula.

REVIEW OF THE GENUS GUARANIDRILUS

i i

|

^

CM

<*-

i

r. rarwj

o

ro 3
1 O

^ R 5

1

Eo

^ V-i

3 o

W> Q.

I g^

G

_ .2

e "c3 **
G > o *-

- T .g g

w

c3

" ~

>.S

X o e- g

vt 5 "3 ed

II
Q O

c -o

"" <U

60

f

1

C t

C o>

'

"3

^

- c

S H '<^ .

HH "5

S

S

_ oo 13 25

> > c "^

.S

^

O <n t X5

M

hn" HH" I I 5b *2

^H" "5

^5

o *o *O

e-

^. o

> > > 'i 1

X c 2

^e- e-

n

60 <-

2

a
g

M

X

S 60

4^

Q

,^

cd ^~*

1

i

o

1

O

a a

o ~

c ^

o

60

a
8

\b 2 3
1 T O

o .A ^

-3

**-

J > ^ c

r> ^ > 1)

V 1 " '"" l HH "*

S- 1

C ""

o o S

00 M

c/3

'3

Dp

60

.9

&5

_,_,

i

-.S

Is

2S

3

Q

j < ^ S

^

>

^

^ ^

H H B

m

rt ^

'^

1

0\ g 1

to l

<*-!

> > > c

IrN C

^ ** ^

^

^j G

T*- 1

3 O

- ~- - o

^3- <J ^

J^ g J3 1)

^

fN D,

I <A ^

"c _L

u* Q.

|-H H HH J3

n ^i T 1 x>

K 2 a *

'5

O v>

tS N -M

- T *'

>- '*2

> > > rt

X o ts rt <a

<< *-" (n Cd

g

o

i

I

X G

*o

|

3

.,_,

O

HH ^

a

~||

v>

S >

g

i O

HH ^^

H

> >

<a

c o

p_ QQ 5P

^

cd j_,

^ ^ O

5 1 W5 'tj

1

H> f!

^ *

T "T

.a

3 o

^r HH _r 2 >

X <C v-> ^

u "o 13 &

a

1/3

tt &

a ;* S

1

2

> > > -S

X -3 e- 4

li II

H

3

-

X

o>

60

^^

S

+y

^

N

,. *- c

I

5-7
30-35
throughou

n

-i-T ^

^2
lf

o vi

o

s

C3 ^_

3 o

2?.&
t-> <->

3 g^

'K

5 e

S^ 1 3

^r <S "S
_H o3 "^ >n u
ly 1 T5 ' S

> = ^

.a|l*

8 -S 5 8

C B cd

c

3T

g

_o

c

u

1

>

o

00

S

*.

c

w

S,

S

"13 ^ on

C 3 TD

i

<s

M

rt -^

'S O ^ N

>

s 8

e s

^2

&f 5

"O O

s ' -a

. S > 2 -

13 -5 ID
S2 3 ^^ o

^ * 1 * 1

S 1

6 J 2

<= 5 a
4)

" ' c o>

D B. Oj

.2 i? u

Cutaneou

Clitellar j
Head por

08 flB M A n

J3 C 3 X

cu ;: c 8 a
o "? E o o

5 i ^ 1

o 8 o

'o x <H J= --

s c I 3 e i 1 1
S-aJJ |l e

? JS D, u

O PQ co co

03 3

S |o 5

M C 3 U

W at TD <u
CO

10

B. HEALY

02mm

50 \im

VIM

sp

sv

c)

100

10

(d)

Fig. 1 Guaranidrilus cernosvitovi sp. nov. : (a) Segments V-IX. ed-ecta\ duct of spermatheca,
og - oesophageal gland, id - intestinal diverticulum, sp - spermathecal ampulla, sv- seminal
vesicle; (b) Clitellar glands; (c) Transverse section of ectal duct; (d) Sperm funnel and part of
duct; (e) Lymphocytes; (f) Spermatheca.

Guaranidrilus cernosvitovi sp. nov.
Guaranidrilus fridericoides nom. nud. Cernosvitov, 1937a : 282.

DESCRIPTION. Length (preserved specimens) 12-1 3 mm, width 0-4-0-5 mm. Segments 47-48 (6
specimens). Setae straight with weak ental hook, approximately equal, 50-60 um. Cutaneous
glands indistinct. Clitellum extending over XII-|XIII with squarish glands arranged in trans-
verse rows (Fig. Ib). Head pore at the tip of the prostomium. Three pairs of septal glands, all
with ventral lobes and united dorsally. Brain incised posteriorly. Transition between oesophagus
and intestine not sudden, represented by a slight enlargement at VII/VIII. A pair of more or
less spherical glands attached to the oesophagus dorso-laterally in VI, 80-100 um in diameter,
highly vascularized (Fig. la). Paired, hollow, bilobed intestinal diverticula in VII, 230 um in
length, their point of attachment just in front of the oesophago-intestinal transition (Fig. la).
Chlorogogen cells about the same size as the lymphocytes or a little smaller. Lymphocytes
nucleated, finely granulated, 25-35 um, mostly spindle-shaped, sometimes oval or circular

REVIEW OF THE GENUS GUARANIDRILUS \ \

(Fig. le). Nephridia all with a large anteseptal portion about one-third the total length of the
nephridium; efferent duct terminal, short. Origin of dorsal vessel at X/XI.

Sperm funnel large, 350 um in length, width varying between 75 and 200 um with a wide,
much-lobed collar (Fig. Id). Sperm duct long and coiled into a spiral. Seminal vesicle very large,
occupying IX-XI. Penial bulb small. Spermatheca large with a thick-walled ectal duct (Fig. lc)
extending through V and VI and a large cylindrical ampulla, 150-230 jum in diameter, extending
through VII and VIII (Fig. If). Ectal glands absent.

TYPE MATERIAL. Holotype: 1 microslide (unstained whole mount prepared by the author),
British Museum (Natural History) Reg. No. 1949.3.1. Collected 12.vi.32; collector unknown.
Paratypes: 3 microslides (whole mounts), 3 microslides (vertical longitudinal sections), 2 micro-
slides (transverse sections), British Museum (Natural History) Reg. Nos 1949.3.1b-i. Spirit
material (about 17 specimens).

TYPE LOCALITY. Loreto, near Santa Ana, Misiones, Argentina. Habitat unknown.

REMARKS. The species is well defined and readily distinguished by its large size and large sexual
organs, i.e. sperm funnels, seminal vesicle and spermathecae. There appears to be little vari-
ation in segment number or in the form of the internal organs.

It is not evident why Cernosvitov failed to describe so distinctive a species, especially as
more material was available to him than for either G. glandulosus or G. rams. Two of the micro-
slides in his collection (longitudinal and transverse sections) are excellent and show most diag-
nostic characters clearly.

Guaranidrilus europeus sp. nov.

DESCRIPTION. Small species, length 3-6 mm, width 100-140 um. Segments 23-29. Setae straight
with weak ental hook, two per bundle throughout, approximately equal, about 20 um (Fig. 2d).
Setae absent in XII. Occasional single setae in the coelom. Cutaneous glands indistinct, elon-
gated, 3-5 rows per segment (Fig. 2a). Clitellum extending over XII/^XIII with 20-33 rows of
squarish glands (Fig. 2b). Head pore near the tip of the prostomium. Three pairs of septal
glands, all united dorsally and with ventral lobes (Fig. 2h). Brain about twice as long as broad,
the posterior margin deeply cleft (Fig. 2e). Sudden transition between oesophagus and intestine
usually at VII/VIII but the position can vary according to the state of the intestinal diverticula.
When these are expanded, the intestine may be constricted so that the swelling appears in tX.
The two lateral intestinal diverticula communicate at VII/VIII and extend both forwards and
backwards (Fig. 3a). They usually measure 150-160um and occupy VII and VIII but may
extend anteriorly into VI and posteriorly into XI. They are hollow and highly contractile but do
not pulsate; some variations in shape are shown in Fig. 3. Chlorogogen cells about the same
size as the setae. Lymphocytes circular or oval, finely granular and nucleated, 10-20 um (Fig.
2c). Nephridia similar throughout, the anteseptal portion large with coils on the canal, the
efferent duct terminal (Fig. 2i). Dorsal vessel originating in XII or XIII, blood colourless.

One or two mature eggs present at a time. Sperm funnel pear-shaped, about 1| times longer
than wide, length 40 um (Fig. 2g). Sperm duct long and coiled. Seminal vesicle absent. Sperma-
theca free, its length 2-2 1 times the diameter of the worm. The thin-walled ectal duct has small
scattered glan 1 cells. It expands gradually into a terminal ampulla in VI. There is a slight expan-
sion of the ectal duct near its orifice but no ectal glands (Fig. 2f).

The description is based entirely on live specimens. The worms were transparent when viewed
microscopically and all characters could be readily observed. Thirty-two specimens were
examined from two widely separated populations; intraspecific variation appears to be slight.

TYPE MATERIAL. Holotype: 1 microslide (whole mount) British Museum (Natural History)
Reg. No. 1978.39.1. Collected by the author 21.x. 77. Paratypes: 1 microslide (whole mount),
British Museum (Natural History) Reg. No. 1978.39.2. 2 microslides (whole mounts), Museum

12

B. HEALY

10 (i m

h)

If)

Fig. 2 Guaranidrilus europeus sp. nov. : (a) Preclitellar segment showing cutaneous glands; (b)
Clitellar glands; (c) Lymphocytes; (d) Setae; (e) Brain; (f) Segments V-VI showing septal glands
and spermathecae ; (g) Sperm funnel and part of duct; (h) Septal glands, lateral view; (i) Nephri-
dium.

National d'Histoire Naturelle, Paris, Reg. Nos AP 700-701. 7 microslides (whole mounts)
and 5 specimens fixed in Bouin's in the author's collection. Collected by the author 2 1.x. 77.

TYPE LOCALITY. Near Arcachon, S.W. France. Beside a small pond in a pine-wood clearing at
the edge of the road from Le Corneau to Pilat. Wet sandy peat and peaty sand, pH 4-0-5-7.

OTHER MATERIAL EXAMINED. 9 specimens from Roncesvalles, Spanish Pyrenees, collected 31.vii.75.
Marshy pasture, wet-very wet, pH 6-0-7-2. 1 Specimen from a pasture near Arcachon, S.W.
France, collected 6.vii.75. Moist sandy loam, pH 5-9-6-1.

REMARKS. The diagnostic characters of the species are its small size and the absence of both
oesophageal glands and seminal vesicle.

Since it is unlikely that so distinctive a species could have been overlooked in northern Europe
where the Enchytraeidae are quite well known, this new species is presumably confined to the
south where it may represent a sub-tropical element in the European fauna. The enchytraeid
fauna of southern Europe has received little attention; the only well-studied regions are northern
Italy and Romania where Guaranidrilus has not been found. Hence, the genus may be confined
to south-west Europe being representative of a Lusitanian fauna. In south-west France, G.
europeus is rare. In July-August 1975 and October 1977 collections were made at 37 localities

REVIEW OF THE GENUS GUARANIDRILVS

13

(b)

(C

Fig. 3 Guaranidrilus europeus sp. nov. : (a) Intestinal divicula and oesophago-intestinal transition ;
(b)-(d) Various forms of the intestinal diverticula.

in the Landes, 12 in the Dordogne region and 11 in the Atlantic Pyrenees but G. europeus was
found at only two sites. At the type locality the population was small and restricted to the edge
of a small pond. At the other locality, only one specimen was found in spite of much searching.
The population in the Spanish Pyrenees, discovered in July 1975, also appeared to be very
localized. In October 1977, no specimens could be found at this locality. Generally, the enchy-
traeid fauna of Spain and Portugal is poorly known. Before the status of G. europeus can be
established in Europe, it will be necessary to collect extensively in these regions and in the
Mediterranean region of France.

Guaranidrilus columbianus (Michaelsen, 1913)
Henlea columbiana Michaelsen, 1913 : 211.

TYPE MATERIAL. Syntypes : 2 spirit specimens (1 mature, 1 immature), Zoologisches Museum,
Hamburg, Reg. No. V 7739. Collected by O. Fuhrmann, date unknown.

TYPE LOCALITY. Medellin, Medellin Lake at Ufer, Colombia. Altitude 2000 m. Habitat unknown.

REMARKS. Michaelsen's description, although covering the principal features, is imprecise in
places and is not illustrated, hence the species is not clearly defined. He referred it to the genus
Henlea on the morphology of the oesophago-intestinal transition which he observed just behind
the septum in X with the dorsal vessel also originating in this position and large, hollow 'pep-
tonephridia' in VII-IX. The spermathecae were seen to communicate with the oesophagus, a
characteristic feature of Henlea. However, Cernosvitov (1937ft) considered that many features
such as setal number, brain, septal glands, clitellar glands and spirally coiled sperm duct, linked
the species with Guaranidrilus and that the 'peptonephridia' in VII-IX were in fact similar to
the intestinal diverticula of G. glandulosus. He also observed that in Guaranidrilus the sperma-
thecal ampulla often adheres to the oesophagus making it difficult to trace the spermatheca
over its entire length, he concluded that Michaelsen may have been mistaken in his interpretation
of this organ. In support of this view it is worth noting that Michaelsen seems to have experienced
difficulty in distinguishing both spermathecae and 'peptonephridia' because he states that in
one serially sectioned specimen the 'peptonephridia' appeared to extend forward as a narrow
tube into V, terminating close to the opening of the spermathecal duct. It seems more likely
that the tube was actually the spermathecal duct and that the spermathecal ampulla was not
developed, as, for example, in an unfertilized specimen.

14 B. HEALY

A re-examination of Michaelsen's type material led to the conclusion that Cernosvitov was
correct in placing the species in Guaranidrilus. Although the attachment of the intestinal diver-
ticula could not be clearly seen in either of the specimens, the organs are more similar in general
appearance to those of other members of the genus than to those of the known species of Hen lea.
Unfortunately, the proximal regions of the spermathecae were not distinct in the single mature
specimen so the uncertainty concerning the communication with the oesophagus remains. It
was possible, however, to see the nephridia which Michaelsen did not describe fully. They are
clearly of the Guaranidrilus-type with a large anteseptale consisting of nephrostome surrounded
by coils of the canal and a terminal efferent duct, quite different from the Henlea-type in
which the anteseptale consists of nephrostome only and the efferent duct originates in the anterior
position just behind the septum.

From the information available, the distinguishing characters of G. columbianus are its size,
the absence of oesophageal glands and seminal vesicle and the position of the intestinal diverticula
and oesophago-intestinal transition. In both type specimens the intestinal transition is at VIII/IX,
not at IX/X as stated by Michaelsen. However, the specimens are somewhat contracted and
there may have been displacement of some organs. In the immature specimen, intestinal diverti-
cula on both sides are confined to VII and VIII but in the mature specimen one of the sacs occupy-
ing VIII extends back into IX. There may therefore be variations in the size of the diverticula.

Guaranidrilus lamottei Omodeo, 1958
Guaranidrilus lamottei Omodeo, 1958 : 11, fig. 1.
TYPE MATERIAL. Not located. Collected 8-9.viii.51.
TYPE LOCALITY. Zouguepo, Mount Nimba, Ivory Coast. Path in mountain grassland, 1360 m.

REMARKS. The description was based on immature specimens only and is thus incomplete. The
author referred the species to Guaranidrilus with some hesitation noting that several features
such as the position of the head pore at O/I and the ciliated inner surface of the intestinal diver-
ticula are more characteristic of the Henlea group. There may be some doubt about the position
of the head pore since the author admitted that it was barely perceptible. The bilobed diverticula
at the oesophago-intestinal transition, the nephridia with terminal efferent ducts, the presence
of free spermathecae and coiled sperm ducts are characteristic of Guaranidrilus.

G. lamottei possesses some unusual features not shown by other members of the genus, notably
the absence of dorsal setae in all segments posterior to VIII and the length and point of attach-
ment of the oesophageal glands. The almost straight posterior border of the brain also distin-
guishes it from other species. In view of these pecularities and the immaturity of the specimens,
the systematic position of the species must remain uncertain.

Guaranidrilus rarus Cernosvitov, 1937
Guaranidrilus rarus Cernosvitov, 19376 : 151, figs 10-15.

TYPE MATERIAL. Syntypes : 2 microslides (1 whole mount and 1 of horizontal longitudinal sections),
British Museum (Natural History), Reg. Nos 1949.3.1.954-955. Collected 21.xii.31; collector
unknown.

Cernosvitov (19376) noted that only two mature specimens were found. As the two specimens
in the British Museum (Natural History) collection must therefore represent the entire material
underlying the description, it is evident that Cernosvitov did not deposit the specimens in the
Museo Argentine as he stated.

TYPE LOCALITY. Loreto, near Santa Ana, Misiones, Argentina. Under the bark of an old tree.

REMARKS. Cernosvitov distinguished G. rarus from G. glandulosus on the position of the intestinal
dilatation in VII and not in VIII, the smaller size of the intestinal diverticula, the origin of the
dorsal vessel at X/XI instead of at XII/XIII and the absence of cutaneous glands. The size of
the intestinal diverticula cannot be considered significant in view of the contractile nature of

REVIEW OF THE GENUS GUARANIDRILUS 15

these organs while the position of the intestinal swelling may also be variable. A more reliable
taxonomic character is the position of attachment of the intestinal diverticula - clearly seen in
the longitudinal sections. This character, together with the point of origin of the dorsal vessel,
could be sufficient to delineate this species but since observations were made on only two speci-
mens, the consistency of these characters cannot be considered proven. The small size of the
specimens may be diagnostic but it is possible that this is due to ecological factors such as low
moisture content. The segment number falls within the range of G. glandulosus (Table 1) and
both oesophageal glands and spermathecae are similar in the two species. In view of these
similarities, the lack of reliable diagnostic characters and the paucity of material, the validity
of the species must be considered doubtful.

Acknowledgments

My thanks go to R. W. Sims of the British Museum (Natural History) for his co-operation and
for advice on nomenclature and taxonomic problems. Part of this work was carried out during
two surveys of French Enchytraeidae during which I stayed at the Station Biologique, Arcachon.
I wish to thank all those who made my visits so pleasant and particularly Pierre and Bernadette
Lasserre for their kindness. The surveys were funded jointly by the Centre National de la
Recherche Scientifique and the National Science Council of Ireland whose support is gratefully
acknowledged.

References

Brinkhurst, R. O. & Jamieson, B. G. M. 1971. Aquatic Oligochaeta of the World. Edinburgh, Oliver &

Boyd, pp. 1-860.
ternosvitov, L. 1937o. System der Enchytraeiden. Zap. nauchno-issled. OVed. russk. svob. Univ. Praga

5 (10) (Sci. nat. Mat.) No. 34 : 263-295.
19376. Notes sur les Oligochaeta (Naididees et Enchytraeidees) de 1' Argentine. An. Mus. argent.

Cienc. nat. 39 : 135-157.

Michaelsen, M. 1913. Die Oligochaeten Columbias. Mem. Soc. neuchat. Sci. nat. 5 : 202-253.
Omodeo, P. 1958. La reserve naturelle integrate du Mont Nimba. IV. 1. Oligochetes. Mem. inst. fr.

Afr. noire 53 : 9-17.

Manuscript accepted for publication 26 January 1979

Larval development of British prawns and shrimps
(Crustacea : Decapoda : Natantia). 3. Palaemon
(Palaemon) longirostris H. Milne Edwards, 1837
and the effect of antibiotic on morphogenesis

A. A. Fincham

Department of Zoology, British Museum (Natural History), Cromwell Road, London S W7 5BD

Synopsis

The larval stages of Palaemon (Palaemon) longirostris are described from specimens reared in the labora-
tory with ( + AB) and without (- AB) antibiotics. Growth in the early stages of larval development was
enhanced by the use of antibiotics. There was also an increase in intermoult period and a two-moult
delay of metamorphosis from 7 to 9 in + AB larvae. With these exceptions, the effect of antibiotics on
morphogenesis was no more than expected in normal meristic variation. The relationship between
moulting, growth and morphogenesis is examined in relation to genetic input, epigenesis and phenotypic
variation in this estuarine species.

Introduction

The aim of this paper is twofold: to provide a complete description of the larval development
of Palaemon (Palaemon) longirostris H. Milne Edwards, 1837 including the hitherto unknown
zoea 4 and to assess both quantitative and qualitative effects of antibiotics on morphogenesis.
Little attention has been given to the larval development of P. (P.) longirostris with the exception
of papers by Gurney (19246) and Korshelt (1944) in which zoeae 1-3, 5 and post larvae were
described. The adults of this species are widely distributed in estuaries (Gurney, 1923; De Man,
1923) with records from the southern North Sea, west coast of Ireland and as far south as the
Bay of Biscay (Fincham & Williamson, 1978). The species is also present in the Mediterranean
and Black Seas. Larvae from plankton samples, however, are extremely rare (Gurney, 19246;
Williamson, personal communication). Personal observations on offshore, inshore and estuarine
plankton samples which were collected from the North Sea and around the English and Welsh
coasts by MAFF Fisheries Laboratory, Lowestoft, confirm these earlier findings.

Antibiotics are used routinely in the culture of larval decapods at the BM(NH) as an additive
to the sea water to control build-up of bacteria in the compartmented rearing trays (Fincham,
1977, 1979). The effect of antibiotics on the development of brachyuran decapods has been
investigated (Christiansen, 1971) but their effect on morphogenesis in natant decapods is less
well known. Wickins (1972) compared the survival of Palaemon serratus (Pennant, 1777) larvae
and the proportion which metamorphosed when cultured with and without antibiotics.

Materials and methods

Ovigerous Palaemon (Palaemon) longirostris were collected in July 1977 from filtering screens
at the tidal water intake of the West Thurrock Electricity Generating Station on the River
Thames (grid reference: TQ 593768). Rearing techniques similar to those reported previously
(Fincham, 1977, 1979) were used with the following modifications:

1 . The controlled temperature room was at 18 0-5 C, equivalent to ambient water tempera-
ture at the time of collection.

Bull Br. Mus. not. Hist. (Zool.) 37 (1) : 17-46 Issued 29 November 1979

17

18

A. A. FINCHAM

.

.2

1

I

S

'>

o
o

1

o

N

+ +CS

+

'

cs + -i- +

\. X ' *^_^/ '

+

+

o _. o i i o

\^S ^3 N^ ' S ^ * ' S.^

o ^o

-f

^

+

o ^

3+ + +

o o o

32 >- i-

*~^. c^ 03

> C C

<u

T3

0) O

^ c
cs cs

U O

3 fc

^ CS O

c.2^

rt .> i

J

u- u

J"l (l
SJ-o +

zl^

rt

CQ

Sgg

-> H S

o

u

JS u

,<u
^

c

a

D

1

'o.

I

'S-

J2

in

+"

n

_3

t/5
U

'5,
M

H

i

lH

co

(U

'S,

"rt

-2-
5

C

O

T3

.c

g

o

"c3

V5

1

H

OJ

2

g

c

*03

-Si

_D,

a

O

-o

SJ

5

l_

c

S

2

<u <+-,

%

CM

o

O

2

IM

3

"c
<u

!i

2

a

3
C/3

6

Z

oo

O

g

>

u

t

N

Stylocerite +/-

+

oo
>>

O

o
S

C/5

A U

I

I

I

C

segmen

No

flag

< %

i

c

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

19

VV (-*) vo i i

^O S ro I +

^ <N

^ I

I fi.

v^o3

v^

I d, oo SSooSS n-SmO

V^

O (N

C^SSSS. TtScoCW

W "

O *"*

j r

3 " + +

"t Sm Crf Sm

m ^ i i

i<

\O o

*S

ON <F
6 f .

i

+ H

3 -

4 C 1 ^ +

i i sL ^ Sr- 6

^*\

< i
( > i

^.^ o

Tt S

(N

1
N r

'?

/

*

-H

j. -4-

vo r

" -^ I ' +

i i d, ^? St- S

_t *^f rr r^i . Tt* r

C^t- 5^-

i 1 ^/ \Q

^3" o ^

TtS

Tt-9

(N

fT

+ ' "

- c

r> ^

+

ON, | J^ +

i i ^ vos^S

Tt Sm C^r Sm

er> <o ""*

"- 1 O>o

S

* ^

5 S2

O J2

CO

U PQ

C

(U

'

|

-2

1

u

distal segme

plumose setc

CO

! 1

o3 co

x %

Q pi

i

4)

- 1

^ m 09

O 'co 'co "O
rj o3 o3 C
PQ CQ W

e - Exopodi

rnal margin

proximo-
lateral
setae

f segments

o

f|

^H

r- Ol

03

w

CM

o

ex

o

O

fe '2

01

Jj

'u

CO

.g

6

03

CO

6

6

t? '3
W J^

& ^

1

U

CO

c

6

i

*o

60

Exopodite -

Mandible - ]

O

c

3^

^ *0

^5

03 Z

s

3

a
'o

6
Z

03

.S"^

03 Z

Exopodite -

[axilliped 2
Endopodite

20

A. A. FINCHAM

a.

1 1

"^ 4- '

\_^ s_x

,

T

+ . +

4. 4.

^ s_^ 1

oS + i + i

+ . +

4. 4.

+ . +

4. 4.

1 ^ V T

4.

+ . +

4. 4.

I s_x '

+i++i

52 + CL +

+ . +

o
N

I

+

<s |

<D 1

^ "

'

i VAX

00

^

00

*

I 1

11

5 "

o. w

o =2

18 i I

o g >o

1

o
3

5

ramou

X t>

W c<o

W

X
U V2

'S S W

i

n

|

rti

1

i

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS 21

4

+I ' ++I I

'' ^ 1

+ io s i

o I

+

1 8 I

1 _e "^ 2 "

~1 -S I x-s V, D< O

+ 1 i ~i J

8 g + S g & ww

aS ? I i 2 i

5| ^ 8 8 , S 1

2 i S .S - - 8
-I 1 -HrS S a i+ t o

~~- ~ . ;3 U i t_ ra O (/j

+ + ^3^ i S a 2"3^^ i o

a ' ' i 8 K i 1

H s ill | I i 5 it

0,6 c 'a 'a "5w)<u feosO 'Ooo

s -S II 1 I -S & ? 1 1 o |

3^^ < IS <& S X S^

22 A. A. FINCHAM

2. Salinity of the water in the circulating system and the water used in the compartmented
trays was adjusted to the conditions prevailing in the estuary (salinity varied from 10 to
17% at West Thurrock - laboratory salinity maintained at 14% ).

3. In 6 rearing trays, each with 18 compartments, a mixture of 50 i.u. benzylpenicillin (sodium)
and 0-05 mg streptomycin sulphate per ml sea water was used instead of streptomycin
sulphate alone.

4. A replicate series of trays was set up with no antibiotic in the sea water. The length of
intermoult period, growth rate and meristic characters could then be compared with
larvae reared with antibiotics.

Larval material has been deposited in the Crustacea collection of the BM(NH), registration
number 1978: 307.

Palaemon (Palaemon) longirostris H. Milne Edwards, 1837

Palaemon longirostris H. Milne Edwards, 1837.

Palaemon edwardsii Heller, 1863.

Leander edwardsii Czerniavsky, 1884.

Leander longirostris De Man, 1915.

Leander longirostris Gurney, 19246.

Palaemon (Palaemon) longirostris Holthuis, 1950.

SYNOPSIS OF LARVAL DATA FROM PUBLISHED WORK. Leander longirostris Gurney, 19246: Figs 1-7
(zoeae 1-3, 5, post larvae 1, 2) laboratory reared and plankton; Leander longirostris Korschelt,
1944: Figs 810, 811 (zoeae 1-3) (after Gurney).

In the following short descriptions of the key characters of the larval stages, all setal counts
have been omitted but these are recorded in Table 1.

DESCRIPTION OF LARVAL STAGES
Key characters are printed in italic type and are useful for separating stages of British species.

ZOEA 1 (Fig. 1) 3-5 mm (3-3-3-8 mm)*.

Head (Figs la, b): eyes sessile.

Carapace (Figs la, b): without spines, rostrum straight, tapering distally, ventral margin with
minute retrorse teeth distally; rostrum as long as or longer than antenna 1 (excluding terminal
aesthetascs and setae).

Antenna 1 (Fig. Ic): peduncle bearing single flagellar segment, with three aesthetascs distally,
one wide and two narrow.

Antenna 2 (Fig. Id): exopod as a broad lamina divided into 5 (occasionally 6) segments
distally; endopodite of I segment.

Mandibles (Fig. le): asymmetrical.

Maxillipeds 1-3 (Figs Ih-j): with natatory exopods.

Pereiopods 1 , 2 (Figs 1 k, 1) : rudimentary, biramous.

Pereiopod 3 (Fig. 1m): minute biramous bud.

Pereiopods 4, 5 (Figs In, o): minute uniramous buds.

Abdomen (Figs la, b): somite 5 with posterior margin rounded, not produced into spines;
somite 6 continuous with telson. No trace of pleopods.

Telson (Figs Ip, q): fans out distally, posterior margin bears 7 + 7 plumose spines, inner pairs
with longitudinal rows of scale-like spinules, decreasing in size on the plumose setae towards
the edge of the fan; minute spines between four innermost pairs of plumose setae.

ZOEA 2 (Fig. 2) 3-9 mm (3-8-4-1 mm)

Head (Figs 2a, b): eyes stalked.

Carapace (Figs 2a, b): one dor so-medial and a pair of supraorbital spines all bent forward with
small retrorse teeth, rostrum without teeth, downturned at end to form a small hook.

* Lengths of larvae reared without antibiotic quoted throughout; for comparison of larval growth rates with and
without antibiotic see Fig. 18.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

23

Fig. 1 Zoea 1: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles;
(f) maxilla 1; (g) maxilla 2; (h) maxilliped 1; (i) maxilliped 2; (j) maxilliped 3; (k) pereiopod 1;
(1) pereiopod 2; (m) pereiopod 3; (n) pereiopod 4; (o) pereiopod 5; (p) telson; (q) detail from
inner pair of telsonic spines. Bar scales: a, b 0-5 mm; c, d, h-p = 0-2 mm; f , g = 0-1 mm;
e, q = 0-05 mm.

24

A. A. FINCHAM

Fig. 2 Zoea 2: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles; (f)
maxilla 1; (g) maxilla 2; (h) maxilliped 1; (i) maxilliped 2; (j) maxilliped 3; (k) pereiopod 1;
(1) pereiopod 2; (m) pereiopod 3; (n) pereiopod 4; (o) pereiopod 5; (p) telson. Bar scales: a, b =
0-5 mm; c, d, h-p = 0-2 mm; f , g = 0-1 mm; e = 0-05 mm.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS 25

Antenna 1 (Fig. 2c): two peduncle segments, stylocerite forming on proximal external margin
of first segment; single flagellar segment with four terminal aesthetascs, two wide and two narrow.

Antenna 2 (Fig. 2d): exopodite with distal part divided into 4 or 5 segments; wide gap between
9 distal setae and new setae added at this moult.

Pereiopods 1, 2 (Figs 2k, 1): developed, with natatory exopodite.

Pereiopod 3 (Fig. 2m): rudimentary, biramous.

Pereiopods 4, 5 (Figs 2n, o): rudimentary, uniramous.

Abdomen (Figs 2a, b) : somite 5 with posterior margin produced into a pair of conspicuous
spines, somite 6 continuous with telson.

Telson (Fig. 2p): developing uropods visible beneath exoskeleton alongside telson proper;
in central group of small spines, one pair longer than others.

ZOEA 3 (Figs 3, 4) 4-5 mm (4-2-4-7 mm)

Carapace (Fig. 3a, b): two dorso-medial spines and pair of small fronto-lateral spines at
edge of carapace beneath eyes, former with retrorse teeth; rostrum still downturned at end to
form a small hook.

Antenna 1 (Fig. 4a): conspicuous spine medially, stylocerite more pronounced; distal segment
of peduncle bearing first segment of internal flagellum, single segment of external flagellum
bearing 3 wide aesthetascs distally.

Antenna 2 (Fig. 4b): expedite with distal part divided into 3 short segments; endopodite of
3 segments.

Pereiopods 1, 2 (Figs 4f, g): endopodite with internal distal margin of propodus produced slightly
forward (will become fixed finger of chela).

Pereiopod 4 (Fig. 4i): rudimentary, uniramous.

Pereiopod 5 (Fig. 4j): developed, uniramous.

Abdomen (Figs 3a, b, 4k): somites 1-5 with rudimentary pleopods, somite 6 divided from
telson by suture. Uropod endopodite with no marginal setae, exopodite with marginal setae.

Telson (Fig. 4k) : narrower but still broader distally, outer pair of spines on posterior margin
considerably reduced.

ZOEA 4 (Figs 5, 6) 5-2 mm (4-7-5-5 mm)

Carapace (Figs 5a-d): three dorso-medial spines with retrorse teeth; rostrum still downturned
at end to form a small hook.

Antenna 1 (Fig. 5h): rudiment of circular statocyst visible on first segment of peduncle;
single segment of external flagellum bearing 3 wide and 1 narrow aesthetascs distally.

Antenna 2 (Fig. 5i): endopodite of 5 segments and as long or longer than the exopodite;
distal part of exopodite no longer divided into short segments but with a stout terminal tooth on
outer distal edge.

Pereiopods 1, 2 (Figs 6d, e): endopodite with internal distal margin of propodus produced
forward to about half length of dactylus (excluding terminal setae).

Pereiopod 3 (Fig. 6f): developed, with natatory exopodite.

Pereiopod 4 (Fig. 6g): developed, exopod absent or bud-like.

Abdomen (Figs 5a, b, 6i): uropod exopodite and endopodite both with marginal plumose setae.

Telson (Fig. 6i): a little broader distally than proximally; spine formula as in Zoea 3 but further
reduction in size of two outer pairs of spines.

ZOEA 5 (Figs 7, 8) 5-8 mm (4-8-6-1 mm)

Carapace (Figs 7a, b): rostrum still downturned to form small hook.

Antenna 1 (Fig. 7c) : single external flagellum with 4 aesthetascs distally, with additional group
of 2 aesthetascs on internal margin.

Antenna 2 (Fig. 7d): endopodite of 6-8 segments, longer than exopodite.

Pereiopods 1, 2 (Figs 8d, e): endopodite with internal distal margin of propodus produced forward
to over half length of dactylus (excluding terminal setae).

Pereiopod 4 (Fig. 8g) : exopod absent.

Abdomen (Fig. 7b) : pleopods on somites 1-5 rudimentary, biramous.

26

A. A. FINCHAM

Fig. 3 Zoea 3: (a) dorsal view; (b) lateral view; (c) mandibles; (d) maxilla 1; (e) maxilla 2. Bar
scales: a, b = 0-5 mm; c = 0-05 mm; d, e = 0-1 mm.

Telson (Fig. 8n): almost parallel sided; spine formula [3] 4 + 4 [3], some of small outer spines
occasionally missing, posterior margin weakly concave, length about one-third overall length
of telson.

ZOEA 6 (Figs 9, 10) 6-3 mm (6-1-6-6 mm)

Carapace (Fig. 9a): short plumose seta in angle of anterior dorso-medial spine, rostrum still
down turned to form small hook.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

27

Fig. 4 Zoea 3: (a) antenna 1; (b) antenna 2; (c) maxilliped 1; (d) maxilliped 2; (e) maxilliped 3;
(f) pereiopod 1; (g) pereiopod 2; (h) pereiopod 3; (i) pereiopod 4; (j) pereiopod 5; (k) telson.
Bar scale = 0-2 mm.

28

A. A. FINCH AM

Fig. 5 Zoea 4: (a) dorsal view; (b) lateral view; (c) antero-lateral teeth of carapace; (d) first dorso-
medial carapace spine; (e) mandibles; (f) maxilla 1; (g) maxilla 2; (h) antenna 1; (i) antenna 2.
Bar scales: a, b = 0-5 mm; c-e = 0-05 mm; f , g = 0-1 mm; h, i = 0-2 mm.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

29

Fig. 6 Zoea 4: (a) maxilliped 1 ; (b) maxilliped 2; (c) maxilliped 3 ; (d) pereiopod 1 ; (e) pereiopod 2;
(f) pereiopod 3; (g) pereiopod 4; (h) pereiopod 5; (i) telson. Bar scale = 0-2 mm.

30

A. A. FINCHAM

Fig. 7 Zoea 5: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles;
(f) maxilla 1 ;(g) maxilla 2. Bar scales : a, b = 0-5 mm;c, d = 0-2 mm;e = 0-05 mm;f, g = 0-1 mm.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

31

Fig. 8 Zoea 5 : (a) maxilliped 1 ; (b) maxilliped 2; (c) maxilliped 3 ; (d) pereiopod 1 ; (e) pereiopod 2;
(f) pereiopod 3; (g) pereiopod 4; (h) pereiopod 5; (i) pleopod 1; (j) pleopod 2; (k) pleopod 3;
(1) pleopod 4; (m) pleopod 5; (n) telson. Bar scale = 0-2 mm.

32 A. A. FINCHAM

Antenna 1 (Fig. 9c): internal flagellum of two segments.

Antenna 2 (Fig. 9d): endopodite of at least 9 segments.

Maxilliped 1 (Fig. lOa): exopodite with 3 (usually) proximo-lateral plumose setae.

Pereiopods 1, 2 (Figs lOd, e): endopodite with immovable finger of propodus produced forward
to almost length ofdactylus (excluding terminal setae).

Abdomen (Figs 9b, lOi-m): pleopods with rudimentary setae on margins of endopodite and
exopodite, endopodite of pleopods 2-5 with rudimentary appendix interna.

Telson (Fig. lOn): narrowing distally, posterior margin weakly convex and length just over a
quarter overall length of telson; small spines between large spines reduced in number to a central
pair.

ZOEA 7 (Figs 11-13) 6-5 mm (6-3-6-8 mm)

A few specimens have already metamorphosed to post larvae at this moult. Meristic characters

from these post larvae have not been included in Table 1 nor in the following description.

Carapace (Fig. 11 a): rostrum with weak distal hook or straight, occasionally a single ventral
tooth.

Antenna 1 (Fig. 12a): internal flagellum of 3 segments, external flagellum of 2 segments.

Antenna 2 (Fig. 12b): endopodite of at least 12 segments.

Maxilliped 1 (Fig. 12f): exopodite with 5 (usually) proximo-lateral plumose setae.

Pereiopods 1, 2 (Figs 13a, b): propodus 'palm 1 almost as wide as long (0-8 : 1).

Telson (Fig. 13k): no small spines between large spines.

POST LARVA 1 (Figs 14-17) 6-7 mm (6-3-7-1)

More than 50% of specimens have metamorphosed to post larvae at this moult. Meristic charac-
ters from the post larvae only have been included in Table 1.

Carapace (Figs 14a, b, 15b): rostrum usually with 6 dorsal and 2 ventral teeth, rostral tip
straight (see intermediate condition of rostrum Fig. 15a) supraorbital spines missing.

Antenna 1 (Fig. 15c): internal flagellum of 6-9 segments and external flagellum of 6 segments,
distal group of aesthetascs reduced to three.

Antenna 2 (Fig. 15d): endopodite multisegmented, several times longer than the scaphocerite.

Mandible (Figs 16a, b): divided into pars incisiva and pars molaris, lacinia mobilis no longer
present and palp (three-jointed in adult) not yet appeared.

Maxilla (Fig. 16c): up to 16 setae on basal endite.

Maxilla 2 (Fig. 16d): complete loss of setae on coxal endite and a reduction to 1 seta on endopo-
dite, increases in setae on basal endites.

Maxilliped 1 (Fig. 15e): large increase in setae on internal margin of basis (zoea 7: 7-8, post
larva 1 : 27-30).

Maxilliped 2 (Fig. 15f): endopodite with dactylus, propodus and merus flattened, exopodite
shortened and with reduced plumose setae.

Maxilliped 3 (Fig. 15g): endopodite dactylus shortened, exopodite reduced to less than half
the length of endopodite.

Pereicpods 1, 2 (Figs 16e, f): ischium, merus and carpus lengthened, exopodite reduced to
less than length of ischium of endopodite, plumose setae much reduced.

Pereiopod 3 (Fig. 16g): endopodite dactylus tapering evenly distally, other joints lengthened,
exopodite reduced, extending less than halfway along ischium of endopodite, plumose setae
much reduced.

Pereiopods 4, 5 (Figs 16h, i): endopodite dactylus tapering evenly distally, other joints
lengthened.

Pleopod 1 (Fig. 17a): ratio of endopodite to exopodite 1 : 3-5, endopodite bearing terminal
plumose setae, exopodite fringed with long plumose setae.

Pleopods 2-5 (Figs 17b-e): endopodite over half length of exopodite, both with long, marginal
plumose setae, endopodite with appendix interna bearing well-developed intero-distal coupling
hooks.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

33

Fig. 9 Zoea 6: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles; (f)
maxilla 1 ; (g) maxilla 2. Bar scales: a, b 0-5 mm; c, d = 0-2 mm; e = 0-05 mm; f, g = 0-1 mm.

34

A. A. FINCH AM

n

Fig. 10 Zoea 6: (a) maxilliped 1 ; (b) maxilliped 2; (c) maxilliped 3; (d) pereiopod 1 ; (e)pereiopod 2;
(f) pereiopod 3; (g) pereiopod 4; (h) pereiopod 5; (i) pleopod 1 ; (j) pleopod 2; (k) pleopod 3; (1)
pleopod 4; (m) pleopod 5; (n) telson. Bar scale = 0-2 mm.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

35

Fig. 11 Zoea 7: (a) dorsal view; (b) lateral view. Bar scale = 0-5 mm.

Abdomen (Fig. \4b): fifth abdominal somite truncated, point of stages 2-7 much reduced.

Telson (Fig. 17f): extremely narrow, posterior margin tapering to a point and bearing a pair of
small (outer) and large (inner) spines and medially a pair of plumose setae. Two pairs of stout
spines developed on lateral margins of telson (see intermediate condition found in specimens
with intermediate rostral form Fig. 17g).

Effect of antibiotic

The effect of antibiotic on the length of intermoult period on two series each of 108 larvae
kept at 18 C is shown in Fig. 17a. Larvae reared without antibiotic ( AB) moult in less than
5 days to stage 3 compared with larvae reared with antibiotic ( + AB) which take an average of

36

A. A. FINCHAM

f

Fig. 12 Zoea 7: (a) antenna 1 ; (b) antenna 2; (c) mandibles; (d) maxilla 1 ; (e) maxilla 2; (f) maxil-
liped 1 ; (g) maxilliped 2; (h) maxilliped 3. Bar scales: a, b, f-h = 0-2 mm; c 0-05 mm; d, e =
0-1 mm.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

37

Fig. 13 Zoea 7: (a) pereiopod 1 ; (b) pereiopod 2; (c) pereiopod 3; (d) pereiopod 4; (e) pereiopod 5;
(f) pleopod 1; (g) pleopod 2; (h) pleopod 3; (i) pleopod 4; (j) pleopod 5; (k) telson. Bar scale
= 0-2 mm.

38

A. A. FINCHAM

Fig. 14 Post larva 1 : (a) dorsal view; (b) lateral view. Bar scale = 0-5 mm.

nearly 7 days. This advantage is maintained throughout the remaining period of larval develop-
ment. Post larval stage is reached by at least 50% of AB larvae at moult 7-8; metamorphosis
is delayed until moult 9-10 in +AB larvae.

Fig. 17b shows the growth of the two series of larvae. Faster growth is shown by +AB larvae
until stage 4 when AB larvae overtake them in size. This increase is maintained throughout
larval development with the two series coming close at moult 8. From this moult the difference
in growth widens markedly, AB larvae further increasing their average size over + AB larvae.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

39

Fig. 15 Intermediate stage: (a) rostrum. Post larva 1: (b) rostrum; (c) antenna 1; (d) antenna 2;
(e) maxilliped 1; (f) maxilliped 2; (g) maxilliped 3. Bar scale = 0-2 mm.

40

A. A. FINCHAM

Fig. 16 Post larva 1 : (a) mandibles and musculature; (b) mandible with pars incisiva (above) and
pars molaris (below); (c) maxilla 1 ; (d) maxilla 2; (e) pereiopod 1 : (f) pereiopod 2; (g) pereiopod 3;
(h) pereiopod 4; (i) pereiopod 5. a, e-i 0-2 mm; c, d =0-1 mm; b = 0-05 mm.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

41

f

Fig. 17. Post larva 1: (a) pleopod 1; (b) pleopod 2; (c) pleopod 3; (d) pleopod 4; (e) pleopod 5;
(f) telson. Intermediate stage: (g) telson. Bar scale = 0-2 mm.

42

A. A. FINCHAM

Moults

10i

8-

6-

4-

<m

10

Length
(mm)

8-

6-

4-

2-

b

PL

/-^*x- 1

* x- PL

15 20 25

Days

PL

,o

^ H--' U i

" i

o cr

PL

6

i i
8

10

30

Moults

Fig. 18 Effect on (a) intermoult period, (b) growth rate of larvae reared with (o O) and

without ( ) antibiotic. Figures averaged from 10 individuals at each stage.

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS

43

genotype

Food, hormones,
light, salinity,
temperature

Juveniles

Hormones etc

i>

Adults

Fig. 19 Larval development of decapod Natantia showing the relationship between genetic input,
epigenesis and the internal and external environmental factors which affect the semi-independent
developmental processes of moulting, growth and morphogenesis. This results in larval phenotpyic
variation in morphometric and meristic characters and in the number of stages. The system
stabilizes in the post larva after metamorphosis. Bar scale = 0-5 mm.

Discussion

Effect of antibiotic on morphogenesis

The results of detailed comparisons of moults and larvae at each stage from specimens reared
with ( + AB) and without ( AB) antibiotic are shown in Table 1. There was no appreciable
effect on normal phenotypic variation. However, there was a noticeable effect on the average
intermoult period, which in AB larvae was 2-9 days while in +AB larvae it was 3-5 days
(Fig. 18a). This is the reverse of the situation in the brachyuran Hyas araneus (Linnaeus) in which
the intermoult period was shortened at both 10 and 15 C at the same concentration of anti-
biotics used in the present work (Christiansen, 1971). Christiansen also found that mortality
of laboratory reared larvae was lower when antibiotics were used. This effect was not found in
P daemon (P daemon) longirostris; 92% survival was recorded in AB larvae but only 71 % in
+ AB larvae. In two experiments made with two replicate populations of 100 and 25 larvae of
Palaemon serratus, Wickins (1972) found that stocking densities did affect total percentage
survival but that the addition of antibiotics had little effect on survival (100 larvae/litre: AB
59% survival, +AB 58%; 25 larvae/litre: -AB 88%, +AB 86% from Wickins, 1972). The
moult to post larvae in P. (P.} longirostris in the present work was delayed by about 10 days

44 A. A. FINCHAM

after two further moults had occurred in +AB larvae. Wickins also found that during the first
5-7 days of metamorphosis in each trial, more (average 28 %) larvae had metamorphosed at
any given time in the cultures without antibiotics. It was considered that either the presence of
antibiotics or their bactericidal action was responsible for retarding the development of the
larvae.

In the present work it was found that the effect of antibiotics on growth changed during the
course of larval development (Fig. 18b). The initial increase in size by +AB larvae over AB
larvae may be a direct result of bacterial infection in the cultures. In the open sea the numbers
of bacteria are low and it seems likely that this is correlated with a lack of solid surfaces on
which to settle (Walne, 1958). In the compartmented rearing trays used in this study there is a
high surface to volume ratio and bacteria are introduced into the trays both on the larvae them-
selves and on the Artemia nauplii used as food. It seems likely, therefore, that the unnatural
stress introduced by the build-up of bacteria is successfully counteracted initially by the use of
antibiotics and actually enhances growth compared with the AB larvae. That this increased
growth is not maintained may be due to a number of factors but one which might repay further
investigation would be an assessment of the effect on the populations of gut flora. This would
adversely affect food absorption and consequently growth. At the eighth moult both the +AB
and AB larvae are similar in size. This apparent slowing of growth in AB larvae occurs
after the majority of specimens have moulted to post larvae. Much internal reorganization takes
place in connection with the change in the swimming mode. The upside down, tail first, thoracic
exopodite swimming and planktonic mode of life is replaced by a basically benthic mode, and
when swimming occurs the post larva is the right way up, head first and uses abdominal pleopods.
The internal changes have been monitored by Campillo et al. (1975) who found that during
larval development the growth process in Palaemon serratus (Pennant) alternates between periods
of hyperplasy (stages 1-4, 5-7) and hypertrophy (stages 4-5, 7 to post larva). Prior to the post
larval moult any growth is by an increase in cell size rather than from an increase in cell number.
Regnault (1971) found that free amino acids had two peaks during the larval development of
Crangon septemspinosa Say at stage 3 and last larval stage. These biochemical and cellular
changes tie in with two major periods of morphogenesis, i.e. the development of uropods and
the loss of the thoracic swimming exopods and their replacement with abdominal swimming
pleopods. Morphogenesis of the mouthparts (compare Figs 12c, d, e and 16a, b, c, d), necessitated
by the change in diet, accompany these other changes in the mode of propulsion at the post
larval moult. Hence at this moult there is unlikely to be much spare energy for growth. Once
this first post larval moult is passed, growth rate increases and is faster in these AB individuals
than at any time during the development of either +AB or AB larvae.

The process of development

Moulting, growth and morphogenesis in the early stages of natant development are synchronous.
For the first few moults phenotypic variation is small or even absent, for example the number
of plumose setae on the exopodite of antenna 2 in zoea 1 is 9, while in maxilla 2 the figures are
5 and 7 for zoeae 1 and 2 respectively (Table 1). Both size and morphology of these early stages
are largely predictable. However, in later larval development there is some independence of
the three development processes (Fig. 19). This results in variable numbers of stages and larvae
of intermediate stage morphology in both plankton and laboratory reared material. This
variability has been recorded in the Palaemonidae by Faxon (1879), Gurney (1924#), Gurney &
Lebour (1941), Broad (19570), Tsurnamal (1963), Little (1969) and Fincham (1977, 1979).

Knowlton (1974) in his hierarchical representation of energy flow in caridean larval develop-
ment gives high priority to general maintenance and survival, moulting is placed next followed
by the synthetic processes of growth, morphogenesis and regeneration which are given equal
weight. The genetic content of the fertilized egg is fixed and the embryos and larvae develop
through a series of 'causal relations' (Waddington, 1953). This epigenesis may be influenced
by a number of factors including food (Broad, 19576; Reeve, 1969; Knowlton, 1974), hormones
(Hubschman, 1963; Costlow, 1968; Little, 1969), light (Knowlton, 1974), salinity and tempera-
ture (Sandifer, 1973; Knowlton, 1974). The virtual independence of the development processes

LARVAL DEVELOPMENT OF PALAEMON LONGIROSTRIS 45

is shown diagrammatically in Fig. 19. Larvae of similar moulting history may be quite different
in size within certain limits (see size ranges in section on description of larval stages) and in general
morphology. In contrast, the moult is a definitive event and this is represented on the left of the
diagram by single lines at the time of each ecdysis. In the centre of the diagram growth is shown
with a limited range of values at a given moult. The right of the diagram is evenly divided after
stage 3 indicating the continuous process of morphogenesis, 'arrested' at intervals by the hard
exoskeleton.

This model for prawn and shrimp development does not apply to all decapods. Knowlton
(1974) states that 'Instances of variability are less frequently encountered in Macrura and
Anomura, rarely in Brachyura'. The short larval phase of these groups produces a fixed number
of moults which are usually morphologically distinct, comparable with the early stages in natant
development. Knowlton further speculated that, in general, the degree of larval variation is an
index of primitiveness. In the past the phylogeny of the Euphausiacea and Decapoda has been
based largely on adult characters but in the light of more larval data it will be possible to re-
examine existing classifications and test Knowlton's hypothesis.

Acknowledgments

It is my pleasure to thank the Central Electricity Generating Board for permission to collect
on their property, Mr D. Rickard of the Thames Water Authority for his assistance at the
West Thurrock site and Miss Ann Gurney for efficiently running the larval rearing unit.

I am most grateful also to the staff at MAFF Fisheries Laboratory, Lowestoft, for allowing
me to examine unique collections of inshore plankton samples and in particular I wish to thank
Dr D. Harding and Messrs J. Riley and J. Nicholls.

References

Broad, A. C. 1957 a. Larval development of Palaemonetes pugio Holthuis. Biol. Bull. mar. biol. Lab.

Woods Hole 112 (2) : 144-161.
19576. The relationship between diet and larval development of Palaemonetes. Biol. Bull. mar. biol.

Lab. Woods Hole 112 (2) : 162-170.
Campillo, A., Regnault, M. & Luquet, P. 1975. Evolution de acides nucleiques au cours du developpe-

ment larvaire de la crevette rose Palaemon serratus (Pennant). Rev. Trav. Inst. Peek, marit. 39 (3) :

333-342.
Christiansen, M. E. 1971. Larval development of Hyas araneus (Linnaeus) with and without antibiotics

(Decapoda, Brachyura, Majidae). Crustaceana 21 (3) : 307-315.

Costlow, J. D. 1968. Metamorphosis in crustaceans: 3-41. In: Etkin, W. & Gilbert, L. I. (eds), Meta-
morphosis: a problem in developmental biology. North Holland Publishing Co. Amsterdam.
Cviic, V. The bacterial and bacteriostatical action of antibiotics on marine bacteria. 1. Penicillin and

streptomycin. Acta adriat. 5 (7) : 135-166.
Czerniavsky, W. 1884. Crustacea Decapoda Pontica littoralia. Materialia ad Zoographiam Ponticam

comparatam. II. Trans. Soc. Univ. Kharkow 13 (suppl.) : 1-268.
Edwards, H. Milne 1837. Histoire naturelle des Crustaces. II. Paris 531 pp.
Faxon, W. A. 1879. On the development of Palaemonetes vulgaris. Bull. Mus. comp. Zool. Harv. 5 :

303-330.
Fincham, A. A. 1977. Larval development of British prawns and shrimps (Crustacea : Decapoda :

Natantia). 1. Laboratory methods and a review of Palaemon (Palaeander) elegans Rathke, 1837.

Bull. Br. Mus. not. Hist. (Zool.) 31 (1) : 1-28.
1979. Larval development of British prawns and shrimps (Crustacea : Decapoda : Natantia).

2. Palaemonetes (Palaemonetes) varians (Leach, 1814) and morphological variation. Bull. Br. Mus.

nat. Hist. (Zool.) 35(2) : 163-182.

& Williamson, D. I. 1978. Crustacea, Decapoda: Larvae. VI Caridea, Families Palaemonidae and

Processidae. Fich. Ident. Zooplancton, 159/160 : 8 pp.
Gurney, R. 1923. Some notes on Leander longirostris M. Edwards, and other British prawns. Proc. zooL

Soc. Lond. 97-123.
1924fl. The larval development of some British prawns (Palaemonidae) - 1. Palaemonetes varians.

Proc. zool. Soc. Lond. 297-328.

46 A. A. FINCH AM

19246. The larval development of some British prawns (Palaemonidae). II. Leander longirostris and

Leander squilla. Proc. zool. Soc. Lond. 961-982.
& Lebour, M. V. 1941. On the larvae of certain Crustacea Macrura, mainly from Bermuda. /. Linn.

Soc. Lond., Zool. 41 : 89-181.
Heller, C. 1863. Die Crustaceen des siidlichen Europa. Crustacea Podophthalmia. Wilhelm Braumuller,

Wien. x, 336 pp.
Holthuis, L. B. 1950. The Palaemonidae collected by the Siboga and Snellius expeditions with remarks

on other species. 1. Subfamily Palaemoninae. Siboga Exped. 39a : 1-268.
Hubschman, J. H. 1963. Development and function of neurosecretory sites in the eyestalks of larval

Palaemonetes (Decapoda : Natantia). Biol. Bull. mar. biol. Lab. Woods Hole 125 : 96-113.
Knowlton, R. E. 1974. Larval developmental processes and controlling factors in decapod Crustacea,

with emphasis on Caridea. Thalassia jugosl. 10 (1/2) : 138-158.
Korschelt, E. 1944. Ontogenie der Decapoden. Bronns Kl. Ordn. Tierreichs 5 (1) Bk 7, Decapoda No. 6 :

671-861.
Little, G. 1969. The larval development of the shrimp, Palaemon macrodactylus Rathbun, reared in

the laboratory, and the effect of eyestalk extirpation on development. Crustaceana 17 : 69-87.
Man, J. G. De 1915. On some European species of the genus Leander Desm., also a contribution to the

fauna of Dutch waters. Tijdschr. ned. dierk. Vereen. Ser. 2, XIV : 115-179.
1923. Leander longirostris (H. M.-Edw.) var. robusta nov. van, the common prawn of the estuary

of the Meuse and of the Hollandsch Diep. Tijdschr. ned. dierk. Vereen. Ser. 2, XIX (1) : 1-9.
Reeve, M. R. 1969. Growth, metamorphosis and energy conversion in the larvae of the prawn, Palaemon

serratus. J. mar. biol. Ass. U.K. 49 : 77-96.
Regnault, M. 1971. Acides amines libres chez les larves de Crangon septemspinosa (Caridea). Variation de

leur taux de 1'eclosion a la metamorphose. Leur role au cours du developpement et leur importance

dans la nutrition. Mar. Biol. Berlin 11(1): 35-44.

Sandifer, P. A. 1973. Effects of temperature and salinity on larval development of grass shrimp, Palaemon-
etes vulgaris (Decapoda, Caridea). Fishery Bull. natn. ocean, atmos. Adm. 71 : 115-123.
Tsurnamal, M. 1963. Larval development of the prawn Palaemon elegans Rathke (Crustacea Decapoda)

from the coast of Israel. Israel J. Zool. 12 (1-4) : 117-141.

Waddington, C. H. 1953. Epigenetics and evolution. Symp. Soc. exp. Biol. 7 : 186-199.
Walne, P. R. 1958. The importance of bacteria in laboratory experiments on rearing the larvae of Ostrea

edulis (L.). /. mar. biol. Ass. U.K. 37 : 415-425.
Wickins, J. F. 1972. Developments in the laboratory culture of the common prawn Palaemon serratus

(Pennant). Fishery Invest.. Lond., Ser. 2, 27 (4) : 23 pp.

Manuscript accepted for publication 20 November 1978

The larval development of the spider crab Rochinia
carpenteri (Thomson) [Oxyrhyncha : Majidae] with a
review of majid subfamilial larval features

R. W. Ingle

Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD

Synopsis

Two zoeal and a megalopal stage of the N.E. Atlantic deepwater spider crab Rochinia carpenteri are
described from laboratory-reared material. Subfamilial larval features of the Majidae are reviewed
and the larval affinities of R. carpenteri are discussed.

Introduction

The deepwater spider crab Rochinia carpenteri (Thomson) has been recorded from S. Iceland
to N.W. Africa (Christiansen, 1969 : 123) and in depths from about 180 to 1258m (Hansen,
1908 : 12) on sandy chalk mud (Thomson, 1874 : 176).

In October 1977, five ovigerous crabs were trawled from the Rockall Trough/Hebridean
Terrace region of the N.E. Atlantic ocean and transported to the British Museum (Natural
History). The larvae hatched by two of these females were successfully reared to the megalopal
stage and provide material for the first description of the larval stages of this genus and species.

Materials and methods

The ovigerous crabs were trawled in the Rockall Trough/Hebridean Terrace region of the N.E.
Atlantic ocean in the following positions: (a) 5623-l'N: 0918-2'W, 1010-1030 m, Agassiz
trawl towed 30 min on 20.10.1977; (b) 5631-7'N: 0913-2' W-5629-4' N: 0911-3'W, 780-
710 m, single warp trawl towed 2 h 28 min on 22.10.1977.

Larvae were reared using methods described by Rice & Ingle (1975 : 104) and Ingle & Clark
1977). Material was fixed in Steedman's solution (Steedman, 1976 : 148) and later transferred
to 70% IMS. Drawings and measurements were made with the aid of a camera lucida. Measure-
ments given under Descriptions are: (a) distance between tips of dorsal and rostral spines (T.T.);
(b) carapace length, from between the eyes to the posterio-lateral carapace margin (C.L.). The
material has been registered as BM(NH) 1978: 282-3.

Results

Hatching occurred between 16 and 20 January 1978. An average of 36 days elapsed between
hatching and the appearance of megalops. Only a small number of second-stage zoeae successfully
moulted to the megalopal stage.

Descriptions

First zoea

DIMENSIONS: T.T. 2-3-3-5 mm; C.L. 1-0-1-2 mm.

Bull. Br. Mus. not. Hist. (Zool.) 37 (1) : 47-66 Issued 29 November 1979

47

48

R. W. INGLE

Fig. 1 Rochinia carpenteri: (a)-(c) zoea I; (d)-(f) zoea II; (a), (d) lateral aspect, scale A = 0-5 mm;
(b), (e) antennule, scale B = 0-1 mm; (c), (f) antenna, scale C = 0-1 mm; (g), (h) megalopa,
antennule & antenna respectively, scale D = 0-1 mm.

LARVAL DEVELOPMENT OF R. CARPENTERI

49

Rochinia carpenteri: ventral aspects of: mandible (a) zoea I, (b) zoea II, (c) megalopa;
maxillule (d) zoea I, (e) zoea II, (f) megalopa; scale 0-1 mm.

CARAPACE (Fig. la): Surface punctate, dorsal and rostral spines well developed, slightly curved,
narrowing distally; lateral spines well developed. A prominent, broad, median posterio-dorsal
tubercule and a pair of posteriorly placed setae; dorso-median elevation of carapace well de-
veloped; posterio-lateral margin with a few long spinules arising from inner surface.

EYES: Partly fused to carapace.

50 R. W. INGLE

ANTENNULE (Fig. Ib): Unsegmented, with 2 terminal aesthestascs and one long seta.

ANTENNA (Fig. Ic): Distal half of spinous process spinulate; exopod reaching well into distal
half of spinous process, with 3 terminal spines, longest reaching to apex (or slightly beyond)
of spinous process; endopod slightly more than half length of exopod.

MANDIBLE (Fig. 2a): Incisor and molar processes well developed, palp absent.

MAXILLULE (Fig. 2d): Endopod 2-segmented, with 1, 6 setae respectively; distal margin of basal
endite with 5 setose spines and 1 seta, inner margin with 1 seta; coxal endite with 7 setae.

MAXILLA (Figs 3a, b) : Endopod with large outer and small indistinct inner lobe with 3 -f 3 setae ;
basal endite with large outer and smaller inner lobe with 4 + 5 setae; coxal endite bilobed with
4 + 5 setae; scaphognathite with 12 marginal setae.

FIRST MAXILLIPED (Fig. 4a): Basis with 10 setae (arranged 2, 2, 3, 3); endopod 5-segmented
with 3, 2, 1, 2, 4+ 1 setae; exopod incipiently 2-segmented (in some specimens) with 4 terminal
natatory setae.

SECOND MAXILLIPED (Fig. 4c): Basis with 3 setae; endopod 3-segmented, 1st and 2nd segments
indistinct, in some specimens, with 0, 1 , 4+1 setae ; exopod incipiently 2-segmented, with 4
terminal natatory setae.

THIRD MAXILLIPED AND PEREIOPODS: Represented as incipient buds.

ABDOMEN (Figs 5a, b): 5-segmented + telson; second segment with a pair of outwardly directed
broad-based, curved, dorso-lateral processes; posterio-lateral processes on segments 3-5 acute,
decreasing slightly in size on successive segments, those on fifth angled downwards. A pair of
minute setae near posterio-dorsal margin of each segment. Telson forks long, curved and spinu-
late, each with one long lateral spine; inner posterior margin of telson straight or slightly convex,
with 3 pairs of long, plumed setae.

Second zoea

DIMENSIONS: T.T. 3-0-3-2 mm; C.L. 1-3-1-5 mm.

CARAPACE (Fig. Id): Dorsal spine with few setae near base; posterio-dorsal tubercule larger

than in first stage.

EYES: Stalked.

ANTENNULE (Fig. le): With 5 aesthetascs and one short seta.

ANTENNA (Fig. If): Longest of 3 spines on exopod reaching well beyond distal extremity of
spinous process; endopod much more than half length of exopod.

MANDIBLE (Fig. 2b): Molar process longer than in first stage; unsegmented mandibular palp
developed.

MAXILLULE (Fig. 2e): Outer margin of basal endite with prominent plumose seta, distal margin
with 6 spines and 2 setae, inner margin with 2 setae; coxal endite with 6-7 setae.

MAXILLA (Fig. 3c): Endopod with 4 + 3 setae; basal endite with 8 + 5 setae; scaphognathite with
21-23 setae.

FIRST MAXILLIPED (Fig. 4b): Exopod with 6 terminal natatory setae.

SECOND MAXILLIPED (Fig. 4d): Exopod with 6 terminal natatory setae.

THIRD MAXILLIPED AND PEREIOPODS: Represented as incompletely segmented buds.

ABDOMEN (Figs 5c, d): 6-segmented, 6th segment with acute posterio-ventral margin. Additional
pairs of minute dorsal setae on segments 3-4 and sometimes on 5; pleopod buds well developed
on segments 2-5.

Megalopa

DIMENSIONS: C.L. 2-3-2-5 mm.

LARVAL DEVELOPMENT OF K. CARPENTERI

51

Fig. 3 Rochinia carpenteri: maxilla (a) zoea I, (b) terminal portion from another specimen, (c)
zoea II, (d) megalopa; (e) telson and left uropod and (f) 1st pleopod of megalopa; scale's =
0-1 mm.

52

R. W. INGLE

Fig. 4 Rochinia carpenteri: 1st maxillipeds (a) zoea I, (b) zoea II, (f) megalopa; 2nd maxillipeds
(c) zoea I, (d) zoea II, (e) megalopa; scales = 0-1 mm.

LARVAL DEVELOPMENT OF R. CARPENTERI 53

CARAPACE (Figs 7c, d): Longer than broad, anteriorly setose, rostrum long; hepatic regions
inflated, each protogastric region with a cristate tubercle; a longitudinal carina on each epi-
branchial region ending posteriorly in a cristate process; a pair of carinae on metagastric region;
each epibranchial/mesobranchial region with an obliquely placed carinae; a very long, stout,
median cardiac spine and a small median marginal intestinal tubercle.

EYES : Large, with well-developed cornea.

ANTENNULE (Fig. Ih): Peduncle 3-segmented, 2nd segment with two distally placed setae;
exopod with 4-5 incomplete segments, second with 6 aesthetascs, 3rd with 4, 4th with 3, 5th
with one aesthetasc and one seta; endopod 2-segmented, distal segment with 3 setae.

ANTENNA (Fig. Ig): Peduncle 3-segmented, first segment with outer lower distal margin produced
as acute process, inner upper distal margin with sub-acute expansion; second and third peduncle
segments each with 2 distal dorsal median setae; flagellum 4-segmented, third and fourth seg-
ments each with 3 setae.

MANDIBLE (Fig. 2c): Incisor process expanded as broad plate with sharp margins, molar process
reduced; mandibular palp large, 2-segmented, proximal segment longer than distal, distal with
5 plumose setae.

MAXILLULE (Fig. 2f): Endopod reduced, segmentation indistinct, with 4 setae; distal margin of
basal endite with 5-6 setosed spines and 4 setae; coxal endite with 10 setae.

MAXILLA (Fig. 3d): Endopod reduced, basal endite with 6 + 6 setae, coxal endite with 3-4 + 9
setae.

FIRST MAXILLIPED (Fig. 4f): Coxa with 6, basis with 10 setae arranged longitudinally. Exopod
2-segmented, proximal with one and distal segment with 4 terminal setae; endopod reduced,
unsegmented; epipod well developed, with 6 long setae.

SECOND MAXILLIPED (Fig. 4e) : Exopod 2-segmented, distal segment about half length of proximal
and with 4 terminal setae; endopod 4-segmented, 2nd segment (carpus) with one and third
(propodus) with 4 setae, 4th segment (dactylus) with 4 spines and 3 setae; epipod small.

THIRD MAXILLIPED (Fig. 6a): Coxa/basis with a transverse row of setae. Endopod 5-segmented,
inner margin of ischium with irregular serrations, outer surface with 13-15 short setae; outer
margin of merus with 2, ventral with 4-5 setae, carpus and propodus with 5 distally placed
setae respectively, dactylus with 4 terminal setae. Exopod 2-segmented, distal segment about
half length of proximal and with 6-7 terminal setae. Epipod with 5 long setae.

PEREIOPODS (Figs 6b-d, 7a, b): Chelipeds moderately stout, setosed, inner propodal margin with
several cristate teeth, inner dactylar margin cristate. Pereipods 2-5 moderately stout, setosed,
dactylus terminally thin and curved, inner margin with 3 spines; pereiopods 2-4 with a coxal
and ischial spine.

ABDOMEN (Fig. 5e): 6-segmented + telson; with 1-2 pairs of posterio-dorsal setae; posterio-
lateral margins of segments rounded, first often with 1-2 setae. Telson broader than long (Fig.
3e), with a pair of dorsal and ventral median setae. Five pairs of pleopods, exopod of first (Fig.
3f) to fourth with 14-16 marginal plumose setae; endopod with 2 coupling hooks; exopod of
fifth pleopod (uropod) with 5 long plumose setae.

Review of subfamilial larval features of the Majidae

Larval relationships of genera and species in the Majidae have been discussed by Cano (1893),
Lebour (1928, 1931), Aikawa (1937), Bourdillon-Casanova (1960), Yang (1968), Kurata (1969),
Italo Campodonico & Leonardo Guzman (1972) and Yang (1976); some studies still await
publication (Yang, 1967). There would seem little point in attempting further phylogenetic
evaluations until descriptions of the larvae of many more majids become available, particularly
of species belonging to the subfamily Majinae, of which only five species have been described,
and of the Ophthalmiinae, of which the larval stages of only one species is known. Despite this

54

R. W. INGLE

Fig. 5 Rochinia carpenteri: abdomens, lateral and dorsal aspects (a), (b) zoea I, (c), (d) zoea II
(lower inset of another specimen), (e) megalopa lateral aspect; scales = 0-1 mm.

LARVAL DEVELOPMENT OF R. CARPENTER!

55

Fig. 6 Rochinia carpenteri: megalopa (a) 3rd maxilliped, (b) cheliped, (c) 2nd pereiopod, (d)

3rd pereiopod; scale = 0-1 mm.

56

R. W. INGLE

Fig. 7 Rochinia carpenteri: megalopa (a) 4th pereiopod, (b) 5th pereiopod, (c) dorsal and (d)
lateral aspects of carapace; scales, upper - 0-1 mm, lower = 0-5 mm.

LARVAL DEVELOPMENT OF R. CARPENTERI 57

incompleteness it would seem useful to attempt an evaluation of larval characteristics described
to date for the Majidae to discover the degree of larval homologies that exist within the currently
accepted subfamilial (see Griffin, 1966) classification of the adults. In this appraisal only features
are considered that have been sufficiently well illustrated or described for the majority of majid
zoeae and megalopae. Some of these (i.e. setation on basal segment of the maxillipeds of zoeae
and the presence or absence of pereiopod coxal spines of the megalopae) as well as other features
that are not listed here (i.e. setal armature of the maxillule and of the first and second maxilliped
endopod of the zoeae) may eventually prove to be of considerable importance, when used in
combination with other features, for separating the subfamilies. In general, these above-mentioned
features are poorly documented for the majority of previously described majid larvae and only
a limited amount of material has been available for verifying some of these features during the
course of this present study. A larval bibliography of the Majidae, resulting from this literature
appraisal, is given on pp. 59-63.

Subfamily OREGONIINAE

ZOEA: Carapace lateral spines present; dorsal and rostral spines very long; more than one spine 1
on each telson fork; dorso-lateral processes on 2nd and 3rd abdominal segments; posterio-
lateral abdominal processes on 3rd-5th segments often very long; basal segment of 2nd maxil-
liped with 4-5 setae; antennal exopod with terminal setae. 2

MEGALOPA: Rostrum present, not strongly deflected ventrally; submedian spines (carapace
spines behind eyes) present; a single or a pair of prominent cardiac spines; basal segment of
2nd pereiopod with a spine; uropods present.

GENERA: Chionoecetes, Hyas, Oregonia.

Subfamily ACANTHONYCHINAE

ZOEA: Carapace lateral spines absent; dorsal and rostral spines very short or absent; only one
spine 1 on each telson fork ; dorso-lateral processes on 2nd, rarely on 3rd (? Menaethius) abdominal
segment; posterio-lateral abdominal processes on 3rd-5th segments short or absent; basal
segment of 2nd maxilliped with 2-3 marginal setae; antennal exopod with sub-terminal 3 setae.

MEGALOPA: Rostrum present, strongly deflected ventrally or reduced in size; submedian spines
absent; cardiac spine sometimes absent or reduced; basal segment of 2nd pereiopod? without
a spine; uropods present.

GENERA: Acanthonyx, Epialtus, Huenia, Menaethius, Pugettia, Taliepus.

Subfamily INACHINAE

ZOEA (group I): Carapace lateral spines absent; dorsal spine of moderate length, sometimes
long; rostral spine absent, rarely present (Naxid); only one spine 1 on each telson fork; dorso-
lateral processes on 2nd, rarely on 3rd (Stenorhynchus) abdominal segment; posterio-lateral
abdominal processes on 3rd-5th segments sometimes long; basal segment of 2nd maxilliped
with not more than 3 marginal setae; antennal exopod with sub-terminal 3 setae.
MEGALOPA (group I): Rostrum absent (present only in Stenorhynchus and strongly deflected
ventrally); submedian spines often well developed (except in Anasimus); cardiac spines small
or absent (except in Stenorhynchus}; some abdominal segments with dorsal spinules or spinules
(except in Stenorhynchus); basal segment of 2nd pereiopod ? with or without a spine; uropods
absent or vestigial (except in Stenorhynchus).
GENERA : A chaeus, Anasimus, Inachus, Macropodia, Naxia, Stenorhynchus.

1 Distinction is not made here between lateral and dorsal spines or spinules as this difference is not always clearly
shown in some illustrations.

2 This implies that the exopod terminates in 2 or 3 spines or setae of equal thickness but not necessarily of equal

3 This implies that the exopod terminates in a short movable or non-movable stout spine from the base of which
arise one or two setae, usually shorter than the length of the spine.

58 R. W. INGLE

ZOEA (group II): Carapace lateral spines present; dorsal spine of moderate length; rostral spine
of moderate length often short; more than one spine 1 on each telson fork; dorso-lateral processes
on 2nd and 3rd abdominal segments; posterio-lateral abdominal processes on 3rd-5th segments
sometimes short ; basal segment of 2nd maxilliped with not more than 3 marginal setae ; antennal
exopod with terminal 2 setae.

MEGALOPA (group II): Rostrum present, strongly deflected ventrally or straight; submedian
spines developed or reduced; cardiac spine short or long; abdominal segments without dorsal
spines; basal segment of 2nd pereiopod with or ? without spines; uropods present.

GENERA: Eurypodius, Macrocheira, Pleistacantha, Camposcia.

Subfamily PISINAE

ZOEA: Carapace lateral spines absent; dorsal spine of moderate length; rostral spine of moderate
length often short; only one spine 1 on each telson fork; dorso-lateral processes on 2nd abdominal
segment, rarely on 3rd (Herbstid); posterio-lateral processes on 3rd-5th abdominal segments
often short; basal segment of 2nd maxilliped with not more than 3 marginal setae; antennal
exopod with sub-terminal 3 setae.

MEGALOPA: Rostrum present often straight; submedian spines absent; cardiac spine absent
(except in Libinia); basal segment of 2nd pereiopod without a spine; uropods present.

GENERA : Eurynome, Herbstia, Hyastenus, Libidoclaea, Libinia, Lissa, Naxioides, Pisa, Pisoides,

Sub family MAJINAE

ZOEA (group I): Carapace lateral spines present; dorsal spine often well developed and usually
of moderate length; rostral spine prominent, of moderate length; more than one spine 1 on each
telson fork; dorso-lateral processes on 2nd and 3rd abdominal segments (? absent on 3rd in
Maja verrucosa); posterio-lateral processes on 3rd-5th abdominal segments prominent sometimes
long; basal segment of 2nd maxilliped with not more than 3 marginal setae; antennal exopod
with terminal 2 setae.

MEGALOPA (group I): Rostrum present and moderately well developed, slightly deflected ven-
trally; submedian spines or processes present; cardiac spine not prominent; basal segment
of 2nd pereiopod with spines; uropods present.

GENERA: Maja, Schizophrys.

ZOEA (group II): Carapace lateral spines absent; dorsal spine sometimes reduced or absent;
rostral spine sometimes reduced; more than one spine 1 on each telson fork; dorso-lateral
processes only on 2nd abdominal segment; posterio-lateral processes on 3rd-5th abdominal
segments sometimes short; basal segment of 2nd maxilliped ? with not more than 3 marginal
setae; antennal exopod with terminal 2 setae.

MEGALOPA (group II): Rostrum present, sometimes small; submedian spines or processes
sometimes absent; cardiac spine small; basal segment of 2nd pereiopod with a spine; uropods
present.

GENERA: Leoptomithrax, Acanthophrys.

Subfamily OPHTHALMIINAE

ZOEA: Carapace lateral spines present; dorsal spine very short; rostral spine very short; more
than one spine 1 on each telson fork; abdominal dorso-lateral processes absent; ? 3 prominent
dorso-lateral setae on 1st abdominal segment; posterio-lateral processes on 3rd-5th abdominal
segments absent; basal segment of 2nd maxilliped ? without setae; antennal exopod with
terminal 2 setae.

MEGALOPA: Rostrum present, well developed, straight; submedian spines or processes absent;
cardiac spine absent; basal segment of 2nd pereiopod ? without spines.

GENERA: Stilbognathus

LARVAL DEVELOPMENT OF R. CARPENTERI 59

Subfamily MITHRACINAE

ZOEA (group I): Carapace lateral spines absent; dorsal spine of moderate length; rostral spine
short; only one spine 1 on each telson fork; dorso-lateral process on 2nd abdominal segment;
posterio-lateral processes on 3rd-5th abdominal segments sometimes short; basal segment of
2nd maxilliped ? without setae; antennal exopod with sub-terminal 3 setae (? except in Mithrax).
MEGALOPA (group I): Rostrum present, deflected slightly ventrally; submedian processes present;
cardiac spine reduced; basal segment of 2nd pereiopod ? without spines; uropod present.
GENERA: Microphrys, Mithrax, Tiarinia.

ZOEA (group II): Carapace lateral spines present; dorsal spine absent; rostral spine of moderate
length; more than one spine 1 on each telson fork; dorso-lateral processes on 2nd-3rd abdominal
segments; posterio-lateral abdominal processes on 3rd-5th segments short; basal segment of
2nd maxilliped ? without setae; antennal exopod with sub-terminal 3 setae.
MEGALOPA (group II): Rostrum present, deflected slightly ventrally; submedian processes
absent; cardiac spine absent or reduced; basal segment of 2nd pereiopod with a spine; uropods
present.

GENUS: Micippa.

The subfamilial larval features listed above reveal considerable degrees of apparent phylogenetic
homology within three of the six subfamilies (i.e. Oregoniinae, Acanthonychinae and Pisinae)
of the Majidae. This homology is less clear within the remaining subfamilies except, perhaps, the
Inachinae in which two groups can be recognized that correspond to the divisions of this sub-
family proposed by Balss (1929), the Camposcioidea (group I) and the Macrocheiroidea (group
II). These divisions were partly (see Garth, 1958) and totally (see Griffin, 1966) rejected by sub-
sequent workers because of the difficulties in interpreting homologies of orbital spines in genera
and species assigned to one or the other group. Similar relationships of larvae attributed to the
subfamilies Majinae and Mithracinae are not apparent at present and no doubt reflect our limited
larval knowledge of these two subfamilies. Nevertheless, this present evaluation seems to suggest
multiple phylogenetic lineage for both subfamilies and perhaps similar to that shown for the
Inachinae.

Larval affinities of Rochinia carpenteri

In having prominent lateral spines on the carapace and ten setae on the basis of the first maxil-
liped, the first zoea of R. carpenteri shows affinities to the Oregoniinae rather than to the other
subfamilies of Majidae in which the lateral spines are sometimes absent and in which the first
maxilliped basis often has less than ten setae. The single spine on each fork of the telson and the
sub-terminal setae on the antennal exopod place the zoea of R. carpenteri near to the Acan-
thonychinae, some group I Inachinae and to the Pisinae. Its affinities to the Pisinae are further
strengthened by the setal armature of the first maxilliped basis (10) and the moderately developed
spinous processes of the posterio-lateral margins of the third to fifth abdominal segments. The
megalopa of R. carpenteri is more difficult to assign but the well developed rostral spine, the
absence of submedian spines or processes and the prominent cardiac spine all suggest tenuous
affinities to some inachinid (i.e. Macrocheird) and pisinid (i.e. Libinia) megalops.

Larval bibliography of the family Majidae

OREGONIINAE

Chionoecetes bairdi Rathbun

Haynes, 1973 : 769, figs 1-2 (prezoea, 1st zoea); Jewett & Haight, 1977 : 459 (megal.)

Chionoecetes japonicus Rathbun

Motoh, 1970 : 7 (descript prezoea & 1st zoea); Motoh, 1976 : 533, figs 1-4 (1st, 2nd zoeae,

megal.)

60 R. W. INGLE

Chionoecetes opilio (Fabricius)

non Stephensen, 1935 : 40, fig. 16 ( = ? Geryon sp.); Aikawa, 1935 : 222, PI. I, fig. 5 (prezoea);

Aikawa, 1937 : III, fig. 17 (prezoea); Kurata, 1963 : 25, fig. 1 (1st, 2nd zoeae, megal.); Kon,

1967 : 727, fig. 1, PI. I (prezoea, 1st zoea); Ito, 1968 : 91 (descript. prezoea); Kuwatani et al,

1971 : 32, Pis 1-3 (prezoea, 1st zoea as C. opilio elongatus); Kuwatani et al., 1973 : 93, fig. 1

(prezoea); Haynes, 1973 : 774, fig. 21, m (1st zoea); Motoh, 1973 : 1223, figs 1-4 (1st, 2nd zoea,

megal.)

Hyas araneus (Linnaeus)

Bjorck, 1913 : 22, figs 1, 2 (megal. as H. coarctatus); Williamson, 1915 : 526, figs 424-5 (megal.
after Bjorck as Inachus coarctatus); Lebour, 1928 : 544, PI. XIV, fig. 10 (1st crab as H. coarctatus);
Lebour, 1931 : 93, PI. II, fig. 1 (1st zoea); Christiansen, 1973 : 67, figs 1A-19A (1st, 2nd zoeae,
megal. 1st, 2nd crab)

Hyas coarctatus Leach

Williamson, 1911 : 13, PI. I, figs 1, 2, PI. V, figs 70-81, 83 (1st, 2nd zoea, megal. 1st crab as
H. araneus); Stephensen, 1912 : 127, fig. 33 (1st zoea as Brachyurid-larvae); Williamson, 1915 :
521, figs 409-417 (1st, 2nd zoeae as Inachus araneus); Williamson, 1915 : 526, figs 420-423,
426-429 (1st, 2nd zoeae); Stephensen, 1917 : 241, fig. 1 (2nd zoea); Lebour, 1928 : 544, fig. 4
(14-15), fig. 5 (29-30) PI. II, fig. 9, PI. XIV, figs 8-9 (1st, 2nd zoeae, megal); Kurata, 1963 : 28,
fig. 2 (1st, 2nd zoea, megal. as H. coarctatus alutaceus); Christiansen, 1973 : 67, figs 1B-19B
(1st, 2nd zoea, megal. 1st, 2nd crab)

Hyas lyratus Dana

Hart, 1960 : 542, figs 29-38 (1st, 2nd zoeae, megal.)

Oregonia gracilis Dana

Hart, 1960 : 540, figs 1-28 (1st, 2nd zoeae, megal.)

ACANTHONYCHINAE

Acanthonyx lunulatus (Risso)

Cano, 1893 : 539, Tav. 35, figs 60-63 (2nd zoea as Acanthonyx sp.); Boraschi, 1921 : 8, Tav. I,
fig. 4 (1st zoea); Bourdillon-Casanova, 1960 : 214, fig. 73 (1st, 2nd zoeae); Heegaard, 1963 : 482,
figs 1 12-1 18, PI. XVII, fig. 21 (1st zoea)

Acanthonyx petiverii H. Milne Edwards
Lebour, 1944 : 120, fig. 10 (1st zoea)

Epialtus dilatatus A. Milne Edwards

Yang, 1968 : 181, figs 1-8 (1st, 2nd zoeae, megal. 1st crab)

Huenia pro tens de Haan

Aikawa, 1935 : 220, PI. I, fig. 4 (1st zoea); Aikawa, 1937 : 108, fig. 15 (1st zo.ea); Kurata, 1969 :

98, fig. 11 (1st zoea)

Menaethius monoceros (Latreille)

Gohar & Al-Kholy, 1957 : 194, PI. VIII (1st, 2nd zoeae, megal.)

Pugettia gracilis (Dana)

Forss & Coffin, 1960 : 4, Pis III, IV (1st zoea, megal.)

Pugettia incisa (de Haan)

Kurata, 1969 : 96, figs 9, 10 (1st zoea, megal.)

Pugettia quadridens (de Haan)

Aikawa, 1927 : 270, PI. I (1st zoea); Aikawa, 1929 : 38, PI. Ill, fig. 19, PI. IV, figs 25, 33 (1st
zoea); Kurata, 1969 : 94, figs 7, 8 (1st zoea, megal.); Iwata, 1970 : 189, fig. 1, PI. I-II (prezoea,
1st zoea)

Taliepus dentatus (A. Milne Edwards)

Fagetti & Campodonico, 1971 : 1, figs 1-3 (1st, 2nd zoea, megal.)

LARVAL DEVELOPMENT OF R. CARPENTERI 61

INACHINAE

Achaeus cranchii Leach

Bocquet, 1954 : 50, figs 1-4 (1st, 2nd zoeae, megal.)

Achaeus tuber culatus Miers

Kurata, 1969 : 87, figs 2, 3 (1st zoea, megal. 1st crab)

Achaeus spp.

Aikawa, 1935 : 218, 219, PI. I, figs 1, 2 (1st zoeae); Aikawa, 1937 : 107, 108, figs 13, 14 (1st

zoeae); Bourdillon-Casanova, 1960 : 220, figs 76-77 (megal. 1st crab)

Anasimus latus Rathbun

Sandifer & van Engel, 1972 : 141, figs 1-4 (1st, 2nd zoeae, megal.)

Camposcia return Latreille

Gohar & Al-Kholy, 1957 : 189, figs 1-3 (1st, 2nd zoeae, megal.)

Eurypodius latreillei Guerin

Italo Campodonico & Leonardo Guzman, 1972 : 233, figs 1-4 (1st, 2nd zoeae, megal.)

Inachus dorsettensis (Pennant)

Claus, 1876: Taf. X, fig. 8 (1st zoea as /. scorpio); non Cano, 1893: Tav. 35, fig. 71 (1st zoea
as /. scorpio = I. thoracicus); non Williamson, 1915 : 530, fig. 430 (1st zoea after Cano = /.
thoracicus); Lebour, 1927 : 802, Pis I-IV (prezoea, 1st, 2nd zoeae, megal. 1st crab); Lebour,
1928 : 546, PL III, fig. 5, PI. XV, figs 4-5 (prezoea, 1st, 2nd zoeae, megal. lst-3rd crab); Heegaard,
1963 : 471, figs 70-76, PI. XVII, fig. 15 (1st zoea); Ingle, 1977 : 331, figs 1-10 (1st, 2nd zoeae,
megal. lst-3rd crab)

Inachus leptochirus Leach

Lebour, 1928 : 548, PI. Ill, fig. 7, PI. XIV, figs 11, 13, PI. XV, figs 1, 2 (1st, 2nd zoeae, megal.

1st crab)

Inachus phalangium (Fabricius)

Gourret, 1884 : 17, PI. I, figs 5-6 (1st zoea as /. dorynchus}; Williamson, 1915 : 531, figs 431-432
(1st zoea as /. dorynchus after Gourret); Lebour, 1928 : 547, PI. Ill, fig. 6, PI. XIV, fig. 12,
PI. XV, fig. 3 (1st, 2nd zoeae, megal. as /. dorynchus)

Inachus thoracicus Roux

Heegaard, 1963 : 474, figs 77-83, PI. XVII, fig. 16 (1st zoea)

Inachus spp.

Bourdillon-Casanova, 1960 : 215, figs 74-75 (megals, 1st crab)

Macrocheira kaempferi de Haan

Aikawa, 1941 : 119, fig. 2 (1st zoea); Tanase, 1967 : 303, figs 1-3 (1st, 2nd zoeae, megal.);

Kurata, 1969 : 89, figs 4-5 (1st zoea, megal.)

Macropodia deflexa Forest

Lebour, 1928 : 550, PI. Ill, fig. 9, PI. XV, fig. 6, PI. XVI, figs 1, 2, 7 (prezoea, 1st, 2nd zoeae,

megal. 1st crab as M. egyptid)

Macropodia rostrata (Linnaeus)

Thompson, 1836 : 371, fig. le (1st zoea as M. phalangium}; Cano, 1893: Tav. 35, figs 70-74,
77-83 (1st zoea, megal. crab stage as Stenorhynchus phalangium) ; Lo Bianco, 1904 : 33, Taf. XII,
fig. 43 (1st crab as S. phalangium); Lebour, 1928 : 550, PI. Ill, fig. 10, PI. XV, fig. 7, PI. XVI,
figs 3, 5, 8 (prezoea, 1st, 2nd zoeae, megal. 1st crab)

Macropodia tenuirostris (Leach)

Lebour, 1927 : 806, PI. I, figs 2, 8, PI. II, fig. 2, PI. Ill, figs 2, 4, 5 (prezoea, 1st, 2nd zoeae, megal.

1st crab as M. longirostris)

Naxia hystrix Miers

Aikawa, 1935 : 222, PI. I, fig. 6 (1st zoea); Aikawa, 1937 : 109, fig. 16 (1st zoea)

62 R. W. INGLE

Paratymolus pubesc ens Miers
Aikawa, 1937 : 106, fig. 12 (1st zoea)

Pleistacantha sancti-johanis Miers

Aikawa, 1935 : 220, PI. I, fig. 3 (1st zoea); Aikawa, 1937 : 105, fig. 11 (1st zoea); Kurata, 1969 :

92, fig. 6 (1st zoea)

Stenorhynchus seticornis (Herbst)

Yang, 1976 : 158, figs 1-9 (1st, 2nd zoeae, megal. 1st crab)

Stenorhynchus sp.

Yang, 1976 : 168, figs 11-13 (1st zoea)

PISINAE

Doclea gracilipes Stimpson
Chhapgar, 1959 : 48, fig. 11 (megal.)

Eurynome aspera (Pennant)

*Kinahan, 1858 : 233 (prezoea); *Kinahan, 1860 : 73, PI. 9, figs 4-6 (? 1st zoea); Cano, 1893 :
Taf. 35, figs 57-59 (1st zoea as Eurynome sp.); *Gurney, 1924 : 433, figs 1-2 (prezoea, 1st
zoea); "Lebour, 1928 : 543, fig. 5 (31), PI. II, fig. 8, PI. XIV, figs 2-5 (1st, 2nd zoeae, megal.
1st crab); Bourdillon-Casanova, 1960 : 204, fig. 67a (megal.)

Herbstia condyllata (Herbst)

Cano, 1893 : Tav. 35, figs 52-56 (2nd zoea, megal.); Bourdillon-Casanova, 1960 : 205, fig. 68

(1st zoea)

Hyastenus diacanthus (de Haan) Kurata, 1969 : 101, figs 13-14 (1st zoea, megal.)

Libidoclaea grandaria H. Milne Edwards & Lucas
Fagetti, 1969 : 131, figs 1-4 (1st, 2nd zoeae, megal.)

Libinia dubia H. Milne Edwards

Sandifer & van Engel, 1971 : 18, figs 1-4 (1st, 2nd zoeae, megal.)

Libinia emarginata Leach

Johns & Lang, 1977 : 831, figs 1-5 (1st, 2nd zoeae, megal.)

Libinia setosa Lockington

Rathbun, 1923 : PI. XXXVI, fig. 1 (megal.)

Libinia spinosa H. Milne Edwards

Boschi & Scelzo, 1968 : 170, figs 1-42 (1st, 2nd zoeae, megal.)

Lissa chiragra (Herbst)

Cano, 1893 : Tav. 35, figs 45-51 (1st, 2nd zoeae, megal. 1st crab as Lissa sp.); Boraschi, 1921 : 8,
Tav. 1, fig. 15 (1st zoea); Bourdillon-Casanova, 1960 : 212, fig. 72 (1st zoea); Heegaard, 1963 :
480, figs 105-111, PI. XVII. fig. 20 (1st zoea)

Naxioides histrix (Miers)

Kurata, 1969 : 100, fig. 12 (1st zoea)

Naxioides serpulifera (Guerin)

Rathbun, 1914 : 661, PI. II, figs 9-10 (1st, 2nd crab, direct develop.)

Pisa armata (Latreille)

Lebour, 1931 : 94, PI. II, fig. 2, PI. I, figs 2-5 (1st, 2nd zoeae as P. biaculeata); Bourdillon-
Casanova, 1960 : 207, fig. 69 (megal. as P. gibbsi); Heegaard, 1963 : 476, figs 84-90, PI. XVII,
fig. 1 7 ( 1 st zoea)

Pisa corallina (Risso)

Gourret, 1884 : 15, PI. II, figs 3-5 (1st zoea)

* Identity uncertain because of confusion with E. spinosa (see Hartnoll, 1961).

LARVAL DEVELOPMENT OF R. CARPENTERI 63

Pisa nodipes Leach

Heegaard, 1963 : 479, figs 98-104, PI. XVII, fig. 19 (1st zoea)

Pisa tetraodon (Pennant)

Heegaard, 1963 : 477, figs 84-90, PI. XVII, fig. 18

Pisa spp.

Cano, 1893 : Tav. 35, figs 41-44 (1st, 2nd zoeae, megal.? crab stage); Bourdillon-Casanova

1960 : 210, fig. 70 (1st zoeae, megal.)

Pisoides edwardsi (Bell)

Fagetti, 1969a : 160, figs 1-3 (1st, 2nd zoeae, megal.)

Pisoides ortmanni (Balss)

Kurata, 1969 : 103, figs 15, 16 (1st zoea, megal.)

MAJINAE

Acanthophrys longispinosus (de Haan)

Kurata, 1969 : 111, figs 22, 23 (1st zoea, megal.)

Leptomithrax bifidus Ortmann

Kurata, 1969 : 106, figs 18, 19 (1st zoea, megal.)

Leptomithrax edwardsi (de Haan)
Kurata, 1969 : 105, fig. 17 (1st zoea)

Maja squinado (Herbst)

Schlegel, 1911 : 480 (1st, 2nd zoeae, megal. descript. only); Lebour, 1927 : 809, PI. I, figs 3, 9,

PI. II, figs 3, 9, PI. II, figs 3, 9, PI. IV, figs 9-15 (prezoea, 1st, 2nd zoeae, megal. not crab stages =

portunids); Lebour, 1928 : 542, figs 2 (1-2), PI. II, fig. 7, PI. XIV, figs 6-7 (1st, 2nd zoeae, megal.

not lst-3rd crab = portunids); Bourdillon-Casanova, 1960 : 203, fig. 66 (megal.)

Maja verrucosa H. Milne Edwards

Cano, 1893: Tav. 34, figs 26-28 (not fig. 29 = Sir pus, see Gordon 1953), fig. 30, Tav. 35, fig. 86

(1st, 2nd zoeae, megal. crab stage); Bourdillon-Casanova, 1960 : 202 (cf 1st, 2nd zoeae with

M. squinado); Heegaard, 1963 : 484, figs 119-126, PI. XVII, fig. 22 (1st zoea)

Schizophrys aspera (H. Milne Edwards)

Kurata, 1969 : 108, figs 20, 21 (1st zoea, megal.)

Majinae

Rice & Williamson, 1977 : 41, fig. 17 (2nd zoea)

OPHTHALMIINAE

Stilbognathus erythraceus von Martens
Al-Kholy, 1959 : 240, figs 1-21 (2nd zoea, megal.)

MITHRACINAE

Micippa thalia (Herbst)

Kurata, 1969 : 113, figs 24, 25 (1st zoea, megal.)

Microphrys bicornutus (Latreille)

Lebour, 1944 : 122, fig. 12 (1st, 2nd zoeae); Hartnoll, 1964 : 241, figs 1-3 (prezoea, 1st, 2nd

zoeae)

Mithrax forceps A. Milne Edwards
Lebour, 1944 : 121, fig. 11 (1st zoea)

Mithrax spinosissimus (Lamarck)

Provenzano & Brownell, 1977 : 735, figs 1-7 (1st, 2nd zoeae, megal. 1st crab)

Tiarinia cornigera Latreille

Aikawa, 1937 : 112, fig. 18 (1st zoea); Kurata, 1969 : 115, fig. 26 (1st zoea)

64 R. W. INGLE

Acknowledgments

I wish to thank Dr John Gordon for allowing me to participate in Leg 1 of Cruise 16/17 of RRS
Challenger, for providing facilities on board and for his generous help throughout the cruise.
I also wish to express my sincere gratitude to Capt. Maw and his officers for their willing and
able assistance in obtaining this material in spite of the difficult circumstances that frequently
prevailed during this particular cruise. Finally, I wish to thank Dr A. L. Rice for his valued
discussions throughout the preparation of this manuscript.

References

Aikawa, H. 1927. On the morphology of crab zoeal stages. Proc. sclent. Fishery Ass., Tokyo 4 : 270-296

(not seen).

1929. On larval forms of some Brachyura. Rec. oceanogr. Wks Japan 2 : 17-55.

1935. [Inachidae zoea: Larval characters of the family Inachidae and allied forms.] Zool. Mag.

Tokyo 42 : 217-227. (In Japanese.)

1937. Further notes on brachyuran larvae. Rec. oceanogr. Wks Japan 9 : 87-162.

1941. Additional notes to brachyuran larvae. Rec. oceanogr. Wks Japan 12 : 117-120.

Al-Kholy, A. A. 1959. Larval stages of four brachyuran Crustacea (from the Red Sea). Publs mar. biol.

Stn Ghardaga 10 : 239-246.
Balss, H. 1929. Decapoden des Roten Meeres. IV. Oxyrhyncha und Schlussbetrachtungen in: Expedition

S.M. Schiff"Pola" in das Rote Meer. Zool. Ergenb. 36 : 1-30, Wien.
Bjorck, W. 1913. Biologisch-faunistische Untersuchungen aus dem Oresund. I. Pantopoda, Mysidacea

und Decapoda. Acta Univ. lund. 9 2 : 1-39.
Bocquet, C. 1954. Developpement larvaire d'Achaeus cranchii Leach (Decapode Oxyrhynque). Bull.

Soc. zool. Fr. 79 : 50-56.
Boraschi, L. 1921. Osservazioni sulle larve dei Crostacei Decapodi. Memorie R. Com. talassogr. ital.

87 : 1-32.
Boschi, E. E. & Scelzo, M. A. 1968. Larval development of the spider crab Libinia spinosa H. Milne

Edwards, reared in the laboratory (Brachyura, Majidae). Crustaceana (Suppl. 2) : 170-180.
Bourdillon-Casanova, L. 1960. Le meroplancton du Golfe de Marseille: Les larves de crustaces decapodes.

Reel. Trav. Stn. mar. Endoume 30 18 : 1-286.

Cano, G. 1893. Sviluppo e Morfologie degli Oxyrhynchi. Mitt. zool. Stn. Neapel 10 : 527-583.
Chhapgar, B. F. 1959. On the breeding habits and larval stages of some crabs of Bombay. Rec. Indian

Mus. 54 : 33-52.
Christiansen, M. E. 1969. Marine invertebrates of Scandinavia, No. 2. Crustacea, Decapoda, Brachyura.

Universitetsforlaget, Oslo 1-143, Oslo.
1973. The complete larval development of Hyas araneus (Linnaeus) and Hyas coarctalus Leach

(Decapoda, Brachyura, Majidae) reared in the laboratory. Norw. J. Zool. 21 : 63-89.
Claus, C. 1876. Untersuchungen zur Erforschung der Genealogischen Grundlage des Crustaceen-Systems.

i-ciii +114 pp., Wien.
Fagetti, E. 1969. The larval development of the spider crab Libidoclaea granaria H. Milne Edwards &

Lucas under laboratory conditions (Decapoda, Brachyura : Majidae, Pisinae). Crustaceana 12 :

131-140.
1969a. Larval development of the spider crab Pisoides edwardsi (Decapoda, Brachyura) under

laboratory conditions. Mar. Biol. Berlin 4 : 160-165.

& Campodonico, I. 1971. Desarrollo larval en el laboratorio de Taliepus dentatus (Milne-Edwards)

(Crustacea Brachyura : Majidae, Acanthonychinae). Revta Biol. mar. 14 : 1-14.
Forss, C. A. & Coffin, H. G. 1960. The use of the brine shrimp nauplii, Artemia salina, as food for the

laboratory culture of decapods. Walla Walla Coll. Publs Dep. biol. Sci. 26 : 1-15.
Garth, J. S. 1958. Brachyura of the Pacific coast of America Oxyrhyncha. Allan Hancock Pacif. Exped.

21 : 1-854.
Gohar, H. A. F. & Al-Kholy, A. A. 1957. The larvae of some brachyuran Crustacea (from the Red Sea).

Publs mar. biol. Stn Ghardaga 9 : 145-202.

Gordon, I. 1953. On Sirpus, a genus of pigmy Cancroid Crabs. Bull. Br. Mus. nat. Hist. (Zool.) 2 : 43-65
Gourret, P. 1884. Considerations sur la faune pelagique du Golfe de Marseille. Premiere partie. Annl.

Mus. Hist. nat. Marseilles 11 (2) : 12-101.
Griffin, D. J. G. 1966. A review of the Australian majid spider crabs (Crustacea, Brachyura). Aust. Zool.

13 : 259-298.

LARVAL DEVELOPMENT OF R. CARPENTERI 65

Gurney, R. 1924. The zoea of Eurynome aspera. J. mar. biol. Ass. U.K. 13 : 433-436.
Hansen, H. J. 1908. Crustacea Malacostraca. I. In: The Danish Ingolf-Expedition 32 : 1-122. Copen-
hagen.
Hart, J. F. 1960. The larval development of British Columbia Brachyura. 2. Majidae, subfamily

Oregoniinae. Can. J. Zool. 38 : 539-546.
Hartnoll, R. G. 1961. A re-examination of the spider crab, Eurynome Leach from British waters. Crusta-

ceana 2 : 171-182.
1964. The zoeal stages of the spider crab Microphrys bicornutus (Latr.). Ann. Mag. nat. Hist. (13)

7 : 241-246.
Haynes, E. 1973. Descriptions of prezoeae and stage I zoeae of Chionoecetes bairdi and C. opilio (Oxy-

rhyncha, Oregoniinae). Fishery Bull. natn. ocean, atmos. Adm. 71 : 769-775.
Heegaard, P. 1963. Decapod larvae from the Gulf of Napoli, hatched in captivity. Vidensk. Meddr dansk

naturh. Foren. 125 : 449-493.
Ingle, R. W. 1977. The larval and post-larval development of the scorpion spider crab, Inachus dorset-

tensis (Pennant) (Family : Majidae), reared in the laboratory. Bull. Br. Mus. nat. Hist. (Zool.) 30 :

331-348.

& Clark, P. F. 1977. A laboratory module for rearing crab larvae. Crustaceana 32 : 220-222.

Italo Campodonico, G. & Leonardo Guzman, M. 1972. Desarrollo larval de Eurypodius latreillei Guerin

en condiciones de laboratorio. (Crustacea Brachyura : Majidae, Inachinae). An. Inst. Patagonia 3 :

233-247.
Ito, K. 1968. Observations on the primary features of the newly hatched zoeal larvae of the zuwai-crab.

Bull. Japan sea reg. Fish. Res. Lab. 20 : 91-93 (in Japanese).
Iwata, F. 1970. Studies on the development of the crab Pugettia quadridens (de Haan): I. Hatching and

zoeae. Publs Seto mar. biol. Lab. 18 : 189-197.
Jewett, S. C. & Haight, R. E. 1977. Description of megalopa of snow crab Chionoecetes bairdi (Majidae,

subfamily Oregoniinae). Fishery Bull. natn. ocean, atmos. Adm. 75 : 459-463.
Johns, D. M. & Lang, W. H. 1977. Larval development of the spider crab, Libinia emarginata (Majidae).

Fishery Bull. natn. ocean, atmos. Adm. 75 : 831-841.

Kinahan, J. R. 1958. Remarks on the zoe of Eurynome aspera, and the habits of the animal in confine-
ment. Ann. Mag. nat. Hist. 1 : 233-235.
1860. Remarks on the zoe of Eurynome aspera, and the habits of the animal in confinement. Proc.

nat. Hist. Soc. Dublin 2 : 77-79.
Kon, T. 1967. Fisheries biology of the Tanner crab, Chionocetes opilio -I. On the prezoeal larva. Bull.

Jap. Soc. sclent. Fish. 33 : 726-730.
Kuarata, H. 1963. Larvae of Decapod Crustacea of Hokkaido. 2. Majidae (Pisinae). Bull. Hokkaido reg.

Fish. Res. Lab. 27 : 25-31.
1969. Larvae of decapod Brachyura of Arasaki, Sagami Bay. IV. Majidae. Bull. Tokai Reg. Fish.

Res. Lab. 57 : 81-127.
Kuwatani, Y., Wakui, T. & Nakanishi, T. 1971. Studies on the larvae and post-larvae of a Tanner crab,

Chionoecetes opilio elongatus Rathbun. Bull. Hokkaido reg. Fish. Res. Lab. 37 : 32-40.
1973. Studies on the larvae and post-larvae of a Tanner Crab, Chionoecetes opilio

elongatus Rathbun. A correction to previous report (1971) on the external features of protozoeal

larva. Bull. Hokkaido reg. Fish. Res. Lab. 39 : 93-94.
Lebour, M. V. 1927. Studies of the Plymouth Brachyura. I. The rearing of crabs in captivity with a

description of the larval stages of Inachus dorsettensis, Macropodia longirostris and Maia squinado.

J. mar. Biol. Ass. U.K. 14 : 795-820.

1928. The larval stages of the Plymouth Brachyura. Proc. zool. Soc. Lond. (1928) : 473-560.

1931. Further notes on larval Brachyura. Proc. zool. Soc. Lond. (1931) : 93-96.

1944. Larval crabs from Bermuda. Zoologica 20 : 113-128.

Lo Bianco, S. 1904. Pelagische Tiefseefischerei der 'Maja' in der umgebung von Capri. Beitrage zur

Kenntnis des Meeres und seiner Bewohner. Erster Band i-vi + 92 pp., Jena.
Motoh, H. 1970. Spec. Publs Pref. mar. Cult. Sta. Ishikawa 7-11 (not seen).
1973. Laboratory-reared zoeae and megalopae of zuwai crab from the Sea of Japan. Bull. Jap.

Soc. sclent. Fish. 39 : 1223-1230.

1976. The larval stages of Benizuwai-gani, Chionoecetes japonicus Rathbun reared in the laboratory.

Bull. Jap. Soc. sclent. Fish. 42 : 533-542.
Provenzano, A. J. & Brownell, W. N. 1977. Larval and early post-larval stages of the West Indian spider

crab, Mithrax spinosissimus (Lamarck) (Decapoda : Majidae). Proc. biol. Soc. Wash. 90 : 735-752.
Rathbun, M. J. 1914. Stalk-eyed crustaceans collected at the Monte Bello Islands. Proc. zool. Soc. Lond.

(1914) : 653-664.

66 R. W. INGLE

1923. The brachyuran crabs collected by the U.S. Fisheries Steamer 'Albatross' in 1911, chiefly

on the west coast of Mexico. Bull. Am. Mus. nat. Hist. 48 : 619-637.
Rice, A. L. & Ingle, R. W. 1975. The larval development of Carcinus maenas (L.) and C. mediterraneus

Czerniavsky (Crustacea, Brachyura, Portunidae) reared in the laboratory. Bull. Br. Mus. nat. Hist.

(Zool.)28 : 101-119.
& Williamson, D. I. 1977. Plaoktonic stages of Crustacea Malacostraca from Atlantic Seamounts.

'Meteor' Forsch.-Ergenb., D 26 : 28-64.
Sandifer, P. A. & van Engel, W. A. 1971. Larval development of the spider crab, Libinia dubia H. Milne

Edwards (Brachyura, Majidae, Pisinae), reared in laboratory culture. Chesapeake Sci. 12 : 18-25.
1972. Larval stages of the spider crab, Anasimus latus Rathbun, 1894 (Brachyura, Majidae,

Inachinae) obtained in the laboratory. Crustaceana 23 : 141-151.
Schlegel, C. 1911. Sur le developpement de Maia squinado Latr. C. r. hebd. Seanc. Acad. Sci. Paris 153 :

480-482.
Steedman, H. F. 1976. General and applied data on formaldehyde fixation and preservation of marine

zooplankton (pp. 103-154). In: Steedman, H. F. (ed.), Zooplankton fixation and preservation. Mono-
graphs on oceanographic methodology. UNESCO Press. 350 pp. Paris.

Stephenson, K. 1912. Report on the Malacostraca collected by the 'Tjalfe'-Expedition, under the direc-
tion of cand. mag. Ad. S. Jensen, especially at W. Greenland. Vidensk. Meddr. dansk naturh. Foren.

64 : 57-134.
1917. Zoogeographical investigation of certain fjords in southern Greenland, with special reference

to Crustacea, Pycnogonida and Echinodermata, including a list of Alcyonaria and Pisces. Meddr

Gr0nland53 : 231-378.

1935. Crustacea Decapoda. The Godthaab Expedition 1928. Meddr Gr0nlandSQ : 1-94.

Tanase, H. 1967. Preliminary note on zoea and megalopa of the giant spider crab, Macrocheira kaempferi

de Haan. Publs Seto mar. biol. Lab. 15 : 303-309.
Thompson, J. V. 1836. On the double metamorphosis in Macropodia phalangium, or spider crab; with

proofs of the larvae being zoea in Gegarcinus hydrodromus, Thelphusa erythropus, Eriphia carribaea,

and Grapsus pelagicus. Ent. Mag. 3 : 370-375.

Thomson, C. W. 1874. The Depths of the Sea. i-xvi + 527 pp. London.
Williamson, H. C. 1911. Report on the larval and later stages of Portunus holsatus, Fabr., Portunus

puber, L; Portunus depurator, Leach; Hyas araneus (L.); Eupagurus bernhardus L; Galathea dispersa,

Spence Bate; Crangon trispinosus (Hailstone); Cancer pagurus, L. Scient. Invest. Fishery Bd Scotl.

(1909) : 1-20.

1915. Crustacea Decapoda larven. Nord. Plankt. 18 : 315-588.

Yang, W. T. 1967. A study of zoeal, megalopal, and early crab stages of some oxyrhynchous crabs

(Crustacea: Decapoda). Doctoral Dissertation, University of Miami.
1968. The zoeae, megalopa, and first crab of Epialtus dilatatus (Brachyura, Majidae), reared in the

laboratory. Crustaceana (Suppl.) 2 : 181-202.
1976. Studies on the western Atlantic Arrow crab Genus Stenorhynchus (Decapoda Brachyura,

Majidae). I. Larval characters of two species and comparison with other larvae of Inachinae). Crust-
aceana 31 : 157-177.

Manuscript accepted for publication 16 November 1978

On the spider genus Cynapes (Araneae : Salticidae)

F. R. Wanless

Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD

Introduction

The genus Cynapes Simon, 1900 at present includes two known species, the type species Cynapes
wrightii (Blackwall) from the Seychelles and C. baptizatus (Butler) from Rodriguez. Two species
previously included in the genus are removed elsewhere: C. albolineatus (Peckham & Peckham)
from Madagascar is returned to Bavia Simon, 1877, the genus in which it was originally described,
while C. canosus Simon from Mauritius is transferred into Marengo Peckham & Peckham,
1892 .

Cynapes baptizatus and Marengo canosa comb. nov. are described below. C. wrightii has not
been included because it has recently been described elsewhere (Wanless, in press). Bavia albo-
lineata comb. rev. will be dealt with when Bavia is revised as preliminary observations suggest
that this placement is doubtful, unfortunately I am unable to propose a more satisfactory solu-
tion at the present time.

Wanless (in press) has suggested that Cynapes may be synonymous with Baviola Simon,
1897. The structure of the genitalia shows that the genera are closely related, but they can be
separated by the shape of the carapace and eye formula. These latter characters were extensively
used by early salticid workers for defining genera and the failure to consider genitalic characters
has largely resulted in the present unsatisfactory taxomony of the family. Somatic characters
can nevertheless provide evidence of relationships particularly in cases when the genitalia are
relatively simple and of similar forms. For the present, Cynapes is therefore considered to be a
valid genus, but it may ultimately be synonymized with Baviola or, as seems more likely, be
given a subgeneric rank. Subgenera have not often been used in this family and they cannot for
the moment be justified as too many tropical Salticidae are poorly known, and it is impossible
to present an overall view of the relationships.

Genus CYNAPES Simon

Cynapes Simon 1900 : 393. Type species Salticus wrightii Blackwall, by original designation.

Simon, 1901 : 462, 467, 469, 471. Petrunkevitch, 1928 : 186. Bonnet, 1956 : 1334. Wanless

(in press).

DEFINITION. Medium to large spiders ranging from about 5-0 to 8-0 mm in length. Species
sexually dimorphic. Distinctive colour markings present, not hirsute. Legs I heaviest especially
in females. Carapace: longer than broad, moderately low with convex lateral margins; fovea
in shallow depression behind PL. Eyes: anteriors subcontiguous, apices more or less level or
slightly recurved; posterior median eyes small, slightly closer to AL than to PL; posterior and
anterior rows subequal in width; quadrangle length between 40 and 50 per cent of carapace
length. Clypeus: low. Chelicerae: robust, moderately porrect, sometimes with an apophysis on
anterior surface; promargin with 3 or 4 teeth, retromargin with 6 or 7. Maxillae: moderately
elongate, parallel or slightly convergent. Labium: elongate tongue-shaped. Sternum: scutiform
to elongate scutiform, sometimes slightly narrowed anteriorly. Abdomen: elongate with a pattern;
spinnerets closely grouped, posteriors slightly longer than the rest, anteriors more robust; tracheal
spiracle a transverse slit just in front of the anterior spinnerets or positioned midway between
the middle of the abdomen and the anterior spinnerets (internal system not examined). Legs:
legs I heaviest, sometimes with scanty ventral fringes; not scopulate, spines numerous and strong
on legs I-II but fewer and weaker on legs III-IV. Epigyne: simple and of similar forms; orifice

Bull. Br. Mas. not. Hist. fZooU 37 (1) : 67-72 Issued 29 November 1979

68 F. R. WANLESS

rounded with obscure lateral openings leading to convoluted fertilization ducts. Female palp:
long and slender, clothed in hairs. Male palp: simple, of variable form; tegulum with distal
coiled or partly coiled embolus; tibial apophysis slender, unmodified.

DIAGNOSIS AND AFFINITIES. Cynapes is closely related to Baviola, but may be distinguished by
the low carapace and the presence of a broad space between the posterior lateral eyes and cara-
pace margin; the space being much greater than the diameter of the posterior eye.

List of species in the genus Cynapes Simon, 1900

Cynapes wrightii (Blackwall, 1877)
C. baptizatus (Butler, 1876)

Cynapes baptizatus (Butler)
(Fig. 1A-K)

Salticus baptizatus Butler, 1876 : 440, ?. Holotype ?, Rodriguez (BMNH. 1876.13) [Examined]. Butler

1879 : 501, 507, 509, pi. LII.
Cynapes baptizatus: Simon, 1901 : 467. Roewer, 1954:980. Bonnet, 1956: 1335. Proszyriski, 1971:

395. Wanless (in press).

DIAGNOSIS. C. baptizatus can be readily distinguished from C. wrightii, the only other species
in the genus, by having the tracheal slit positioned just in front of the anterior spinnerets. In
C. wrightii the slit is clearly positioned a short distance away from the spinnerets, with the
intervening space clothed in short hairs.

MALE FROM RODRIGUEZ (previously undescribed). Carapace (Fig. IE): orange-brown with faint
sooty markings, foveal region and posterior half of quadrangle pale yellow; from below AL to
posterior lateral margin a white haired band with scattered white hairs in pale yellow areas
(mostly rubbed in specimens at hand). Eyes: with black surrounds; anteriors subcontiguous
with apices more or less level, fringed in white hairs. Clypeus: white haired. Chelicerae (Fig. IF):
with frontal apophyses; orange-brown; promargin with 3 teeth, retro margin with 6. Maxillae
and labium: orange-brown. Sternum (Fig. 1G): pale yellow. Abdomen: pale yellow with black
bands above and blackish lateral markings, venter light yellow tinged black; spinnerets whitish
yellow. Legs: legs I heaviest, tarsi pale yellow, other segments yellow-brown to orange-brown
tinged black with scanty ventral white haired fringes on tibiae and patellae; other legs pale yellow.
Spination of legs I: metatarsi with 2 pairs of ventral spines, tibiae with 3 pairs, patellae with
1 spine, femora with 2 dorsal and 2 prolaterals. Palp (Fig. 1C, H): yellow-orange heavily black-
ened, but cymbium white tipped.

Dimensions (mm): total length 5-68; carapace length 2-36, breadth 2-06; abdomen length
3-36; eyes, anterior row 1-6, middle row 1-38, posterior row 1-64; quadrangle length 1-08.
Ratios: AM : AL : PM : PL :: 12-5 : 7 : 1-1 : 6; AL-PM-PL :: 7-9.

FEMALE FROM RODRIGUEZ. Body form and colour markings (Fig. 1 A, D), more or less as in c?,
except for the following: Chelicerae: lacking apophyses; promargin with 4 teeth, retromargin
with 6. Sternum (Fig. II): more attenuate in front (to allow for the heavier coxae I). Legs: legs I
massive; spines: metatarsi with 2 ventral pairs, tibiae with 3, on both segments prolateral spines
rather small, patellae with 1 spine, femora with 3 dorsal and 2 distal prolaterals; other leg spines
few and generally weak. Palps: light yellow. Epigyne (Fig. U, K): pale with indistinct circular
orifice and posterior fertilization ducts.

Dimensions (mm): total length 8-16; carapace length 3-44, breadth 2-96; abdomen length 5-0;
eyes, anterior row 2-10, middle row 1-88, posterior row 2-2; quadrangle length 1-44. Ratios:
AM : AL : PM : PL :: 16 : 8 : 1-5 : 7; AL-PM-PL :: 9-13.

VARIATION. Another male measures 5-0 mm total length, 2-60 mm carapace length. Females
vary from 6-5 to 8-1 mm total length, 2-7 to 2-96 mm carapace length (5 specimens). The con-
formation of the fertilization ducts is inconsistent, but the general appearance of the epi-
gynes does not show much variation. Some individuals are paler, probably as a result of long
preservation.

SPIDER GENUS CYNAPES

69

K

Fig. 1 Cy napes baptizatus, <5: C, palp, lateral view; E, carapace, lateral view; F, chelicerae; G,
sternum; H, palp, ventral view. ?: A, dorsal view; B, leg I; D, carapace, lateral view; I, sternum;
J, vulva; K, epigyne.

70 F. R. WANLESS

DISTRIBUTION. Rodriguez.

MATERIAL EXAMINED. Holotype ?, Rodriguez, Transit of Venus Exped., 1874-75, (G. Gulliver)
(BMNH. 1876.13). RODRIGUEZ: x. 1918, 2. &?, 5 ?? (H. P. Thomasset & H. J. Sm>//)(BMNH.
1952.3.6.95-104).

Marengo canosa (Simon) comb. nov.

(Fig. 2A-K)

Cynapes canosa Simon, 1900 : 393, <J, ?. LECTOTYPE <J, PARALECTOTYPE ? (here designated)
Mauritius (MNHN, Paris, no. 20353) [Examined]. Simon, 1901 : 462, 463, 467. Roewer, 1954 : 980.
Bonnet, 1956 : 1335. Proszyriski, 1971 : 396.

The female here designated paralectotype is smaller than the lectotype male and as such does
not agree with the original description (Simon, 1900). However, as the remaining data are in
agreement it is assumed that Simon's measurement is erroneous, and that the specimens are
syntypes.

AFFINITIES AND DIAGNOSIS. Uncertainties as to the affinities of this species arise from the suspicion
that the specimens described below may not be conspecific, principally because of differences in
the position of the posterior lateral eyes in relation to the carapace width (Fig. 2A, G). The
structure of the genital ia is in broad agreement with the genus as defined in Wanless (1978).
However, legs I are not grossly enlarged, the chelicerae are not set well back and, furthermore,
in the male they are moderately porrect with low lateral keels. These characters serve to distinguish
M. canosa from all other known species of Marengo, but the generic placement will have to be
reconsidered when more is known of tropical salticids.

MALE LECTOTYPE. Carapace (Fig. 2A, B): Dark reddish brown with an iridescent sheen in eye
area; clothed with short scattered white hairs forming an obscure band encircling eye region,
fovea lacking. Eyes: with black surrounds; anteriors subcontiguous with apices slightly recurved;
fringed in white hairs. Clypeus: low; orange-brown edged black with scattered light brown hairs.
Chelicerae (Fig. 21): dark orange, shiny, lateral margins black forming slightly raised keels;
promargin with 3 teeth, retromargin with 6 followed by a blunt apophysis. Maxillae: brown.
Labium: brown with blackish lateral depressions. Sternum (Fig. 2F): orange-brown. Abdomen:
yellow-brown lightly mottled black with a dark reddish brown scutum sparsely clothed in fine
silky white hairs (mostly rubbed). Legs: legs I heaviest, dark orange-brown with scanty ventral
fringes on tibiae and patellae; other legs yellow-brown to light orange-brown. Spination of legs I:
metatarsi with 2 pairs of stumpy spines; tibiae with 3 distal pairs of normal spines, but proximal
pair offset; patellae with 1 spine; femora with 3 distal dorsal and 2 prolateral spines. Other leg
spines weaker and fewer especially on legs III-IV. Palp (Fig. 2D, E): orange-brown; embolus
apparently with 3 spirals.

Dimensions (mm): total length 5-16; carapace length 2-4, breadth 2-16; abdomen length 2-84;
eyes, anterior row 1-64, middle row 1-58, posterior row 1-84; quadrangle length 1-04. Ratios:
AM : AL : PM : PL :: 12 : 7 : 1-2 : 6-3; AL-PM-PL :: 7-8.

FEMALE PARALECTOTYPE. Carapace (Fig. 2G): dark reddish brown, weakly iridescent in eye
region; thinly clothed in short whitish hairs; fovea lacking. Eyes: with black surrounds; anteriors
subcontiguous, apices more or less level, fringed in white hairs. Clypeus: low; orange edged
black, sparsely white haired. Chelicerae (Fig. 2H): finely rugose; yellow-orange, weakly iridescent;
promargin with 3 teeth, retromargin with 5. Maxillae: brownish orange. Labium: blackish tipped
brownish orange. Sternum: orange-brown tinged black. Abdomen: pale yellow-brown with
faint blackish mottling; spinnerets blackish. Legs: legs I only slightly heavier than the rest.
Yellow-brown. Spination of legs I: metatarsi with 2 pairs of ventral spines; tibiae with 3 pairs;
patellae with 1 spine; femora with 3 dorsals and 1 prolateral. Other leg spines generally weak
and fewer except for the femoral spines. Palp: yellow-brown. Epigyne (Fig. 2K): vulva not
examined.

SPIDER GENUS CYNAPES

71

K

Fig. 2 Marengo canosa, lectotype cj: A, dorsal view; B, carapace, lateral view; C, leg I; D, palp,
lateral view; E, palp, ventral view; F, sternum; I, chelicera. Paralectotype ?: G, dorsal view;
H, chelicera; J, leg I; K, epigyne.

72 F. R. WANLESS

Dimensions (mm): total length 4-36; carapace length 1-78, breadth 1-54; abdomen length
2-72; eyes, anterior row 1-36, middle row 1-28, posterior row 1-52; quadrangle length 0-84.
Ratios: AM : AL : PM : PL :: 10 : 6 : 1-1 : 5; AL-PM-PL :: 6-6-5.

VARIATION. Unknown.
DISTRIBUTION. Mauritius.

MATERIAL EXAMINED. Lectotype & paralectotype $, Mauritius (Ch. Alluaud) (MNHN, Paris,
no. 20353).

Acknowledgments

I wish to thank M. M. Hubert, Museum National d'Histoire Naturelle, Paris, for the loan of
specimens and Mr D. Macfarlane, Commonwealth Institute of Entomology, London, for
reading the manuscript.

References

Blackball, J. 1877. A list of spiders captured in the Seychelles Islands by Professor E. Perceval Wright,

M.D., F.L.S. with descriptions of species supposed to be new to arachnologists. Proc. R. Ir. Acad.

3 : 1-22.

Bonnet, P. 1945-61. Bibliographia Araneorum 3 vols. Imprimerie Douladoure, Toulouse.
Butler, A. G. 1876. Preliminary notice of new species of Arachnida and Myriapoda from Rodriguez

collected by Messrs George Gulliver and H. H. Slater. Ann. Mag. nat. Hist. (4) 17 : 439-446.
1879. Myriopoda and Arachnida. (in) Zoology of Rodriguez. An account of the petrological,

botanical and zoological collections made in Kerguelen's Land and Rodriguez during the Transit

of Venus Expeditions, carried out by order of Her Majesty's Government in the years 1874-75. Phil.

Trans. R. Soc. 168 : 497-509.
Peckham, G. W. & Peckham, E. G. 1892. Ant-like spiders of the family Attidae. Occ. Pap. Wis. nat.

Hist. Soc. 2 (1) : 1-83.

Petrunkevitch, A. 1928. Systema Aranearum. Trans. Conn. Acad. Arts Sci. 29 : 1-270.
Proszyriski, J. 1971. Catalogue of Salticidae (Aranei) specimens kept in major collections of the world.

Annls zool. Warsz. 28 : 367-519.
Roewer, C. F. 1954. Katalog der Araneae. 2, Abt. B : 924-1290. Institut Royal des Sciences Naturelle

de Belgique, Bruxelles.
Simon, E. 1877. Etudes arachnologiques 19e Memoire. XXVII. Arachnides recueillis a Assinie (Afrique

occidentale) par MM. Chaper et Alluaud. Annls Soc. ent. Fr. (6) 7 : 261-276.
1897. Etudes arachnologiques 29e Memoire (1) XLVI. Arachnides recueillis en 1895 par M. le

Dr A. Brauer (de 1'Universite de Marburg) aux iles Sechelles. Annls Soc. ent. Fr. 66 : 370-388.

1900. Descriptions d'arachnides nouveaux de la famille des Attidae. Annls Soc. ent. Belg. 44 : 381-

407.

1901. Histoire Naturelle des Araignees, 2 (3) : 381-668. Paris, Roret: Libraire Encyclopedique.

Wanless, F. R. 1978. A revision of the spider genus Marengo (Araneae : Salticidae). Bull. Br. Mus.

nat. Hist. (Zool.) 33 (4) : 231-296.
(in press). The salticid spiders (Araneae : Salticidae) of the Seychelles. Revue zool. afr.

Manuscript accepted for publication 22 January 1979

A new species of Phthiracarus (Acari,
Cryptostigmata) from Austria

B. W. Parry

Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD

Introduction

A small collection of mites from oak litter in the Vienna Woods, Austria was examined by
Sheals in connection with a computer study of phenetic affinity within the Phthiracaroidea
(Sheals, 1969). The material was found to contain several specimens of an undescribed phthira-
carid mite, the notogastral integument of which was particularly striking and apparently unique
amongst the known members of the superfamily. Moreover, although the species showed an
overall similarity to Steganacarus, its anal chaetotactic pattern corresponded to that of Phthira-
carus. Classification of the mite was deferred at this time and Sheals excluded it from his study.
However, in view of its interesting morphological features a detailed description was considered
desirable and is given below, based on the seven available specimens.

Family PHTHIRACARIDAE Perty, 1841

Phthiracarus papillosus sp. nov.

Aspis (Figs 2-4; PI. Ib): 242-282 urn long and with a greatest width of 181-212 urn. All the
dorsal setae are rather long, stout and procumbent. The interlamellar setae (//) are about 1-5
times the length of the lamellars (la) and extend two-thirds of the distance between the bases of
the interlamellar and rostral setae (ro). The latter almost reach the anterior aspal margin. The
sensilli are 40-55 um long, lanceolate and striated. Three finger-like tracheoles are associated
with each bothridium and there is a single pair of short exobothridial setae (ex). In front of the
il-la setae there is a pronounced median keel and on each side three lateral keels arranged in
an oblique row (PI. Ib). Of the lateral keels, the paraxial pair is the longest and the antiaxial
pair the shortest. The prodorsal integument is not strongly ornamented and appears to be finely
punctate as in other Phthiracarus species. Behind the sensilli the integument is raised into a
number of low longitudinal ridges.

Notogaster (Figs 1, 5; Pis la; 2d): 464-606 um in length along a line through c^-h^ and with
a greatest depth of 333-484 urn. All the setae are stout, slightly recurved and more or less equal
to the distance c^-d^ (PI. la). Seta c 3 is inserted on the posterior margin of the collar and setae
(Va submarginally. Vestigial f v is located just below the seta h v . The fissures ip and ips are present.
In one paratype ip and ips were only present on one side. The integument is densely and irregularly
papillate (PI. 2d).

Ano-genital region (Fig. 9; PI. If): There are two pairs of comparatively long anal setae (an^)
located marginally and three pairs of adanals (ad^) forming an oblique row submarginally,
the longest in the row being ad 2 which is approximately equal in length to the anal setae. Scanning
electron micrographs show that all the setae on the anal plates are weakly serrated. On each
genital plate there are nine simple genital setae arranged in two rows. Setae g!_ 5 are minute
and located on the paraxial border while the other four pairs (g 6 _ 9 ) are moderately short and
submarginal. A single aggenital seta ag is located antiaxially in the genital furrow. The integu-
ment of the genital and anal plates is weakly papillate (PI. If) with the exception of the areas
bearing setae (an^*,) where the integument has no distinct ornamentation. There are three
pairs of genital papillae, the anterior pair being rather small.

Bull. Br. Mus. not. Hist. (Zool.) 37 (1) : 73-79 Issued 29 November 1979

73

74

B. W. PARRY

1

Figs 1-5 Phthiracarus papillosus: (1) notogaster, lateral; (2) aspis, dorsal; (3) aspis, lateral; (4)
sensillus and bothridium; (5) notogaster, dorsal.

A NEW SPECIES OF PHTHIRACARUS

75

6

Figs 6-9 Phthiracarus papillosus: (6) pedipalp; (7) chelicera, paraxial; (8) chelicera, antiaxial; (9)

ano-genital region.

Infracapitulum: This is typically phthiracaroid in form (see for example, Parry, 1979). Of the
three pairs of adoral setae, the anterior pair are brush-like distally and the two posterior pairs
weakly serrated.

Pedipalps (Fig. 6) : Three-segmented with the setal formula (2-2-7). Four of the tarsal setae
(acm, ul\ ul" and sul) are eupathidial, sul being rather short.

Chelicerae (Figs 7, 8): The movable digit has four teeth and the fixed digit carries five. The
latter are arranged in two rows, an outer one of two teeth and an inner row of three. The principal
segment carries 12-21 sharply pointed spines on the paraxial surface and 13-26 conical spines
antiaxially. Although cheliceral spines have been observed in other species of Phthiracarus
(see Parry, 1979), the greatest number has always been observed on the paraxial surface. Setae
cha and chb are both serrated, cha being somewhat longer than chb. The cheliceral integument
is finely punctate.

Legs (Figs 10-13; PI. Ic-e): The solenidial formulae for the legs are I (2-1-3); II (1-1-2);
III (0-1-1) and IV (0-1-0). On tarsus I solenidion co 2 is coupled with a small distal seta. The
latter is conical in shape and apparently lacks the longitudinal constriction observed in certain
other species of Phthiracarus (Parry, 1979) and Steganacarus (Parry, 1978). On tibia I the reduced

10

Figs 10 & 11 Phthiracarus papillosus, posterolateral aspect of leg IV: (10) tarsus; (11) tibia to

trochanter.
Figs 12 & 13 Phthiracarus papillosus, posterolateral aspect of leg I: (12) tarsus; (13) tibia to

trochanter.
(Figs 10 and 12 are drawn at the same magnification.)

A NEW SPECIES OF PHTHIRACARUS 77

dorsal seta coupled with solenidion <p is rather long and prominent (PI. le) but on tibiae II-IV
it is much shorter. Solenidion a^ on genu I is coupled with a small posterolateral seta /". The
formulae for the leg setae are I (1-4-2-5-16-1); II (1-3-2-3-12-1); III (2-2-1-2-10-1) and
IV (2-1-1-2-10-1). On all four tarsi the setal arrangement closely resembles that found in other
species of the genus. On tarsus I six of the setae ( (it), (p), s and a') are eupathidial and apparently
blunt distally. The famulus e is short, rugose and closely associated with co 1 . Seta ft' is some-
what unusual in being hooked distally (this seta is generally straight on all four tarsi in Phthira-
carus species). Setae (tc) and () on tarsus I (PI. le) and (tc], (u), (p) and s on tarsi II-IV are
covered with whorls of rather blunt spicules in the middle third. On femur I seta d is thickened
and densely serrated (PI. Id), the serrations being blunt distally (in other species of the genus
the serrations are fewer and sharply pointed). All the tarsi terminate in a single claw bearing
two ventral teeth and an antero- and posterolateral row of serrations.

MATERIAL: Holotype, BMNH reg. no. 1978.11.10.1, and six paratypes, BMNH reg. no. 1978.
11.10.2-7, from oak litter, Leopoldsberg, Austria. The material was collected by Professor W.
Kiihnelt, 4 July 1948 and 13 November 1948.

REMARKS : In relation to the general shape of the aspis and the integumental ornamentation of
the notogaster, P. papillosus is atypical of the genus Phthiracarus, showing affinities with Stegana-
carus and Tropacarus. In these two genera the prodorsal integument posterior to the il-la setae
is raised into a number of longitudinal ridges, while anteriorly there is a median keel and, in
certain species, a pair of lateral keels. However, the presence of seven keels in P. papillosus is
apparently unique in the Phthiracaridae. An ornamented notogastral integument is also present
in Tropacarus and Steganacarus. In the former genus the integument may be described as being
irregularly stellate-papillate (PI. 2a, b) while in Steganacarus, although usually punctate (PL 2f),
a rather striking ornamentation (raised reticulate) has been observed in Steganacarus clavigerus,
a species described by Berlese (1904) from the Boboli Gardens, Italy (PI. 2c). In addition, the
notogastral setae of P. papillosus are rather long and stout as in Steganacarus species while the
serrated nature of the setae in the ano-genital region is characteristic of both Steganacarus and
Tropacarus.

Despite these affinities the Austrian species has been classified in Phthiracarus. Procumbent
interlamellar setae and a 2 + 3 setal arrangement on the anal plates are sufficient to distinguish
Phthiracarus from other phthiracarid genera. However, to accommodate P. papillosus in Parry's
(1979) definition of the genus, the following modifications must be made: The integument of
the dorsal and ventral shields is densely punctate and/or papillate while that of the infracapi-
tulum, chelicerae and appendages is punctate.' ' . . . aspis usually without a median keel or lateral
keels.' Although the integument of other Phthiracarus species appears smooth when seen with
the light microscope, scanning electron micrographs have revealed that in a number of British
species the integument of the genital plates is weakly papillate and punctate (PI. 2e).

References

Parry, B. W. 1978. A new species of Steganacarus (Acari, Cryptostigmata) from Israel. Bull. Br. Mus.

nat. Hist. (Zool.) 33 (4) : 279-285.
1979. A revision of the British species of the genus Phthiracarus Perty, 1841 (Cryptostigmata :

Euptyctima). Bull. Br. Mus. nat. Hist. (Zool.) 35 (5) : 323-363.
Sheals, J. G. 1969. Computers ii acarine taxonomy. Acarohgia 11 : 376-396.

Manuscript accepted for publication 11 December 1978

78

B. W. PARRY

Plate 1 Phthiracarus papillosus: (a) lateral aspect, x70; (b) aspis, dorsal aspect, X400; (c) tectal
and unguinial setae on tarsus I, x3500; (d) dorsal seta on femur I, xSOOO; (e) solenidion and
associated seta on tibia I, dorsolateral aspect, x3650; (f) genital plates, x480.

A NEW SPECIES OF PHTHIRACARUS

79

Plate 2 Integument: notogaster (a-d, f); genital plate (e). (a) Tropacarus pulcherrimus (Berlese),
X 550. (b) Tropacarus carinatus (C. L. Koch), x 700. (c) Steganacarus clavigerus (Berlese), x 1 800.
(d) Phthiracarus papillosus, x 1 100. (e) Phthiracarus affinis (Hull), x 8000. (f) Steganacarus magnus

(Nicolet), x 700.

An Important New Title From British Museum
(Natural History) Monographs

British Marine Amphipoda: Gammaridea

by R. J. Lincoln

1979 658pp 2,300 figures 4to hard bound
ISBN 0565 008 18 50.00

Amphipods are both numerous and diverse in numbers of genera and species in
British coastal waters, but in the absence of any form of modern systematic
synopsis or key this group of crustaceans has acquired the reputation of being
notoriously difficult to identify. This monograph, which is the first comprehensive
and illustrated text on British gammaridean amphipods to be published in more
than a century, should go a long way towards solving the problem.

Titles to be published in Volume 37

Miscellanea

The echinoderms of Aldabra and their habitats. By N. A. Sloan,
Ailsa M. Clark & J. D. Taylor

The Fellodistomidae (Digenea) of fishes from the northeast Atlantic.

By Rodney A. Bray & David L. Gibson

The anatomy, phylogeny and classification of bariliine cyprinid fishes.
By G. J. Howes

Printed by Henry Ling Ltd, Dorchester

Bulletin of the

British Museum (Natural History)

The echinoderms of Aldabra and
their habitats

N. A. Sloan, Ailsa M. Clark &
J. D. Taylor

Zoology series Vol 37 No 2 20 December 1979

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,

British Museum (Natural History),
Cromwell Road,

London SW7 5BD,
England.

World List abbreviation: Bull. Br. Mus. nat. Hist. (Zool.)

Trustees of the British Museum (Natural History), 1979

ISSN 0007-1498 Zoology series

Vol 37 No 2 pp 81-128
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 20 December 1979

The echinoderms of Aldabra and their habitats

N. A. Sloan

Department of Zoology & Comparative Physiology, Queen Mary College, University of London,
Mile End Road, London El 4NS. Interim address: Bermuda Biological Station for Research,
St George's West 1-15, Bermuda

Ailsa M. Clark & J. D. Taylor

Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD

Contents

Synopsis ............. 81

Introduction ............ 81

Shallow water habitats of Aldabra ......... 82

Seaward habitats 83

Lagoon and Channel habitats ......... 83

Comparison between the echinoderm faunas of exposed and sheltered rocky shores 84

Exposed rocky shore near Point Hodoul (407-080) ..... 84

Sheltered rocky shore on He Picard (056-100) 85

Ophiocomid microhabitats on Aldabra ........ 89

Associates of Ophiocoma species ........ 93

Systematic account ........... 93

Class Crinoidea ........... 95

Subclass Asteroidea .......... 95

Subclass Ophiuroidea .......... 99

Class Echinoidea . . . . . . . . . . . 117

Class Holothurioidea . . . . . . . . . . 121

Acknowledgements ........... 125

References ............. 125

Synopsis

This first full account of the echinoderm fauna of Aldabra Atoll, western Indian Ocean, includes a des-
cription of the shallow-water habitats and two special ecological studies dealing with the microhabitats
of the Ophiocomidae - the dominant family of coral-reef ophiuroids - and a comparison between the
echinoderms of an exposed and a sheltered part of the shore. The remainder consists of an annotated
fauna list, with details of habitats, colours in life, parasites, commensals and other observations on the
individual species. It is based primarily on recent collections and observations by Sloan, but also takes
account of previous work by others at the Royal Society's base on the atoll. Nearly 130 species are
included, of which 20% are recorded for the first time from the islands of the western Indian Ocean
(excluding the Mascarene Islands).

Introduction

Aldabra Atoll is a slightly elevated group of four main limestone islands which form a seaward
rim with a maximum dimension of 34 km from east to west (see Fig. 1). It lies at 924' S and
4620' E, some 650 km off the African mainland. With the exception of Assumption and Cosmo-
ledo, Aldabra is isolated from all land by an oceanic basin. Its east and south seaward shores are
the most exposed, being directly in the path of the SE trades, which blow constantly from April
to November. During the other months more variable winds blow from the northwest. Detailed

Bull. Br. Mus. Nat. Hist. (Zool.) 37 (2) : 81-128 Issued 20 December 1979

81

82

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Fig. 1 Map of Aldabra showing localities mentioned in the text and the one kilometre

square grid.

information on the climate, topography, tidal regime and many other aspects of Aldabra is avail-
able in Westoll & Stoddart (1971).

Collections of echinoderms from Aldabra were made during most phases of the Royal Society's
Aldabra Expedition in 1967-1968, and between November 1977 and May 1978. Most of the collec-
tions have been deposited in the British Museum (Nat. Hist.). Thus there is sufficient material to
warrant a report, as detailed reports on marine groups from this area are generally lacking.
Zoogeographically Aldabra is somewhat isolated by being surrounded by the deep sea and com-
parisons between its echinoderm fauna and that of mainland Africa, the Mascarene Islands and
Madagascar may be of interest.

This paper was initiated by one of us (N. A. S.) who made the observations, provided the
photographs, the comparative study of two rocky shores and the section on ophiocomid micro-
habitats, as well as drawing up the systematic account after identification, or at least confirmation
of identification, had been made by others. A. M. C. augmented the systematic account in some
cases, confirmed or made all identifications except the majority of the holothurians, and super-
vised the literature references. J. D. T. provided the maps, some habitat information, and overall
continuity of the habitat information. Dr F. W. E. Rowe identified most of the holothurians
while on a study visit to London.

The ranges of 30 species are extended to the 'Islands of the western Indian Ocean' in the sense
of Clark & Rowe (1971) (i.e. exclusive of the Mascarene Islands), plus a further nine species need-
ing corroborative records. A new subspecies of Ophiarachnella macracantha is described and two
subspecies of Ophiolepis cincta are recognized, prompting the revival of the long-synonymized
name O. garretti Lyman, 1865.

The important study of Cherbonnier & Guille (1978) on the ophiuroids of Madagascar came
to hand only after completion of the typescript of this paper but a few references to it have been
inserted. It gives keys, descriptions, illustrations and habitat notes for a large proportion of the
ophiuroid species now recorded from Aldabra.

Shallow water habitats of Aldabra

The general relationships of the major shallow water marine habitats are briefly described below
and illustrated in Figs 1 & 2.

A nearly continuous land rim of Pleistocene limestones, breached in four places by narrow
channels, surrounds a large shallow lagoon. On the seaward side of the Atoll the land is terminated
by steep 4 m high cliffs, beyond which an intertidal or shallow sublittoral platform extends up to
450 m and ends abruptly at the steep seaward slope of the Atoll. The lagoon shoreline is extremely

ECHINODERMS OF ALDABRA

83

km
5

>'v#i'

'-w

rock, silt &
algae ^

- marine angiosperms

:- '.' '

sand

\ sublittoral rock platfor

coral

=r| intertidal rock & silt

!i

coral & sand I

mangroves

Fig. 2 Map of Aldabra showing the distribution of the major shallow water habitats.

irregular and almost completely fringed by mangroves growing upon rock or silty substrates. The
rest of the lagoon is very shallow with a water depth rarely exceeding 5 m and with a rock bottom
covered with a variable thickness of fine sands and silt. However, near the channels into the lagoon
the hydrodynamics are much more complex and a mosaic of habitats is developed.

Seaward habitats

Most of the Atoll is fringed by intertidal cliffs which are ramp shaped on the exposed southerly
and easterly shores but vertical or undercut at the more sheltered north and easterly sides. Sub-
aerial, marine and biological erosion has produced severe dissection and topographic complexity
on these cliffs and further details of the habitats may be found in Taylor (1971) and Trudgill
(1976). Along the only large intertidal beach on the Atoll, at the Settlement area on He Picard,
there is an extensive development of beachrock over 1000 m long and 20-30 m wide. It occupies
most of the upper eulittoral zone and the seaward dipping slabs of rock form a habitat for a
diverse assemblage of echinoderms.

The base of the cliffs is continuous with the narrow seaward platform which virtually surrounds
the Atoll; the platform is widest at the western and northern sides of the Atoll and narrowest in
the east. At the western end where the platform is about 450 m wide it is mostly covered by a thin
veneer of sediment and colonized by stands of marine grasses particularly Thalassia and Thalas-
sodendron, but Halodule and Cymodocea also occur. The grasses grow upon medium grade sand,
but cobbles and coral debris may be abundant. Small coral colonies may be common in the deeper
parts of the platform, particularly Millepora and Porites. Large dune-like sand bodies migrate
along the platform with the seasonal wind changes. Towards the seaward edge of the platform
there is a belt about 30 m wide consisting of boulders and cobbles thrown up by wave action.
The boulders rest upon calcareous algae covered rock or upon sand and rubble colonized by
Thalassodendron. On the south and east shores the narrower platform is generally sediment free
and colonized by an algal turf with Laurencia, Dictyosphaera, Turbinaria, Cladophoropsis and
calcareous red algae.

Lagoon and Channel habitats

Except for the mangrove fringe, much of the central lagoon presents a generally monotonous,
uniform habitat. The water is very turbid and the shallow more or less silt covered bottom is

84 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

covered by sometimes extensive growths of algae; particularly common are Halimeda, Cystoseira,
Hydroclathrus, Sargassum and Caulerpa, with abundant epilithic sponges.

However near the four lagoon channels, complex habitat mosaics are developed in response to
the increased circulation produced by regular tidal flushing. Main Channel, the largest channel,
has extensive and diverse coral growths along the edges of the channel and its complicated system
of tributaries. More patchy habitats of sand and isolated coral colonies form an arc around the
'catchment' area and this habitat passes transitionally into that of the central lagoon, mentioned
above. The smaller channel, Passe Houareau, has limited coral growth along its edges with more
or less concentric arcs of marine grasses mixed sand and coral, and sandy habitats. Passe Gionnet,
a small channel, has areas of coral, mixed coral and sand, and sand.

The greatest complexity of habitats is seen in the West Channels area where the land rim
has been breached several times by narrow, shallow channels, only one of which, Passe Dubois,
is deep enough to breach the seaward platform. Immediately within the lagoon behind the channels
an extremely complex patchwork of habitats is developed in relation to the catchment of each of
the minor channels. The main habitat types present include beds of marine angiosperms, par-
ticularly Thalassia and Thalassodendron, patches of abundant algae particularly Gracilaria,
Laurencia and Halimeda, sand patches, mixed sand and coral habitats, and areas consisting of
algally coated cobbles and boulders. Coral microatolls and growths ofPorites, Goniastrea, Pavona
and Millepora are common on the grass and sandy areas, whilst more extensive coral growth
including Millepora platyphylla, branching Porites and Goniastrea occur within the channels.

Comparison between the echinoderm faunas of exposed and
sheltered rocky shores

Exposed rocky shore near Point Hodoul (407-080)

This site was an extremely exposed shore in the form of an intertidal rock platform or bench which
is described in more detail in Taylor (197 la). It projected about 65 m to the seaward from a small
pocket sand beach and it was surrounded on the landward by an undercut limestone cliff. The
rock platform was eroded quite smooth with numerous erosion pits, some of which contained
sand retained by overlying trapped boulders. There were no boulders overlying the platform itself.
This habitat sheltered an echinoderm fauna low in numbers but relatively high in species for
this exposed shore. Holothuria arenicola was the most common holothurian and it occurred as
infauna in the trapped sand, numbering up to three per pit. As cryptofauna under the boulders
over the sand the following species: Afrocucumis africana, Holothuria cinerascens, H. hilla, H.
impatiens, H. leucospilota and H. pardalis, Echinoneus cyclostomus, Echinometra mathaei, Macro-
phiothrix longipeda and Ophiocoma scolopendrina were fairly common, as was Holothuria cineras-
cens in larger crevices. The abundance of crevice-dwelling holothurians increased considerably
while the numbers of O. scolopendrina diminished to seaward. Also in the lower eulittoral, small
epibenthic Actinopyga mauritiana became more common in the rock pools. Some were only about
40 mm long, which is interesting as Bakus (1968) had difficulty in finding small specimens of this
species on the Marshall Islands. Most noteworthy, however, were the great numbers of echinoids
in burrows at the seaward end of the platform.

About 10 m landward from the low water spring level an escarpment about 60 cm high and
1 10 m long ran parallel with the water level. A strip along this escarpment 20 m wide on the land-
ward and 10 m wide on the seaward side contained the area of highest echinoid and holothurian
density. Fifty quadrats of 0-25 m 2 were sampled by random casting over the shoulder along this
strip on each side of the escarpment and 25 vertically-orientated quadrats were sampled along the
escarpment itself using random number tables. The results of the echinoid counts are listed in
Table 1. The vertical escarpment face was dominated by the larger echinoid Stomopneustes
variolaris while the horizontally-orientated burrows were usually filled with Echinometra mathaei.
Table 2 shows the size differences of the burrows of these two species. Almost every burrow was
occupied so the availability of burrows in this area could be a limiting factor affecting the popula-
tion size of the echinoids. Their burrowing activities probably contribute to bioerosion on this

ECHINODERMS OF ALDABRA 85

Table 1 Numbers of echinoids per 0-25 m 2 quadrat

No. of
quadrats
Echinoid counts 1 23 4 5 6 7 8 9 10 11 12 13 14 sampled

Echinometra mathaei above escarpment 30 12 2 6 50\

Stomopneustes variolaris above escarpment 46 3 1 - 50 f

E.

mathaei

on escarpment

1

-

3

10

5

3

2 1 -

25 \

S.

variolaris

on escarpment

-

-

1

4

3

3

- 5 1 3 3 1

1 25 /

E.

mathaei

below escarpment

23

16

6

3

2

_

_____ _

50^)

S.

variolaris

below escarpment

45

2

2

1

-

-

_____ _

50>

Table 2 Sizes of echinoid burrows.

Mean depth

Mean diameter

No. of burrows

(mm) S.D.

(mm) S.D.

Echinometra mathaei

40

67-4 5-1

39-7 3-4

Stomopneustes variolaris

40

117-0 8-5

85-5 7-0

shore (Taylor, 1971a; Trudgill, 1976). The echinoids remain permanently in their burrows which
trap wave-transported drift algae and sea grasses.

There were two holothurian species, Afrocucumis africana and especially Holothuria cineras-
cens, which were common in this very pock-marked habitat. Up to 15 H. cinerascens were found
tightly packed in a single vacant S. variolaris burrow and individuals were usually found wherever
crevice or overhang spaces were available. A. africana was restricted to the smaller crevices
although the two species were sometimes found together. Both these species are passive suspension
feeders that trap suspended food on their sticky tentacles. The holothurians and the echinoids
did not, however, coexist in the same burrows. The E. mathaei from the exposed south coast were
larger than anywhere else on the atoll. Similarly, Khamala (1971) in Kenya and Russo (1977) in
Hawaii both reported that Echinometra mathaei from areas of the strongest wave action were the
largest. Interestingly, there were no Echinostrephus molaris in the lower eulittoral here although
they were present on more sheltered seaward shores of the west and north coasts.

In summary, this exposed shore was dominated by echinoids in the lower eulittoral although
two species of holothurians were quite common. A small population of six holothurians, two
echinoids and only two ophiuroids occurred in the mid and upper eulittoral. The conspicuous
dearth of ophiuroids in this habitat is noteworthy and is probably due to the exposed nature of
this rocky shore.

Sheltered rocky shore on He Picard (056-100)

This shore consisted of a series of eroding slabs of beach-rock in the upper eulittoral with a sandy
beach to the landward side and a grass pool to seaward. This area is described in detail in Taylor
(1971, 1976) and Sloan (in press). There were boulders overlying intact bedrock here, unlike
the exposed shore, probably because of the less violent wave action. These boulders provided
shelter for cryptofauna and numerous holothurians occurred under them. The most common
species were the suspension feeder Afrocucumis africana and the four deposit feeders, Polycheira
rufescens, Holothuria parva, H. impatiens and H. leucospilota. The first three of these species were
quite common and detailed information on the distribution and abundance of this fauna is pro-
vided in Sloan (in press). Other uncommon holothurians from the beach-rock were H. cinera-
scens, H. atra, H. rigida, H. arenicola and H. moebii. Echinoids were rare in this area and a total

86

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

8

03

si

at c u 13

O

o

.^

^r

^ S

*o

o

c

o

C

]>

c "a
c
i: BJ

Bj
M

C

3
X)

3

^

IM

Zi O C " * C

s sl 8 |E .

|
u

T3

12
1

1

is, in Thalassodend
id coral heads
1 in small crevices

O

g

5 "^

j

!s
c

P
^ >

o
Z

E

o
T

the microhabitats

Hi ill i Hi

5|l S|| ^ 87S
|1B S, u - g *K*?|

! 8 .8 'Sc ! a 1 C -g 2

f *% KHS ^ s'S S"
Sos .S * * 1 9 8 ?^

c

13
o

'a

1

E
o

(5

1

C^

C
1

"V

CJ > O

<u b "5

ol-o
coc

Bj c 3

^ ~ "I

r; u-
o "O c^

>?

1

IE

c

^c

2

o
xi

t!

o
o

fe

o

3

</5

T3
J5

cS
o

Bj

"O

slands: crevice dwelling on seaward platforrr
aon platforms covered by Thalassodendron ai
ill Islands: in the lower eulittoral under cora

12

O
o

c _

(U
M T3

e

8,'s

(N *2

o

i_

(U

</)
3

U

2

d "i

s^X ^.

a

3 oo
O *^
60 1>

3 _

IX I

S O
1^, O

-o ^
^ u

T3
^

CO

a> *
H

1 ^
II

(^N

o

a
u
o

BJ

15

GO

c

<u

T3

^= 3
c
u. X)

- u

T3

3 Bj

-O f

t- m
^ u
"5 ^>

O <u

c
|

5
>.

_s
1

'Lo
U

c
_o

~c
u

^
r~

ON
>

> si

imon' in heads of Pocillopora, Acropora and
ider these heads, a 'more active species than
ed. Stressed the 'complete separation of habi

>lands and Lord Howe Island: mentioned thi
:h O. scolopendrina, often found in 'dead and
.mbique: considered a 'rare' species in coral ii
in that area
mn O. scolopendrina and O. brevipes on grass
, fronts and algal ridges
cryptofauna on seaward platform and the a
ly some confusion here between O. erinaceus
i extended intertidal beach-rock plates densiti
ach-rock the population is compressed into a
s this species as using the feeding methods w

H* r^ ^

C/^

G

^ C/J r*

*^ . .

T 1

-^

^N "^

fc=x> Ni-E9.Mt3iS6.fi. goo

1

*o

a -a "S
c c -a
J3

boulder

J

III

S

|

11

C

c c

BS O

go

BJ -a

8iis& s ll- R ||ii41

> ? .23.5e'2, C 5S2Jg3

h < "O -i; 3 oo fc : v- o Q. ,Bj .ti ^

> u c^2<H o S S 8 ..<-

<L>

CO

O

~"c 2

i

5 H

-S'-o "^

'c

1 c

"o O

~ BJ

& '$

. . S as Si B _r*S. v 6 i: if Q -S *<

lj* J -C^*J~O^ 3 ri l ^C nS - o '^'r- -Q

r" < O ' r\ d ~ rtra^ "^^

JS

2 </5

r;

8 yL

y c ^

aj

3 'c/5

^ 'S

s '^

^^ ^>

rt o"o2rtis'""a'22 3-S

, including t

on
<u

O

Z

a> 3 ^
SS.E

Hawaii; u

u. "5
60^

8 5 1 IS

II

'u <U

D J2

I. <-

Eniwetok
O. doederl

r\ ^

"S *

s

u C/3

si

iii'jtimraftj

JillJfliljtislii

2

IS

o

R

03

/-^ x-^

0\

- - 2

o

<N *

; ^'

s s

o

a\ as

- -

- ^ rT

^^

r7

^v

ON ON ^ .22

^J"

I-H ^^ r** > -^T^ r^.

E

UN UN

2

S

1^ 1,

g

o\

r^

o\

^ ^ 0\ 00 u 2 <3N

'g

</i

^~^

2

o?i ~~ -*

' '

^

^~^

BJ Bj s -' KJ ^ c

uu

o'

s ~'

^^ O
u. u O

U

g

u"

^ u j? r;!rs

o

o

J J

5

rt

Bj

BJ

j j s -Ills

a
o

3

KK

a

H H U

1

U

1

. ? r3 rd ?3 ri -
I I 05 H H H UH

c

o

2

rt

.

T3

c

^

a

?5

-5

*

(U

C/5

f

1

1

^

Ik

i

c

3

-O

^

cu

s

g

o

^

o

*w

C/5

.Si

.5

^

&

'G

a
(in

O*

i

1

ECHINODERMS OF ALDABRA

87

72) Eniwetok Atoll : the most abundant and widespread Ophiocoma species, maximum densities in
heads of Heliopora coerulea of 6 per 0-002 m 3 (or 6 x 10 3 per m 3 ), ajso found in Pocillopora
heads and occasionally in Acropora and Millepora, most numerous in proximal region of the
coral heads. Heads of 0-01 m 3 frequently had 15 O. erinaceus present in crevice habitats, up to
5 per m 2
4) SE Polynesia: in live and dead coral heads like Acropora and Pocillopora and under boulders,
less common on sandy and more common on more solid substrates
) Jeddah, Saudi Arabia : 'replaces' O. scolopendrina to the seaward on reef flats

1921) Torres Strait and Lord Howe Island: not common; coral head infauna like O. erinaceus
1946) Torres Strait: not common, 'very secretive habitats' in coral heads
7) Inhaca Island : seen only on 'more tropical' reef at Ponta Torres
0) Hawaiian Islands and others in that area: most common in heads of Pocillopora meandrina var.
nobilis, smaller specimens very common in coral whereas larger individuals can be found under
boulders, coral rubble and living coral
72) Eniwetok Atoll: common in living and dead Heliopora coerulea, but preferred Pocillopora
elegans, and occasionally under rocks and in crevices. Less common than O. erinaceus in coral
heads, and tended to occupy the distal areas of coral heads in contrast to O. erinaceus
4) SE Polynesia: confirmed strong association with P. meandrina and suggested that young may
settle in these heads preferentially, adults may be found under coral and in 'non coral' habitats

7) Inhaca Island; a 'rare' member of the coral infauna but more common on the 'more tropical'
reefs and in dead coral heads
0) Eniwetok and SE Polynesia; in shallow sublittoral areas occurs as coral infauna, in both live
and dead coral and under coral rubble, 'confined to the shallow sublittoral zone within a coral

substratum'
4) SE Polynesia: a specimen from Hawaii on black coral Antipathes from 70 m, generally found in
shallow areas in or under live or dead coral

1921) Torres Strait: very common 'occupies a restricted and peculiar habitat near high- water mark' as
crevice fauna, first descriptions of the characteristic 'swaying arms' of this species at the air-
water interface during a flood tide and suggested at least a respiratory function if not a feeding
function for this
7) Inhaca Island : 'abounds' under algal mat at the fringe of Cymodocea (Thalassodendron) beds on
sand flats
Egyptian and other tidal Red Sea coasts : as crevice fauna on sheltered rocky shores between
middle and low tides, up to 50 per m 2 . The first worker to describe in detail the unique feeding
methods of this species at the air-water interface during flood tides
0) General distribution : confirms 'restricted littoral habitat' of this species
/is (1970) Mahe: 'abundant' under coral boulders and in the bases of living corals, possible confusion
here with O. erinaceus

fj\

r~ r~ ^^ ^ ' >o t~-

ON r-~ "">

^.

f^.

s_^

""> In t~~ *

Wl

i-H

ON r~- ^ >> ON ON

c, 2} a a OC-

JD ON ON

ON

ON

g

ON ON

I

a? c/i U U j<" *

"* ^ ^

?

<U

U

X 3 ^

t

_c . ^ ^

C t/")

% 2

C
>

c
(*j

J

00 c Bo
a S 1 >"

6

<U 3 . . "J3 d>

a & &

&

u

Q

*

35

2 Q H

.

i

R

^

1

1

1

1

5,

^

&S

55

Q

Q

g

^

g

o

2

.0

._

88

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

C

U

a

C/5 Q)

XI

B

O

c

C
<U

_

Is

"3 =
en

rt

'1

oo

03

'c

<*. *

1/3

a T7

U

"O

v^

'C

wT "

4> g-

X>

^> ^

C

<L>

c x:

3 ~
X J

S
o

1

-3

"O

\~ o

W c^

to

8

J2 "u

(L> -C

1 C/5
U

>uld be some confusion betwe*

00 to

'C c

a 8

rt f)

s I

II

o

c 2

O

na on rocky shores sheltered f

5

n

r3

c:
u_

1

c

o

N

15
f2

o

c

ntertidal of reef flats'

f Cymodocca (Thalassodendron]
lower eulittoral but common

idges' seaward of grass beds

ocks and coral fragments on 1
e, 'common under coral colon

al colonies on the . . . reef flat

bble cryptohabitat and a coar

O 'a>
u

"eb 3

1 1

o Z_

00 (jj

c ~~

CM

OS
C 3
O O

E E

C OS
n a)

C/J

O

Z

o
o

1

c

IM

X)
03

oj

X>

<*-
O

"3
O

species here
Diego Garcia Atoll : on seaward platform, 'abundant'
lagoon platforms covered with Thalassodendron and c

O. erinaceus here
Eniwetok Atoll: 'upper intertidal zone' as crevice fau
surf, up to 100 per m 2 but 20 per m 2 more common

^c
-a

L-i

i

p

c

cs
_0

JO
3
_

15

2

"5

UJ
C>0

reef platforms, densities of up to 50 per m 2
Jeddah: 'abundant under loose stones on the higher i

Inhaca Island; not common in algal mat of the fringe o
once among coral, rare as cryptofauna under rocks of

tropical' reef area
Mahe: common on grass beds and 'sand and cobble i

Green Island, Queensland: 'very common . . . under i
SE Polynesia: in live coral and under coral and rubbl
sand' along with O. dentata and O. doederleini

Torres Strait : 'quite common in dead portions of cor

i Florida keys: turbulent reef flats, areas with ample ru

>

X)

03

CS

S
-5

O

!_.

03

OO

1)

OS

E

) Florida keys : as for O. echinata and pumila but less c<
turbulent areas
[Abundance of the last three species increases with th

ON

ON

ON

q

fN

^

_O

JD

p;

^

r7

?

r v

r^

2 ^

ON

">>

c3

01

ON

~

ON

o7

ON

0?
NO

* ' r-

^

H

H

In

C

ON

^4

,

C-

ON

u- O

oj

^

4

y

O

to

^

o

O

>>

c

8

<yj

^^

U w

60

60

60
C

Xj

x:

O

c

OS

r3

60

jD

l o3
>

J

~7>

"5

^jj

3

.S2

o3

3

"oj

OS*

U

C/3

.22

.

<

UH

H

U

PC

H

IQ

X

3

^

*

J$

.53

S

Ci,

<3

^

<3

o

a

Q

55

5;

60

^^

_c

Si

*t#

Si

Si

53

5

1

Q

i

"S

i

H

>

s

w

S

a

s

"vl

^

<3

Q

g

s

1

5

CO

o
tj

g

o

O

5

5

.0

2

2

2

2

^

CJ

<y

s:

5

5

a

GO

1

1

1

ECHINODERMS OF ALDABRA 89

of eight individuals of the following species were found; Echinoneus cyclostomus, Eucidaris
metularia, Tripneustes gratilla and Echinothrix calamaris. Ophiocoma scolopendrina, the only
ophiuroid in this area, occurred in great numbers, especially under boulders over rubble at the
seaward end of the beach-rock, see site 1 in Table 4.

The difference in the echinoderm fauna between the exposed and sheltered beach-rock sites is
noteworthy although the latter site occupied only the upper eulittoral which the former occupied
both the upper and lower eulittoral. Firstly, there was the predominance of deposit-feeding holo-
thurians over suspension-feeding ones on the sheltered shore. Although Afrocucumis africana
was the most common species on the sheltered shore in terms of numbers, this small species was
greatly overshadowed in relation to total holothurian biomass and H. cincerascens was rare (only
three found). Secondly, echinoids were much less common on the sheltered shore, although they
are not generally well adapted to life in the upper eulittoral. (For instance, Stomopneustes
variolaris was not present on the sheltered shore.) There was no comparable dominance by
echinoids in the lower eulittoral of the whole seaward platform of the sheltered shore of the west
coast although echinoids, at least six species, were more common under the boulders of the lower
eulittoral overlying sandy gravel than in the upper eulittoral overlying the bedrock. Finally,
ophiuroids were much more abundant on the sheltered rocky shore. Also their species diversity,
at least ten species, increased in the lower, non-rocky, eulittoral of the sheltered shore platform.
Ophiuroids, being less robust in construction compared to echinoids and holothurians, find
exposed wave-washed shores less amenable habitats.

The few species able to exploit successfully the rigorous conditions of the exposed shore clearly
find it a productive environment. The two echinoids make their own cryptohabitat by burrowing
while the two holothurians exploit crevice space and vacant burrows. There are no appreciable
quantities of deposited food but plenty of suspended material is made available by wave action.
The relative lack of cryptohabitat on the vigorously scoured exposed shore compared to the
boulder-strewn sheltered shore is an important factor in the establishment of a deposit-feeding
holothurian cryptofauna on the sheltered shore.

Ophiocomid microhabitats on Aldabra

The most prominent family of shallow-water tropical ophiuroids is the Ophiocomidae and one of
its genera, Ophiocoma Agassiz, 1836, dominates among coral reef ophiuroids (A. M. Clark,
19766). All ophiocomids are cryptic and live in crevices, in or under boulders and coral heads,
and at the bases of sea grasses and algae. They can be present in great numbers and dominate the
local echinoderm cryptofauna and indeed all the cryptofauna. With the exception of Charlock's
(1972) study of 'niche separation' among seven Ophiocoma species from Eniwetak Atoll, Marshall
Islands, observations on the ecology of this important group are scattered and generally anec-
dotal. Some interesting recent observations on ophiocomid microhabitats from the literature,
starting with H. L. Clark's (1921) Torres Strait paper, are listed in Table 3. These will be compared
with recent findings on the Aldabran ophiocomid species.

Table 4 lists a quantitative assessment of the microhabitats of Aldabran ophiocomids from the
eulittoral and shallow sublittoral. Deeper areas were not sampled extensively as they were much
less productive. The microhabitats sampled were:

1. (056-100) An upper eulittoral site where the slabs of beach-rock overlying rubble had been
eroded from a band of intact beach-rock to the landward. These loose slabs remained immersed
throughout neap tides since a seaward pool, formed by a further seaward sand bar, could not
completely drain during these tides. The sand bar, however, was quite mobile, so the size and
depth of this pool, and therefore the immersion time of slab-over-rubble microhabitat, was quite
variable. The slabs were measured, drawn to scale on graph paper, the cut-out weighed, and thus
their area calculated. The ophiuroids under these slabs were identified and counted. They were
also weighed moist but free of excess water and debris.

2. (054-091) A lower eulittoral site consisting of a dense stand of Thalasso dendron ciliatum (For-

90

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

\

\

H

5

\

!/)

\

a

\

ro

\

SO

E

* \

"^ * *

H-Ol

i t*5

J2

2

1 ^ V>

'

r^* fN ^^

OO
\ <N ON

o c
if 3

9

^

fS <N ON

\ -

, If

ON

^2

~

\ "

(N

\ -

o~s

r?, -V

|

s

*

t "5

Q S

^

i

oo

<: j
S..S

o ^*

o

se

ooo

dl

r-

1

<N

'C

d SJ

o

OOO

J

V*

^

p

.5

^

O-^

"f

(S

o

O

t ^i

0-*

006

%

is

5 . *

5

2 i-S

* bX

so 22

2 i-S

<J

a <a

o 8"!.

m <N

* *s^

f^

<s m

ex. ^ .P

oo

sp

oSt.

o o

O O SO

O

1

u

&j

c

Q

*

H <

<;

B

'0

(N

o

3

00

S

00 fi

dl

o 6

f^4

6 o

d

5 ^

^

.c

S -S

\

o

1

1 1

O ^i

1

1

o

"2 . "s
8 *

\ o o

6

d

ON ^|-

P -S

* *

* -4

< <1

w **1

oo so

oo r-.

O v->

5

2

^ d|

SO ^"

6 6

^9 o
f^

6 6

1

J5

*

#

* <i

^

^

o 22

2

tr 53

W . e

| O'S,

o
6 o

o>

O ^,

6 o

d

ro ^5

o

d 5^

8

VI

u g

S.

oo ""*

T3
|

E

c^ . .5

1 o t

* *

S3
6 6

* <

oo o\
o -^

fs co

O SO

d

(N

E

o.

.2 '5

j

O

& |

5

O ^

o

* *

* ^

^

E

^? rO

8

o ^*

oo -<t
m fN

3 >~-

O op

00

o

d

s?

I

u.

I 1

o

^ 6

O O rs

6 o

c a g

L,

j

c .2-S-

* *

* <

&

j2

efl * &

SO -<t

oo <s

6

00

*

!,{

2 O -o

r 17-

O fS

O S

606

o o

-

^oo

-f~- -H

d

fN fN

T3

1 ed S3

*7

ON
S

ss

*

s2 ifl

VO

I i

1 O ON

o

a
1 ^

rt

15

**" ^

2 !

OS

O ^/

?

ry^ '"^

|

^S

1

rN 2-

_u

llll

laal

*

* *

<N O

SO O

<
O 1 m

o o

.3

i

e

K

r- oo

"a. c

C l>

on 2

ECHINODERMS OF ALDABRA 91

skaal) on a wave-washed platform with occasional erosional pits filled with boulders. Ophiuroids
were common on the platform only under the occasional boulder and rubble patches that overlay
sandy gravel and some small rubble. The sea-grass root mat was very strong indeed and thus
maintained the integrity of this habit. To landward the platform consisted of mixed stands of
Thalassia on sandy, higher profile areas and Thalassodendron on rubble and in depressions with
standing water - as far as the sandy beach.

3. (054-091) A very sandy site adjacent at the lower eulittoral to site (2). This is an area of numer-
ous boulders where the sea-grass, probably Thalassodendron, although reduced to a worn stubble,
still provided the substrate with some degree of stability. Under the boulders was fairly fine white
sand. The reduced state of the sea-grass could be due to gradual inundation by sand as, to the
landward, there was a large sand bar bordering sand flats with little sea-grass to prevent erosion
of the substrate during rough weather. Ophiuroids were present only as cryptofauna.

4. (054-091) An area on the other side of site (2) of firmly placed boulders over coarse rubble with
some trapped sand. The boulders were 'cemented' together with deposited gravel in irregular
groups that contributed a marked third dimension to this habitat, unlike sites (2) and (3) above.
To landward a Thalassodendron platform was present, being elevated on an approximately 25 cm
escarpment. Ophiuroids were present only as cryptofauna.

5. (065-090) This lagoon habitat, a short distance from the northernmost western channel, was a
sand flat covered with a dense stand of Thalassia hemprichii (Ehrenb.) with abundant algal
Halimeda sp. and Gracilaria sp. (see Hughes & Gamble, 1977, p. 336). The Thalassia formed a
substrate for a considerable epiphytic algal mat. This area drained for up to 3 h after low tide but
never completely dried out as there was always a residual 2-3 cm of standing water present.
Interestingly, on slightly higher sandy substrates in this area that drained fully at low tide, there
was sea-grass but much less algae present and a greatly reduced epifauna. Ophiuroids were present
as cryptofauna at the bases of the sea-grass and under or in the algae.

The next three microhabitats come from the same area. This was another lagoon habitat
(068-083) adjacent to Passe Dubois of the West Channels. It consisted of discrete patches of coral
heads in a matrix of dense sea-grasses, mostly Thalassodendron and some Thalassia. This was a
sublittoral habitat as a residual 10-15 cm of water was present at the lowest ebb of the tide. The
Ophiuroids here were either in or under the three types of coral head microhabitat.

6. Microatolls of Porites lobata Dana up to 1 -20 m in diameter. It should be noted, however,
that other potentially microatoll forming Porites species occur on Aldabra (Rosen, pers. comm.).
They are P. sp. cf. P. luta Edwards and Solander, P. australiensis Vaughn, and P. somaliensis
Gravier. The microatolls formed a cryptic microhabitat covering sandy gravel.

7. Coral heads of Porites nigrescens Dana of up to 55 kg. Ophiuroids occurred here mainly as
infauna.

8. Coral heads of Millepora exacea Forskaal of up to 7-60 kg. Ophiuroids occurred almost com-
pletely as infauna. The relatively small heads of this hydrocoral had much smaller interstitial
spaces than P. nigrescens above. Although relatively barren of ophiocomids, this was a production
microhabitat for ophiotrichids such as Ophiothrix (Keystonea) propinqua.

General distribution notes on these ophiocomid species around the whole of Aldabra are given
in the systematic account. All the species were most common on the more sheltered shores like
the seaward platforms of the coast and lagoon flats, so their microhabitats were examined in
these areas. The only ophiocomid excluded here is the diminutive Ophiocomella sexradia, since
estimates of its abundance are unreliable due to sampling error.

Ophiocoma brevipes: This relatively small species was most common at site 3 where it occurred
partly buried in sand under boulders in the lower eulittoral. Up to 14 specimens were found under
one boulder, compared with a population of only two per m 2 in Charlock's (1972) optimum
habitat. This is the most sand-loving Ophiocoma species on Aldabra, as well as in other parts of

92 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

its range, as listed in Table 3. Interestingly, O. brevipes was uncommon both under microatolls
over sandy substrates in the shallow sublittoral of the lagoon (site 6) and also relatively rare
throughout the sublittoral, regardless of substrate.

Ophiocoma scolopendrina: This species is widely reported as common in the upper eulittoral of
rocky shores throughout its range. Aldabra is no exception, the species occurring in crevices or
under boulders in large numbers in the He Picard beach-rock at site 1. This was the only ophiuroid
present in an upper eulittoral habitat and was only rarely found in the lower eulittoral or the
sublittoral.

Ophiocoma valenciae: Like O. scolopendrina, this species was very particular, abundant in its
chosen habitat and rare elsewhere. O. valenciae dominated at site 5 on the sheltered grass flats in
the bases of sea-grass covered with algal mat and standing water throughout low tide. It was rarely
found elsewhere in the eulittoral or the sublittoral. In 1968 Taylor reported that this species was
common in areas seaward of grass beds as well as in the grass beds themselves.

Ophiocoma doederleini: In the lower eulittoral this species was abundant on coarser substrates
such as gravel or rubble under boulders but even more numerous, however, in the shallow sub-
littoral under the microatolls at site 6. There are few positive records from the literature because
of confusion with O. dentata (formerly insularia) but the two references do list a similar micro-
habitat of sandy gravel under boulders in shallow water. Unlike the findings of Charlock (1972)
on Eniwetok, on Aldabra this species is much less solitary, up to 16 specimens being found under
a single microatoll.

Ophiocoma erinaceus: A large species and the most ubiquitous ophiocomid in the lower eulittoral
and shallow sublittoral. In the lower eulittoral it dominated under boulders in the most rocky
habitat, site 4. It was common under the microatolls and abundant as coral head infauna in site
7 (P. nigrescens) and indeed in all corals with large interstitial spaces. This species is well known
in the literature as coral head infauna and cryptofauna under boulders, although less mention is
made of its marked preference for rocky rather than sandy substrates. Also, there has been some
confusion between it in the past and the upper eulittoral rocky shore species O. scolopendrina.

Ophiocoma pica: Not a common species on Aldabra where it was generally restricted to the
shallow sublittoral in association with coral. Fairly common in heads of P. nigrescens but abun-
dant under the microatolls in site 6; reported in the literature as coral head infauna rather than
cryptofauna. Devaney (1970) noted that larger specimens could be found as cryptofauna and this
is also true on Aldabra.

Ophiocoma pusilla: This little species is suspected to be more common than appears from
Table 2. Being small and highly active when disturbed makes thorough collecting difficult. On
Aldabra O. pusilla is rather an ubiquitous species in the lower eulittoral, although less common
on the sandy substrates. Therefore it was not confined to the shallow sublittoral, as Devaney
(1970) reported it to be at Eniwetok and S.E. Polynesia. In the sublittoral of Aldabra it was
abundant under the microatolls, although present also as coral head infauna.

Ophiomastix caryophyllata: The dominant ophiocomid under the microatolls at site 6 and com-
mon also as coral head infauna. This rather fragile species is interesting as it could be an impor-
tant competitor with Ophiocoma species in coral microhabitats in the shallow sublittoral of
Aldabra. Restricted to the sublittoral and reported elsewhere as a dead coral infauna species.

Ophiomastix koehleri: The only specimen was found under a microatoll. Probably a sublittoral
species but this requires confirmation from other areas.

Ophiomastix venosa: The largest ophiocomid on Aldabra. Basically restricted to the sublittoral
although occasionally found under rubble in standing water, like Ophiarthrum elegans below, in
the eulittoral of seaward platforms. It was collected particularly under microatolls at site 6
although it could not be considered a common species. No references were found to the habitats
of this species.

ECHINODERMS OF ALDABRA 93

Ophiarthrum elegans: A fairly common species whose sublittoral microhabitat compares with
the two references in Table 3. It was most common at site 6 under the microatolls but uncommon
as coral head infauna.

The greatest diversity of ophiocomids and indeed ophiuroids as a whole was under the sub-
littoral microatolls in the lagoon at site 6. The area is one of persistent tidal currents providing
quantities of suspended food yet it is sheltered from disruptive wave action. The patches of coral
heads were surrounded by grass beds in which the ophiuroids could perhaps forage at night while
protected from predatory fish like wrasses.

The success of Ophiocoma species throughout the tropics could be related to their robust form
and adaptability. They can occupy less benign eulittoral microhabitats as well as the sublittoral
ones from which other related genera are excluded. Indeed, no other ophiuroid genus, of any
family, is so well represented on tropical eulittoral shores. Having the resilience to cope with the
rigours of eulittoral existence may have enhanced the competitive ability of Ophiocoma species in
the shallow sublittoral, although Ophiomastix, as a successful specialist in this area, must be
contended with. The monopolization by O. scolopendrina and O. valenciae of their particularly
rigorous microhabitats is clearly a highly successful strategy. Temperature tolerance, up to 40 C
for O. scolopendrina (Charlock, 1972), at low tide at midday would be an essential prerequisite
for life in both these microhabitats. A tolerance to lowered salinity due to periodic tropical down-
pours would also be important. Both species belong to different morphological subgroups of
Ophiocoma (Devaney, 1970) so that the evolutionary potential to withstand physiological stress
is perhaps a characteristic of the genus as a whole. Interestingly, O. anaglyptica Ely, a member of
the scolopendrina subgroup, dominates under dead coral and rubble in the algal ridge (sublittoral
fringe) of seaward platforms at Eniwetok Atoll to the virtual exclusion of other ophiuroids (Char-
tock, 1972). It was present in its preferred area at densities of up to 150 per m 2 (240 g per m 2 ).

Kohn (1971) suggested that co-occurring non-predatory tropical marine invertebrates are more
likely to have a specialized microhabitat than specialized food type, as do predatory gastropods.
Charlock (1972) supported Ihis hypolhesis Ihrough his observalions on Ihe seven Ophiocoma
species al Eniwelok Aloll. He found lhal Ihe ophiuroids were 'apparenlly non-seleclive' general
delrilus feeders wilh significanlly differenl microhabilal specializalions. In Ihis presenl sludy,
each of Ihe Ophiocoma species demonslraled appreciable differences in Iheir microhabilal or
combinalion of microhabitat choices. It seems likely lhal on Aldabra, like Eniwelok, Ihe availa-
bilily of and compelilion for suilable cryplic microhabilals influences the presence of ophiocomids.

Associates of Ophiocoma species

A wide range of macroscopic animals occur in association wilh species of ophiocomids, some
recorded from the present collection and olhers from Ihe lileralure. References lo Ihose found wilh
Indo-Wesl Pacific species of Ophiocoma are given in Table 5.

Systematic account

II should be menlioned lhal only Ihree ophiuroid species oul of Ihe 41 lisled in Hughes & Gamble
(1977) can be confirmed by us. Their specimens were incineraled lo oblain Ihe ash-free dry masses
lo eslimale biomass per melre 2 for each species. They do, however, cile Dr F. W. E. Rowe as
having checked Iheir hololhurian idenlificalions.

Dr R. N. Hughes has kindly supplied us wilh an unpublished appendix upon which Ihe paper
of Hughes & Gamble is based, while Dr W. F. Humphreys has also made available lo us his
Aldabra specimens and an unpublished lisl of preliminary idenlificalions of Ihem made in 1971.
His contribulion has added len confirmed species new to the fauna of Aldabra.

Throughoul Ihis accou.nl, references are kepi lo a minimum by ciling major works when pos-
sible, such as Clark & Rowe (1971) or Rowe & Doly (1977), from which Ihe original species de-
scriptions can be Iraced. After Ihese cilalions Ihere are in some cases biological references lo Ihe
species. These are included as an indication of the available biological information.

94

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Oph

U

^

a.

o

-3

_0

H

pe

JJIiiilllf Hull!!

M

a

CJ

CO

o
o

a

!

s:

,

Cf

5 ^

53 CL

.60131^53

-o

'-

||

2 SJ

^ -o

d

.. >

U

-

tj
d

dd

~ s

nigropuncta,
lems and tJ

E

O O

II

0.5
* s

ECHINODERMS OF ALDABRA 95

Class CRINOIDEA

COMASTERIDAE

Comanthus sp. juv.

One specimen taken from within the coral frame-work on a patch reef in the Main Channel lagoon

drainage system (104-110).

MARIAMETRIDAE

Lamprometra klunzingeri (Hartlaub, 1 890)
See: A. H. Clark, 1941 : 527; A. M. Clark, 1972 : 102.

One specimen; taken during daytime from under a coral overhang 25-30 m deep off the research
station beach (053-091).
Range extended from East Africa and Madagascar.

Stephanometra indica (Smith, 1876)

See: A. H. Clark, 1941 : 436; A. M. Clark, 1972 : 107.

During daytime specimens were collected from under overhangs of patch reefs in the Main

Channel drainage system (104-110), under coral heads on the lagoon grass flats near the West

Channels (068-083), in Passe Dubois (060-080), 25-30 m deep off the research station (053-091),

and commonly on rubble at the entrance of Passe Houareau (316-122).

COLOBOMETRIDAE

Oligometra serripinna (Carpenter, 1881)

See: A. H. Clark, 1947 : 216; A. M. Clark, 1972 : 129.

One specimen taken from under rubble in Passe du Bois (060-080).

ANTEDONIDAE

Dorometra mauritiana (A. H. Clark, 1911)

See: A. H. & A. M. Clark, 1967 : 69; A. M. Clark, 1972 : 141.

Abundant in the coral frame-work on patch reefs, especially Acropora patches in the Main

Channel drainage system (104-110) and (120-080) during the daytime.

Colour in life: uniform pale green, mauve, orange or purple.

Range extended from the Mascarene Islands.

Subclass ASTEROIDEA

ASTROPECTINIDAE

A stropecten polyacanthus phragmorus Fisher, 1913

See: A. M. Clark, 1974 : 433.

Only reported by Hughes & Gamble (1977) (as A. phragmorus) and listed as coming from the

east side of Passe Houareau (318-118). We think that this specimen was more likely to be A.

polyacanthus, recorded below.

Astropecten polyacanthus Miiller & Troschel, 1842

See: Fisher, 1919 : 63; A. M. Clark & Rowe, 1971 : 44.

One found washed up on West Point (056-106) (R = 59 mm) and another from sand patch in the

West Channels (060-070 to 060-090) (R = 56 mm).

GONIASTERIDAE

Stellaster sp. juv.

One specimen from algae (Gracilaria & Laurencia) on Thalassia in the West Channels (060-070

to 060-090).

OREASTERIDAE

Culcita schmideliana (Retzius, 1805)

See: Jangoux, 1973a : 18; Thomassin, 1976 : 51 (feeding and ecology).

96 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Juvenile (R= 12-15 mm) and subadult (R = 33 & 34-5 mm) found in the lower eulittoral under
coral rubble on the seaward platforms of the west (054-091) and north (140-125) (241-127)
coasts. Adult forms (R = 44 & 48 mm) found under rubble in standing water in the depression of
the Anse Malabar cliff base (241-127).

Colour in life of juveniles: medium green with occasional orange blotches, subadults pale
orange and adults pale green with rich blue spines.

Protoreaster lincki (de Blainville, 1834)

See: Jangoux, 1973a : 23; Thomassin, 1976 : 51 (feeding & ecology).

Interestingly, all were juveniles (mean R = 28-2 mm for the eight specimens) washed up on the
sandy beach in front of the research station (054-091). The animals must have come from the
seaward grass flats although intensive searching among the grass beds failed to yield specimens.
Colour in life: cryptic mottled blue-green, in contrast to the adult colour of red on pink or
pinkish-grey ground.

OPHIDIASTERIDAE

Dactylosaster cylindricus (Lamarck, 1816) Fig. 4

See: H. L. Clark, 1921 : 85.

Two specimens (R = 54 & 57-5 mm) found under boulders over coarse rubble in the lower
eulittoral off the research station (054-091).
Colour in life: the shiny skin uniform burgundy.

Fromia milleporella (Lamarck, 1816)

See: H. L. Clark, 1921 : 40; Marsh, 1977 : 257.

One specimen (R = 25 mm) found on coral at 15 m depth off the research station (052-091).

Colour in life: deep purple above, much paler below.

Range extended from the Mascarene Islands, East Africa and Madagascar.

Leiaster sp. juv.

One specimen from algal mat on Thalassia in West Channels (060-070 to 060-090).

Linckia guildingi Gray, 1840

See: H.L.Clark, 1921 : 67.

Six specimens (mean R = 33 mm) taken from Porites and Goniastrea microatolls in Passe Houareau

(316-120).

BMNH records (unpublished, see A. M. Clark & Rowe, 1971 : 36) include specimens from
Curieuse, Platte and Moyenne Is of the Seychelles.

Linckia laevigata (Linnaeus, 1758)

See: Jangoux, 1973a : 29; Laxton, 1974 : 47 (feeding & ecology); Yamaguchi, 19770 : 13 (growth,

reproduction, population structure).

Two specimens (R= 161, 169 mm) taken from coral flats in the Main Channel drainage system in
the lagoon (108-120) where it was fairly common.

Colour in life: uniform light blue; no other colour forms as seen at Palau (Marsh, 1977) were
observed.

Hughes & Gamble (1977) list this species but their specimen came from grass beds in Passe
Houareau (316-120) and we think it may instead have been L. multifora.

Linckia multifora (Lamarck, 1816)

See: Jangoux, 1973a : 32; Rideout, 1978 : 287 (asexual reproduction).

Abundant on coral rubble and sand among patch reefs in the Main Channel drainage system of

the lagoon (120-080). One particular patch reef which consisted mostly of Acropora acuminata

Verrill, a species new to Aldabra (B. Rosen, pers. comm.), had densities of 2-3 L. multifora per m 2

(R=114mm maximum). In the same habitat were found extreme densities of the holothurian

Stichopus chloronotus Brandt of up to 12 per m 2 , although their usual density was around 2 per

m 2 . Elsewhere in the lagoon L. multifora was found under boulders around He Esprit (1 14-063),

beneath Porites and Goniastrea microatolls in Passe Houareau (316-120), among algae in north

ECHINODERMS OF ALDABRA

97

Figs 3, 4 Ophidiaster hemprichi M tiller & Troschel and Dactylosaster cylindricus (Lamarck),
each viewed from above and below. (Scale = 10 mm.)

98 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

and south lagoon shores. Less common on seaward coasts but present on the algal covered sub-
littoral fringe on the south coast at Dune Jean Louis (275-039) and near Point Hodoul (407-080)
and on the north coast at Anse Malabar (241-127).

Some specimens were infested with the parasitic gastropod Thyca crystallina (Gould, 1846).

Colour in life: pale flesh tone with numerous purple blotches and pale blue arm tips.

Nardoa sp.

Reported only by Price (1971 : 169) from the grass beds of the West Channels. The identification,

however, was made in the field by another worker and may not be reliable.

Neoferdina offreti (Koehler, 1910)

See: Jangoux, 19736 : 778; Marsh, 1977 : 263.

Two specimens (R = 25-5 & 31 mm) were found on coral 1-0 m deep at low tide along the sides of

the drainage channels of Main Channel in the lagoon (104-1 19) (108-120); another (R = 28-5 mm)

was found on coral in 15 m off the research station (052-091).

Colour in life: disc centre pink, some plates with large violet spots, the rest off-white, pink
below.

Ophidiaster hemprichi Miiller & Troschel, 1842 Fig. 3

See: de Loriol, 1885: 22 (as O. purpureus Perrier); A. M. Clark & Rowe, 1971 : 61.
Fairly common under boulders over coarse rubble (range of R = 23-45 mm for 9 specimens) at
the lower eulittoral off the research station (054-091); in similar microhabitats of the seaward
platforms at Passe Gionnet (140-125) and Anse Malabar (241-127).

Colour in life: burgundy with irregular white flecking.

The considerable similarity of appearance of this species and Dactylosaster cylindricus promp-
ted a suspicion that the two may represent growth stages of a single species. The skin of D.
cylindricus, however, is obvious and shiny even in the dried specimens whereas the skin of O.
hemprichi is not evident. Both have coarse granulation in the middle of the arm plates but in
O. hemprichi this is surrounded by fine interstitial granulation rather than the thick skin of
Dactylosaster. Both species are also sympatric in Mauritius from which a small specimen of un-
doubted D. cylindricus exists in the BMNH collections.

Tamaria lithosora H. L. Clark, 1921

See: H. L. Clark, 1921 : 90.

Reported only by Hughes & Gamble (1977). Listed as coming from a thick stand of Halimeda

on the west side of Passe Houareau (316-119).

ASTEROPSEIDAE

Asteropsis carinifera (Lamarck, 1816)

See: de Loriol, 1885 : 67 (as Gymnasteria carinifera (Lamarck)); A. M. Clark, 1967 : 37; A. M.

Clark & Rowe, 1971 : 65.

Quite common (R = 20-5-68 mm for 11 specimens) under boulders over rubble or gravel in the
lower eulittoral of the seaward platform off the research station (054-091), Passe Gionnet (140-
125) and Anse Malabar (241-127). Also found in heads ofPorites nigrescens Dana from the lagoon
grass flats (068-083) near West Channels.

Colour in life: pale grey with irregular green marking; always red-brown after preservation.

ASTERINIDAE

Asterina burtoni Gray, 1 840

See: James & Pearse, 1969 : 84; A. M. Clark & Rowe, 1971 : 68; Achituv, 1969 : 329 (reproduc-
tion & distribution).

Fairly common (R = 3-13 mm for 7 specimens) in heads of Porites and Acropora in the lagoon
(068-083, 120-080, 115-067). Individuals with 6-7 arms were the most common. Also recorded
from the bases of Thalassia on lagoon grass flats (063-089), associated with the algal mat on the

ECHINODERMS OF ALDABRA 99

sea-grass of this area, and under boulders on the seaward platform off the research station (054-
091) and West Channels (060-070 to 060-090).

Dr Humphreys' specimens include 5 five-armed ones and 25 fissiparous multiradiate or newly-
divided ones with up to 8, usually 7 arms. The maximum R is 7-5 mm. The spinelet form appears
similar in the two lots of specimens, the abactinal ones slightly more elongate and tapering than in
Red Sea individuals (see Clark & Rowe, 1971, fig. 17a), more like those of the specimen from
Lord Howe Island, Australia (fig. 17g), usually with only three points at the tip. However, it is
notable that the five-armed specimens have no more than three furrow spines, whereas the multi-
radiate ones of similar size have four or sometimes five.

It needs to be emphasized that the two syntypes of A. burtoni are either fissiparous (the six-
armed one) or potentially so (the five-armed one), since both have multiple madreporites. Use of
the name A. wega Perrier for fissiparous Red Sea specimens is therefore untenable since that must
be a synonym of A. burtoni. If any second name should be needed for five-armed specimens with
single madreporites from this area having squat abactinal spinelets (as opposed to the attenuated
spinelets of A. burtoni cepheus M tiller & Troschel), then a new name must be proposed.

ACANTHASTERIDAE

Acanthaster planci (Linnaeus, 1758)

See: H. L. Clark, 1921 : 101; Sale et al., 1976 (review of biology); Yamaguchi, 19776 : 283

(zoogeography).

Found on coral in the lagoon from: West Channels (060-070 to 060-090), nearby grass (068-023),
and along the edges of the Main Channel drainage system (104-1 10) (108-120) where it could be
quite common. Found on coral off the seaward platform of the west coast (052-091) and the
north coast at Passe Gionnet (140-125) where it was very common.

Colour in life: reddish-brown.

MITHRODIIDAE

Mithrodia clavigera (Lamarck, 1816)

See: Marsh, 1977 : 276.

Two specimens (R = 89-5 & 91 mm) were found under boulders over rubble at the lower eulittoral

off the research station (054-091), another was among coral and sponge in Passe Femme (060-087),

and one under a boulder in Thalassodendron in Passe Houareau (316-120). Unpublished BMNH

records included a specimen from Mahe, Seychelles.

Colour in life: pale flesh tone with large brick-red patches, especially proximally, arm tips
black.

Subclass OPHIUROIDEA

Note: all measurements are of dried specimens and thus the discs in particular are likely to be
smaller than they would be in life. Abbreviations: d.d. = disc diameter; a.l.=arm length.

OPHIOMYXIDAE

Ophiomyxa australis Lutken, 1869 Figs 5, 6

See: Devaney, 1974 : 115.

All 6 specimens (maximum d.d. = 23mm; mean a.l. :d.d.=4-9 : 1) were found under Porites
microatolls in the lagoon grass flats (068-083) near West Channels.

With the exception of the oral papillae the whole animal is covered with smooth skin. In life
the disc is very soft and amorphous.

Colour in life: ranging from uniform light orange to uniform blood-red to disc blood-red with
arms distally banded with yellow, in fact similar to that of two specimens collected by Dr M.
Yamaguchi from Palau, Caroline Islands (BMNH reg. nos. 1975.11.25.45, 46).

One Aldabra specimen had on its disc an undescribed parasitic gastropod species of the genus
Hemiliostraca Pilsbry, 1917 (A. Waren, in preparation). This is the first recorded host for this
group of Eulimacea.

100

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

6

Figs 5, 6 Ophiomyxa australis Liitken, red specimen with banded arms and uniform
orange-coloured specimen. (Scale = 10 mm.)

ECHINODERMS OF ALDABRA 101

AMPHIURIDAE

Amphiodia (Amphiodid) dividua Mortensen, 1933

See: Mortensen, 1933 : 176; Cherbonnier & Guille, 1978 : 92-93, fig. 41.

Taken from under, as well as in, coral alongside and at the bottom of Passe Dubois (060-080).
Range extended from the Mascarene Islands and Madagascar.

Amphioplus (Amphioplus) impressus (Ljungman, 1867)
See: A. M. Clark, 1970 : 63; Cherbonnier & Guille, 1978 : 60-62, fig. 23.
Coral crypto- and infauna along both sides of Passe Dubois from Thalassodendron flats.
Range extended from Madagascar and the western Pacific.

Amphioplus (Lymanelld) hastatus (Ljungman, 1867)

See: H. L. Clark, 1939 : 75; A. M. Clark & Rowe, 1971 : 102.

Reported by Hughes & Gamble (1977) only. Listed as coming from grass flats in the lagoon
(063-089) near West Channels.

If correct, this would extend the range from East Africa and Madagascar.

Amphioplus (Lymanelld) integer (Ljungman, 1867)
See: Balinsky, 1957 : 1 1 ; Cherbonnier & Guille, 1978 : 86-87, fig. 38.

Coral crypto- and infauna along the edges of Passe Dubois (060-080) in the Thalassodendron flats
Range extended from East Africa and Madagascar.

Amphipholis squamata (Delle Chiaje, 1829)

See: Devaney, 1974 : 125; Hughes & Gamble, 1977 : 335 (microhabitat information).
Mentioned in Hughes & Gamble (1977) as coming from numerous sites and listed in their appen-
dix from Passe Houareau (315-118) (319-1 19), the seaward platform off West Channels (056-090),
in the lagoon (060-089) near West Channels and the seaward platform of Anse Malabar (241-
127). Particularly abundant in sheltered muddy habitats and associated with dense algal turf on
sea-grass in sheltered habitats of these areas. Also found in the sediments of land-enclosed tidal
pools at the eastern end of He Malabar (297-1 18) and He Picard (059-094).

We agree with Hughes & Gamble (1977) on the abundance of this ubiquitous little species of
sheltered shores.

Amphiura (Amphiurd) inhacensis Balinsky, 1957

See: Balinsky, 1957 : 11.

Infauna taken from boulders and dead coral heads in the lower eulittoral off the Settlement

(055-108).

Range extended from Inhaca Island, southern Mozambique.

OPHIACTIDAE

Ophiactis picteti (de Loriol, 1893a)

See: A. M. Clark & Rowe, 1971 : 104; Cherbonnier & Guille, 1978 : 123-125, fig. 56.
Associated with sponge in the bases of heads of Porites nigrescens and Millepora exacea Forskaal
from lagoon grass flats (068-083) near West Channels. Common in dead coral boulders at the
seaward end of Passe Dubois (060-080); d.d. up to 6-1 mm.

Colour in life: disc uniformly brown or with irregular brown markings on white, arms banded
white and brown with some dark spots on the white bands, ventrally the disc and arms white.

This is an extension of range from Madagascar and from Amboina, Indonesia - the type locality.
Additionally, several specimens from Watamu, southern Kenya and the Pemba Channel, Tan-
zania, collected by W. F. Humphreys, are referable to this species. These, with the Aldabra speci-
mens, were initially determined as Ophiactis hemiteles H. L. Clark (type locality Torres Strait)
following Balinsky's record from Inhaca, Mozambique. However, the presence of only six, rarely
seven, arm spines at d.d. 5 mm, rather than seven or eight and the markedly truncated form of all
but the uppermost spine, with the tip more or less bihamulate, does not agree with H. L. Clark's
description of the spines of hemiteles as sharp. Also these specimens from the western Indian
Ocean all have a narrow median distal lobe to the oral shields, which H. L. Clark describes as
simply elliptical in hemiteles.

102 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

In the key given in Clark & Rowe, 1971, O. picteti (with O. sinensis Mortensen) and O. hemiteles
were poorly distinguished by the adoral shields being relatively large and interradially contiguous
only in hemiteles, the shields of picteti being obscured by skin. In the present specimens the skin
is sufficiently transparent to show the limits of the shields, which are rather variable in extent
interradially; even in the same specimen some may be contiguous while others are not. Nor is the
relative arm length reliable in distinction. The arms were broken in the holotypes of both species.
The estimates of their length against d.d. were given as x 8 for hemiteles and x 6 for picteti but de
Loriol noted that his estimate of the arm length was very approximate. The arms are broken in
most of the present specimens but a small one has a.l. 33 mm, d.d. 3-5 mm, a ratio of 9-4/1 ; in
another of similar size it is 8-4/1. One unusual feature that they show is a pair of slight longitu-
dinal grooves on each ventral arm plate, just as in de Loriol's fig. 2b (1893a) from the holotype
of O. picteti. This character in conjunction with the blunt arm spines supports recognition of the
Aldabra and East African material as O. picteti rather than O. hemiteles and is in accord with the
identification by Cherbonnier & Guille (1978) of a malagasy specimen as O. picteti.

It may be noted that preserved specimens with the colour muted have considerable resemblance
to the occasional five-armed specimen of Ophiactis savignyi, especially in the relatively large radial
shields, two distal oral papillae each side of the jaw, elliptical dorsal arm plates and truncated
rugose arm spines, also the light-coloured patch on the distal part of each pair of radial shields.
However, the absence of the small median distal lobe on the dorsal arm plates (so characteristic
of O. savignyi) and the isolation of the scales in the skin of the central part of the disc in specimens
of O. picteti of d.d. more than c. 4-5 mm, should serve to distinguish them.

Ophiactis savignyi Muller & Troschel, 1842
See: Balinsky, 1957 : 14; A. M. Clark & Rowe, 1971 : 103.

Commonly found associated with sponge in the bases of heads of P. nigrescens and M. exacea
from lagoon grass flats (068-083) near West Channels, extremely common on sponge on rubble in
the central intake of the land-enclosed tidal pool on He Picard (059-094), common in the inter-
stices of Halimeda in Passe Houareau (316-114), in algal mat and sea-grass bases in lagoon grass
beds (063-089) near West Channels. A. 1. max. = 15 mm; a.l. : d.d. =4-7 : 1.

Colour in life: basically green, disc with fairly regular light and dark green markings, arms
banded with light and dark green, ventrally disc and arms white with some distal green banding
on the arms.

Ophiactis versicolor H. L. Clark, 1939

See: A. M. Clark, 1967 : 43 (as Ophiactis cornea but as O. versicolor in footnote).

Taken from coral along the edge of Passe Dubois (060-080) where it occurred as both coral crypto-

and infauna.

Range extended from East Africa, Madagascar and the Mascarene Islands.

OPHIOTRICHIDAE

Macrophiothrix demessa (Lyman, 1861)
See: Devaney, 1974 : 139.

Occasionally found in or under Porites heads or microatolls in the lagoon grass flats (068-083)
near the West Channels. D.d. max. = 12-5 mm; a.l. :d.d. = ll : 1.
Colour in life very variable but not spotted like M. longipeda below.

Macrophiothrix longipeda (Lamarck, 1816)
See: Devaney, 1974 : 140.

Usually found as solitary individuals under boulders and coral heads over sand or sandy gravel
with their arms extended upward along crevices in the covering rock or coral. Found in the lower
eulittoral of wasted platforms off the research station (054-091), Anse Malabar (241-127) and
near Pt Hodoul (407-080). Occurs sublittorally off the research station (052-091), in lagoon grass
flats (068-083) near West Channels, Main Channel drainage area (108-120), and in Passe Houareau
(316-119) associated with Halimeda besides coral heads and boulders. As growth proceeds, the

ECHINODERMS OF ALDABRA 1 03

arm length increases in relation to the disc diameter, e.g. d.d. max. = 9-5 mm, a.l. : d.d. = c. 18 : 1 ;
d.d. 17mm, a.l. :d.d = 27 : 1.

Colour in life: bluish with dark blue spots, the spinose disc with large spotted radial shields,
ventrally lighter blue with dark spots.

Ophiothrix (Key stoned) prop inqua Lyman, 1861

See: Balinsky, 1957 : 21; A. M. Clark & Rowe, 1971 : 107.

Exceedingly common in coral heads with small interstitial spaces like Millepom exacea in the

lagoon grass flats (068-083) near the West Channels. Also abundant in dead coral boulders as well

as live coral in the West Channels (060-070 to 060-090), Main Channel drainage system (108-120),

and Passe Houareau (316-120). D.d. max. = 8 mm, a.l. : d.d. = 8 : 1.

Colour in life: very variable as noted by Balinsky (1957) but overall dark bluish with darker
arm banding.

Ophiothrix trilineata Lutken, 1869

See: Balinsky, 1957 : 30; Devaney, 1974 : 150.

Fairly common in heads of Millepora and Porites from lagoon grass flats (068-083) near West

Channels, in coral in the Main Channel drainage system (108-120) of the lagoon and in coral

along Passe Houareau (316-120). D.d. max. = 8 mm; a.l. : d.d. = 5 : 1.

Colour in life: disc always dark blue, arms basically blue or green with five characteristic
median lines alternating white and dark blue; arm spines glassy, reddish and sometimes with a
dark line, ventrally white overall.

Ophiothrix (Acanthophiothrix) purpurea von Martens, 1867

See: Devaney, 1974 : 141.

Commonly found epizoic on Millepora tenera Boschma from 5 to 30 m on the sides of Passe Du-

bois (160-180), along Main Channel drainage system of the lagoon (104-110), at 15-30 m off the

research station (052-091), at 10-30 m off Passe Gionnet (140-125) and along Passe Houareau

(316-120). D.d. max. = 1 mm; a.l. : d.d. = 9-5 : 1. Very delicate and difficult to collect intact.

Much more common at night (Humphreys, pers. comm.).

Colour in life: disc patterned in red, arms dark red with narrow red median line outlined in
white, arm spines red, ventrally lighter red with a dark median line along the arms.

Ophiothela tigris Lyman, 1871 Fig. 7

See: A. M. Clark & Rowe, 1971 : 116; A. M. Clark, \916a : 111 (epizoic habits).
Epizoic on the stinging hydroid Aglaophenia cupressina Lamouroux, 1816 from the coral flats in
the Main Channel drainage system of the lagoon (104-1 10); also found on heads of P. nigrescens
and M . exacea in the lagoon grass flats (068-083) and near West Channels on coral along Passe
Houareau (3 16-1 20). D.d. max. = 7 mm; a.l. :d.d. = c. 3-5 : 1 but relative arm length very variable.
Colour in life: disc rich green but centrally with pattern of black and yellow concentric mark-
ings within pentagonal outline, arms uniform green but spines distally white and proximally
blue, ventral colour light blue overall.

OPHIOCOMIDAE

A more detailed account of the microhabitats of this family on Aldabra is given earlier
(pp. 89-93), also a table summarizing the known associates of Ophiocoma species.

Ophiarthrum elegans Peters, 1851
See: H. L. Clark, 1921 : 139; Devaney, 1974 : 150.

Occurs in pools at the bases of Thalassodendron in rubble on the seaward platforms of the west
(058-091) and north (241-127) coasts, under rubble in the lower eulittoral of these seaward
platforms, under coral over sandy gravel in the lagoon grass flats (068-083) near West Chan-
nels, along Main Channel drainage system (108-120) and along Passe Houareau (316-120). D.d.
max. = 22 mm; a.l. : d.d. =4-5-9-5 : 1, the arms extremely variable in length.

104

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Fig. 7 Ophiothela tigris Lyman. (Scale = 10 mm.)

Colour in life : disc uniform black, arms pale cream with red bands always fading after preserva-
tion, arm spines pale cream with black spots sometimes forming irregular annulations, ventrally
overall white.

Ophiocoma brevipes Peters, 1851

See: Devaney, 1974 : 151.

Prefers sandy substrates under boulders on seaward platforms of the west (058-091) and north

(241-127) (360-112) coasts. Also found at the bases of Thalassodendron and Halimeda in these

areas as well as Passe Houareau (316-120) and West Channels (060-070 to 060-090). D.d. max. =

19mm; a.l. :d.d. = 3-6 : 1.

Colour in life: disc off-white with irregular green markings, arms off-white with pale green
banding, ventrally off-white overall but sometimes stained by amber-coloured gut regurgi tat ions.

Commonly found carrying the polychaete associate Holopidella nigropunctata (Horst, 1915),
occasionally with Gyptis ?sp., and once with a specimen of Lepidasthenia sp. as well but this
could have been only incidental (Dr P. E. Gibbs, pers. comm.).

Ophiocoma doederleini de Loriol, 1899 Figs 8-10

See: Devaney, 1970 : 14; 1974 : 154.

Previously recorded from Aldabra as Ophiocoma dentata Miiller & Troschel, 1842.

Fairly common under boulders over gravel and rubble and in the bases of Thalassodendron in
the lower eulittoral of seaward platforms of the west (058-091) and north (241-127) coasts. More
common in the shallow sublittoral under coral microatolls, especially Porites, over sandy gravel
in lagoon grass beds (068-083) near West Channels and along Passe Houareau (316-120).
D.d. max. = 27 mm; a.l. : d.d.=4-2 : 1.

Colour in life: basically mid-grey, disc grey with fine black reticulating lines or, less commonly,
with white-ringed black spots, or speckled with light spots, arms commonly with spotted white
bands all the way round, arm spines always annulated white and grey, ventrally the same grey as
dorsally, oral shields white peripherally with large irregular grey blotches.

Over 80% of all specimens were infested with the new pycnogonid species Anoplodactylus
ophiurophilus Stock (1979). Some O. doederleini were found carrying the polynoid H. nigro-
punctata.

ECHINODERMS OF ALDABRA

105

8

Figs 8, 9 Ophiocoma doederleini de Loriol, five- and four-armed specimens. (Scale = 10 mm.)

106

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Fig. 10 Ophiocoma doederleini, detail of disc and arm bases, from below. (Scale = 10 mm.)

Ophiocoma erinaceus Miiller & Troschel, 1842 Figs 11, 12

See: Devaney, 1974 : 155.

The most ubiquitous ophiocomid on Aldabra. Common in coral heads in all the channel areas of
the lagoon and off the seaward platforms of the west and north coasts. Present on all the seaward
platforms on all coasts of the atoll and particularly associated with rock and rubble rather than
sandy substrates. D.d. max. = 30 mm; a.l. : d.d. = 4-5 : 1.

Colour in life: all black, some had ventral arm plates with some white trim, occasional indi-
viduals had bright orange podia.

The least infested of the three Ophiocoma species which bear the pycnogonid A. ophiurophilus.

Ophiocoma pica Miiller & Troschel, 1842

See: Devaney, 1970 : 25; 1974 : 159.

Found in coral like O. erinaceus, but less commonly, wherever coral was found in the lagoon

channel areas or off the seaward platforms on the west and north coasts. Not common in the

eulittoral under rubble. D.d. max.= 18 mm; a.l. : d.d. = 4-0 : 1.

Colour in life: disc with thin gold radiating lines on a dark brown background, long dark brown
arm spines, dumb-bell-shaped gold and cream bands on the dorsal arm plates which were other-
wise dark brown, ventral arm plates with a similar pattern, oral shields dark brown with white
lateral edges.

This species has an infestation level by A. ophiurophilus between that of O. doederleini and
O. erinaceus.

Ophiocoma pusilla (Brock, 1888)

See: Devaney, 1970 : 25; 1974 : 160.

Common in the lower eulittoral under rubble on the coastal platforms of the west and north

coasts, under coral heads of lagoon grass beds (068-083) near West Channels, Passe Houareau

(316-120) and Main Channel (108-120). D.d. max. = 7 mm; a.l. : d.d. = 2-7 : 1.

Colour in life: disc uniform dark brown, occasionally spotted with white-rimmed dark spots,
arms dark brown with lighter banding, ventrally overall red-brown.

Ophiocoma scolopendrina (Lamarck, 1816) Fig. 13

See: Devaney, 1974 : 161.

Externally common in the upper eulittoral zone of all rocky seaward shores around the atoll.
Uncommon in grass beds of seaward and lagoon platforms or in lower eulittoral or shallow sub-
littoral. Common around the bases of lagoon islets. With increasing size the arms become rela-
tively longer, e.g. d.d.= 13-4 mm; a.l. : d.d.=4-6 : 1 ; d.d. = 27 mm; a.l. : d.d. = 6-0 : 1 ; d.d.=
29mm, a.l. :d.d. = 10-2 : 1.

ECHINODERMS OF ALDABRA

107

72

Figs 11, 12 Ophiocoma erinaceus Muller & Troschel, six-armed specimen from above and
four-armed specimen from below. (Scale = 10 mm.)

108

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

13

14

Figs 13, 14 Ophiocoma scolopendrina (Lamarck) and six-armed specimen of O. valenciae
Miiller & Troschel. (Scale = 10 mm.)

ECHINODERMS OF ALDABRA 1 09

Colour in life: patterns extremely variable, disc usually dark with irregular off-white markings,
dorsal arm plates brownish with irregular distal white spots, arm spines sometimes annulated,
sometimes spotted, sometimes orange-tipped, ventrally overall off-white, oral shields occasionally
with grey blotches.

Occasionally white triclads were found on the ventral sides of specimens.

Ophiocoma valenciae Miiller & Troschel, 1842 Fig. 14

See: H. L. Clark, 1921 : 131; Devaney, 1970 : 29.

The dominant echinoderm at the bases ofThalassia that is heavily covered with algal mat, among
which Halimeda and Gracilaria is prominent. Such habitats occurred on sheltered grass beds that
never completely dried at low tide in the lagoon (065-090) near West Channels and Passe Houareau
(316-118). Rarely found in lower eulittoral of seaward platforms or in shallow sublittoral under
coral. D.d. max. = 22 mm; a.l. : d.d. = 5-5 : 1.

Colour in life: disc uniform dull green, arms dull green with occasional darker banding, arm
spines dull green, ventrally overall pale green. During preservation the green colouration usually
turns a light tan brown.

Ophiocomella sexradia (Duncan, 1887)

See: A. M. Clark & Rowe, 1971 : 99 & 118; Devaney, 1974 : 162.

Not too common as coral-head infauna of P. nigrescens and M. exacea from the lagoon grass

flats (068-083) near West Channels, in sea-grass bases and coral along Passe Houareau (316-118).

May be a sponge associate like the similarly fissiparous and usually six-armed Ophiactis savignyi.

D.d. max. 4 mm; a.l. : d.d. = 3-8 : 1.

Colour in life: disc dark green, arms dark green with light green banding, ventrally paler green
overall.

This confirms the unspecified extension of range noted in A. M. Clark & Rowe (1971 : 86)
to the islands of the western Indian Ocean from East Africa and Madagascar.

Ophiomastix caryophyllata Lutken, 1869 Fig. 15

See: H. L. Clark, 1921 : 137; A. M. Clark & Rowe, 1971 : 120.

Common in coral heads in all the channel areas of the lagoon and off the seaward platforms of the
west and north coasts. More abundant even than O. erinaceus under Porites microatolls on the
lagoon grass flats (068-083) near West Channels. D.d. max. = 19 mm; a.l. : d.d. = 7-0 : 1.

Colour in life: the spiny disc white with large dark purple spots outlined with white, both basal
and ventral arm plates with a distal V-shaped white band superimposed on dark purple, oral
shields outlined white with a central labyrinthine purple design.

Ophiomastix koehleri Devaney, 1977 Fig. 16

See: A. M. Clark & Rowe, 1971 : 118 (as Ophiocoma wendti: Koehler); Devaney, 1977 : 275;

Cherbonnier & Guille, 1978 : 186-188, pi. 11, figs 1, 2.
Previously recorded from Aldabra as Ophiocoma wendti : Koehler by Hughes & Gamble (1977).

One specimen only, found under a Porites microatoll over sandy gravel in the lagoon grass
beds (068-083) near West Channels. D.d. = 15 mm; a.l. = 106 mm; a.l. :d.d. = 7-0 : 1.

Colour in life: the spineless disc uniform dark purple with white-marked edges, dorsal arm
plates basically white with large irregular purple blotches which may occupy most of the plate so
that about two consecutive segments out of five are almost completely dark, forming an arm
band, large dorsal clavate arm spines pale purple or mottled, small arm spines annulated white
and purple, ventral arm plates white with large proximal purple blotches that can occupy almost
the whole plate in positions corresponding to the dark areas dorsally, tentacle scales banded, oral
shields with large dark purple blotches, podia red.

Ophiomastix venosa Peters, 1851

See: H. L. Clark, 1921 : 138.

Not common; under boulders, coral heads, and Porites microatolls in the lagoon grass beds

(068-083) near West Channels, along Passe Houareau (316-118), in depression at cliff base of

Anse Malabar (241-127), seaward platform of the south coast when it occurred under boulders

110

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

\

\ \

-

\ X

\

15

Figs 15, 16 Ophiomastix caryophyllata Lutken and O. koehleri Devaney. (Scale = 10 mm.)

ECHINODERMS OF ALDABRA 1 1 1

(345-037). D.d. max. = 41 mm;a.l. :d.d. = 5-5 : 1, the ratio not increasing with absolute disc size.
Colour in life: basically bright yellow, the almost completely spineless smooth disc yellow with
dark lines outlined in white, arm spines, including the enlarged clavate upper ones, yellow with a
dark line, dorsal arm plates with an ill-defined darker yellow median band, ventrally pale yellow
overall, podia yellow.

OPHIONEREIDIDAE

Ophionereis dubia (Muller & Troschel, 1842)
See: A. M. Clark & Rowe, 1971 : 122.

Recorded by Hughes & Gamble (1977) and listed as common in a thick stand of Halimeda in
Passe Houareau (316-117) and from dense Thalassia on the west coast seaward platform (056-
090). Specimens of confirmed identity have been found in and under coral along the edge of
Passe Dubois (060-080). Owing to poor preservation, no measurements or colour notes can be
given.
Range extended from East Africa, Madagascar and the Mascarene Islands.

Ophionereis porrecta Lyman, 1860

See Devaney, 1974 : 174.

Fairly common in the crevices of boulders in the lower eulittoral of the west and north coast

seaward platforms, in heads of P. nigrescens and M. exacea from lagoon grass flats (068-083)

near West Channels, in heads of Millepora platyphylla Ehrenberg from the bottom of Main

Channel. D.d. max. = 15 mm; a. 1. :d.d. = 9-0 : 1.

Colour in life: disc greyish with irregular dark markings, the dark pattern much more promi-
nent in young specimens, arms mottled white and grey with dark spots, the short spines white,
ventrally white with grey blotches overall.

OPHIODERMATIDAE

Ophiarachna affinis Liitken, 1869 Figs. 17, 18

See: A. M. Clark & Rowe, 1971 : 123; Devaney, 1974 : 175.

One specimen found under a Porites microatoll over sandy gravel in lagoon grass flats (068-083)
near West Channels.

Colour in life: strongly resembling that of Ophiocoma doederleini, basically mid-grey, disc light
grey with widely-spaced black spots and a pale ramifying linear pattern, arms mid-grey with white
bands, arm spines annulated white and grey, including the elongated ventralmost spines, ventrally
mid-grey, oral shields white with irregular grey markings. The overall impression is that this speci-
men is less robust in construction than O. doederleini. D.d. = 25 mm; a.l. = 97 mm; a.l. : d.d.=
3-9 : 1.

Range extended from Indonesia.

Ophiarachnella gorgonia (Muller & Troschel, 1842)

See: H. L. Clark, 1921 : 141; A. M. Clark & Rowe, 1971 : 125.

Three specimens found under Porites microatolls over sandy gravel in the lagoon grass flats

(068-083) near West Channels. D.d. max. = 21 mm; a.l. : d.d. = 5-l : 1.

Colour in life: disc off-white with irregular light brown markings, radial shields mottled white
and grey with small black spots, arms off-white with mid-brown bands one to three segments
wide, ventrally overall off-white with arms darker distally.

Ophiarachnella macracantha aldabrensis subsp. nov. Figs 19-21

See: H. L. Clark, 1909 : 126; A. M. Clark & Rowe, 1971 : 126 (for O. macracantha).
Holotype: B.M. reg. no. 1978.9.1.1, from under a Porites microatoll over sandy gravel in the
lagoon grass flats (068-083) near West Channels. D.d. = 19 mm, a.l. = 83 mm; a.l. : d.d.=4-4 : 1.
Colour in life: disc blood-red with radial shields the same colour, arms basically blood-red with
a lighter linear pattern distally, also with pale transverse bands superimposed on this distal linear
pattern, ventrally the arms paler, oral shields not white, unlike the other specimens mentioned
below, but with a very broad red band that may cover up to half the oral shield.

112

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

17

$5-

18

Figs 17, 18 Ophiarachna affinis Liitken, detail of disc and arm bases from below,
whole animal from above. (Scale = 10 mm.)

This specimen, still vividly red after being dry for 6 months, was compared with two Pacific
examples of O. macracantha, a dried one from Palau, western Caroline Islands, collected by Dr
M. Yamaguchi, and a spirit specimen from Fiji, collected by the Challenger (named by Lyman
Pectinura rigida-a. synonym of the closely-related O. septemspinosa (Miiller & Troschel)). Arm
length/disc diameter in the two last is c. 100 mm/26 mm ( = 3-8/1) and 110 mm/21-22 mm
( = 5-1/1) respectively. Both have brown bands on the arms and the Palau specimen still has the
disc pink and grey, as when first dried. The maximum arm spine number in the three specimens is
7, 8 and 9 respectively but the smaller number in the Aldabra specimen can be attributed to its
smaller size. Its enlarged lowest arm spines are particularly flattened and spatulate in form, as

ECHINODERMS OF ALDABRA

113

19

20

Figs 19, 20 Ophiarachnella macracantha aldabremis subsp. nov. (Scale = 10 mm.)

114 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

21

Fig. 21 Ophiarachnella macracantha aldabrensis, detail of disc and arm bases from below.

(Scale =10 mm.)

broad at the tip as the base, whereas in the Pacific specimens the tips are somewhat thickened
(though not quite cylindrical) and more or less narrowed. The longest of these spines just exceed
twice the segment length, equalling 3-0-3-5 mm. It should be noted in all the specimens that only
one in every two or three of the lowest spines is markedly enlarged, a feature which was either
overlooked by H. L. Clark or absent in the type material of O. macracantha from Ponape, eastern
Carolines, where up to nine arm spines were found at R 24 mm. The enlarged spines mostly alter-
nate on the two sides of every second (or sometimes third) segment from about segment eight
onwards. The size of the disc granules and exposed parts of the radial shields and the structure of
the jaws are similar in all three and the minor difference in shape of the arm spines seems to be the
only morphological difference to support the vivid colouration in distinguishing the Aldabra
specimen from the Pacific ones. However, since the brownish colouration of O. macracantha
provides the main distinction from the very dark O. septemspinosa, it seems consistent to treat
this specimen from Aldabra as at least subspecifically distinct from the Pacific material.

There is a parallel precedent for this in the reddish-orange subspecies erythrema Devaney,
1974, from south-eastern Polynesia, of the confusingly similarly-named Ophiarachna megacantha
H. L. Clark (from Australia), patterned with grey or brown (though H. L. Clark did not see any
live specimens).

Ophiarachna megacantha (with O. robillardi de Loriol from Mauritius) and Ophiarachnella
macracantha (with O. septemspinosa) form an interesting convergence. Both pairs of species have
a combination of small bare patches of radial and supplementary oral shields, not found elsewhere
in Ophiarachna, where most species of Ophiarachnella have larger bare shields and also have more
numerous oral papillae and appressed arm spines. Ophiarachna megacantha and robillardi are
distinguished from Ophiarachnella macracantha and septemspinosa by the much longer and fewer
arm spines, numbering no more than six in the first two and that only at disc diameter 25 mm or
more, when the longest lower arm spines are as much as 7 mm long.

The range of Ophiarachnella macracantha is extended from the Caroline Islands and Fiji.

Ophiarachnella septemspinosa (Miiller & Troschel, 1842)

See: H. L. Clark, 1909 : 126; A. M. Clark & Rowe, 1971 : 126.

Occasionally found under coral rubble along Passe Dubois (060-080), in the lower eulittoral off

the research station (054-091), along the sides of Main Channel (108-120). Most common under

Porites microatolls over sandy gravel in lagoon grass flats (068-083) near West Channels where

groups of up to eight individuals were found. D.d. max. = 38 mm; a.l. :d.d. = 4-l : 1, only a

ECHINODERMS OF ALDABRA 1 1 5

slight increase in relative arm length with disc size. The largest specimen weighed 29-3 g whereas
the largest Ophiomastix venosa weighed 23-6 g (d.d. = 41 mm).
Colour in life: uniform dull grey or grey-green overall, radial shields dark brown.

Ophiochaeta hirsuta LUtken, 1869 (with synonym O. boschmai A. H. Clark, 1964)

See: A. M. Clark & Rowe, 1971 : 127; Gibbs, Clark & Clark, 1976 : 129.

One specimen under a Porites microatoll over sandy gravel in the lagoon grass flats (068-083)

near West Channels, two more from coral along Passe Houareau (316-118). D.d. max. = 8-2 mm;

a.l. :d.d. = 3-8 : 1.

Colour in life: uniform grey-brown with pale brown banding on the arms, ventrally off-white
overall.

The varied occurrence of spinelets on the discs of the four specimens from Aldabra and two
from Palau, Caroline Islands, indicate that Ophiochaeta boschmai A. H. Clark, 1964 (type locality
Molucca Islands) is a synonym of O. hirsuta LUtken, 1869. One Aldabra and one Palau specimen
have peripheral and ventral spinelets only (intermixed among the indented granules), as thought
to be characteristic of O. boschmai. The other specimens have spinelets in various degrees of
frequency and length also on the upper sides of the discs. The length of the coarser armament on
the convex marginal plates of the disc is also variable, being usually almost granuliform but
sometimes distinctly elongate.

Cherbonnier & Guille (1978 : 219) have described a new species of Ophiochaeta, O. crinita,
on the basis of a single specimen from Madagascar with the disc armament entirely spiniform
except for a few granules near the genital slits.

Ophioconis permixta Koehler, 1905
See: A. M. Clark, 1965 : 63.

Under boulders and coral in the shallow sublittoral off the Settlement (056-100).
Range extended from East Africa and Madagascar.

Ophiopeza fallax Peters, 1851

See: de Loriol, 18936: 4; A. M. Clark & Rowe, 1971 : 127.

One specimen found under a Porites microatoll over sandy gravel in lagoon grass flats (068-083)

near West Channels. D.d. = 13 mm, a.l. =43 mm; a.l. : d.d. = 3-3 : 1.

Colour in life: disc covered with smooth grey-brown skin, arms the same colour with red-brown
bands up to two segments wide, ventrally off- white overall.

Range extended from East Africa, Madagascar and the Mascarene Islands.

OPHIURIDAE

Ophiolepis cincta Miiller & Troschel, 1842 Fig. 22

See: Balinsky, 1957 : 28; Cherbonnier & Guille, 1978 : 232.

Under boulders, particularly buried in sandy gravel in the lower eulittoral of seaward platforms
of the west and north coasts, under coral heads and microatolls in the lagoon grass flats (068-083),
near West Channels and along Passe Houareau (316-112). D.d. max. 15 mm.

This species occurs in two distinct colour forms. The majority are predominantly dark brown
on the disc (sometimes all brown - Fig. 22 right) with some white spots or larger patches, which in
three specimens forms a pentaradiate pattern with a more or less regular white patch in each
interradius (Fig. 22 top); the upper side of the arms is brown interrupted by white bands in
proportion to the extent of white on the disc, the white bands numbering 3-7 per arm. (When
dried, the brown usually turns to black or dark grey and the underside is paler, especially the disc.
Under the microscope the white areas of the upper side are seen to be finely marbled with dark
grey.) In contrast, some specimens had the colour in life uniformly ochre or grey-green on the
upper side.

These two colour forms are also correlated with differences in the relative arm length. In life
the seven drab uniform specimens collected were estimated (by Sloan) to have a.l. : d.d. = 5-5 : 1,
compared with 2-7-3-8 : 1 for 15 of the boldly coloured specimens. (Following some shrinkage
of the disc in drying, the ratios have fallen slightly to means of 5-2 : 1 and 3-6 : 1.)

116

N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

22

Fig. 22 Ophiolepis cincta garretti Lyman (pale specimen on left) and O. cincta cincta.

(Scale =10 mm.)

Evidently the patterned form is not restricted to Aldabra since Balinsky (1957) records O. cincta
at Inhaca, Mozambique, as 'rather brownish with irregular white spots on the disc and white
cross-bands on the arms, these white areas being finely marbled with greyish'. Also Cherbonnier
& Guille (1978) found specimens from Madagascar (and Dar-es-Salaam, Tanzania) to have the
disc 'marron fonce et beige clair . . ., marron fonce dominant' and the arms banded; they give the
a.l. as c. 3xd.d. In a sample of 38 specimens in spirit in the BMNH collections from Zanzibar
the disc colour is now mainly dull dark brown or grey patterned with light grey marbled spots or
patches, usually interradially and centrally and sometimes making a regular pentaradiate pattern
as in the top specimen in fig. 22. The range of a.l. : d.d. in these specimens is 2-2-4-6 : 1, mean
3-2 : 1 but the sample also includes a single relatively long-armed and uniformly dull brownish
coloured specimen with a.l. : d.d. = 4-6 : 1. Three similar drab-coloured specimens from Aldabra
collected by Taylor and two from Mauritius have a.l. : d.d. 4-4-5-9 : 1.

The type locality of O. cincta is the Red Sea (no details). Miiller & Troschel do not describe the
disc colour but note that the arms are banded dark and light; a.l. : d.d. they give as only 2-5:1.
Eleven specimens in the BMNH collections from the Gulf of Aqaba and the southern Red Sea
(5-15 years in alcohol) still show broad bands on the arms; their discs are mainly brownish or
grey with light grey spots or larger patches, varying in size, shape and position but often with an
enlarged central light patch while sometimes the smaller patches are interradial and in one case
form a pentaradiate regular pattern. They have a.l. : d.d. 3-1-4-1 : 1, mean 3-7 : 1.

Although all these specimens show the distinctive regular arrangement of small platelets almost
encircling the smooth disc plates and bordering each dorsal arm plate on its distal side, thought
to be characteristic of O. cincta, it is clear that two taxa (best ranked as subspecies in the opinion
of A. M. C.) can be distinguished, at least in the western Indian Ocean, one relatively long-armed
and drab-coloured and the other shorter-armed and with a bolder and usually patterned coloura-

ECHINODERMS OF ALDABRA 1 1 7

tion. The latter is clearly the nominate subspecies, O. cincta cincta, of which the Aldabra specimens
may be recognizable as a colour form with a strong tendency for developing a regular pentaradiate
pattern.

In seeking a name for the drab-coloured relatively long-armed subspecies, a possible candidate is
Ophiolepis garretti Lyman, 1865, type locality Kingsmills (i.e. Gilbert) Islands in the Pacific.
The holotype and only specimen had a.l. : d.d. 55 : 9 = 6-1 : 1. In life the colour of the disc was
uniform brick red but the arms were banded with paler areas. Lyman then thought it distinguish-
able from O. cincta not only by the longer arms but also by the shorter oral shields and the rougher
texture of the dorsal arm plates. However, in his Challenger report (1882) he synonymized it with
O. cincta. Unfortunately, no good samples from the Pacific are available. In spite of the banded
arms of the holotype of O. garretti, the long arms and uniformly coloured disc justify reviving the
name for the subspecies showing these characters.

The apparent absence of the widespread species Ophiolepis superba H. L. Clark from Aldabra,
despite intensive collecting, is notable. This conspicuous ophiuroid with d.d. commonly 15-25 mm
has been recorded in the western Indian Ocean from East Africa, Madagascar, Mauritius and
from Mahe. It usually has a bold pentaradiate disc pattern of purple (in life) on light brown but in
this case the light areas are radial, not interradial.

Ophioplocus imbricatus (Muller & Troschel, 1842)

See: H. L. Clark, 1921 : 143; A. M. Clark & Rowe, 1971 : 128.

Fairly common partially buried in sand or sandy gravel under boulders in the lower eulittoral of

the west and north coast seaward platforms, occasionally at the bases of Thalassodendron on

these platforms as well. D.d. max. = 19-5 mm, arm length very variable in every specimen but

a.l. : d.d. usually more than 4:1.

Colour in life: darker grey-green overall than Ophiolepis cincta garretti, the disc with ill-defined,
coarse, dark grey reticulations, arms banded with dark grey, ventrally more grey still with dark
grey oral shields.

Class ECHINOIDEA

CIDARIDAE

Eucidaris metularia (Lamarck, 1816)

See: Mortensen, 1928 : 386.

Common under boulders and at the bases of sea-grass on the seaward platforms of the west and

north coasts, in and under coral in lagoon grass flats near West Channels and Passe Houareau. A

ubiquitous species wherever coral and sea-grass occur with the exception of the exposed seaward

east and south coasts.

Phyllacanthus imperialis (Lamarck, 1816)

See: Mortensen, 1928 : 504.

Abundant on sponge-covered rocks, along with a large population of Ophiactis savignyi, at the

tidal intake to the large land-locked marine pool (059-094) on He Picard. For a description of this

unusual marine pool habitat see Taylor (197 la : 196).

Prionocidaris baculosa (Lamarck, 1816)

See: Mortensen, 1928 : 437.

Found under algae and coral amongst Thalassodendron on seaward platforms of and in old Stomo-

pneustes variolaris burrows at the cliff base of Cinq Cases (396-055).

Prionocidaris verticillata (Lamarck, 1816)

See: Mortensen, 1928 : 428 (as Plococidaris).

At the bases of Thalassodendron on the west and north coast seaward platforms, under coral heads

in lagoon grass flats near West Channels and Passe Houreau. Never as common as Eucidaris

metularia and rarely found as coral head infauna. Only small specimens were found, e.g. greatest

test diameter 20 mm.

118 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

DIADEMATIDAE

Astropyga radiata (Leske, 1778)

See: Mortensen, 1940 : 187.

Epifaunal on a sand patch as a group of 20 in the western lagoon (097-105).

Diadema savignyi Michelin, 1845

See: Mortensen, 1940 : 265; A. M. Clark & Rowe, 1971 : 153.

Under boulders in the lower eulittoral of the west and north coast seaward platforms, more com-
mon in, but especially under, coral heads in the lagoon grass flats (068-083) near West Channels
and in Passe Houareau (316-118). Specimens found were never large, maximum test diameter
50 mm.

Diadema setosum (Leske, 1778)
See: A. M. Clark & Rowe, 1971 : 153; Dart, 1972 : 50 (feeding & ecology); Herring, 1972 : 169

(feeding & ecology).

Reported by Hughes & Gamble (1977), found among sea-grasses on lagoon flats near West
Channels (063-089) and by Price (1971), incorrectly identified, from the same area and habitat.

We feel the record needs confirmation as the differences between D. setosum and D. savignyi
are so subtle (see A. M. Clark, 1967 : 49) and D. setosum appears to be rare in the Mascarene
Islands and East Africa, though it is the dominant or only species in the Red Sea. If correct,
it would provide an extension of range.

Echinothrix calamaris (Pallas, 1774)

See: Mortensen, 1940 : 285; Herring, 1972 : 169 (feeding & distribution).

Under boulders in the lower eulittoral of the west and north coast seaward platforms, most

common, like D. savignyi, in the more sheltered lagoon habitats like the grass flats near West

Channels and Passe Houareau where they occur under coral heads and microatolls.

STOMECHINIDAE

Stomopneustes variolaris (Lamarck, 1816)

See: Mortensen, 1935 : 507; Taylor, 19710 : 183 (ecology on Aldabra); Herring, 1972 : 169

(feeding & ecology).

Abundant in holes along the lower eulittoral and especially the sublittoral fringe of the exposed
east and south coastal platforms. Interestingly, this species was present under rubble on the rock
platform at Dune Jean Louis (275-039) whereas the relatively more exposed rock platform near
Point Hodoul (407-080) had no rubble present and the animals were found only in holes in intact
bedrock.

In the burrows of this species, two species of crustaceans were found: the gnathophyllid shrimp
Gnathophyllum americanum Guerin, 1856 (normally free-living predators on asteroids and ophiu-
roids - Dr A. J. Bruce, pers. comm.) and the porcellanid crab Petrolisthes virgatus Paulson, 1876
(not reported previously as associated with echinoids and a new locality record for this species -
Dr J. Haig, pers. comm.). The crabs at least were found as male/female pairs.

Although unrecorded from the islands of the western Indian Ocean by A. M. Clark & Rowe
(1971), this species was in fact collected by J. D. T. at Mahe and reported in Taylor (1968). It is
also common in the Mascarene Islands and on the East African coast.

TOXOPNEUSTIDAE

Tripneustes gratilla (Linnaues, 1758)

See: Mortensen, 19430 : 500; Herring, 1972 : 169 (feeding & ecology).

Sometimes found under boulders in the lower eulittoral of the west and north coast seaward

platforms, but common under microatolls and coral heads on the lagoon grass flats near West

Channels and Passe Houareau. As with some other Aldabra echinoids, only small individuals

were taken, maximum test diameter 52 mm.

Toxopneustes pileolus (Lamarck, 1816)

See: Mortensen, 19430 : 472.

Reported only by Hughes & Gamble (1977) - from a sheltered site in Passe Houareau (319-118).

ECHINODERMS OF ALDABRA 1 19

PARASALENIIDAE

Parasalenia gratiosa A. Agassiz, 1863
See: Mortensen, 19436 : 269.

From rubble in the bottom of Passe Dubois, at about 10 m.
Range extended from East Africa and Madagascar.

ECHINOMETRIDAE

Colobocentrotus atratus (Linnaeus, 1758)

See: Mortensen, 19436 : 434.

On the lower eulittoral cliff face between Point Hodoul and Anse Cedres (390-103).

Echinometra mathaei (de Blainville, 1825)

See: Mortensen, 19436 : 381 ; Khamala, 1971 : 167 (ecology); Dart, 1972 : 50 (feeding & ecology) ;

Herring, 1972 : 169 (feeding & ecology); Russo, 1977 : 693 (distribution).
The most ubiquitous echinoid on Aldabra, found under boulders, coral heads and in holes on all
the seaward shores, in or under coral heads or microatolls in all the channel areas of the lagoon.
Most common and reaches its largest size in holes in the lower eulittoral of exposed east and south
coast rocky shores, see Taylor (197 la : 183) for habitat details.

As with S. variolaris, on the south coast E. mathaei shares its burrow with the shrimp G.
americanum or the crab P. virgatus.

Echinostrephus molaris (de Blainville, 1825)

See: Mortensen, 19436 : 304; Campbell et al., 1973 : 155 (feeding & ecology).
Common in burrows in coral boulders at places where there is plenty of current like any of the
channels leading out of the lagoon, also found off the seaward platforms of the west and north
coasts, not common in the eulittoral.
Range extended from East Africa, Madagascar and the Mascarene Islands.

Heterocentrotus trigonarius (Lamarck, 1816)

See: Mortensen, 19436 : 420.

Spines and test fragments of this species were found among corals at 1 5 m deep off the research

station (052-091), found live in a crevice at the base of cliffs at Point Hodoul (404-105).

ECHINONEIDAE

Echinoneus abnormalis de Loriol, 1883

See: Mortensen, 1948a : 80.

In sand under coral heads in the lagoon grass flats (068-083) near West Channels, under coral

alongside Passe Dubois (060-080), under boulders in the shallow sublittoral off the research

station (052-091).

Range extended from the Mascarene Islands.

Echinoneus cyclostomus Leske, 1778

See: Mortensen, 1948a : 75; Rose, 1978 : 199 (ecology).

A shallow burrower in sand under boulders on the seaward platforms of all coasts around Aldabra,

also under boulders and coral heads in the lagoon at all the channel areas. Never common.

CLYPEASTERIDAE

Clypeaster fervens Koehler, 1922

See: Mortensen, 19486 : 86.

Reported only by Hughes & Gamble (1977) - infauna at sea-grass bases on the seaward platform

at Anse Malabar (241-127) on the north coast.

Clypeaster reticulatus (Linnaeus, 1758)
See: Mortensen, 19486 : 71.

In sand among sea-grass roots in the lagoon (065-090) near West Channels, on the seaward plat-
form of the west coast, and on sand in the lagoon near lie Chalen (080-074).

120 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

FIBULARIIDAE

Echinocyamus crispus Mazetti, 1896
See: Mortensen 19486 : 185.

In sand and coarse gravel along Passe Dubois (060-080) among Thalassodendron beds.
Range extended from East Africa and Madagascar.

Fibularia ovulum Lamarck, 1816
See: Mortensen, 19486 : 208.

In sand and coarse gravel on the north side of Passe Dubois (060-080) with plentiful Thalassoden-
dron. As with the other channel areas, this is a habitat of strong and persistent currents.

Clark & Rowe (1971 : 144) give no record of this species from the islands of the western Indian
Ocean but their proposed neotype (p. 168) is in fact from Mahe. This is the first record from
Aldabra.

Fibularia volva A. Agassiz, 1846
See: Mortensen, 19486 : 213.

Reported only by Hughes & Gamble (1977) - infauna in soft mud at a very sheltered site in the
Passe Houareau area (319-119).
If correct, this record extends the range from the Red Sea.

ECHINOLAMPADIDAE

Echinolampas ovata (Leske, 1778)
See: Mortensen, 19480 : 275.

Reported only by Hughes & Gamble (1977) - infauna from a muddy, sheltered site in Passe
Houareau (3 19- 11 8).
If correct, this record extends the range from the Mascarene Islands.

SPATANGIDAE

Maretia planulata (Lamarck, 1816)

See: Mortensen, 1951 : 27.

Found in sand in Passe Magnan (063-074) of the West Channels and in sand under boulders at the

edge of the seaward platform on the north coast (308-124).

Pseudomaretia alta (A. Agassiz, 1863)

See: Mortensen, 1951 : 58.

Reported only by Price (1971) - from the grass flats on the seaward side of West Channels.

SCHIZASTERIDAE

Schizaster lacunosus (Linnaeus, 1 758)
See: Mortensen, 1951 : 300.

Reported only by Hughes & Gamble (1977) - infauna from under a thick stand of Halimeda in
Passe Houareau (316-119).
Range extended from Natal.

BRISSIDAE

Brissus latecarinatus (Leske, 1 778)

See: Mortensen, 1951 : 514.

Infauna associated with sand under sea-grass in Passe Houareau (316-118), in sand under boulders

on seaward platforms of the west and north coasts.

Metalia dicrana H. L. Clark, 1917

See: Mortensen, 1951 : 546.

Reported only by Hughes & Gamble (1977) - infauna under sea-grass in the lower eulittoral of

the seaward platform at Anse Cedres (360-1 12).

ECHINODERMS OF ALDABRA 121

Metalia spatagus (Linnaeus, 1758)

See: Mortensen, 1951 : 540.

In sand patches amongst Thalassodendron on seaward platforms of the west and north coasts,

under boulders at the edges of these platforms, and in sand in all the channels.

Metalia sternalis (Lamarck, 1816)

See: Mortensen, 1951 : 535.

In a sand pocket under boulders on west coast seaward platform (063-128); a large test (126x

114 mm) was found on the beach at Anse Polymnie (108-124) in the lagoon.

Class HOLOTHURIOIDEA

HOLOTHURIIDAE

Some useful colour illustrations to many of the species below are given in the recent paper by
Rowe & Doty (1977).

Actinopyga sp. cf. A. bannwarthi Panning, 1944
See: Cherbonnier, 1955 : 136.

One specimen only, found on the lagoon grass flats (068-083) near West Channels.
If this specimen is A. bannwarthi, it extends the range from the Red Sea.

Actinopyga echinites (Jaeger, 1833)

See: Cherbonnier, 1955 : 136; Rowe & Doty, 1977 : 228.

Reported only by Hughes & Gamble (1977) - from the lower eulittoral among seagrass on the

seaward platform at Anse Cedres (360-112).

Actinopyga mauritiana (Quoy & Gaimard, 1833)

See: Rowe & Doty, 1977 : 228; Bakus, 1968 : 24 (ecology).

An ubiquitous epifaunal species on hard substrates at the edge of the seaward platforms on the

west and north coasts and more widespread through the eulittoral on the rock benches of the

exposed seaward east and south coasts. Interestingly the A. mauritiana on the exposed coasts

were much smaller, but more numerous than those on the more sheltered coasts. Occasionally

seen on the coral flats in the Main Channel drainage system in the lagoon. The only epifaunal

holothuroid on the eulittoral rock benches of the exposed coasts.

Large specimens commonly carried the polychaete Gastrolipidia davigera Schmarda, 1861.
Some specimens had a parasitic gastropod, Melanella muelleriae (Sturany, 1904), projecting from
the body wall.

Actinopyga miliaris (Quoy & Gaimard, 1833)

See: Panning, 1944 : 47 (as A. lecanora miliaris).

Found on the sand in Passe Dubois (060-080) and on grass flats in the lagoon (083-085).

Bohadschia marmorata Jaeger, 1833
See: Rowe & Doty, 1977 : 229.

Found on lagoon grass flats (068-083) or under the overhang of isolated coral heads near West
Channels, usually covered with dead fronds of sea-grass.
Range extended from East Africa and Madagascar.

Labidodemas rugosum (Ludwig, 1875)

See: Theel, 1886 : 226 (as Holothuria rugosa).

Common under boulders at the edge of the west coast seaward platform.

Holothuria (Cystipus) rigida (Selenka, 1867)
See: Rowe & Doty, 1977 : 234.

Not common, found in sand under boulders in the lower eulittoral of the seaward platforms of the
west and north coasts, in sand in the lagoon north east He Esprit (1 15-067).
Range extended from East Africa and Madagascar.

122 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Holothuria (Halodeimd) atra Jaeger, 1833

See: Rowe & Doty, 1977 : 230; Bonham & Held, 1963 : 305 (ecology); Bakus, 1968 : 24 (ecology).

Common throughout the whole lagoon but not in the great densities reported in the literature from

other sheltered reef and lagoon flats. Occasional on the seaward platforms of the west and north

coasts.

Holothuria (Lessonothuria) pardalis Selenka, 1 867

See: Rowe & Doty, 1977 : 233.

Found under boulders over sandy gravel or sand on all the seaward platforms around Aldabra,

especially on the more sheltered west and north coast seaward platforms, also in the channel areas

of the lagoon under coral heads and microatolls over sandy gravel and sand. It should be noted

that the spicules of various specimens which have been attributed to H. pardalis show a wide

range of form and the species is therefore in need of review (Rowe, pers. comm.).

Holothuria (Lessonothuria) verrucosa Selenka, 1867

See: Semper, 1968 : 90 (as Holothuria immobilis).

One specimen only collected from a sand and coral patch in one of the West Channels.

Holothuria (Mertensiothurid) leucospilota (Brandt, 1835)

See: Rowe & Doty, 1977 : 233; Bonham & Held, 1963 : 305 (ecology).

Occasionally found under boulders over sand in standing water in the upper eulittoral of the He

Picard beach-rock (056-100); more common in the lower eulittoral of all the coastal seaward

platforms around Aldabra, especially the sheltered west and north coasts; also under coral heads

over sand in the channel areas of the lagoon.

Holothuria (Mertensiothurid) pervicax Selenka, 1867
See: Rowe & Doty, 1977 : 234.

Not too common, usually found under coral over sand in channel areas of the lagoon like Main
Channel (104-110), West Channels (068-083) and Passe Houareau (316-118); less commonly
seen under coral at the edge of seaward platforms like Dune Jean Louis (275-039).
One specimen had parasitic gastropods, Melanella muelleriae, projecting from the body wall.

Holothuria (Microthele) nobilis (Selenka, 1867)
See: Rowe & Doty, 1977 : 231.

Most individuals were of the black form with the large lateral bumps white; others were uniformly
grey. On lagoon grass flats (068-083) and among coral heads near West Channels, near He Chalen
(080-072). On 23 April 1978 a specimen with one end elevated off the substrate was noted as emit-
ting a milky substance and was probably spawning.

Holothuria (Platyperond) difficilis Semper, 1 868

See: Rowe & Doty, 1977 : 232; Bakus, 1968 : 23 (ecology).

One specimen found under a coral head over sandy gravel on lagoon grass flats (068-083) near

West Channels, one specimen found under coral over muddy sand in a mangrove creek at Passe

Gionnet (137-120) in the lagoon, one specimen found under boulders over sand on the seaward

platform at Anse Var (068-122).

Holothuria (Selenkothuria) moebii Ludwig, 1883

See: H. L. Clark, 1946 : 426.

One specimen only, found under a boulder in the upper eulittoral of the He Picard beach-rock

(056-100).

Range extended from the Mascarene Islands.

Holothuria (Selenkothuria) parva Krauss in Lampert, 1885
See: Cherbonnier, 1952 : 503.

Common under boulders over intact bed rock in the upper eulittoral beach-rock of He Picard
(056-100), also found in sand pockets at the cliff base of Dune d'Messe (265-038) on the exposed
south coast. (The identification is by A. M. C.)
Range extended from East Africa and Madagascar.

ECHINODERMS OF ALDABRA 123

Holothuria (Semperothurid) cinerascens (Brandt, 1835)
See: Rowe & Doty, 1977 : 230.

Abundant in crevices and holes in the same lower eulittoral band with the echinoids Echinometra
mathaei and Stomopneustes variolaris on the exposed south and east coasts at Dune Jean Louis
(275-039) and near Point Hodoul (407-080); uncommon in the sheltered upper eulittoral beach-
rock at lie Picard (056-100) on the more sheltered west coast.
One specimen had several of the parasitic gastropods, Melanella muelleriae, on the body wall.

Holothuria (Theelothuria) sp. cf. H. hamata Pearson, 1913
See: Pearson, 1913 : 51; Cherbonnier, 1955 : 156.

Two specimens taken from amongst algae and sea-grass on the lagoon side of West Channels.
If these specimens are H, hamata, the range would be extended to the western Indian Ocean
from the Red Sea.

Holothuria (Theelothuria} maculosa Pearson, 1913

See: Pearson, 1913 : 53 (Aldabra is the type locality for this species).

Recorded only by Hughes & Gamble (1977) - from sea-grass flats in Passe Houareau (316-118).

Holothuria (Thymiosycid) arenicola Semper, 1868
See: Rowe & Doty, 1977 : 232.

A common, ubiquitous species in sand or sandy gravel under boulders in the lower eulittoral of
the seaward platforms on all coasts around Aldabra; also in sand under coral heads and sea-
grass in the channel areas of the lagoon.

Some specimens were infested with the parasitic gastropod Melanella muelleriae on the body
wall.

Holothuria (Thymiosycid) hilla Lesson, 1830

See: Rowe & Doty, 1977 : 232.

Found under boulders over sandy gravel or coarse rubble in the lower eulittoral of the seaward

platforms all around Aldabra, also under coral heads in all the channel areas of the lagoon.

Holothuria (Thymiosycid) impatiens (Forskaal, 1775)

See: Rowe & Doty, 1977 : 233.

The most common of the ubiquitous cryptic holothurian species, under boulders over sand and

sandy gravel on all seaward platforms around Aldabra and, like all the others, more common on

the relatively sheltered coastal platforms of the west and north coasts, common under coral heads

in the channel areas of the lagoon and occasionally in the coral heads as well.

Holothuria (Thymiosycid) remollescens Lampert, 1888

See: H. L. Clark, 1946 : 437.

Three specimens were taken from under rubble among sea-grasses on the seaward platform of

the west coast.

Range extended from the Red Sea.

STICHOPODIDAE

Stichopus chloronotus Brandt, 1835

See: Rowe & Doty, 1977 : 227; Yamanouchi, 1956 : 347 (feeding & ecology).

The most common epifaunal holothurian species on hard substrates in the sublittoral fringe and

shallow sublittoral of the west and north coastal platforms, also a common species on the coral

flats and patch reefs in the Main Channel drainage system of the lagoon. Not seen on the exposed

rock benches of the east and south coasts.

Humphreys & Lutzen (1972) described a new parasitic gastropod, Megadenus cantharelloides,
from an Aldabra specimen of S. chloronotus, living on the inner face of the body wall of S.
chloronotus.

Stichopus sp. cf. S. horrens Selenka, 1867
See: Rowe & Doty, 1977 : 227.

124 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Found under Porites microatolls on the lagoon grass flats (068-083) near West Channels and under
boulders at the edge of the seaward platform of the west coast.

The length of the larger specimen as preserved is 85 mm, whereas S. horrens may exceed 200 mm.
This may account for the absence of the large tack-like tables with pointed spires so characteristic
of large specimens of 5". horrens and present in a specimen from Egmont Reef, Chagos Archipelago,
recorded under 'Maldive area' by Clark & Rowe (1971). If correct, this new record would extend
the range still further. However, Cherbonnier (1967) has recorded a new species, Stichopus pseud-
horrens, from Eilat, northern Red Sea, the holotype (length 220 mm) also having some tack-like
tables. It has some conspicuous conical warts similar to but probably forming more rows than in
the Aldabra specimens and also differs in being a dark chestnut colour.

Stichopus sp. cf . S. variegatus Semper, 1 868

See: James & Pearse, 1969 : 102; Yamanouchi, 1956 : 347 (feeding & ecology).

One specimen of an unusual red colouration found under a Porites microatoll on the lagoon grass

flats (068-083) near West Channels.

Thelenota ananas (Jaeger, 1833)
See: Rowe & Doty, 1977 : 227.

Found on sand at the bottom of channels between patch reefs in the Main Channel drainage sys-
tem of the lagoon and on sand off the seaward platform of the north and west coasts at greater
than 5 m depth.

Range extended from the Maldive Islands and the Mascarene Islands (unpublished from the
BMNH collections) as well as from the West Pacific.

CUCUMARHDAE

Orbithyone megapodia H. L. Clark, 1938
See: H. L. Clark, 1946 : 396.

Reported only by Hughes & Gamble (1977) from under rubble off the west coast seaward platform.
The species is poorly distinguished, the holotype from northern Australia having spicules only
in the tentacles, not the body wall, and measuring only 15 mm in length, so the record is very
dubious.

PHYLLOPHORIDAE

Afrocucumis africana (Semper, 1863)

See: Rowe & Doty, 1977 : 226.

Abundant under boulders in the upper eulittoral of the He Picard beach-rock (056-100) and

common in crevices on the exposed seaward rock benches of the east and south coasts, less

common under boulders over gravel on the seaward platforms of the west and north coasts.

SYNAPTIDAE

Euapta godeffroyi (Semper, 1868)

See: Rowe & Doty, 1977 : 235.

Under boulders and at the bases of sea-grass in the lower eulittoral of the sea-grass platforms

of the west and north coasts, under coral heads on the grass flats in the lagoon at Passe Houareau

and West Channels.

Range extended from the Mascarene Islands.

Polyplectana kefersteini (Selenka, 1 867)
See: Rowe & Doty, 1977 : 235.

Some collected in the same areas as Euapta godeffroyi above.
Range extended from the Red Sea.

Synapta maculata (Chamisso & Egsenhardt, 1821)
See: Rowe & Doty, 1977 : 234.

ECHINODERMS OF ALDABRA 125

Found on Thalassia beds on the seaward platform of the west coast and on the seaward side of
the West Channels.

CHIRIDOTIDAE

Chiridota stuhlmanni Lampert, 1896
See: Heding, 1931 : 676.

Collected under the upper eulittoral beach-rock of He Picard (056-100) and from rubble in Passe
Houareau.
Range extended from East Africa.

Chiridota violacea (J. Miiller, 1850)

See: Heding, 1928 : 296; Hughes & Gamble, 1977 : 334 (habitats & distribution on Aldabra).
Can be the dominant member of the infauna in certain sheltered, sandy areas in the lagoon (315-
117) near Passe Houareau as well as other similar areas.
Range extended from East Africa and Madagascar.

Polycheira rufescens (Brandt, 1835)

See: Heding, 1928 : 306.

Common under boulders on the upper eulittoral beach-rock of He Picard (056-100), also one

specimen was found under boulders on the lagoon side (063-080) of Passe Dubois.

Acknowledgements

N. A. S. would like to thank the Royal Society for the award of the John Murray Travelling
Studentship and for the provision of facilities at its research station on Aldabra, Professors N. B.
Marshall and J. D. Pye and Dr A. C. Campbell for their support at Queen Mary College, London,
Mr L. U. Mole of the Royal Society for his support and consideration, and especially Ailsa Clark
and John Taylor whose encouragement and support made this project possible.

We thank Dr A. J. Bruce (Heron Island Research Station, Queensland) for the shrimp identifi-
cation; Dr P. E. Gibbs (Marine Biological Station, Plymouth) for the polychaete identification;
Dr J. Haig (Allan Hancock Foundation, California) for the crab identification; Professor
J. Stock (Zoological Museum, University of Amsterdam) for the pycnogonid identification;
Dr A. Waren (Zoological Institute, Goteborg University) for the parasitic mollusc identification;
Mr G. L. J. Paterson (BMNH) for valuable assistance with the collection; Dr F. W. E. Rowe
(Australian Museum, Sydney) for identification of most of the holothurians. Dr W. F. Humphreys
(University of Bath) and Dr R. N. Hughes (University College of North Wales at Bangor) kindly
provided unpublished material from their Aldabra fieldwork.

References

Achituv, Y. 1969. Studies on the reproduction and distribution of Asterina burtoni Gray and A. wega

(Asteroidea) in the Red Sea and the Eastern Mediterranean. IsraelJ. Zool. 18 : 329-343.
Bakus, G. J. 1968. Defensive mechanisms and ecology of some tropical holothurians. Mar. Biol. Berlin

2 : 23-32.

Balinsky, J. B. 1957. The Ophiuroidea of Inhaca Island. Ann. Natal Mus. 14 : 1-33.
Bonham, K. & Held, E. E. 1963. Ecological observations on the sea cucumbers Holothuria atra and H.

leucospilota at Rongelap Atoll, Marshall Islands. Pacific Sci. 17 : 305-314.
Campbell, A. C., Dart, J. K. G., Head, S. M. & Ormond, R. F. G. 1973. The feeding activity of Echino-

strephus molaris (de Blainville) in the central Red Sea. Mar. Behav. Physiol. 2 : 155-169.
Chartock, M. A. 1972. The role of detritus in a tropical marine ecosystem: niche separation in congeneric

ophiuroids, food partitioning in cryptic invertebrates and herbivore detritus production at Eniwetok,

Marshall Islands. Ph.D. thesis, University of Southern California, Los Angeles. 177 pp.
Cherbonnier, G. 1952. Les holothuries de Quoy et Gaimard. Mem. Inst. r. Sci. nat. Belg. 44 : 1-50.
1955. Resultats scientifiques des campagnes de la 'Calypso'. Les holothuries de la Mer Rouge.

Annls Inst. oceanogr., Monaco 30 : 129-183.

126 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

1967. Deuxieme contribution a 1'etude des Holothuries de la Mer Rouge collectees par des Israeliens.

Bull. Sea Fish. Res. Stn Haifa 43 : 55-68, 6 figs.

& Guille, A. 1978. Echinodermes: Ophiurides. Faune de Madagascar. No. 48 : iv + 272 pp.

Clark, A. H. 1941. A Monograph of the Existing Crinoids. I (4a). Superfamily Mariametrida (except the

family Colobometridae). Bull. U.S. natn. Mus. 82 (4a) : 1-603.
1947. A Monograph of the Existing Crinoids. I (4b). Superfamily Mariametrida (concluded - the

family Colobometridae) and Superfamily Tropiometrida (except the families Thalassometridae and

Charitometridae). Bull. U.S. natn. Mus. 82 (4b): 1-473.

1964. Description of two new species of Ophiuroidea collected during the Snellius Expedition.

Zool. Meded., Leiden 39 : 385-390.

& Clark, A. M. 1967. A Monograph of the Existing Crinoids. I (5). Suborders Oligophreata (con-

cluded) and Macrophreata. Bull. U.S. natn. Mus. 82(5) : 1-860.
Clark, A. M. 1965. Japanese and other ophiuroids from the collections of the Munich Museum. Bull. Br.

Mus. not. Hist. (Zool.) 13 : 37-71.
1967. Echinoderms from the Red Sea. 2. Crinoids, ophiuroids, echinoids and more asteroids. Bull.

Sea Fish. Res. Stn Israel 41 : 26-58.

1970. Notes on the family Amphiuridae (Ophiuroidea). Bull. Br. Mus. not. Hist. (Zool.) 19 : 1-81.

1972. Some crinoids from the Indian Ocean. Bull. Br. Mus. not. Hist. (Zool.) 26 : 421-487.

1974. Notes on some echinoderms from southern Africa. Bull. Br. Mus. not. Hist. (Zool.) 26(6) :

423-487.

19760. Tropical epizoic echinoderms and their distribution. Micronesica 12 : 111-117.
- 19766. Echinoderms of coral reefs. In: O. A. Jones & R. Endean (Editors). Biology and Geology of
Coral Reefs. 3. Biology 2 : 95-123. New York.

& Rowe, F. W. E. 1971. Monograph of the Shallow-Water Indo-West Pacific Echinoderms. 1-238 pp.,

31 pis. London.

Clark, H. L. 1909. Notes on some Australian and Indo-Pacific echinoderms. Bull. Mus. comp. Zool.
Harv. 52 : 107-135.

1921. The Echinoderm fauna of Torres Strait. Pap. Dep. mar. biol. Carnegie Instn Wash. 10 : 1-233.

1939. Ophiuroidea. Sclent. Rep. John Murray Exped. 6 : 29-136.

1946. The Echinoderm fauna of Australia. Publs Carnegie Instn No. 566 : 1-567.

Dart, J. K. G. 1972. Echinoids, algal lawn and coral recolonization. Nature 239 : 50-51.

Devaney, D. M. 1967. An ectocommensal polychaete associated with Indo-Pacific echinoderms, primarily
ophiuroids. Occ. Pap. Bernice P. Bishop Mus. 23 : 287-304.

1970. Studies on Ophiocomid brittlestars. I. A new genus (Clarkcoma) of Ophiocominae with a re-
evaluation of the genus Ophicoma. Smithson. contr. Zool. No. 51 : 1-41.

1974. Shallow-water asterozoans of southeastern Polynesia. 2. Ophiuroidea. Micronesica 10 : 105-

204.

1977. Ophiomastix koehleri, a new ophiocomid brittlestar (Echinodermata : Ophiuroidea) from the

western Indian Ocean. Proc. biol. Soc. Wash. 90 : 274-283.
Fishelson, L. 1971. Ecology and distribution of the benthic fauna in the shallow waters of the Red Sea.

Mar. Biol. Berlin 10 : 113-133.
Fisher, W. K. 1919. Starfishes of the Philippine seas and adjacent waters. Bull. U.S. natn. Mus. 100 :

1-711.
Gibbs, P. E. 1969. Aspects of polychaete ecology with particular reference to commensalism. Phil. Trans.

R. Soc. ser. B 255 : 443-458.

1971. The polychaete fauna of the Solomon Islands. Bull. Br. Mus. not. Hist. (Zool.) 21 : 99-211.

Clark, A. M. & Clark, C. M. 1976. Echinoderms from the northern region of the Great Barrier

Reef. Bull. Br. Mus. not. Hist. (Zool.) 30 : 103-144.
Heding, S. G. 1928. Synaptidae. Vidensk. Meddr dansk naturh. Foren. 85 : 105-323.

1931. Ober die Synaptiden des Zoologischen Museums zu Hamburg. Zool. Jb. (Syst.) 61 : 637-696.

Herring, P. J. 1972. Observations on the distribution and feeding habits of some littoral echinoids from

Zanzibar. /. nat. Hist. 6 : 169-175.
Hughes, R. N. 1977. The biota of reef flats and limestone cliffs near Jeddah, Saudi Arabia. /. nat. Hist.

11 : 77-96.
& Gamble, J. C. 1977. A quantitative survey of the biota of intertidal soft substrata on Aldabra Atoll,

Indian Ocean. Phil. Trans. R. Soc. Ser. B. 279 : 327-355.
Humphreys, W. F. & Liitzen, J. 1972. Studies on parasitic gastropods from echinoderms. I. On the structure

and the biology of the parasitic gastropod Megadenus canthrelloides n. sp., with comparisons on

Paramegadenus n.g. Biol. Skr. 19 : 1-27.

ECHINODERMS OF ALDABRA 127

James, D. B. & Pearse, J. S. 1969. Echinoderms from the Gulf of Suez and the Northern Red Sea./. Mar.

Biol. Ass. India 11 : 78-125.
Jangoux, M. 1973<7. Les asteries de 1'Ile d'Inhaca (Mozambique) (Echinodermata : Asteroidea). l.Les

especes recoltees et leur repartition geographique. Annls Mus. r. Afr. cent. (Ser. 8 Sci. zool.) No. 208 :

1-50.
19736. Le genus Neoferdina Livingstone (Echinodermata, Asteroidea : Ophidiasteridae). Rev. Zool.

Bot. afr. 87 : 775-794.
Khamala, C. P. M. 1971. Ecology of Echinometra mathaei (Echinoidea : Echinodermata) at Diani Beach,

Kenya. Mar. Biol. Berlin 11 : 167-172.
Kissling, D. L. & Taylor, G. T. 1977. Habitat factors for reef-dwelling ophiuroids in the Florida keys. In:

Taylor, D. L. (Editor). Proceedings of the Third International Coral Reef Symposium. University of

Miami, pp. 225-231.

Kohn, A. J. 1971. Diversity, utilization of resources, and adaptive radiation in the shallow-water marine
invertebrates of tropical ocean islands. Limnol. Oceanogr. 16 : 332-348.

Laxton, J. H. 1974. A preliminary study of the biology and ecology of the blue starfish Linckia laevigata
(L.) on the Australian Great Barrier Reef and an interpretation of its role in the coral reef ecosystem.
Biol. J. Linn. Soc. 6 : 47-64.

Loriol, P. de. 1885. Catalogue raisonne des Echinodermes recueillis par M. V. de Robillard a 1'Ile Maurice

2. Stellerides. Mem. Soc. Phys. Hist. nat. Geneve 29 : 1-84.

1893a. Echinodermes de la Baie d'Amboine. Revue suisse Zool. 1 : 359-426.

18936. Catalogue raisonne des Echinodermes recueillis par M. V. de Robillard a 1'Ile Maurice.

3. Ophiurides et Astrophytides. Mem. Soc. Phys. Hist. nat. Geneve 32 : 1-63.

Lyman, T. 1865. Ophiuridae and Astrophytidae. Illust. Cat. Mus. comp. Zool. Harv. No. 1 : 1-200.
1882. Ophiuroidea. Rep. sclent. Results Voy. Challenger. Zool. 5 : 1-386.

Magnus, D. B. E. 1967. Ecological and ethological studies and experiments on the echinoderms of the

Red Sea. Stud. Trop. Oceanogr. 5 : 635-664.

Marsh, L. M. 1977. Coral reef asteroids of Palau, Caroline Islands. Micronesica 13 : 251-281.
Mortensen, T. 1928. Monograph of the Echinoidea. 1. Cidaroidea. 1-551. Copenhagen.
1933. Biological observations on ophiurids, with descriptions of two new genera and four species.

Vidensk. Meddr dansk naturh. For en. 93 : 171-194.

1935. Monograph of the Echinoidea. 2. Bothriocidaroida, Melonechinoida, Lepidocentroida, and

Stirodonta. 1-647. Copenhagen.

1940. Monograph of the Echinoidea. 3 (1). Aulodonta. 1-370. Copenhagen.

1943a. Monograph of the Echinoidea. 3 (2). Camarodonta 1. 1-533. Copenhagen.
19436. Monograph of the Echinoidea. 3 (3). Camarodonta 2. 1-446. Copenhagen.

1948o. Monograph of the Echinoidea. 4 (1). Holectypoida and Cassiduloida. 1-371. Copenhagen.

19486. Monograph of the Echinoidea. 4 (2). Clypeastroida. 1-471. Copenhagen.

1951. Monograph of the Echinoidea. 5 (2). Spatangoida 2. 1-593. Copenhagen.

Panning, A. 1944. Die Trepangfischerei. Mitt. zool. Stlnst. Hamb. 49 : 1-76.
Pearson, J. 1913. Notes on the Holothuroidea of the Indian Ocean. Spolia zeylan. 9 : 173-190.
Price, J. H. 1971. The shallow sublittoral marine ecology of Aldabra. Phil. Trans. R. Soc. Ser. B 260 :
123-171.

Rideout, R. S. 1978. Asexual reproduction as a means of population maintenance in the coral reef asteroid
Linckia multifora on Guam. Mar. Biol. Berlin 47 : 287-295.

Rose, E. P. F. 1978 [1976]. Some observations on the recent holectypoid echinoid Echinoneus cyclostomus
and their palaeoecological significance. Thalassia jugosl. 12 : 299-306.

Rowe, F. W. E. & Doty, J. E. 1977. The shallow-water holothurians of Guam. Micronesica. 13 : 217-250.

Russo, A. R. 1977. Water flow and the distribution and abundance of echinoids (genus Echinometra') on
an Hawaiian reef. Aust. J. Mar. Freshwater Res. 28 : 693-702.

Sale, P. F., Potts, D. C. & Frankel, E. 1976. Recent studies on Acanthaster planci. Search 7 : 334-338.

Semper, C. 1868. Holothurien. Reisen in Archipel der Philippinen. 2. Wissenschaftliche Resultate. 1-288.
Weisbaden.

Sloan, N. A. 1979. A pycnogonid-ophiuroid association. Mar. Biol. Berlin 52 : 171-176.

(in press). Microhabitat and resource utilization in cryptic rocky intertidal echinoderms at Aldabra

Atoll, Seychelles. Mar. Biol. Berlin

Stock, J. 1979. Anoplodactylus ophiurophilus n. sp., a sea spider associated with brittle stars in the Sey-
chelles. Bijdr. Dierk. 48 : 156-160.

128 N. A. SLOAN, A. M. CLARK & J. D. TAYLOR

Takeda, M., Miki, M., Asai, M. & Suzuki, K. 1976. Halicarcinus oriental's Sakai (Crustacea : Hymeno-

somatidae) associated with Ophiocoma brevipes Peters (Ophiuroidea : Ophiocomidae) from Hochijo

Island. Bull. natn. Sci. Mus. Tokyo (Zool.) 2 : 33-38.
Taylor, J. D. 1968. Coral reefs and associated invertebrate communities (mainly molluscan) around

Mane, Seychelles. Phil. Trans. R. Soc. Ser. B 254 : 129-206.

1971o. Intertidal zonation at Aldabra Atoll. Phil. Trans. R. Soc. Ser. B. 260 : 173-213.

19716. Observations on the shallow-water marine fauna of Diego Garcia. Atoll. Res. Bull. No. 149 :

31-39.

1976. Habitats, abundance and diets of Muricacean gastropods at Aldabra Atoll. /. Linn. Soc.

(Zool.) 59 : 165-193.
Taylor, J. D. & Lewis, M. S. 1970. The flora, fauna and sediments of the marine grass beds of Mahe,

Seychelles. J. nat. Hist. 4 : 199-220.

Theel, H. 1886. Holothurioidea. Part 2. Rep. sclent. Results Voy. Challenger. Zool. 39 : 1-290.
Thomassin, B. A. 1976. Feeding behaviour of the felt-, sponge- and coral-feeder sea stars, mainly Culcita

schmideliana. Helgolander wiss. Meeresunters. 28 : 51-65.
Trudgill, S. T. 1976. The marine erosion of limestones on Aldabra Atoll, Indian Ocean. Z. Geomorph.

n. f. 26 : 164-200.
WestolI,T. S. & Stoddart, D. R. 1971. (Editors). A discussion on the results of the Royal Society expedition

to Aldabra 1967-1968. Phil. Trans. R. Soc. Ser. B 260 : 1-654.
Yamaguchi, M. 1977a. Population structure, spawning, and growth of the coral reef asteroid Linck ia

laevigata (Linnaeus). Pacific Sci. 31 : 13-30.
19776. Larval behaviour and geographic distribution of coral reef asteroids in the Indo-West Pacific.

Micronesica 13 : 283-296.
Yamanouchi, T. 1956. The daily activity rhythms of the holothurians in the coral reefs of Palao Islands.

Publs Seto mar. Biol. Lab. 5 : 347-362.

Manuscript accepted for publication 23 March 1979

British Museum (Natural History)

also publishes the following works on
Echinoderms

Catalogue of the recent sea-urchins (Echinoidea) in the collection of
the British Museum (Natural History). H. Lyman-Clarke

1925, xxviii + 250 pp, 12 plates, systematic and alphabetical indexes and list of
types, 8-25

Monograph of the shallow-water Indo-West Pacific Echinoderms.

A. M. Clark and F. W. E. Rowe

1971, ix + 238 pp, 31 plates, coloured frontispiece, 100 text figures, 4to, 19-00

The Echinoderms of Southern Africa. A. M. Clark

1976, 276 pp, 276 diagrams, 4to boards, 20-00

Starfishes and related Echinoderms. A. M. Clark.

1968; 3rd ed. 1977, 160 pp, text figures, many colour plates, paper-back 3-00
Joint with T. F. H. Publications Inc. Ltd.

Lists of all BM(NH) publications are available free on request to :

Publications Sales

British Museum (Natural History)

Cromwell Road

London SW7 5BD

Titles to be published in Volume 37

Miscellanea

The echinoderms of Aldabra and their habitats. By N. A. Sloan,
Ailsa M. Clark & J. D. Taylor

The Fellodistomidae (Digenea) of fishes from the northeast Atlantic.

By Rodney A. Bray & David L. Gibson

The anatomy, phylogeny and classification of bariliine cyprinid fishes.
By G. J. Howes

Printed by Henry Ling Ltd, Dorchester

Bulletin of the

British Museum (Natural History)

The anatomy, phylogeny and classification
of bariliine cyprinid fishes

G. J. Howes

Zoology series Vol 37 No 3 31 January 1980

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and
an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers
are works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,

British Museum (Natural History),
Cromwell Road,

London SW7 5BD,
England

World List abbreviation: Bull. Br. Mus. nat. Hist. (Zool.)

Trustees of the British Museum (Natural History), 1980

ISSN 0007-1498 Zoology series

Vol 37 No 3 pp 129-198
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 31 January 1980

r

GENERAL

The anatomy, phylogeny and classification of \
bariliine cyprinid fishes

G. J. Howes

Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD

Contents

Synopsis 129

Introduction . . . . . . . . . . . . .129

Specimens examined . . . . . . . . . . .130

Abbreviations used in the figures . . . . . . . . .130

Anatomical description of Opsariichthys, Zacco, Luciosoma, Barilius and Engraulicypris 1 3 1

Osteology 131

External anatomical characters . . . . . . . . .171

The brain 177

Summary of characters . . . . . . . . . . 1 79

Relationships of the bariliine genera . . . . . . . . .179

Classification of Barilius . . . . . . . . . .180

Classification of Leptocypris . . . . . . . . . .181

Classification of Engraulicypris . . . . . . . . .182

Classification of Rasbora (in part) . . . . . . . . .182

Parluciosoma gen. nov. . . . . . . . . . . .183

Megarasbora Gunther, 1868 183

Interrelationships and classification of the bariliine genera . . . . .183

Interrelationships of the bariliine group . . . . . . . .184

Biogeography of the bariliines . . . . . . . . . .186

Conclusions . . . . . . . . . . . . .187

Acknowledgements . . . . . . . . . . . .188

Appendix 1. Annotated list of bariliine genera and species . . . . .188

Appendix 2. A note on the taxa formerly included in Engraulicypris . . .195
References . . . . . . . . . . . . .196

Synopsis

The osteology, external anatomical features and gross brain morphology of the cyprinid fish genera
Opsariichthys, Zacco, Luciosoma, Barilius and Engraulicypris are described and compared. With the
exception of Zacco a series of synapomorph characters are identified in these genera which enable them
to be related as a monophyletic assemblage termed the bariliine group. The distribution of certain
characters demands a reclassification of the considered genera. Species formerly included in Barilius are
now referred to four genera, Barilius, Opsaridium, Raiamas and Leptocypris. Engraulicypris is recognized
as monotypic and forming the sister group to Leptocypris. The identification of synapomorphies in
Luciosoma and some species of Rasbora requires the establishment of a new genus, Parluciosoma, to
contain the latter. Megarasbora is identified as a member of the bariliine group and as representing a
plesiomorph luciosomine. The geographic distribution of the group is summarized. Relationships of the
bariliines with other monophyletic groups is unresolved, but the chelines are nominated as the likely
sister group. Zacco does not share any of the derived characters that relate the bariliine genera and it is
suggested that its closest relatives are to be found amongst the alburnine cyprinids. Appendices are
included which contain annotated lists of the bariliine taxa and notes on the generic allocation of those
species previously included in Engraulicypris.

Introduction

C. Tate Regan (1911) was the first author to consider Opsariichthys as a primitive or generalized
cyprinid. He did so on the basis of its possessing triserial pharyngeal teeth, a 'complete series' of

Bull. Br. Mus. tint. Hist. (Zool.) 37 (3): 129-198 Issued 31 January 1980

129

130

G. J. HOWES

circumorbitals, large posttemporal fossae, second and third vertebrae separated, a fenestra
between the quadrate and the metapterygoid and on the form of the cleithrum.

Regan's opinion of the primitive nature of Opsariichthys has been accepted by all subsequent
authors who have had cause to comment on cyprinid phylogeny (see, for example, Ramaswami,
1955; Weitzman, 1962; Greenwood et al., 1966; Roberts, 1973). The characters enumerated by
Regan do, indeed, appear to be plesiomorph for the Cyprinidae and as such are of little value in
indicating the relationships of Opsariichthys.

In an earlier paper (Howes, 1978) I described and commented upon some anatomical features
of Opsariichthys and compared them with similar characters in other cyprinid genera. It is now
possible to present a more detailed description of osteological and other anatomical features of
Opsariichthys and to evaluate them in terms of their plesiomorphy and apomorphy. From
my earlier studies (Howes, 1978) I formed the opinion that Opsariichthys was related to Barilius
and, furthermore, that Luciosoma also shared characters linking it with these taxa. (It must be
pointed out here that my earlier remarks concerning Barilius (Howes, 1978) mostly refer to B.
bola and B. microcephalus, taxa which this present study shows to be derived members of the
genus.) As this study progressed it became evident that Engraulicypris sardella was also allied to
Barilius but that it had no relationship with other species assigned to Engraulicypris. Thus an
anatomical description of this taxon is also included.

Previous authors had regarded Zacco as the closest relative of Opsariichthys. Contrary to their
opinion, the studies reported herein show that Zacco does not belong to the same monophyletic
assemblage as Opsariichthys; in order to show why this is so, the anatomy of Zacco is described
along with that of the other genera.

The data are presented in the form of character headings under which appear descriptions for
each genus studied, followed by a comparative analysis at the end of each section.

An annotated list of the taxa comprising the bariliine group is given in Appendix 1 (p. 195), and
a list of those taxa formerly included in Engraulicypris in Appendix 2 (p. 196).

Specimens examined

Skeletal and alizarin material of representatives of all bariliine genera has been examined.
Dissections have been made on a total of c. 50 species and all specimens have been radiographed.
All the type specimens of species currently assigned to Engraulicypris and Chelaethiops have been
examined. A complete list of specimens used in this study is deposited in the Fish Section of the
British Museum (Natural History).

Abbreviations used in the figures

Al Outer division of adductor mandibulae EPO

muscle ES

AA Anguloarticular EXO

AHF Anterior hyomandibular fossa F

AL Axial lobe of pectoral fin FC

AP Apophysial platform FF

AS Axial scale FG

BO Basioccipital FO

BOP Basioccipital plate FR

CE Cerebellum Fill

CL Cleithrum FIV

CLA Claustrum FIX

COR Coracoid HMF

CU Cavum utriculus HY

EB Epibranchial HYP

EM Exit of the posterior myodome 1C

EP Epural IF

Epioccipital

Extrascapular

Exoccipital

Frontal

Foramen for carotid artery

Frontal fossa

Frontal groove

Optic foramen

Frontal ring

Foramen for oculomotor nerve

Foramen for trochlear nerve

Foramen for 9th cranial nerve

Hyomandibular fossae

Hypural (numbered)

Hypurapophysis

Intercalar

Infrapharngobranchials (numbered)

BARILIINE CYPRINID FISHES

131

INC Incus OT

IO Infraorbitals (numbered) PA

KE Kinethmoid PAP

LC Lateral commissure PC

LE Lateral ethmoid PE

LJB Lateral jugular bridge PH

LKE Ligament connecting kinethmoid to PHF

mesethmoid (divided) PLP
LKM Ligament connecting kinethmoid to

maxilla PMX

LKP Ligament connecting kinethmoid to PPJ

premaxillary ascending process PRO

LP1 Lateral process of 1st vertebra PS

LP2 Lateral process of 2nd vertebra PSFF

LPE Ligament connecting palatine with

ethmoid PTE

LPEK Ligament connecting palatine with PTEP

ethmoid and kinethmoid PTF

LPM Ligament connecting palatine with PTS

maxilla PTT

LTF Lateral temporal fossa PU1 + UI

LOC Lateral occipital fenestra RA

MC Mesocoracoid RAD

ME Mesethmoid SC

MEC Mesencephalon SCA

MMP Maxillary medial process SCP

MX Maxilla SE

MYO Posterior myodome SO

MYR Posterior myodome roof SOR

N Nasal SP

NC Neural complex STF

NP2 Neural plate of 2nd vertebra TEL

NP3 Neural plate of 3rd vertebra TR

NS4 4th neural spine UN

OL Olfactory lobe V

OS Orbitosphenoid VF

OSE Orbitosphenoid septum VO

OSS Os suspensorium

Olfactory tract

Parietal

Parasphenoid ascending process

Postcleithrum

Preethmoid

Parhypural

Posterior hyomandibular fossa

Vomerine process for attachment of

palatine ligament
Premaxilla

Posterior opening of pars jugularis
Prootic
Parasphenoid
Foramen connecting posttemporal

and subtemporal fossae
Pterotic

Pterotic process
Posttemporal fossa
Pterosphenoid
Posttemporal

Fused preural and ural centra
Retroarticular
Radial

Supracleithrum
Scaphium
Scapula
Supraethmoid
Supraoccipital
Supraorbital
Sphenotic
Subtemporal fossa
Telencephalon
Tripus
Uroneural
Vertebra

Vomerine foramen
Vomer

Note on the figures

All anatomical drawings captioned Opsariichthys are of specimens of Opsariichthys uncirostris
bidens. Unless otherwise stated the scale on all figures indicates 3 mm.

Anatomical description of
Opsariichthys, Zacco, Luciosoma, Barilius and Engraulicypris

Osteology

Ethmo-vomerine region

Opsariichthys

The supraethmoid (SE, Figs 1-3) is a wide bone with a flat dorsal surface and an irregular posterior
border sutured to the frontals. The lateral margin of the supraethmoid is gently concave and along
this concavity lies the nasal. The supraethmoid's anterior border is thickened and bears a concave
notch.

The mesethmoid (ME, Figs 1-3) underlies the supraethmoid and is composed of two lateral
walls fused anteriorly bearing a sloped concavity (a continuation of the supraethmoid notch)

132 G. J. HOWES

which accommodates the kinethmoid. Antero-ventrally the mesethmoid is produced into a wedge-
shaped surface that provides part of the preethmoid fossa. The posterior border of the meseth-
moid wall is deeply notched so as to form part of the olfactory foramen.

PE

SE

Fig. 1 Opsariichthys. Dorsal view of ethmoid region.

Each preethmoid (PE, Figs 1-3) is a large irregular hemispherical bone which lies in the fossa
formed from the mesethmoid and vomer. The preethmoid does not lie flush with the edges of
these two bones but leaves a wide margin of each exposed. In a dissection it is found that the
cartilage which covers the preethmoid fills this surrounding area. In a specimen of 38-5 mm SL
the preethmoids are small, partially ossified elements occupying only the vomerine part of the
fossa.

ME

PS
Fig. 2 Opsariichthys. Lateral view of ethmo-vomerine region.

Each lateral ethmoid (LE, Figs 2 & 3) contacts the dorsal and ventral parts of the mesethmoid,
the space between the bones forming part of the olfactory foramen. Laterally, the bone is produced
as a thick wall with a wide triangular base. The dorsal surface of the wall is covered by the
frontal. The orbital face of the lateral ethmoid is concave and its posterior margin bears a notch
through which passes the superficial ophthalmic nerve. The posterior margin of the orbital part

BARILIINE CYPRINID FISHES

133

of the bone is sutured to the orbitosphenoid, and the dorsal part to the frontal. Medially, the
lateral ethmoids meet above the parasphenoid. There is no anterior myodome.

The kinethmoid v^hen viewed laterally is seen as an elongated S (Fig. 4A). The dorsal curvature
expands into two heads upon which are attached the ligaments joining the bone to the pre-
maxillaries. A ledge along the lateral face of the bone also supports a ligament. This ligament
runs from the maxilla and becomes bipartite, one part inserting on the dorsal surface of the ledge,
the other, on its ventral surface (Fig. 4B). Basally the kinethmoid is strongly curved towards the
ethmo-vomerine complex and it is this part of the bone which is seated in the anterior groove of
the vomer. Connection with the ethmo-vomerine block is via a bipartite ligament which inserts on
a tissue sheath that encloses the ventral half of the kinethmoid (Fig. 4C).

N

ME

PLP

PE

Fig. 3 Opsariichthys. Ventral view of ethmo-vomerine region.

The vomer (V, Figs 1-3) is short, its posterior margin extending to a line of the lateral ethmoid
walls. Anteriorly it curves ventrad and is transversely convex. The anterior border is concave
and bears a wide groove to accommodate the kinethmoid. The ventral surface of the vomer bears,
on either side of the midline, two small processes (PLP) onto which the palatine ligament attaches.
Laterally the bone forms into a wedge-shaped platform which provides the lower part of the
preethmoid fossa. The vomer is overlain anteriorly by the ethmoid and posteriorly by the para-
sphenoid. There is a cavity between the posterior part of the vomer and the parasphenoid into
which insert the ligaments attaching the dorsal borders of the suspensorial elements to the length
of the parasphenoid. Postero-laterally the vomer is sutured to the bases of the lateral ethmoids.

Zacco

Zacco differs from Opsariichthys in that the mesethmoid is narrower and deeper with a ventro-
lateral nasal cavity; the dorsal surface of the supraethmoidis strongly curved rostrad; the olfactory
foramen scarcely indents the posterior border of the mesethmoid. The depth and curvature of
the ethmoid block is particularly noticeable in Zacco platypus (Fig. 5). The lateral ethmoids

134

G. J. HOWES

LKP

LKM

LKE

Fig. 4

Opsariichthys. Kinethmoid in A, lateral, and B, ventral view. C. Ligamentous connections,

ventral view.

are thicker basally and laterally truncated; medially they are indented to form a shallow anterior
myodome. The vomer is thinner than that of Opsariichthys, contributes less to the preethmoid
fossa, extends further forward and bears a deeper anterior notch. The preethmoids are smaller;
the kinethmoid is longer and more tubular with a gutter-like groove along its anterior face, its
dorsal surface is expanded into two rounded heads which curve forward and are attached to the
premaxillaries by ligaments (Fig. 8A).

Fig. 5 Zacco platypus. Lateral view of ethmo-vomerine region.

Luciosoma

Luciosoma differs from Opsariichthys in the following features : the supraethmoid is more extensive
its dorsal surface having a bowl-shaped depression for accommodating the kinethmoid (Fig. 6);
the kinethmoid is S-shaped but more compressed; the preethmoids are exposed ventrally, not being
completely covered by the vomer; the walls of the lateral ethmoids are curved anteriorly and
thickened basally, they contain the major part of the olfactory nerve foramen which is produced
anteriorly as a bony tube. The orbital part of each lateral ethmoid is inflated. The vomer is deeply
concave and not swollen as in Opsariichthys.

BARILIINE CYPRINID FISHES

135

ME

VO

B

Fig. 6 Luciosoma bleekeri. Ethmo-vomerine region in A, dorsal and B, ventral views.

Barilius

Barilius displays considerable variation in the morphology of the ethmo-vomerine region and,
as in this and other characters to be discussed, at least three major groups of species can be
distinguished. For the purposes of the descriptive anatomical section these are designated as
Groups A, B and C. Formal taxonomic treatment of these groups is given in pp. 189-193.
Differences are as follows :

Group A. The mesethmoid narrow and deep; dorsal surface of supraethmoid strongly curved
rostrad; kinethmoid rod-like with a groove along the anterior surface; lateral ethmoids laterally
truncated, not extending beyond the frontal margin; preethmoids reduced; vomer short and
deeply notched anteriorly.

Examples of this group of species are Barilius barila and B. ornatus (see p. 189).

Group B (Fig. 7). The mesethmoid is wide and shallow, the dorsal surface of the supraethmoid
is flat or gently sloped rostrad; the kinethmoid bears greatly expanded dorsal heads which give the
bone a triangular appearance (Fig. 8B). These heads articulate closely with the premaxillaries,
connection being effected through a broad, ligamentous sheet. Whereas in the two Asian species
assigned to this group the ventral aspect of the kinethmoid is curved backwards as in Opsariichthys
in the Africa species it is rounded or curved forward. The vomer is short but widely flared anter-
iorly, medially it is deep and rises steeply beneath the ethmoid.

Examples of species in this group are Barilius bola, B. loati and B. salmolucius (see p. 193).

Group C. The mesethmoid is wide and deep, the dorsal surface of the supraethmoid slopes
rostrad; the kinethmoid is rod-like and the vomer greatly thickened anteriorly.

136

G. J. HOWES

ME

Fig. 7 Barilius bola. Lateral view of ethmo-vomerine region.

Examples of species in this group are Barilius microcephalus, B. zambesensis and B. ubangensis
(see p. 191).

Engraulicypris

Engraulicypris differs markedly from the genera described above. The ethmoid block (supra-
ethmoid + mesethmoid) is elongate and shallow (Figs 9A-E), its anterior border deeply indented
in the shape of a horseshoe. Each arm is broadly bevelled ventro-laterally so as to accommodate
the underlying vomer. The anterior face of each arm is broad and sloped backwards at 45 to
the vertical, on its medial tip there is a prolongation. The supraethmoid has a slightly convex
lateral margin and a shallow medial depression, its posterior border forming an irregular suture
with the frontals. Each large preethmoid (Fig. 9C) lies antero-laterally to its respective ethmoid
arm, ventrally it is supported by the vomer. Each lateral ethmoid (Fig. 9E) scarcely extends from
below its respective frontal. The olfactory foramen is contained almost entirely within the bone's
medial face.

A , , B

Fig. 8 Kinethmoid of A, Zacco macrophthalmus, B, Barilius bola, in dorsal and lateral views.

BARILIINE CYPRINID FISHES

137

The kinethmoid is long and somewhat rod-shaped, its ventral tip articulating with the anterior
vomerine notch. Its distal tip is slightly bifurcated and a ligament runs from each bifurcation to
the tip of the maxillary. From either side of the kinethmoid shaft a ligament attaches to the
dorsal rim of the maxillary (see p. 162).

LE

1mm

Fig. 9 Engraulicypris sardella. Ethmo-vomerine region. A, dorsal and B, ventral views of the
ethmoid block. C, dorsal view of vomer. D, exploded lateral view of ethmo-vomerine block. E,
dorsal view of entire ethmo-vomerine region showing articulation of the kinethmoid.

The vomer (Figs 9C & D) is short and broad, its posterior tip lying below the posterior border
of the lateral ethmoids. The anterior half of the vomer is perforated by an ellipsoidal foramen
(VF, Fig. 9C), the boundary of which coincides with the rim of the overlying ethmoid. The lateral
border of the vomer is slightly raised to form a support for the preethmoid and its anterior rim
is deeply notched.

Orbital region

Opsariichthys

The orbitosphenoids (Fig. 10) are deep, anteriorly expanded bones. Ventro-medially they are
joined to form a thick interorbital septum which contacts the parasphenoid. The pterosphenoids
(Figs 10 & 12) provide the walls of the optic foramen and the roof to the anterior part of the
posterior myodome. Anteriorly each pterosphenoid is bordered by the orbitosphenoid, dorsally
it is overlain by the frontal and posteriorly by the sphenotic and part of the prootic. The junction

138

G. J. HOWES

with the prootic is along a narrow medially directed portion of that bone which forms the dorso-
medial wall of the trigeminofacialis foramen. Ventrally the pterosphenoid is produced into a
narrow stem which contacts the medial face of the ascending lateral wing of the parasphenoid.
The orbital face of the pterosphenoid bears a groove for the passage of the supraorbital nerve
trunk. Ventrally the groove enters the pars jugular is of the trigemino-facialis chamber of the
prootic. The posterolateral face of the pterosphenoid bears a deep concave fossa from which runs
a tendinous sheet to underlie the levator arcus palatini muscle (PTSF, Figs 10 & 12). Part of the
anterior hyomandibular fossa extends onto the dorsal surface of the pterosphenoid.

Fl PTSF

AHF

PRO

OSE
Fig. 10 Opsariichthys. Lateral view of orbital region.

The parasphenoid (Figs 11 & 12) is horizontally aligned, thin and grooved anteriorly where it
overlies the vomer. On either side, posteriorly are ascending wings each of which contacts the
pterosphenoid wing along a narrow front. The posterior dorsal edge of the wing is sutured to the
prootic. This suture is interrupted by the carotid foramen (hypophysial foramen of Ramaswami,
1955}. Below the prootic the parasphenoid widens to form a platform against which articulate the
infrapharyngobranchials, the edges of this apophysis are formed by the prootics (parasphenoid
platform of Howes, 1978). A prominent ridge runs along the ventral midline of the parasphenoid
but becomes lessened posteriorly. The posterior ventral part of the bone is rounded and lies below
the basioccipital.

Zacco

The orbitosphenoid differs from that of Opsariichthys in that it is much shallower and has a
deeper interorbital septum. The pterosphenoid (Pig. 16) is much like that in Opsariichthys, bearing
in its orbital face a deep furrow for the supraorbital nerve trunk. A major difference, however,
is that the pterosphenoid does not articulate with any part of the parasphenoid but is sutured
only to the medial projection of the prootic and, furthermore, there is no fossa in its postero-
lateral face.

The parasphenoid differs from that in Opsariichthys in being curved upward instead of horizont-
ally aligned. The carotid foramen is greatly expanded in Zacco to form a large fenestra between
the parasphenoid wing and the prootic (Fig. 16). The fenestra is partially covered by a fascia of
tissue from which extends part of the adductor arcus palatini muscle. There is a moderately
developed dorsal medial crest along the length of the parasphenoid but this never extends
upwards to contribute to the interorbital septum stemming from the orbitosphenoids. As in

BARILIINE CYPRINID FISHES

139

PAP

AP

Fig. 11 Opsariichthys. Ventral view of parasphenoid.

Opsariichthys the parasphenoid bears a ventral groove for half its length. The apophyseal platform
below the prootic is narrow and receives no contribution from the prootics.

Luciosoma

The orbitosphenoid is shallow and contacts the parasphenoid via a narrow interorbital septum.
The pterosphenoid (Fig. 18) closely resembles that of Opsariichthys but whereas in that genus the
bone forms the border of the anterior trigemino-facialis opening, in Luciosoma it is sutured
without interruption both to the prootic and to the parasphenoid ascending process. There is a
strong lateral ridge along the descending limb which makes contact with the parasphenoid and a
shallow partially covered lateral fossa. Part of the anterior hyomandibular fossa extends onto the
parasphenoid.

The parasphenoid ascending process contacts both the pterosphenoid and prootic. The carotid
foramen is small, and the apophyseal platform is extensive with small lateral contributions from
the prootics, features shared with Opsariichthys.

Barilius

Group A. There is some variation in the depth of the orbitosphenoid septum. The pterosphenoid
contacts the parasphenoid ascending process across a very narrow front. In some species (e.g.
B. bendelisis) contact between the two bones is almost prevented by the intrusion of a thin wedge

140 G. J. HOWES

from the prootic. There is no lateral pterosphenoid fossa, and the carotid foramen is sometimes
extensive, as inZacco. The posterior margin of the parasphenoid is deeply indented and does not
extend to meet the posterior border of the basioccipital. Thus, there is a ventral opening leading
into the posterior myodome (see pp. 146-147).

Group B. The orbitosphenoid septum is narrow as in Opsariichthys. In Barilius bola a shallow
ridge runs along the lateral face of the orbitosphenoid, broadening posteriorly into a wide shelf
continuous with that on the pterosphenoid. The pterosphenoid (Fig. 19) contacts the anterior edge
of the prootic without interruption ; its ventral edge is sutured to the ascending process of the
parasphenoid. There is some variation in the extent of contact between these bones, being greatest
in B. bola and B. guttatus. The anterior hyomandibular fossa extends onto the pterosphenoid.
Although a fossa is present on the lateral face of the pterosphenoid, it is very shallow. The
parasphenoid ascending processes are short, but with long dorsal borders contacting the ptero-
sphenoids. There is a depression in the lateral ascending process for the insertion of the adductor
arcus palatini muscle. The carotid foramen is reduced. The apophyseal platform is well developed
and receives lateral contributions from the prootics.

Group C. The orbitosphenoid has a lateral shelf which is confluent with a similar feature on the
pterosphenoid. The pterosphenoid (Pig. 17B) contacts the prootic medially as well as laterally and
its connection with the parasphenoid is also medial across a very narrow spur of that bone.
There is a cavernous fossa on the lateral face of the pterosphenoid (PTSF, Fig. 19 A), it is of the
basin-type like that in Opsariichthys, but is covered laterally so as to become a deep, conical
chamber. The parasphenoid is wide anteriorly. Posteriorly its ascending process forms a wide
connection with the prootic. The carotid foramen is reduced. The prootics contribute only
marginally to an apophyseal platform. The parasphenoid does not provide a floor to the myodome.

Engraulicypris

The orbitosphenoid is produced into a long, shallow interorbital septum. The pterosphenoid
(Fig. 22A) bears a deep lateral fossa (PTF) and it contacts the ascending process of the para-
sphenoid across a narrow front. The parasphenoid ascending process is high and broad with a
lateral depression. Its otic portion is flat, with the posterior border deeply indented and not
entirely flooring the basioccipital.

Otic region

Opsariichthys

The prootic (Figs 10 & 12) is the largest bone in the braincase. The lateral face bears a wide
commissure covering the trigemino-facialis chamber. Anteriorly, contact is made with the
pterosphenoid along a narrow surface just above the anterior opening of the chamber. Ventro-
anteriorly the prootic is sutured to the ascending process of the parasphenoid. Ventrally, the
prootic is flattened and contributes to the apophyseal platform formed on the lateral border of
the parasphenoid. The posterior border of the prootic is contacted by the basioccipital and
epioccipital. The postero-dorsal face of the prootic is depressed and forms part of the medial wall
of the subtemporal fossa. Dorsally, the prootic is bordered by the autosphenotic and the auto-
pterotic. The anterior and posterior hyomandibular fossae are confined to these two latter bones
and do not invade the prootic. The medial face of the prootic bears a shelf extending from below
the trigemino-facialis chamber to meet its partner from the opposite side in the midline and so
forming the roof of the myodome (Figs 12 & 15). A large depression in the medial face of the
prootic forms the cavum utriculus (Figs 12B-C).

The major part of the dilatator fossa lies in the sphenotic (Fig. 13) which is overlapped along
its medial margin by the pterotic and the frontal. Anteriorly the bone extends laterally as a thin
concave wall, the ventral surface of which forms part of the anterior hyomandibular fossa.
Between the dorsal borders of the sphenotic and pterotic is a lateral temporal foramen. The

BARILIINE CYPRINID FISHES

PHF
AHF

141

SR

STF

PTS
PRO

B

PPJ

Fig. 12 Opsariichthys, otic region in A, lateral and B, medial views. C, cross-section through

anterior part of the prootic.

lateral process of the sphenotic is sloped antero-ventrally, its dorsal surface is broad which
narrows ventrally into a lamellar wall.

The pterotic (Figs 12-14) is bordered dorso-medially by the parietal and epioccipital, and
dorso-anteriorly by the frontal and sphenotic. The exposed cranial surface of the pterotic is
narrow. Ventrally it is sutured to the prootic and extends posteriorly as a thick limb terminating
in a ventrally directed spine. The ventral border of the pterotic houses the greater portion of both
hyomandibular fossae. The pterotic contributes to the lateral wall and roof of the subtemporal
fossa and it also provides, posteriorly, the lateral wall of the posttemporal fossa. Medially this
wall is perforated and opens into the subtemporal fossa; thus the subtemporal fossa is connected
to the posttemporal fossa via an intrapterotic tunnel (PSFF, Fig. 13B). Epaxial muscle fibres run
through this tunnel to insert along the lateral wall of the subtemporal fossa.

The exoccipital (EXO, Figs 13-15) is bounded dorsally by the epioccipital with which it forms
the medial wall of the subtemporal fossa; anteriorly it is bordered by the prootic, and ventrally by
the basioccipital. Ventrally, the exoccipital is slightly inflated to form part of the saccular recess,
and laterally it is pierced by the foramen for the vagus nerve. Dorso-posteriorly, that part of the
bone containing the semi-circular canal turns outward to contact the pterotic and the intercalar.
The posterior portion of the epioccipital surrounds the lateral occipital foramen, its dorsal
surface being bounded by the supraoccipital. Medially, a horizontal sheet of bone extends from
each exoccipital to contact one another in the midline and so form the roof of the cavum sinus
imparis.

The lateral occipital fenestra (LOC, Fig. 14) is covered by connective tissue which thickens
around the border of the foramen. There appears to be no insertion of muscle fibres onto this
tissue and it lies as a 'window' covered by epaxial musculature.

142

FF

PTE

PTEP

PSFR

STF.

B

EXO

Fig. 13 Opsariichthys, pterotic region in A, lateral and B, ventral views.

The epioccipital (EPO, Fig. 14) is bordered medially by the supraoccipital, posteriorly by the
exoccipital, anteriorly by the parietal and laterally by the pterotic. Its lateral face forms the upper
medial wall and roof of the posttemporal fossa (PTF, Fig. 14). The dorsal surface of the epioc-
cipital also contributes to the roof of the subtemporal fossa.

The intercalar (1C, Figs 13B & 14) is large and overlaps the suture between the exoccipital and
the pterotic.

The basioccipital (BO, Fig. 15) is bordered dorsally by the exoccipital, anteriorly by the prootic
and ventro-medially by the parasphenoid. Medially, each basioccipital contacts its partner to
form the floor of the cavum sinus imparis and the roof of the posterior myodome (see below).
The walls of the cavum sinus imparis are formed by dorsal extensions of each basioccipital on
either side of the midline, the extension contacting the roof formed by the exoccipital (Fig. 15).
Ventro-posteriorly there is a pharyngeal process which bears a high dorsal ridge; its ventral sur-
face is expanded into a triangular masticatory plate (BOP, Fig. 15) which terminates in a high
blade-like wing. The aortic foramen (AF) lies dorsal to the plate.

The posterior myodome (MYO, Fig. 15) extends far into the basioccipital. The basioccipital
walls of the myodome converge to contact each other and seal the myodome posteriorly. In a
specimen of 38-5 mm SL, however, the basioccipitals do not contact one another and the myodome
is open posteriorly.

The supraoccipital (SO, Fig. 14) bears a medial ridge which runs backwards as a low lamellar

BARILIINE CYPRINID FISHES
SO

EPO PTF

143

PTE

Fig. 14 Opsariichthys, pterotic region, right side, posterior oblique view.

process. Together with the epioccipitals, the supraoccipital forms a narrow post-parietal platform
(see Howes, 1978).

Zacco

The prootic is of a similar shape to that in Opsariichthys, and likewise forms the border of the
anterior foramen to the trigemino-facialis chamber. However, unlike Opsariichthys the lateral
commissure is narrow, and the carotid foramen is greatly enlarged (Fig. 16). The sphenotic
(Figs 16 & 17A) bears a thin lateral process which is curved ventro-posteriorly ; its anterior face
is deeply indented and provides a site of origin for part of the levator arcus palatini muscle. Only
part of the sphenotic contributes to the dilatator fossa, the frontal forming the anterior part. The
anterior hyomandibular fossa extends along the ventral surface of the sphenotic.

EXO

(MYO)

AF

EXO

BOP

I -i

Fig. 15 Opsariichthys, basioccipital region cut through to show posterior myodome in anterior

oblique view.

144

G. J. HOWES

HMF

PTS

STF

Fig. 16 Zacco macrophthalmus otic region in lateral view.

The posterior face of the pterotic forms the rear wall of the subtemporal fossa and the floor of
the posttemporal fossa. There is no connection between the posttemporal and subtemporal fossae.
The ventral surface of the pterotic houses the posterior hyomandibular fossa. The posterior
process is of variable length and diameter, being thick and truncated in Z. platypus but slender
and elongate inZ. macrophthalmus (PTE, Fig. 17A). The basioccipitals house part of the posterior
myodome which is completely floored by the parasphenoid.

PTE

FR

Fig. 17 A. Zacco macrophthalmus pterotic region, lateral view. B. Barilius microcephalus otic region,

lateral view.

BARILIINE CYPRINID FISHES

Luciosoma

145

The prootic is long and relatively shallow when compared with that in Opsariichthys and Zacco
(Fig. 18). Contrary to the condition in those two genera, the anterior trigemino-facialis foramen
is situated within the lateral face of the prootic, some way from its anterior border. The lateral
commissure is wide. The sphenotic (Fig. 18) is long with a thick lateral process. The ventral surface
of the bone bears the long anterior, and part of the posterior, hyomandibular fossa. The dorsal
border of the sphenotic is sutured with the frontal and pterotic but neither of these bones forms
a roof to the dilatator fossa.

PAR

PTE

PTS

PRO

PAP

Fig. 18 Luciosoma bleekeri otic region, lateral view.

The pterotic (Fig. 18) is an elongate bone bearing the greater part of the posterior hyomandib-
ular fossa on its ventral surface. The posterior process is long and spine-like but not so obliquely
angled as in Opsariichthys. The pterotic forms the outer and posterior walls of the subtemporal
fossa and part of the lateral wall of the posttemporal fossa which opens into it by way of an
intramural channel. A posterior myodome is enclosed posteriorly by the basioccipitals and is
floored ventrally by the parasphenoid. The lateral occipital fenestra is almost rectangular, a
feature shared with some Rasbora species. There is no post-parietal platform and the supra-
occipital crest is greatly reduced.

Barilius

Group A. The prootic resembles that of Opsariichthys and Zacco in that the anterior trigemino-
facialis foramen is situated on the border of the bone. The sphenotic forms the major part of the
dilatator fossa, its lateral process varies interspecifically from a thin lamella, as in Opsariichthys,
to a stout spine. The posttemporal fossa contained in the pterotic is connected with the sub-
temporal fossa through a small foramen.

In some species of this group the posterior myodome is open ventrally so as to provide a
passage for part of the eye musculature (see below, p. 147).

146

G. J. HOWES
AHF

PTSF

PHF

PRO

LJB

PTS

B

PTSF

i i LC AP

Fig. 19 Otic regions of A, Barilius microcephalus and B, Barilius bola, lateral views.

Group B (Fig 19B). The prootic bears the anterior trigemino-facialis foramen in its lateral face,
and the lateral commissure is narrow. The carotid foramen is small. The lateral process of the
sphenotic is a stout limb projecting either at right-angles or directed caudad. The anterior hyo-
mandibular fossa is small and confined to the anterior ventral surface of the bone. There is no
dilatator fossa, the lateral face of the sphenotic being almost perpendicular. There is a lateral
temporal foramen shared between the sphenotic and pterotic. The pterotic does not provide a
solid rear wall to the subtemporal fossa but is perforated by the greatly enlarged posttemporal
foramen (see Howes, 1978). The posttemporal 'fossa' does not exist as an intramural channel as
it does in Opsariichthys, Luciosoma and Barilius Group A species. The large subtemporal vault
is roofed by the epioccipital and parietal and partly by the pterotic which also forms the lateral
wall and posterior ventral bridge (Fig. 20).

The posterior myodome is open ventrally in the African species but closed in the two Asian
representatives, B. bola and B. guttatus of this group. The form of the opening is similar to that
described below in species of Barilius Group C but is not as extensive. A trend towards the closure
of the myodome is apparent in various species of this group, the myodome being largely open in
B. loati but the opening much reduced in B. salmolucius. This morphocline is correlated, as are
other characters, with an elongation of the cranium (see p. 181).

Group C (Fig. 19A). The prootic is large and the anterior trigemino-facialis foramen is in its
lateral face, the lateral commissure is long. The sphenotic is short but deep and bears an extensive

BARILIINE CYPRINID FISHES
RTF

147

EP

-EX

STF

Fig. 20 Barilius bola. Ventral oblique view of posttemporal and subtemporal fossae.

lateral process. In Barilius microcephalus there is a large lateral temporal foramen, a feature
shared with species of Barilius Group B and Opsariichthys (LTF, Fig. 17B & Howes, 1978 : 35).
The ventral surface of the sphenotic contains part of the anterior hyomandibular fossa. The
basioccipital is unfloored so that the posterior myodome is open ventrally (Fig. 21). The recti
inferioris eye muscles extend from the medial edges of the basioccipital process to enter the
myodome; the recti exterioris originate from the prootic walls of the myodome. This feature is
discussed further below, p. 153.

Fig. 21 Barilius microcephalus otic region seen ventrally. Dashed lines indicate path of the eye muscles

through the myodome.

Engraulicypris

The prootic (Fig. 22A) is of similar shape to that in Luciosoma. The anterior trigemino-facialis
foramen is situated in the lateral face of the bone, the lateral commissure is narrow and there
is a deep jugular groove crossed by a narrow spur. The sphenotic (Figs 22 & 23) bears a broad,

148

G. J. HOWES

ventrally directed lateral process. Its dorsal surface forms the major portion of the dilatator fossa
and ventrally bears the anterior hyomandibular fossa.

The pterotic (Figs 22-24) has a broad cranial surface and a short, stout, posterior process. Its
medial face forms the lateral wall of the subtemporal fossa which is perforated by a small foramen
leading into the posttemporal fossa (Figs 22 & 24); no muscle element passes through into the
subtemporal fossa. The basioccipital is unfloored so that the posterior myodome is open ventrally
(Figs 22 & 23). The recti inferioris eye muscles extend from a notch in the ventro-lateral border of
the bone to enter the myodome. The basioccipital process is very short with a ventral channel and
the masticatory plate is weakly developed.

PTE

PTS

EXO

PTE

LJB PRO

Fig. 22

B

A. Engraulicypris sardella neurocranium, ventro-lateral view. B. Barilius ( Leptocypris)
niloticus otic region, lateral view.

Comments on the neurocranium

The broad, shallowly notched ethmoid block which is characteristic of Opsariichthys is a feature
shared with Luciosoma and Barilius species of Groups B and C. This broad type of ethmoid appears
to represent a plesiomorph condition amongst the Cyprinidae. In the various groups investigated
so far, it is the most plesiomorph members which possess a wide, shallowly notched supra-
ethmoid-mesethmoid and a vomer that protrudes little beyond the anterior ethmoid border (see
Howes, 1978). With elongation of the cranium and jaws, there is seen a correlated narrowing of
the ethmoid and a posterior elongation of the anterior notch. Such a morphocline can be found

BARILIINE CYPRINID FISHES
PA

149

EPO

EXO

BOP

EM

PTS

PTSF

AHF LJB

STF PSFF

Fig. 23 A. Engraulicypris sardella cranium in A, dorsal and B, ventral views. The left side is un-
shaded for clarity and the limits of the various fossae are indicated by dotted lines.

in the cultrines (Culter-Erythroculter\ aspinines (Aspius-Luciobramd) and hemicultrines (Xenocy-
pris-Ochetobius). (In barbines and labeoines, the most derived members of the respective groups
possess a medial elongation of the ethmoid block. This is presumably an adaptation involved with
the inferior position of the mouth.)

In the taxa reviewed in this paper, the dorsally excavated supraethmoid of Luciosoma is
considered apomorphic. A similar ethmoid morphology is found in some Rasbora species (see
p. 183). The thickened vomer of Opsariichthys and Barilius Groups B and C is also considered a
synapomorphy (see Howes, 1978).

The ethmoid region of Engraulicypris sardella is highly derived, its modification making for a
highly protrusile jaw mechanism (see p. 164). In this respect it shares no feature with any of the
genera described above.

The preethmoids of Opsariichthys, Zacco, all Barilius and Luciosoma are large and articulate
with the palatine. This is presumably the plesiomorph condition for in those taxa which show

150 G. J. HOWES

cranial elongation the preethmoids are reduced in size and in some hemicultrines (e.g. Ochetobius)
the palatine no longer articulates with the preethmoid but has moved posteriorly so as to contact
the edge of the mesethmoid.

There is no anterior myodome in any of the genera discussed here.

A character which appears to have some significance as an indicator of relationship amongst
the genera under discussion is the connection between the pterosphenoid, parasphenoid and
prootic. In Opsariichthys, Luciosoma, some species ofRasbora, all Barilim and Engraulicypris there
is some connection (either medial or lateral or both) between the pterosphenoid and the ascending
wing of the parasphenoid. In an earlier paper (Howes, 1978) I stated that in some species of
Barilius the pterosphenoid did not contact the parasphenoid. Having now re-examined all the
specimens used in that study I find that there is such a connection, albeit a slight one in some
cases. An exception is in Zacco where the pterosphenoid has no connection with the parasphenoid.

PA

EPO

PTF
PTE

BO 2mm

Fig. 24 Engraulicypris sardella cranium, posterior view of left side.

An associated character is the position of the anterior foramen of the trigemino-facialis
chamber. In all genera apart from Luciosoma, Engraulicypris and Barilius Groups B and C it is
formed by the anterior border of the prootic, but in the exceptional taxa the foramen is situated
within the lateral face of the prootic.

In order to determine the significance of these features a morphological survey was made of the
otic region in the Cyprinidae; the following 'classification' emerges.

Type 1. Pterosphenoid makes no contact with the parasphenoid; the anterior trigemino-facialis
foramen is formed by the anterior border of the prootic; lateral commissure variable in
length; myodome of variable depth; hyomandibular fossa sometimes invades the ptero-
sphenoid.

Type 2. Pterosphenoid contacts the medial face and a narrow dorsal area of the parasphenoid
wing; the anterior trigemino-facialis foramen is formed by the anterior border of the
prootic; lateral commissure wide; myodome deep; hyomandibular fossa extends on to the
pterosphenoid.

Type 3. Pterosphenoid contacts the dorsal border of the parasphenoid wing; the anterior
trigemino-facialis foramen is situated within the lateral face of the prootic; lateral com-
missure narrow; myodome shallow; hyomandibular fossa does not invade the pterosphenoid.

BARILIINE CYPRINID FISHES

151

GO

152 G. J. HOWES

Type 4. As type 3 but the anterior trigemino-facialis foramen formed by the anterior border of
the prootic and the hyomandibular fossa invades the pterosphenoid.

Type 5. Pterosphenoid contacts the parasphenoid via a medial parasphenoid process; lateral
commissure excessively narrow; myodome reduced; the pterosphenoid does not bear the
anterior hyomandibular fossa.

This classification is a basic one and a degree of intermediacy exists between the five types.

Type 1 appears to be the plesiomorph condition on the basis of the following evidence: (a) It is
the type of contact found in members of those cyprinid groups so far investigated which are
considered plesiomorphic on other grounds, (b) In juvenile specimens of Opsariichthys (c. 40 mm
SL) the pterosphenoid fails to contact the parasphenoid and the ascending parasphenoid wing
projects anteriorly as a small spur. During ontogeny a ventro-lateral process of the ptero-
sphenoid contacts the parasphenoid spur, (c) In the characoids and presumed 'primitive' teleosts
(i.e. Elopiformes) there is no contact between the parasphenoid and pterosphenoid.

B

Fig. 26 Semi-diagrammatic drawings of the pre-otic neurocranium of A. Cyprinus. B. Catla.
C. Rutilus. D. Alburnus. E. Labeo. The medial part of the prootic is cross-hatched, the ptero-
sphenoid darkly shaded.

As well as being present in Zacco and Barilius Group A, Type 1 morphology is characteristic of
Cyprinus, Carassius, Catla and Xenocypris (Figs 25D; 26A & B).

A derived condition of Type 1 is present in Leuciscus and Rutilus where a lateral strut of the
pterosphenoid extends ventrally but does not make contact with the parasphenoid wing (Fig.
26C). However, in Alburnus, Abramis, Squaliobarbus, Capoetobrama, Oreoleuciscus, Barbus
intermedius and some other taxa such a strut extends to join the parasphenoid (Fig. 26D).

Rognes (1973) described in labrids a strut extending from the pterosphenoid to the prootic;
this he termed the 'internal jugular bridge'. Rognes (1973) illustrated variable conditions in the
development of this strut which did not always contact the prootic. No link with the para-
sphenoid was described. Greenwood (1976) also described an internal jugular bridge in centro-
pomids and again illustrated the variability of contact - only in one case was there found to be a
direct pterosphenoid/parasphenoid contact.

No such variability exists in any of the adult cyprinids examined and the 'incomplete' contact
of the pterosphenoid spur and parasphenoid (e.g. Rutilus) or complete contact (Abramis) are
taxon-defining characters.

Type 2 morphology is confined to Opsariichthys. In this taxon the medial wall of the prootic
does not extend beyond the anterior border of the parasphenoid wing, in contrast to the situation
in Type 1 where the prootic extends well forward (cf. Figs 25B & D).

BARILIINE CYPRINID FISHES 153

Type 3 morphology is confined to Barilius Groups B, C, Engraulicypris and Luciosoma (Fig.
25C); in no other taxa so far examined is the anterior trigemino-facialis foramen placed within
the lateral face of the prootic. This is considered a derived condition and is seen as a 'sequential
development' of Type 2.

Type 4 morphology appears to be a derived Type 1 condition in that there has simply been an
elongation of the pterosphenoid and a lengthening of the parasphenoid wing. Numerous examples
of this condition can be cited (e.g. Barbus barbus; Schizothorax esocinus; Ochetobius elongatus;
Erythroculter mongolicus; Elopichthys bambusd) all of which are amongst the most derived
members of their respective lineages in terms of cranial elongation (see comments on the ethmoid
region, p. 149). This morphotype is best exemplified by Oreoleuciscus (Fig. 27). In this genus there
is an extensive contact between the parasphenoid and pterosphenoid but an elongate fenestra
exists between the pterosphenoid and the prootic, posterior to the parasphenoid ascending wing;
the anterior trigemino-facialis foramen is also contained on the border of the prootic. (It is
interesting to note that the closure of the pterosphenoid-prootic-parasphenoid fenestra would
produce a condition approximating to that in Barilius Group B; cf. Figs 19B & 27.)

PTS

PAS

Fig. 27 Orbital-otic region of Oreoleuciscus pewslowi, lateral view.

Type 5 morphology has been encountered only in Labeo, Barbichthys and Tylognathus and is
considered to be a derived condition (Fig. 26E).

A seemingly derived character shared by Zacco and Barilius Group A is the enlarged carotid
foramen between the prootic and parasphenoid. This feature is also found in Cyprinus, Rutilus
and Abramis. The common condition in the Cyprinidae is a small carotid foramen.

Mention should be made here that Vanderwalle (1974) considers the trigemino-facialis chamber
of Gobio to represent a 'derived' condition because the jugular vein is not enclosed within a
separate channel or tunnel. I have not found such an enclosure in any cyprinid examined,
although in Engraulicypris, Leptocypris and some species of Barilius Group C there is a deep
jugular groove along the lateral face of the prootic. Associated with this feature is a thin bridge
of bone (LJB, Figs 19A; 22 A & B; 23) crossing the posterior foramen of the pars jugularis and
serving to separate the jugular from the trigemino-facialis nerve bundle. In some species of
Barilius this bridge is incomplete and developed only as a thin process arising from the ventral
part of the groove. A similar lateral jugular bridge occurs in alburnines and cultrines. Until the
distribution of this feature is more fully documented for the cyprinids, no polarity can be ascribed
to it.

The pterotic, when viewed dorsally, in Opsariichthys, Luciosoma, Engraulicypris and Barilius
Groups B and C is seen to extend well posteriorly, whereas in Zacco and Barilius Group A, the
posterior margin of the bone is truncated (Fig. 28).

A ventrally open posterior myodome is a feature of Barilius species in Group A, B and C. The
only other cyprinid genera in which I have found an open myodome are Leptocypris and En-
graulicypris sardella. Likewise, in those taxa the morphology of the eye muscles associated with
the myodome is similar to that in Barilius (see p. 147). A ventro-posteriorly open myodome is
present in several groups of characoids, and in these too some of the eye muscles originate from

154 G. J. HOWES

outside the myodome, the most specialized condition being found in the bryconines, acestro-
rhynchines and some characinines (pers. obs.). An open myodome and externally originating eye
muscles also occurs in clupeoids, salmonids and some elopids. Patterson (1975) gave an account
of the possible phylogenetic history of the posterior myodome, believing the open myodome to
be a derived condition. Verraes (1976) showed that in Salmo the myodome opens during ontogeny
and the eye muscles extend posteriorly through it. Thus, it would seem that the open myodome in
cyprinids must be thought of as a derived state and synapomorphic for those genera possessing it.
It was noted (p. 142) that in a small specimen of Opsariichthys the myodome was open, but
closed during later ontogeny. Certainly the morphology of the posterior myodome in Opsari-
ichthys differs quite considerably from that of other cyprinids in that the basioccipital is constricted
below the lagenar capsule. Thus the posterior portion of the myodome is enclosed within the
anterior part of the basioccipital, whereas in other cyprinids it extends to the posterior wall of the
basioccipital. The same myodome morphology is apparent in the most derived species of Barilius
Group B which suggests that the closed myodome is, as in Opsariichthys, secondarily derived,
although I have been unable to ascertain whether in fact an open myodome occurs during
ontogeny in these Barilius species.

The connection between the subtemporal fossa and posttemporal fossa is a feature common to
all the genera presently under consideration with the exception ofZacco. This character has also
been found in Leptocypris and in some species of Rasbora; its occurrence in these genera is
discussed later (p. 180). As no other cyprinid taxon examined has such a connection between the
fossae, this character is considered to be a synapomorphy.

Dermal bones of the skull

Opsariichthys

The frontals (F, Fig. 28) are broad and short with an irregular midline suture. The lateral border
of each frontal is markedly concave where it contacts the supraorbital. Medial to its suture with
the sphenotic there is a deep ridge from which stems the anterior part of the levator arcus palatini
(FF, Fig. 13A). The frontal slightly overhangs the sphenotic to form a medial roof to the dilatator
fossa. It contacts the pterotic across a narrow area.

Each parietal (PA, Figs 13 & 28) is broad and short (contained 4 times in the length of
the frontals) with an irregular midline suture. The lateral border contacts the pterotic and the
posterior border, the epioccipital and supraoccipital. The posterior border of the parietal is
marked by a deep ridge through which runs the laterosensory canal. Each nasal (N, Figs 1-3) is
small with a single dorsal opening. The supraorbital (SOR, Fig. 29) is very wide and does not
contact the 5th infraorbital, the bones being separated by the outward curvature of the frontal.
The 1st, 2nd, 3rd and 4th infraorbitals are large bones (Fig. 29). The 3rd and 4th do not cover the
cheek but leave a naked area between their posterior borders and the preoperculum. The 5th
infraorbital is reduced and contacts the pterotic.

Zacco

The frontals are narrow anteriorly and their lateral margins are gently concave (Fig. 28B). The
parietals are broad and short, their posterior margins truncated. The supraorbitals are long and
narrow. The 1st infraorbital is deep, the 2nd and 3rd shallow, 4th deep, and 5th reduced to an
ossification around the sensory canal; it does not contact the supraorbital. There is interspecific
variability in the size of the infraorbitals which are deeper and longer in Zacco macrophthalmus,
Z. temmincki, Z. barbatus and Z. pachycephalus than in Z. platypus.

Luciosoma

The frontals are broad and long, with their lateral margin gently concave. As in Opsariichthys
there is a deep frontal fossa (FF, Fig. 18). A particular feature of Luciosoma is a deep channel
between the anterior opening of the frontal sensory canal and the posterior opening of the nasal

BARILIINE CYPRINID FISHES
SOR

155

PTE

EPO

Fig. 28 Cranial roof of A. Opsariichthys. B. Zacco platypus. C. Barilius bold.

canal (FG, Fig. 6A). The nasal is short without a central dorsal pore (Fig. 6A). Each wide supra-
orbital contacts the 5th infraorbital. The 1st and 2nd infraorbitals are shallow, the 3rd expanded
to almost cover the cheek, the 4th and 5th long.

Barilius

Group A. The frontals are narrow as compared with Opsariichthys, their lateral margin gently
concave. The dilatator fossa deeply indents the postero-lateral margin of each frontal. The
parietals are long, varying interspecifically from 2-5 to 2-8 times in the length of the frontals
(cf. 3-8-4-0 in Opsariichthys and in other Barilius groups). The nasals are elongate each with 2
dorsal pores. The 1st and 2nd infraorbitals are shallow, the 3rd, 4th and 5th expanded (Fig. 29 B).
The 5th infraorbital contacts the narrow supraorbital .

Group B. The frontals are elongate and narrow, their lateral margin gently concave where the
supraorbital joins. The parietals are slightly longer than those of Opsariichthys (3-8 times in
frontal length) and their posterior lateral margins are extended (Fig. 28C). The 1st and 2nd
infraorbitals are deep, the 2nd being elongate. The 3rd, 4th and 5th infraorbitals are expanded so
as to cover the entire cheek region and contact the preoperculum (Fig. 29C); the 5th contacts the
broad supraorbital.

Group C. The frontals are elongate but broader than those in Group B. In Barilius microcephalus
the ventral frontal fossa is perforated so as to form a lamellar ring (FR, Fig. 17B). Part of the
levator arcus palatini originates from the ventral surface of the frontal anterior to the ring, the
muscle passing through it to join its other section which originates from the deep fossa dorso-
posteriorly to the ring. In other species included in the group there is a deep frontal fossa similar

156

G. J. HOWES

SOR

IO5

Fig. 29 Circumorbital series of A. Opsariichthys. B. Barilius barila. C. Barilius bola.

to that in Opsariichthys (see above). The arrangement of the circumorbital bones is like that in
species of Group A.

Engraulicypris

Thefrontals (Fig. 23) are long and narrow and formed anteriorly into a shallow pit. The sensory
canal contains a single pore midway along its length. Ventrally, each frontal bears a fossa which
serves as a site for the origin of part of the levator arcus palatini muscle. The parietals are long,
being contained twice in the length of the frontals. The nasals are elongate and curved so as to lie in
the concavity of the ethmoid margin; each nasal has two dorsal pores. The supraorbital is long
and broad and contacts the 5th infraorbital. The 1st infraorbital (Fig. 30) is extensive and when
the jaws are retracted, it completely covers the maxilla and premaxilla. The posterior border is

BARILIINE CYPRINID FISHES 157

markedly concave and covers the anterior third of the 2nd infraorbital; the 3rd, 4th and 5th
infraorbitals are long.

Comments

There is no contact between, the supraorbital and the 5th infraorbital in Opsariichthys or Zacco
but such contact is present in Luciosoma, Engraulicypris and all groups of Barilius and in these
taxa the infraorbitals are expanded.

SOR IO5

I I

Fig. 30 Circumorbital series of Engraulicypris sardella.

In the majority of cyprinids the infraorbitals are reduced, presumably a secondary regression
from an expanded ancestral condition (Tretiakov, 1946). In some groups such as the aspinines,
where there has been elongation of the cranium, the infraorbitals are reduced to the canal-bearing
part of the bone, whereas in the long-headed bariliines the infraorbitals are expanded. This
condition appears to be due in part to the lateral expansion of the cranial musculature in bariliines,
as evidenced by the produced sphenotic processes (Fig. 28C) and the need for a protective shield
(p. 158). In the case of Barilius Group B (Fig. 29C), it appears that the expansion of the infra-
orbitals is a derived condition. This is clearly seen in the excessively elongate 2nd infraorbital and
the pronounced orbital narrowing of the 3rd. Indeed, a transformation series of this character
state can be witnessed in species of Group B and the more extreme forms of this morphocline are
represented in various bariliine genera (see Figs 29A-C) where Opsariichthys represents the
plesiomorph state and Barilius bola the derived.

Ramaswami (1955) has already commented on Regan's (1911) statement that Opsariichthys
has a ' . . . complete series of circumorbitals'. Ramaswami thought Regan's use of the word
'complete' ambiguous. Regan may have been referring to a series without interruption, in other
words he believed there to be a contact between the 5th infraorbital and the supraorbital. The
separation is only a narrow one and the interposing frontal could be mistaken for part of the
supraorbital.

A frontal fossa is present in all the genera under discussion with the exception of Zacco. In
Opsariichthys and Luciosoma the fossa is shallow and serves to support the levator arcus palatini
muscle. In Barilius Group A the fossa is a deep anterior extension of the dilatator channel; in
Group C, the fossa is also deep (and in B. microcephalus is penetrated by a foramen) but, as in
Opsariichthys and Luciosoma it provides the site of origin for the levator arcus palatini. In Barilius
Group B the frontal fossa has all but disappeared and exists only as a narrow ventral ridge. Apart

158

G. J. HOWES

MMP

Fig. 31

B

Anterior of maxilla and premaxilla of A. Opsariichthys. B. Zacco macrophthalmus. Ventral

views.

from Leptocypris and Engraulicypris sardella and some Rasbora species, I have not found any
cyprinid with this type of frontal cavity. Many taxa, however, possess a sphenotic fossa, lying
anteriorly to the sphenotic lateral process, from which the levator arcus palatini originates (e.g.
Zacco, Alburnus).

Forey (1975) commented on a frontal ridge in clupeoids where it also serves as a site of origin
for the levator arcus palatini. He also pointed out the apparent correlation between a frontal
ridge and a hyomandibular process (see p. 165).

The jaw bones

Opsariichthys

The jaws have been described in some detail previously (Howes, 1978), therefore it is necessary
here to note only the more pertinent features: The premaxilla is very slender with the anterior
ascending process sloped backwards. The maxilla bears, antero-medially, an axe-shaped process
which underlies the permaxilla and which contacts its partner from the opposite side along a
narrow lateral face (Fig. 31 A). The anterior tip of each maxilla extends almost to the symphysis
of the premaxillaries (Fig. 32A). The dentary is deeply notched anteriorly and has a strong
symphysial process. The ventral border bears 7-8 pores of the sensory canal. Posteriorly there is a
high coronoid process, see Fig. 36E in Howes (1978).

Zacco

The premaxilla is a thick bone with a short ascending process. The anterior tips of each maxilla
are widely separated from one another across the midline. Each anterior medial maxillary

BARILIINE CYPRINID FISHES 159

process is rounded distally and joined to its partner across the midline via a ligament (Fig. 3 IB).
The dentary is shallow with a high coronoid process; a symphysial process is lacking.

In Zacco pachycephalus and Z. barbatus the jaws are like those in Opsariichthys in that (i) the
premaxilla is thin and separated by a slight gap from the maxilla, (ii) the medial maxillary
processes contact each other directly across the midline and not via a ligament and (iii) the
anterior maxillary tips extend well forward.

Luciosoma

The premaxilla is exceedingly shallow, with a wide and short anterior ascending process which
joins its partner to form an almost beak-like jaw. The maxilla is also shallow, overlapping the
premaxilla for that bone's entire length. The anterior medial maxillary process is wide and contacts
its partner along the midline. The outer anterior tip of the maxilla extends almost to the pre-
maxillary symphysis (Fig. 32C). The dentary is shallow, with a short symphysial process. The
lateral wall is curved outward as in Opsariichthys and the coronoid process is high.

Barilius

Group A. The premaxilla and maxilla are narrow, the former with a rather long ascending process
and the latter with its anterior medial process making contact with its partner in the midline. The
anterior tips of the maxillaries are widely separated from each other (Fig. 32B). The dentary is
very shallow with a high coronoid process and there is no symphysial knob or anterior notch.

Group B. The jaws of Barilius bola are described in an earlier paper (Howes, 1978) and those of
other species assigned to this group correspond in having shallow premaxillaries with short, wide,
beak-like anterior ascending processes, shallow maxillaries each bearing a wide palatine notch
and a wide medial anterior process. The anterior tip of each maxilla almost contacts its fellow
across the midline (Fig. 32E). The dentaries are long and shallow with high coronoid processes
and they lack symphysial knobs and anterior notches.

Group C. The jaws are closely similar to those of Group B except for deeper and shorter jaw
bones and the tips of the premaxillary ascending processes being triangular instead of curved
(Fig. 32D).

Engraulicypris

The premaxilla (Fig. 33C) has a short anterior ascending process, which is slightly indented in its
leading edge, and a broad midlateral ascending process. The maxilla (Fig. 33A) bears a high,
narrow midlateral ascending process from which extends a ligament to attach to the medial face
of the 2nd infraorbital. The anterior tip of the maxilla is bifurcated, the medial process extending
ventrally to contact its partner in the midline (Fig. 33B). The posterior part of the maxilla is
rounded and expanded ventrally, this expanded portion is grooved so as to articulate with the
dorsal posterior margin of the premaxilla. The dentary (Fig. 33 D) is deep, with a short, broad,
upright coronoid process, the posterior border of which is deeply concave. The anguloarticular
is large with a long convex dorsal margin.

Hyopalatine arch

Opsariichthys

The hyomandibula possesses two articular heads separated from each other by a shallow depres-
sion. The anterior facet has a long sloped face which abuts against the pterosphenoid wall of the
articulatory fossa. The posterior facet has a narrow surface which fits into the channel-like
pterotic fossa. The anterior border of the hyomandibula is rounded below the anterior facet and

160

G. J. HOWES

RMX

10mm

Fig. 32 Anterior of maxilla and premaxilla of A. Opsariichthys. B. Barilius barila. C. Luciosoma
bleekeri. D. Barilius microcephalus. E. Barilius bola. Dorsal views.

then becomes perpendicular to form a broad shaft. The upper lateral face bears a well-developed
flange across which passes the dilatator operculi muscle and from which part of adductor mandibulae
A 2 originates. The posterior border of the hyomandibula overlaps the dorsal part of the pre-
operculum and its medial face is almost flat and produced dorso-posteriorly as a compressed
condyle with which the operculum articulates.

The metapterygoid has a markedly concave dorsal border, the posterior part of which is
formed into a wide triangular process directed somewhat anteriorly. This process is a point of
attachment for the levator arcus palatini muscle. The ventral border is concave and forms the
dorsal margin of a large fenestra. Posteriorly the metapterygoid extends to overlap the lower
limb of the hyomandibula.

The symplecic is a long curved bone forming part of the ventral border of the fenestra. It is
separated by cartilage from both the lower limb of the hyomandibula and the metapterygoid.

The quadrate has a long ventral border overlapping the symplectic. The lateral face of this part
of the bone projects as a wide shelf and provides a surface for the origin of the adductor mandibulae
A! muscle. The ascending face of the quadrate is axe-shaped, its concave inner border forming
the anterior margin of the fenestra, its outer border is curved slightly backwards.

The entopterygoid has a slightly convex dorsal border and is directed medially to contact the
parasphenoid. It underlies the medial faces of both the metapterygoid and quadrate.

The ectopterygoid is a narrow lamellate bone with a convex border. It articulates with a
narrow ledge along the border of the metapterygoid.

The palatine is triangular in cross-section, its ventral border produced into a lamellate edge. The
bone narrows anteriorly to form into an articular head bearing a lateral process and a medial
facet which abuts against the preethmoid cartilage. Posteriorly the palatine fits into a right-angled
excavation in the border of the entopterygoid; a ligament joins its ventral edge to the ectoptery-
goid (see Fig. 38A in Howes, 1978).

BARILIINE CYPRINID FISHES

MMR

161

B

D

Fig. 33 Engraulicypris sardella, upper and lower jaws. A. Maxilla (lateral view). B. Maxillary
process (medial view). C. Premaxilla (lateral view). C. Dentary (lateral view).

Zacco

The hyomandibula differs from that of Opsariichthys in that there is no separation of the anterior
and posterior facets; the entire dorsal border being a condylar surface. The anterior dorsal edge
slopes forward at a more oblique angle, the upper part of the bone is narrower and has an almost
perpendicular anterior margin and the lateral flange is absent.

Zacco pachycephalus differs from other species in that the hyomandibula bears a well-developed
lateral flange (see p. 189).

Other features in which the hyopalatine arch of Zacco differs from that of Opsariichthys are the
lack of a metapterygoid process (except in Z. pachycephalus ; see p. 1 89), a smaller metapterygoid -
quadrate fenestra (the correlate of a wider quadrate and a deeper metapterygoid) and the absence
of a ventral lamellate ridge on the palatine.

Luciosoma

The hyomandibula closely resembles that of Opsariichthys in having a long dorsal border, shallowly
separated articular facets and well-developed lateral flange. The quadrate lacks the lateral shelf
present in Opsariichthys. The metapterygoid process is absent and the metapterygoid-quadrate
fenestra, although present, is very small. The ventral border of the entopterygoid is longer than in
Opsariichthys and there is a corresponding lengthening of the ectopterygoid. The palatine bears a
rounded ventral ridge.

Barilim

Group A. There is some interspecific variation in the morphology of the hyopalatine arch, viz.:
the hyomandibula bears separate condyles, although the degree of separation is variable, as is the

162 G. J. HOWES

development of the lateral flange. The metapterygoid dorsal process is absent in all species but in
B. bendelisis the dorsal border is markedly concave and shaped into a long posterior process.
A metapterygoid-quadrate fenestra is sometimes lacking.

Group B. The anterior and posterior condyles of the hyomandibula are widely separated, the
anterior extending forward at an angle of 45 to the vertical to articulate in a small sphenotic-
pterosphenoid fossa; the posterior condyle is high, its narrow dorsal surface articulating with the
long pterotic fossa. The lateral face of the hyomandibula bears a long, laterally extending flange
which, in B. bola, is curved upwards so as to form a deep cavity. The anterior hyomandibula
border is concave.

The metapterygoid in all species bears a long anteriorly directed dorsal process. This process is
in the form of a gutter with the open side being dorsad. It provides both an insertion channel for
fibres of the levator arcus palatini muscle and as a point of origin for fibres of the adductor
mandibulae A 3 muscle. This process is developed to the greatest extent in Barilius bola and B.
guttatus, its distal tip almost reaching the same horizontal plane as the top of the anterior hyo-
mandibular condyle. The posterior border of the metapterygoid extends more than halfway along
the lateral surface of the hyomandibular stem (see Fig. 38B in Howes, 1978).

The symplectic is well developed and deep; the quadrate is also deep with a laterally produced
ventral border as in Opsariichthys. The ectopterygoid is very elongate with a strongly concave
lower margin. As in Opsariichthys the palatine bears a strong ventral ridge. A metapterygoid-
quadrate fenestra is absent.

Group C. As in Group B the hyomandibular condyles are widely separated but the anterior
condyle is much thicker than in any of those species and in this respect more closely resembles the
Opsariichthys condition. The lateral flange is weakly developed and the anterior border irregularly
concave. The metapterygoid bears a small posterior lamellate process similar to that in Opsari-
ichthys. The quadrate is shallow, its ventro-lateral border well produced ; the symplectic elongate
and the palatine with a well-developed, knife-edged ventral ridge. There is no metapterygoid-
quadrate fenestra.

Engraulicypris

The hyomandibula has a long upper face, with an almost continuous articulatory dorsal border
and a convex anterior margin. The dorsal posterior edge is extended to form a triangular condyle.
The metapterygoid is directed medially at a sharp angle so as to form an almost flat shelf. The
medial part of the bone is produced into a long lamellar process. The quadrate is deep, its ventral
posterior border produced into a thin spine. The palatine bears a slight lateral process which is
separated from a medial process by a deep channel. A ligament joins the lateral process to the
antero-dorsal part of the maxilla (LPM, Fig. 34). Another ligament (LPE) joins the palatine
shaft to the ethmoid, and a further one (LPEK), which passes over the ethmoid arm, attaches the
medial palatine process to the kinethmoid shaft.

Opercular series

Opsariichthys

The operculum is deep with a short dorsal border, the dorso-anterior process is curved outwards.
The articulatory facet for the hyomandibular condyle is situated dorso-medially; extending
posteriorly from the facet is a thin strut pierced by two foramina. The preoperculum bears four
latero-sensory canal pores along its nearly vertical upright limb and three along the horizontal
part. The suboperculum has a well-rounded ventral border and dorsally is partially covered by
the operculum. The interoperculum is wide and is largely covered by the preoperculum laterally.

Zacco

There is little difference between Zacco and Opsariichthys, except that the anterior dorsal process
of the operculum is feebly developed.

LKP

163

LPM

LKP

LPEK

LPE

Fig. 34 Engraulicypris sardel/a, sequence of jaw protrusion (semi-diagrammatic). A. 1st stage,
mouth partially open. B. Intermediate stage, mouth completely open. C. Final stage, upper jaw
fully extended

164 G. J. HOWES

Luciosotna

The upright limb of the preoperculum is aligned at 45 to the vertical, and its horizontal limb is
short. The operculum has a well-rounded border and is much shallower than in either Opsariichthys
or Zacco; its anterior dorsal process is feebly developed.

Barilius

Group A. The opercular series differs little from that of Opsariichthys or Zacco, but in the species
of Group B the upright part of the preoperculum is inclined at an angle of 30 to the vertical and
the horizontal limb is short. In Barilius loati, steindachneri and kingsleyae the width of the
operculum is greater and the anterior dorsal process more highly developed and anteriorly extended
than in any other species of the group. In Group C the opercular border is slightly concave and its
lower posterior border produced.

Engraulicypris

The operculum has a long, straight dorsal border and its posterior margin is only slightly convex.

Comments on the jaws, hyopalatine and opercular series

In a previous paper (Howes, 1978) I made some comment on the jaw bones of various cyprinids
and considered those of Opsariichthys as relatively primitive. However, I now realize that in some
respects they are highly specialized and that the separation of the premaxilla and maxilla, and
the strongly developed symphysial knob and notch must be considered as derived features.
Only in Macrochirichthys is there a similar development of the symphysial knob, but this is
regarded as a parallelism; Howes (1979). The coronoid process of all the genera under considera-
tion is of the plesiomorph form, i.e. relatively high and without any contribution from the angulo-
articular bone (Howes, 1978).

There is some variation in the degree of development of the anterior ascending premaxillary
processes in the taxa under discussion. In Opsariichthys, Zacco and Barilius Group A, the ascending
processes are long, with concave lateral borders and rounded dorsal tips. In Luciosoma the
processes are shorter with broader tips and in Barilius Group B the processes are much reduced
with a completely convex lateral border. Although in Barilius Group C, the lateral border of the
premaxillary process is concave the dorsal tip is triangular, allowing for closer contact with the
expanded kinethmoid (Fig. 32). Likewise, a transformation series can be seen in the development
of the anterior part of the maxilla where in Barilius Group A, the tips of each maxilla are widely
separated from each other but become progressively more elongate in Opsariichthys and Luciosoma
until, in Barilius Groups B and C, they almost contact each other across the midline.

The curved walls of the dentary are shared features of Opsariichthys and Luciosoma; the
anterior medial maxillary processes make direct contact along the midline in all genera except
Zacco where contact is indirect and via a ligament; this is the condition usually encountered in
the Cyprinidae and is therefore considered to be the plesiomorph one.

Engraulicypris sardella differs from all other genera reviewed here in the marked development
of highly protrusile jaws. The high degree of protrusibility is attained by the prolongation of the
ethmoid arms in conjunction with ventrally directed maxillary processes which enables the
kinethmoid to rotate through more than 180. The ligament connecting the distal tip of the
kinethmoid to the anterior border of the premaxilla is extremely long and when the jaws open
and protrude, the tensed ligament extends almost horizontally. When the jaws close and retract,
the kinethmoid pivots (in the anterior vomer notch, see p. 137) so that it again lies horizontally
between the ethmoid arms. The long ligament curls over and folds into the ethmoid indentation
(Figs 34A-C). The ventrally directed medial maxillary process is a feature shared with Lepto-
cypris (see p. 181).

The most significant features of the hyopalatine series concern the hyomandibula and the
metapterygoid. Only in Zacco and Engraulicypris does the hyomandibula lack the marked separa-
tion of its articular condyles. The most extreme separation is found in Barilius Group B. It could

BARILIINE CYPRINID FISHES 165

be argued that such a separation of the condyles is a correlate of the elongation of the head.
Admittedly this is partly so, for in Barilius Group B there is some lengthening of the otic region.
However, in other cyprinids (such as the aspinines) where there has been considerable elongation
of the skull (see Howes, 1978) there is no correspondingly great separation of the hyomandibular
condyles. Neither have I found such a feature in piscivorous long-headed characins. The sig-
nificance of the bipartite hyomandibula in functional terms is probably to be explained by the
vertical elongation of the suspensorium rather than through the horizontal elongation of the
skull. Whether or not this is so, the feature is certainly a derived one and is synapomorphous for
Opsariichthys, Barilius and Luciosoma. In these genera the hyomandibula also bears a lateral
flange, which again is developed to the most marked degree in Barilius Group B. Although a
character occurring in other cyprinids (see Howes, 1978 and comments by Forey, 1975) it is here
identified as a synapomorphy.

The metapterygoid in Opsariichthys and Barilius Groups B and C bears a dorsal posterior
process. An incipient condition of this process is present in Opsariichthys and Barilius Group C,
whilst an increasing grade of development is manifest throughout Group B, as is also a deepening
of the metapterygoid. I have failed to find such a feature in any other group of cyprinids and
therefore regard this as a synapomorph character linking those genera in which it occurs.

Engraulicypris sardella differs from the other genera both in the form of the hyomandibula and
the metapterygoid. Although there is no marked separation of the hyomandibular condyles, the
posterior condyle is produced into a triangular shape which articulates with a fossa on the
posterior border of the pterotic. In this taxon, unlike Zacco, it is probably a derived feature,
the form of the articulation suggesting a wide lateral movement of the hyomandibula. The
metapterygoid spine in Engraulicypris also differs from that in Barilius Groups B and C in that
it originates from the anterior part of the metapterygoid and serves as a site of attachment for
the adductor arcus palatini instead of the levator arcus palatini as in the other taxa. A deep lateral
shelf along the ventral part of the quadrate is also shared by Opsariichthys and Barilius B and C.

The metapterygoid-quadrate fenestra has been discussed at some length in an earlier paper
(Howes, 1978). It need only be mentioned that of those genera discussed here it is present in
Opsariichthys, Zacco and Luciosoma; its occurrence in other groups of cyprinids (Howes, 1979)
suggests that it is a plesiomorph feature.

The elongate symplectic found in Opsariichthys, Zacco, Barilius Groups A and C and in Lucio-
soma is the 'usual' cyprinid condition and is thus considered to be plesiomorphic. The short,
deep symplectic in Barilius B is a derived feature.

The opercular series differs little between the genera; only in Luciosoma and Engraulicypris is
there any marked elongation of the operculum and shortening of the preoperculum, whilst in
Engraulicypris the interoperculum is finely serrated.

Hyoid and branchial arches

Opsariichthys

There are few features distinguishing these elements from those of other cyprinids (Ramaswami,
1955; Howes, 1978). Attention is drawn to the narrow waisted ceratohyal, reduced ventral
hypohyal and the three broad branchiostegal rays. The basihyals are all short.

Each ceratobranchial bears 8-9 gill rakers. The pharyngeal bone bears three rows of curved
teeth (dental formula, 2.3.5). The 4th epibranchial has a broad, triangular dorso-medial process
posteriorly and a short, rod-like process midway along its length (Fig. 35). There are two ossified
infrapharyngobranchials identified as the 2nd and 3rd. The 3rd is an extensive triangular element
spanning the medial surfaces of epibranchials 3 and 4. A cartilaginous element at the outer edge of
infrapharyngobranchial 3 and articulating with the 4th epibranchial probably represents the 4th
infrapharyngobranchial .

Chu (1935) studied the pharyngeal bones of Opsariichthys and Zacco and noted that those of
the former genus were narrow and elongate, whilst those of the latter were short and broad. A

166

G. J. HOWES

similar distinction can be found within the groups of Barilius, those of Group A resembling Zacco,
with Groups B and C and Luciosoma resembling Opsariichthys.

In Engraulicypris sardella the ceratobranchials are very long and bear 24-25 long, slender gill-
rakers. The pharyngeal bone has 3 rows of compressed teeth (dental formula: 5.3.2) with finely
serrated cutting edges and recurved tips.

Pectoral girdle (nomenclature follows that of Brousseau, 1976)

IF2

EB4
Fig. 35 Opsariichthys. Upper branchial arches, dorsal view.

Opsariichthys

The upright part of the cleithrum is perpendicular and bears a well-defined ridge along its anterior
border (Fig. 36A). This ridge broadens ventrally and fans out to form a broad, horizontal
cleithral lamina; the leading edge of the horizontal limb is rounded. There is a well-developed
ventro-medial cleithral ridge. The posterior cleithral lamina is broad with a straight border. The
anterior fenestra is well developed. The supracleithrum is very elongate. Each coracoid is small
and when viewed laterally is seen to extend only slightly below the ventral margin of the cleithrum.
The coracoids diverge, meeting only anteriorly; the mesocoracoid shaft is long, rounded and
perpendicular. The postcleithrum is short and spatulate, curves mesially with its ventral tip on a
line level with the ventral border of the coracoid (Fig. 36B). There are four proximal radials, the
3rd and 4th being lamellar.

Zacco

The pectoral girdle morphology is basically that of Opsariichthys but the sternohyoid ridge on the
cleithrum is greatly produced anteriorly resembling that in Danio (see Howes, 1979). The supra-
cleithrum is short and its anterior edge straight with a slight distal notch. The postcleithrum is
shorter than in Opsariichthys and its ventral tip pointed instead of spatulate.

Luciosoma

The upright part of the cleithrum is shorter than in Opsariichthys or Zacco and the posterior
lamina is extended (Fig. 37A). The horizontal limb is broad with a moderately developed ridge.

BARILIINE CYPRINID FISHES

167

The supracleithrum is short, the coracoids shallow, meeting each other along the midline and
containing a large anterior fenestra. The most outstanding feature of the pectoral girdle is the
form of the postcleithrum. It is a broad lanceolate bone attached to the edge of the posterior
cleithral lamina extending laterad to the body scales. Its tip is enclosed in a fleshy sheath and this
sheath extends as the pectoral axial lobe (see Fig. 37A & p. 168) to lie between the pectoral fin
and the body. There are no tendons attached to the postcleithrum.

MC

COR
Fig. 36 Opsariichthys. Pectoral girdle in A, lateral and B, medial views.

Barilius

Within Group A there is some variation in the morphology of the pectoral girdle; the upright part
of the clei thrum is generally short with a posteriorly extended lamina; the supracleithrum is
elongate, the coracoids variable in depth but never joining along their medial faces.

In Group B the upright part of the clei thrum is very narrow with a well-developed anterior ridge,
and the supracleithrum is long and narrow (Fig. 37B). The coracoids are very shallow and meet
anteriorly along their ventral margins; the mesocoracoid shaft is broad and lamellar. The develop-
ment of the anterior fenestra is variable, being small in B. loati and extensive in B. bola.

Species of Group C possess a broadly lamellate upright cleithral limb with an expanded pos-
terior edge to which is attached a spatulate postcleithrum. The supracleithrum has only a slight
anterior inclination as compared with other Barilius where it is inclined at 45 to the cleithrum.
The coracoid is well developed, with a curved ventral margin and a large anterior fenestra. The
mesocoracoid shaft is broad and lamellar.

In all Barilius species examined thepostcleithra are small and spike-like, medially directed bones
which unlike the postcleithrum in Luciosoma do not support an axial flap or scale.

Engraulicypris

The upright part of the cleithrum is short, its ventro-posterior edge extended backwards as a
long spine. The coracoids are shallow and meet each other along the midline. A postcleithrum is
absent.

Comparisons

The most notable variations in the pectoral girdle are in the vertical and horizontal extensions

168

G. J. HOWES

PIT

Fig. 37 Pectoral girdles of A, Luciosoma bleekeri, B, Barilius bola.

of the cleithrum. It has been noted previously (Howes, 1978) that in some long-jawed cyprinids
such as Barilius bola the vertical limb is long and narrow. In Luciosoma, on the other hand,
as in the aspinines (Howes, 1978), the upper limb is short and broad and more closely resembles
that in Opsariichthys. The postcleithrum in Opsariichthys and Zacco is elongate and medially
directed as, indeed, it is in the majority of cyprinids examined - and this is taken to represent the
plesiomorph condition. In all Barilius species it is reduced to a spike-like or scale-like bone and in
Luciosoma lies horizontally above the axial aperture and supports the axial lobe. The only other
cyprinids in which the postcleithrum supports the axial lobe is a group of Rasbora species
(represented by R. argyrotaenid)', however, here the postcleithrum is thin and ventrally directed.
In the cheline Securicula the postcleithrum is absent and 'replaced' by a large scale covering the
dorsal border of the axial aperture. In chelines a reduction of the postcleithrum was considered a
derived character but some forms of postcleithral elongation, as in the case of Macrochirichthys,
could also be interpreted as derived (Howes, 1979). In the case of the bariliines I regard reduction
of this element as an apomorphy, likewise in Engraulicypris where it is absent.

Vertebral skeleton

Opsariichthys

Weberian apparatus (Fig. 38). The 1st vertebra bears strong lateral processes which are rounded
proximally and compressed distally. The 2nd vertebra has a thick lateral process which is inclined
upward. The 2nd and 3rd vertebrae are not fused together and the 3rd bears a wide fossa on its
postero-lateral face for the articulation of the tripus. The 3rd vertebra is expanded dorsally into

BARILIINE CYPRINID FISHES

169

the 3rd neural plate which extends laterally and curves forward along the top of the 2nd centrum.
The anterior border of the 3rd neural plate slopes downward and is joined to the 2nd neural plate.
The neural complex is steeply raised, its curved dorsal border being deeply grooved; the anterior
border is rounded and extends forward to above the 2nd neural plate. The ossicles differ little
from those of other cyprinids (Ramaswami, 1955; Howes, 1978) except that the tripus is much
elongated, its posterior tip extending to below the 6th centrum.

NS4

NC

NP2

NP3

OSS

Fig. 38 Opsariichthys Weberian apparatus, lateral view.

Vertebral column

There are 44 vertebrae (21 abdominal, including the 4 comprising the Weberian apparatus + 22
caudal + the fused pre-ural and 1st ural vertebra - counted as a single element in Table 1) and 6-7
supraneurals, the first being an irregularly shaped lamellate element, the others lanceolate.

Caudal fin skeleton (Fig. 39). There are 6 hypurals, none of which are greatly expanded, the 6th
being a small element. The fused pre-ural and ural centrum (PU1 + U1) bears a blunt neural
spine. There is one epural and a pair of slender uroneurals lying above hypural 6. The parhypural
bears a well-developed broad hypurapophysis.

Comparisons

The intergeneric differences in the morphology of the vertebral column are those involving the

anterior vertebrae which form the Weberian apparatus.

In Zacco the lateral processes of the 1st vertebra are reduced, whilst those of the 2nd are greatly
extended. This is similar to the condition in Opsariichthys, and likewise the 2nd and 3rd vertebrae
are separate.

In Barilius Group A the 1st and 2nd vertebrae are of the type encountered in Opsariichthys and
Zacco but the 2nd and 3rd centra are fused. The lateral processes of the 1st vertebra in Barilius
Group B are short and those of the 2nd are elongate with pointed tips and are directed posteriorly
at 45 to the transverse plane. In Barilius Group C the processes of the 1st vertebra are curved
backwards and partly underlie those of the 2nd vertebra. The lateral processes of the 2nd are long
with rounded tips and extend horizontally. The 2nd and 3rd vertebrae of all Barilius species
examined are fused together.

170

G. J. HOWES

HY6

Fig. 39

HYP
Opsariichthys Caudal fin skeleton.

Engraulicypris has short lateral processes on the 1st vertebra and long, posteriorly directed
ones on the 2nd (similar to Opsariichthys). Luciosoma, on the other hand, has long 1st vertebral
processes and expansive, wing-like processes on the 2nd vertebra (Fig. 40), a feature shared
with some species of Rasbora (p. 182).

VP2

TR

Fig. 40 Luciosoma bleekeri Anterior vertebrae, ventral view.

The morphology of the 1st and 2nd vertebrae exemplified by Opsariichthys appears to be
widespread amongst the cyprinids and is taken to represent the plesiomorph condition. The long,
sharply caudally directed 2nd vertebral processes of Raiamas and the expanded type of Luciosoma
are considered apomorphies.

One feature that seems common to all genera studied here is the form of the ossa suspensoria.
These bones are obliquely angled backwards so that their tips come to lie below the centre or
posterior part of the 5th centrum. Also, the lateral extension of the os suspensorium (4th pleural
rib of Ramaswami, 1955) is short and directed somewhat anteriorly. Only in Zacco are the ossa
suspensoria long and vertically directed, their tips meeting below the 4th centrum. This latter
condition is widespread amongst the cyprinids and is therefore reckoned as being the plesiomorph
one.

BARILIINE CYPRINID FISHES

171

Previously (Howes, 1978) I had given the caudal fin ray count in cyprinids as 19 + 91. This is an
error, I have now examined radiographs of many cyprinid genera and am in agreement with
Roberts (1973) that the count is 19 + 81.

Table 1 The modal number of vertebrae of various species used in this study

Opsariichthys uncirostris
Zacco temmincki
Zacco pachycephalus
Zacco platypus
Zacco macrophthalmus
Zacco barbatus
Luciosoma setigerum
Luciosoma bleekeri
Luciosoma trinema
Luciosoma spilopleura
Rasbora argyrotaenia
Rasbora elanga

Group A Barilius barila

Barilius tilheo
Barilius ornatus

Group B Barilius bola

Barilius guttatus
Barilius lujae
Barilius salmolucius
Barilius macrostoma
Barilius kingsleyae
Barilius loati
Barilius moori
Barilius microlepis

Group C Barilius ubangensis

Barilius microcephalus

Leptocypris modestus
Barilius niloticus
Engraulicypris sardella

44
43
40
40
42
42
43
42
39
40
34
42

40

42
41

47
47
41
43
44
41
44
43
47

43
45

40

37
44

(4+17 + 22+1)
(4+18 + 20+1)
(4+17+18+1)
(4+17+18 + 1)
(4+17 + 20+1)
(4+18+19+1)
(4 + 21 + 17+1)
(4 + 21 + 16+1)
(4+18+16+1)
(4+18 + 17+1)
(4+12+17+1)
(4+17 + 20+1)

(4+15 + 20+1)
(4+16 + 21 + 1)
(4+16 + 20+1)

(4 + 20 + 22+1)
(4+18 + 24+1)
(4+16 + 20+1)
(4+16 + 22+1)
(4+17 + 22+1)
(4+15 + 21 + 1)
(4+17 + 22+1)
(4+15 + 23 + 1)
(4+17 + 25+1)

(4+16 + 22+1)
(4+16 + 24+1)

(4+15 + 20+1)
(4+14+18+1)
(4+19 + 20+1)

Note : Those species in which there is a high number of vertebrae are either very elongate fish (B. bola, B. guttatus)
or lacustrine species (B. microlepis).

External anatomical features

Tubercles

Opsariichthys

Keratinized tubercles are present on the head and sometimes on the scales of the flanks and anal
fin rays of all species. They are present in both sexes but are most prominently developed on
males with ripe testes. The arrangement of the tubercles on the head is shown in Fig. 41 A and is:
1st infraorbital, 2 or 3 on the dorsal border, 2 on the ventral; 2nd infraorbital 5; 3rd infraorbital
5 or 6; 4th absent; 5th absent; preoperculum, 5 or 6 along the lower border; premaxilla (upper
lip) 5-8; dentary, 3 rows, 8 dorsal, 12 middle, and 8 ventral; nasal 2-3 situated lateral to the
nasal bone; supraorbital 4-5; cheek, 10-12, small and irregularly scattered.

172

G. J. HOWES

Fig. 41 Arrangement of tubercles on the heads of A. Opsariichthys uncirostris. B. Zacco macroph-
thalmus. C. Tubercle arrangement on the jaws of Barilius microlepis. Scales 5 mm.

The tubercles are conical but are inclined so that on the anterior part of the lower jaw and the
1st infraorbital the cones point caudad; on the 2nd and 3rd infraorbitals they are perpendicular
whereas on the posterior part of the cheek they incline forward. The tubercles are attached to
the skin through thickened pads of tissue and when a tubercle is removed there remains a circular
pit bearing lamellar radii. In specimens where the tubercles have been shed in life, there remains a
shallow, saucer-shaped depression.

Zacco

The tubercles are of the same conical form as in Opsariichthys and are present on the snout, 1st
infraorbital, nasal, cheek below the infraorbitals, lower jaw, preoperculum and operculum (Fig.
4 IB). They are also well developed on the anal fin rays and on the scales both below and above
the lateral line. Banarescu (1968) described the tubercle distribution in the various species of
this genus.

Luciosoma

No tubercles have been found in any species examined. The skin covering the ethmoid region is
deeply grooved by two or three S-shaped channels which extend from the nasal opening to the
midline (see p. 175).

Barilius

Tubercles are present in many species of Barilius of all three groups. In Groups A and B they are
small and in many species appear to be present only on the head. In Barilius gatensis the tubercles

BARILIINE CYPRINID FISHES 173

of the lower jaw extend onto its medial surface (Fig. 42). In a specimen of B. ansorgii (Group B)
tubercles are distributed over most of the dorsal and lateral parts of the body. In ripe males of
Barilius microlepis (Group C) of Lake Malawi, the area covering the tip of the upper jaw, the
infraorbitals and the entire lower jaw are covered with minute tubercles giving the jaws a granular
appearance (Fig. 4 1C). In these specimens tubercles are also present on the anal fin rays. In
other species of this group (B. ubangensis and B. zambesensis) the tubercles are well developed and
in their pattern resemble those of Opsariichthys.

Fig. 42 Barilius gatensis. Tubercle arrangement on the left lower jaw.

Engraulicypris

No tubercles have been found in Engraulicypris sardella.

Pectoral and pelvic lobes

Opsariichthys

An axial pectoral lobe is present in the form of a minute fleshy appendage. There is also a fleshy
triangular lobe lying medial to each pelvic fin.

Zacco

There is some intraspecific variability in the form of the pectoral and pelvic lobes. In Z. platypus
the pectoral lobe is very small and the pelvic lobe is in the form of a scale with a fleshy border.
In Z. temmincki the pectoral lobe is well developed and the pelvic scale has a fleshy border. In
Z. macrophthalmus the pectoral lobe is barely developed and the pelvic one is an elongate scale
without a fleshy margin. In Z. pachycephalus the pectoral lobe is bipartite and the pelvic lobe a
thick fleshy wedge (see p. 189).

Luciosoma

The pectoral axial lobe is an elongate, ventrally curved structure originating from the post-
cleithrum (see p. 167). An elongate pelvic scale is also present.

Barilius

The pectoral and pelvic lobes are developed as fleshy structures only in Group C. This is partic-
ularly so in B. ubangensis and B. microlepis. In the latter species the pectoral lobe in males is

174 G. J. HOWES

expanded and in a specimen of 430 mm SL the lobe is a complexly divided structure (Fig. 43);
in females it is reduced and scale-like.

In Barilius Group A a pectoral scale with a wide fleshy border is developed, and in Group B
the scale is elongate without such a border.

AL

Fig. 43 Barilius microlepis. Axial pectoral fin lobe of a specimen 430 mm SL.

Engraulicypris

Pectoral lobes are absent, but an elongate pelvic scale is present.

Anal fin

In Opsariichthys and Zacco the 3rd, 4th and 5th anal fin rays in both sexes are extended, partic-
ularly so in males. The tips of the rays are expanded and separate into 9-10 branches. The
posterior ventral surface of the caudal peduncle in both genera is somewhat fleshy and partic-
ularly in Zacco becomes ventrally extended into a keel. This keel forms an uninterrupted margin to
the ventral aspect of the body when the long anal fin is folded and presumably aids in streamlining
(Figs 44A & B).

The anal fin is not extended in males of Luciosoma species, although in some individuals a
filament is developed from the first ray.

In Barilius Group C there is an extension of the first four to six rays in males. This is particularly
evident in B. ubangensis which also displays a colour pattern of the dorsal fin remarkably similar
to that of Zacco. The anal fin is not extended in other groups of Barilius.

Comments on tubercles, axial lobes and fins

The tubercles, axial lobes and anal fin are considered together because their development appears
to be correlated. A series of well-developed tubercles on the head, body and anal fins is associated
with marked secondary sexual dimorphism exhibited by the attenuation of the anal fin in the
males, by fleshy pectoral axial and pelvic lobes and distinctively marked dorsal and anal fins in
both sexes. Such a combination of features is shared by Opsariichthys, Zacco and Barilius Group C.
It seems likely that these characters are associated in a reproductive context. It has been
assumed that tubercles play some part in reproductive behaviour, possibly functioning as
stimulatory tactile organs (see Wiley & Collette, 1970). In this regard Nakamura (1969) states that
in Zacco the male makes use of the tubercles to drive away intruders and rival males from its
territory. It must be said that it is not clear from the English translation of this passage whether
Nakamura has actually observed this behaviour or conjectures it. Okada (1960) notes that in
Zacco the male ' . . . moves quickly against rocks' cleaning areas for the deposition of the eggs.
He goes on to say ' ... In this connection, it is interesting to note that the pearl organs of Japanese
Cyprinoid fishes develop mostly about the tip of the snout and on the side of the face'. Presumably
the connection implied, but not stated, is that the tubercles in some way either aid in cleaning the

BARILIINE CYPRINID FISHES

175

10mm

B

Fig. 44 Anal fins of A, Opsariichthys uncirostris bidens (BMNH 1901.3.6 : 6) and B. Zacco platypus

(BMNH 1969.4.15 : 29-31).

rock surfaces or act as a buffer. But then Okada proposes that ' . . . The tubercles probably serve
as contact organs, holding the mating pairs entangled by increasing the roughness of the surfaces
of the body of the male which come into contact with the body of the female'. I cannot see how
the second statement is linked with the first, and there are no published first-hand observations
to justify it.

In contrast to the idea that tubercles act as contact organs, Reid (unpublished thesis and
pers. comm.) has hypothesized that at least in Labeo their presence is more likely to serve some
hydrodynamic function but one nevertheless correlated with reproductive activity in as much as
these fishes move upstream to spawn, thus encountering a particular kind of water turbulence.
This is likely to be the case in Opsariichthys, Zacco and some Barilius. The orientation of the
tubercles on the head of Opsariichthys and Zacco (see above, p. 172) suggest a hydro-mechanical
function. A similar function may be attributed to the transverse grooves on the snout ofLuciosoma

176 G. J. HOWES

(p. 172) which possibly serve to counteract turbulence in that region by improving streamlining
(Reid, pers. comm.).

A fleshy axial pectoral fin lobe is not confined only to the genera considered here, but is also
present in the American cyprinid, Platygobio, some cultrines, alburnines and some species of
Rasbora. In Platygobio it is a small structure not attached to the cleithrum. In the Rasbora species
it is large, usually elongate and attached to the postcleithrum as in Luciosoma. In the cultrines and
alburnines the lobe can be small or elongate, ventrally curved and is attached to the cleithral
lamina. No data are available on the possible use of this organ. It is possibly associated with
reproductive behaviour and could act as (i) a visual stimulant when the male extends his pectoral
fins, and/or (ii) a tactile organ if the female curves her body around that of the male and/or
(iii) as an area in which pheromone-producing cells are concentrated.

Another character, possibly correlated with the development of the anal fin and axial lobes,
is the presence of 'pectoral pads' in some species of Barilius. These pads were first mentioned by
Hora (1921) in Barilius bendelisis and B. barila. He described them in B. bendelisis as: ' ... The
chest is flattened and the scales in this region are poorly developed. There are characteristic
muscular pads in front of the bases of the pectorals'. Dr Tyson Roberts reports (pers. comm.)
that these pads are only evident in large, ripe males of this species, and that the epidermal surface
of the pad is ridged transversely. A similar, but smaller, swollen area is found in some specimens
of Barilius zambesensis and B. ubangensis. It seems likely that these pads may be used to steady
the fish in fast flowing water. Perhaps significant in this respect is the comment on Barilius
bendelisis made by Hora & Mukerji (1936): ' . . . The pectoral fins are fan-like and horizontal
and are never folded even when the fish moves from place to place. Some of the rays are especially
strengthened and it seems likely that they are used either for digging in sand or for holding on to
rocks in somewhat rapid waters.'

In view of the unknown distribution of this character amongst the cyprinids I am unable to
ascribe to it any polarity, but I would suggest that it represents a derived character for a group
of Indian Barilius. If the development of axial lobes is to be regarded as an apomorph condition
(Howes, 1979), then Opsariichthys, Zacco, Luciosoma, Barilius Group C and some species of
Rasbora would be linked by this character. Indeed, the development of the axial lobes and anal
fin, together with the distinctive colour pattern, appear to be the only characters Zacco shares with
the other bariliine and opsariichthine genera.

According to Nakamura (1969) it appears that the extended anal fin is used to mix the eggs and
milt, which would suggest that this is a derived morphological character evolved in concert with
a specialized reproductive behavioural pattern. There is no evidence at present to suggest that the
marked sexual dimorphism apparent in the development of the anal fin is a plesiomorph character.
Thus, if this character or character combination is viewed as synapomorphic, then Zacco must be
united with other opsariichthine and bariliine genera, an action which would disregard other,
perhaps more significant, osteological characters. Present indications are that Zacco is not closely
related to the bariliine group (see p. 180) and so I consider this set of derived characters to be the
result of convergent evolution in reproductive strategy.

Barbels

Opsariichthys

Barbels are present in one species, O. barbatus. This species was formerly assigned to the genus
Zacco; reasons for placing it in Opsariichthys are discussed on p. 189. Only a single pair of
minute posterior maxillary barbels is present.

Luciosoma

Two pairs of maxillary barbels are developed in all but one species. The rostral barbels are long
and lie along the dorso-lateral border of the premaxilla and under the ventral edges of the 1st and

BARILIINE CYPRINID FISHES 177

2nd infraorbitals. Each barbel stems from a thick pad of tissue which is connected by ligamentous
strands to the edge of the 1st infraorbital. The tissue is embedded in the lateral groove of the
maxilla and is intimately connected with the ligamentous bands which join the maxilla to the
palatine. The insertion point of the adductor mandibulae A. l muscle lies beneath the tissue pad
but there is no direct connection and no muscle fibres run into the barbel. The barbel is innervated
by a branch of the maxillary ramus of the trigemino-facialis nerve. The posterior maxillary
barbel also stems from a thick fibrous pad, and is innervated by a branch of the nerve supplying
the anterior barbel. The proximal head of the barbel is swollen and surrounded by bands of
elastic-like tissue which extend along the outer border of the maxilla.

Barilius

In species of Group A barbels are present in one or two pairs, or are entirely absent. In Group B
they are found in only one species, B. guttatus. In this species there is a minute barbel on the
ventral border of the maxilla just below the anterior border of the 1st infraorbital and near the
distal tip of the maxilla. Barbels are absent in species of Group C.

Engraulicypris

Barbels are absent.

Comments

It is perhaps significant that amongst the bariliines (see p. 180), reduction and absence of barbels
is correlated with increased piscivorous specialization (cf. Opsariichthys uncirostris and O.
barbatus; long-jawed Barilius Group B & C with short-jawed omnivorous Barilius Group A).
Thus, the inference is that loss or reduction of barbels is a derived condition. Although barbels
are widespread throughout the Cyprinoidei (again suggesting the plesiomorph condition) they
are not a characteristic feature of the Cyprinidae. Contrary to Roberts' (1973) statement that
'Cyprinidae have one or two pairs of simple maxillary barbels', most of the species comprising
the large assemblages of cultrine, cheline, alburnine, hemicultrine, aspinine and leuciscine groups
do not have barbels. It could be argued that barbels are a derived feature. If this is so, then of those
taxa presently under discussion, all the Indian and South-East Asian bariliines possessing barbels
would have to be more closely related to barbine, or other carp groups, than to the African and
Amur bariliines without barbels. From other anatomical evidence, this is clearly not the case.
Where barbels are present in non-barbine or carp groups they appear, superficially, to be of a
different morphological type. Gilbert & Bailey (1972) discussed the use of barbels in classifying
some American cyprinids and concluded that this character was of little value as ' . . . their
diversity in position and structure points to repeated independent development and loss'. These
authors discovered in American cyprinids at least six structural types, noting them as ' ... perhaps
indicative of as many independent evolutionary lines'. The whole question of barbel morphology
and homology in the cyprinids, and indeed in the ostariophysans, is obviously one to be investig-
ated.

Roberts (1973) states that in cyprinids the movement of barbels is passive. This is not always
so, for from personal observations on aquarium specimens of Luciosoma and Leptobarbus I can
report that there is a definite movement of the posterior barbels which is independent of any
jaw movement.

The brain

The possibility of using the gross morphology of the brain as a phylogenetic character was
suggested by observations made by Dr R. Vari on various characoids and on Opsariichthys
(pers. comm.).

The outstanding feature of the brain in Opsariichthys is the size and shape of the telencephalic
lobes which are elongate, sausage-like structures separated from the mesencephalon by long

178 G. J. HOWES

posterior trabeculae. The usual situation in cyprinids, and it seems the majority of teleosts, is for
the telencephalic lobes to be globular or ovate and closely apposed to the mesencephalon.

The olfactory bulb in Opsariichthys, and other cyprinids, is situated in the lateral ethmoid and
thereby far removed from the telencephalon, the olfactory tracts being extended. Such extension
of the olfactory tracts is also known in some characoids, in the siluroids, mormyroids and gadoids
(Evans, 1940). It is known that in some cyprinids (e.g. Carassius) there is an ontogenetic change
in the olfactory bulbs and tracts (Schnitzlein, 1964); in young specimens the olfactory bulbs are
adjacent to the telencephalic lobes but with growth the olfactory tracts extend forward. As there
is an actual lengthening of the tracts it appears that there is a disproportionate growth of parts
of the cranium.

In small specimens of Opsariichthys the telencephalic hemispheres are large, occupy most of
the orbitosphenoidal cavity and are well separated from the mesencephalon (Fig. 45A). The
olfactory tracts are short and the bulbs located in the lateral ethmoids. In large specimens the
telencephalic lobes, whilst still well separated from the mesencephalon, are more closely apposed
and lie in the pterosphenoid; the olfactory tracts are long.

Thus, the difference between Opsariichthys and Carassius lies in the fact that although in
Opsariichthys the olfactory lobes lie close to the telencephalon in early stages of development, it is
the retreat of the telencephalon which gives rise to extension of the olfactory tracts whereas in
Carassius (? and all other cyprinids) the telencephalon remains adjacent to the mesencephalon
throughout growth. In a specimen of Opsariichthys uncirostris amurensis 82 mm SL the left
telencephalic lobe is elongate and lies in advance of the right lobe, being connected to the mesen-
cephalon by a long trabecula. In this and all other specimens examined the cerebellum is small.

Zacco and Barilius

The morphology of the brain in Zacco and Barilius Groups A & B is little different from that
described in Carassius and Cyprinus by Evans (1940) and Schnitzlein (1964), i.e. the telencephalic
lobes are rounded to ovate and separated from the mesencephalon by a deep recess (Fig. 45B).
This condition appears to be invariable in the species examined. The olfactory tracts are long,
slender and divergent anteriorly; the cerebellum is small.

In Barilius Group C there is a similarity to Opsariichthys. In Barilius zambesensis (Fig. 45D)
and B. ubangensis the telencephalic lobes are elongate and separated from the mesencephalon by
short posterior trabeculae. The same situation is found in B. microcephalus but in this species the
olfactory tracts are thickened and the cerebellum large (Fig. 45C). In all species of this group the
olfactory tracts lie close together and do not diverge anteriorly, the olfactory bulbs lie close
together within the lateral ethmoids.

Engraulicypris is like Barilius microcephalus in the shape and size of the telencephalic lobes
and the cerebellum (Fig. 45E). It should be noted here that the reference made by Evans (1940)
to the particular brain morphotype of Engraulicypris refers only to the species argenteus assigned
herein to Rastrineobola (see p. 195).

In Leptocypris the telencephalic lobes are also elongate but separated from the mesencephalon
by a fissure; the olfactory bulbs are very large and the tracts short and thick. The cerebellum is
larger than in any of the other genera examined.

The telencephalon of Luciosoma is small in comparison with the other genera, and the overall
brain morphology resembles that of Zacco.

Comments

In functional terms the development of the telencephalon in Opsariichthys is unexplained. It is
possible that the particular development of this part of the brain is in some way associated with
integrating patterns of breeding behaviour of increased complexity (see Segaar, 1961 ; Nieuwen-
huys, 1962). The usefulness of forebrain morphology as a phylogenetic character cannot be
assessed, although from the apparently conservative morphological structure of the brain in other
cyprinids, the condition in Opsariichthys possibly may be derived. The elongate telencephalic
lobes, widely separated from the mesencephalon, is a feature shared with Barilius Group C species.

OL OT TEL

BARILIINE CYPRINID FISHES

MEG

179

B

D

Fig. 45 Diagrams of the brain (dorsal views) in A. Opsariichthys. B. Zacco platypus, C. Barilius
microcephalus. D. Barilius zambesensis. E. Engraulicypris sardella. F. Luciosoma setigerum.
The arrowed lines indicate the suture between the orbito- and pterosphenoid.

Summary of character states

From the foregoing descriptions and analysis, the following characters are considered to be
apomorph.

Narrow, deeply notched supraethmoid; shallow mesethmoid; small preethmoid; tubular olfactory
foramen in lateral ethmoid; medially or distally expanded kinethmoid; ventrally thickened
vomer; pterosphenoid contacting parasphenoid ; pterosphenoid lateral fossa; enlarged carotid
foramen; ventrally open posterior myodome; presence of a. jugular channel; enlarged sphenotic
process; reduced dilatator fossa; posteriorly extended pterotic; posttemporal fossa entering
subtemporal fossa; narrow frontal with concave border; elongate nasal with more than 2 dorsal
pores; elongate parietal; elongate supraorbital; narrow infraorbitals; anterior maxillary tips
extended; anterior maxillary medial process ventrally directed; dentary notched anteriorly;
hyomandibular condyles separated; hyomandibular lateral flange; metapterygoid process; small
symplectic; operculum with elongate dorsal border; cleithrum with short upper part and expanded
posterior lamina; postcleithrum reduced; coracoids in medial contact; 2nd and 3rd centra fused;
axial fin-lobes; reduction of barbels; elongate telencephalon.

Relationships of the bariliine genera

Authors who have considered the systematic position and/or anatomy of Opsariichthys have
assumed it to bear a close relationship to Zacco (see, for example, Regan, 191 1 ; Takahasi, 1925;
Ramaswami, 1955, Greenwood et al., 1966; Banarescu, 1968). This supposed relationship has

180 G. J. HOWES

been based upon the close similarity in external morphology of the two taxa which, in fact, re-
semble one another mainly in plesiomorph characters. None of the apomorphies identified here
show Opsariichthys and Zacco to be sister taxa.

Characters identified as apomorphic in Opsariichthys also occur in Barilius, Leptocypris,
Engraulicypris, Luciosoma and in some species of Rasbora. The synapomorphies which serve to
identify these taxa as the BARILIINE GROUP* are:

1. Subtemporal fossa connected with the posttemporal fossa (in its most derived state the
posttemporal 'fossa' is formed into a foramen, p. 146).

2. Posterior myodome open ventrally, with an external origin for part of the eye musculature
(p. 153).

3. Anterior trigemino-facialis foramen situated entirely within the lateral face of the prootic
(p. 153).

4. Frontal fossa (p. 155).

5. Lateral pterosphenoid fossa (p. 140).

6. Hyomandibula with widely separated condyles and lateral flange (p. 164).

7. Metapterygoid with dorsal posterior process (p. 165).

8. Pectoral and pelvic axial lobes or modified scales (p. 176).

Before the relationships of the bariliine genera can be understood, it is first necessary to discuss
the classification of the genera Barilius, Leptocypris, Engraulicypris and Rasbora (in part).

Classification of Barilius

As previously recognized, Barilius has included both Indian and African species and although
subgroups have been recognized (Giinther, 1968; Day, 1877, 1889, see below) the 'monophyly'
of the taxon has remained unquestioned. On the basis of characters here identified as apomorphic,
Barilius is seen as polyphyletic. The following classification is not entirely satisfactory as not all
the Indian species, and virtually none of those from Thailand and Burma, have been examined.
Hopefully, however, the interim scheme presented will permit a more precise systematic evaluation
of the various groups until such time as a complete taxonomic revision is attempted.

In the descriptive section of this paper three major groups (A, B and C) of Barilius are recog-
nized. However, within Group A two subgroups can be identified, viz. :

1 . Barilius. Termed in the descriptive section Group A and represented by the type species of the
genus Barilius barila (Hamilton, 1822). It includes only those species inhabiting the Indian sub-
continent, Thailand and possibly Burma. The genus is characterized by the apomorph features of
a deep, rostrally curved ethmoid region, long nasals, long parietals, reduced lateral ethmoids
and a deep frontal fossa.

Day (1877) recognized three groups of Indian Barilius characterized by the absence or presence
and number of barbels (see p. 189). I recognize two groups, (i) exemplified by Barilius barila in
which the jaws are long, the hyomandibular condyles well separated, scales with many radii,
body shallow, barbels in two pairs, tubercles small and poorly developed; (ii) exemplified by
Barilius gatensis in which the jaws are short, hyomandibular condyles barely separated, scales
with few radii, body deep, a single pair of barbels, or barbels lacking, and tubercles large and
well developed. As yet I have examined too few species to be certain if these groups represent part
of a morphocline. The species contained in Barilius are listed on p. 189.

2. Opsaridium. Termed in the descriptive section Group C and represented by the type species of
the genus Opsaridium zambesense (Peters, 1852). It includes only African species. The genus is
characterized by the apomorph features of: an expanded kinethmoid; reduced premaxillary

* The use of the name bariliine for this assemblage follows Regan (1922) who first recognized the subfamily
Bariliinae (see p. 185).

BARILIINE CYPRINID FISHES 181

ascending process; anteriorly extended maxillary tips; deep frontal fossa; pterosphenoid cavity;
extensive lateral temporal foramen; broad mesocoracoid and well-developed pectoral axial lobes.
The secondary sexual dimorphic characters of an extended anal fin in males, well-developed
axial lobes and distinctively marked dorsal fin (see p. 191) serve to distinguish this genus from
other related taxa. The species contained in Opsaridium are listed on p. 191.

3. Raiamas. Termed in the descriptive section as Group B and represented by the type species of
the genus Raiamas bola (Hamilton, 1822). It includes both African, Indian, Burmese and
Malaysian species. The genus is characterized by the apomorph features of: a greatly expanded
kinethmoid; long shallow jaws; reduced premaxillary ascending process; extensive pterosphenoid-
parasphenoid contact; small dilatator fossa; apophyseal platform formed partly from the
prootics; extensive posttemporal foramen; long post-parietal platform; broad mesocoracoid;
posteriorly extended cleithrum; elongate pectoral axial scale and extensive development of
adductor mandibulae A2 muscle which laterally covers the dilatator operculi and levator arcus
palatini muscles. Within this genus a transformation series represented by R. loati-R. salmolucius-
R. bola is apparent in cranial elongation.

Previously (Howes, 1978) Raiamas bola was used as an example of bariliine anatomy when
making comparisons with other cyprinids. It is now obvious that this species is not 'typical' of
Barilius but represents a highly derived morphotype. Nevertheless, the conclusions reached in
that study concerning the possible phylogenetic history of certain characters in R. bola remain
valid within the hypothesis of relationships advanced here.

The inclusion of two Asian species in an otherwise entirely African genus requires some
explanation. Four characters are present in Raiamas bola and R. guttatus which are lacking in their
African congeners. These are: a closed posterior myodome; a much broader and S-shaped
kinethmoid; extensive contact of the pterosphenoid and parasphenoid and, in R. guttatus, two
pairs of small barbels. Possession of these characters in no way excludes bola and guttatus from
inclusion in Raiamas. These two species are the most extreme members of their lineage in terms
of cranial and jaw elongation, and are approached only by the African species R. salmolucius and
R. longirostris (see p. 193). As mentioned above, a transformation series involving the lengthening
of the pterosphenoid-parasphenoid suture, broadening of the kinethmoid and the secondary
closure of the posterior myodome can be detected amongst the African species of Raiamas and
the highly derived state of these characters in R. bola and R. guttatus is simply a continuation of
this transformation sequence. It could, of course, be argued that the two Asian Raiamas species
are simply the result of parallel evolution from an ancestral group of Indian bariliines. Such an
hypothesis can be tested by identifying synapomorphies between Raiamas and Barilius and
perhaps by showing a similar transformation sequence amongst the Indian bariliines. So far, I
have found no such characters. The species contained in Raiamas are listed on p. 193.

The only species of Barilius occurring in the Middle East, B. mesopotamicus Berg, 1932, is, in
fact, not a member of any of the genera listed above but should, on the basis of synapomorph
characters, be included in the Leucaspius generic complex (see p. 190).

Classification of Leptocypris

Leptocypris has previously been recognized as monotypic; Leptocypris brevirostris Blgr, 1919
and L. clupeoides Pell., 1922 were synonymized with the type species L. modestus Blgr, 1900 by
Poll & Gosse, 1963 and Gosse, 1966, respectively. The genus is characterized by its shallow lower
jaw, the dentaries being joined at the symphysis by a ligament; inferior position of the mouth;
pointed snout; deep, narrow and strongly curved ethmoid region; truncated lateral ethmoid;
long, ventrally directed medial maxillary process; elongate parietal; elongate nasal opening;
absence of, or reduced intermandibularis muscle; absence of, or reduced number of gill-rakers
(3-6), and much elongated pectoral and pelvic axial scales.

Leptocypris possesses none of those apomorph characters which would suggest a close relation-
ship with Opsaridium or Raiamas (i.e. extensive posttemporal fossa, broad kinethmoid and
gutter-like metapterygoid process), but it does share some of those characters listed above with

182 G. J. HOWES

Engraulicypris (see below). Additionally, it shares the features of an elongate supraorbital, and
1st infraorbital covering the anterior part of the 2nd infraorbital.

Two other species, formerly included in Barilius, B. niloticus (Joannis, 1835) and B. weynsii
Blgr, 1899, share with Leptocypris modestus the apomorph characters listed above as character-
izing the genus and are thus transferred here to Leptocypris. The characters distinguishing the
three species are tabulated on p. 190.

Classification of Engraulicypris

As previously recognized, Engraulicypris has included several species of small-sized African
cyprinids. However, the type species of the genus, E. sardella (Giinther, 1868), possesses many
apomorph characters, none of which are shared with any of the species presently included in
Engraulicypris. For this reason the genus is restricted to the type species.

Engraulicypris is included in the bariliine group because it has characters 1, 2, 3, 4, 5 of those
listed on p. 180 as typifying the group. The characteristic features of Engraulicypris are the
extensively modified ethmoid region, trie highly protrusile upper jaw and the long, numerous
gill-rakers (total number 43-45).

The relationships of Engraulicypris with other members of the bariliine group are somewhat
difficult to determine owing to the highly derived nature of many of its cranial features (auta-
pomorphies). Other features, involving the pectoral girdle and vertebral column, are plesio-
morphic and so offer no clues to relationship. On the basis of its possessing a well-developed
pterosphenoid fossa and the trigemino-facialis foramen being situated in the lateral face of the
prootic, Engraulicypris is included in a subgroup of the bariliine assemblage (see below, p. 184).
Engraulicypris shares with Leptocypris elongate parietals; truncated lateral ethmoids; downwardly
directed medial maxillary processes; an extensive 1st infraorbital bone overlapping the anterior
part of the 2nd infraorbital; elongate supraorbital; ligamentous connection of the dentaries and
an elongate pelvic axial scale. In both Engraulicypris and Leptocypris the subtemporal fossa is
narrowly conical with only a small foramen leading into the posttemporal fossa. The small
opening is suggestive of a 'precursory' condition leading to the intramural channel and extensive
foramen found in other bariliine genera.

Engraulicypris and Leptocypris are here regarded as sister genera forming a branch of the
bariliine assemblage (see below). Which of the two can be considered the more divergently
specialized is debatable. Certainly Leptocypris is trophically more generalized. Matthes (1963)
made the point that Leptocypris was probably derived from Barilius-like ancestors and although
more generalized than Barilius did not necessarily represent a primitive form. Matthes' concept
of Barilius differs from that presented here and the 'Barilius'' species he was comparing with
Leptocypris are now considered as members of the relatively derived groups Opsaridium and
Raiamas. I would agree with Matthes that Leptocypris is not 'primitive' (plesiomorph) and
further point out that there is a morphocline in this genus toward a carnivorous type represented
by L. weynsii. The suggestion made by Goren et al. (1973) that Engraulicypris is derived from
Barilius cannot be accepted as these authors too were considering polyphyletic assemblages.

Classification of Rasbora (in part)

Throughout the descriptive section mention has been made of supposedly derived characters shared
between Luciosoma and some species of Rasbora. These characters are: supraethmoid with a
broad bowl-shaped depression; an expanded kinethmoid; a deeply concave vomer; a tubular
olfactory foramen in the lateral ethmoid; an inflated orbital portion of the lateral ethmoid; a
well-developed symphysial knob on the dentary followed by a deep notch; an elongate lateral
occipital fenestra; a long operculum; elongate ceratobranchials, and a pectoral axial lobe
attached to the postcleithrum. As well as these features, the Rasbora species possess characters
1, 4 and 5 of those identifying all bariliine genera. Within this 'Rasbora' complex, two closely
related groups can be identified. One group is referred to a new genus, Parluciosoma (see below),
and the other to Megarasbora Gunther, 1868.

BARILIINE CYPRINID FISHES 183

PARLUC1OSOMA gen. nov.
Rasbora (in part) Hamilton, 1822
TYPE SPECIES: Leuciscus argyrotaenia Bleeker, 1850.

In his revision of Rasbora, Brittan (1954) recognized an assemblage of species which he named
'The argyrotaenia complex'. He distinguished this from other species complexes within the
genus by their medium to large size and by ' . . . conservative structure and color-pattern'.
Brittan (1954) did not explain the precise nature of their 'conservatism' and whilst it is true that
nearly all members of the group display a distinctive colour pattern comprising a black lateral
stripe and dark edges to the caudal fin (in preserved specimens), this character is of little value in
formulating relationships. More significant in this regard are the osteological characters which
this group of ' Rasbora' shares with Luciosoma: ethmoid region broad, the supraethmoid with a
bowl-shaped depression, the kinethmoid compressed and expanded distally; the dentary with a
well-developed symphysial process; the maxilla with an extended anterior tip narrowly separated
from its fellow medially; the premaxilla with a short anterior ascending process; the frontal with
a ventral fossa (and in one species, P. dusonensis, a lateral groove as in Luciosoma; see p. 155);
a bony tube surrounding the olfactory foramen in the lateral ethmoid; elongate lateral occipital
fenestra; the operculum broad and somewhat attenuated, and a well-developed pectoral axial lobe
attached to the postcleithrum. Barbels are lacking.

Parluciosoma differs from Luciosoma in having a shallower supraethmoid depression, a narrow
contact between the parasphenoid and pterosphenoid, shorter lateral processes of the 2nd
vertebra and a long, medially directed postcleithrum. The jaws also tend to be shorter and
deeper with a stronger symphysial notch on the dentary. The species contained in Parluciosoma
are listed on p. 194, and the relationships of Rasbora (sensu stricto) are discussed on p. 185.

MEGARASBORA Gunther, 1868
TYPE SPECIES: Cyprinus elanga Hamilton-Buchanan, 1822.

Megarasbora was used by Gunther (1868) in a subgeneric sense to include Rasbora elanga.
Brittan (1954) noted that R. elanga 'is discontinuous with the remainder of the species in the
genus'. He drew attention to the short pair of rostral barbels and the 'peculiar' striae and ridges
of the scales as well as to the high number of lateral line scales (40-44) compared with other
Rasbora species. However, when compared with Luciosoma these characters are in concordance.

Megarasbora is characterized by a broad ethmoid region, the supraethmoid with a semi-
circular depression; the kinethmoid lamellar; the premaxilla with a long, slender medial anterior
process; the sphenotic with a broad lateral process which contains a fossa in its anterior face; the
dilatator fossa confined to the sphenotic; the dentary with a deeply concave notch posterior to
the symphysial process; the lateral process of the 2nd vertebra containing a ventral fossa; the
postcleithrum short and supporting a poorly developed axial lobe; pharyngeal teeth with strongly
recurved tips (dental formula: 4.3.2); total number of gill-rakers on 1st gill arch, 8-9.

Megarasbora differs from Luciosoma and Parluciosoma in the nature of the characters enum-
erated above and in possessing a pair of short maxillary barbels, more obliquely aligned jaws and
almost complete lack of body and fin markings. The genus is monotypic; see p. 195 for distribu-
tion.

Interrelationships and classification of the bariliine genera

The synapomorph characters identified in the foregoing anatomical descriptions (listed on p. 180)
serve to relate Barilius, Leptocypris, Engraulicypris, Opsariichthys, Opsaridium, Raiamas, Mega-
rasbora, Parluciosoma and Luciosoma as a monophyletic unit termed the bariliine group (see
p. 180). The taxa comprising this group can be classified into two assemblages related as sister
groups :

1. The bariliine assemblage (Barilius, Leptocypris, Engraulicypris, Opsariichthys, Opsaridium and
Raiamas) defined by those characters listed on p. 180.

184 G. J. HOWES

2. The luciosomine assemblage (Luciosoma, Parluciosoma and Megarasbora) which lacks a
ventrally open posterior myodome. In addition to the derived characters shared with the
bariliines (nos 1, 3, 4, 5 and 8 of those listed on p. 180), the luciosomines possess the auta-
pomorphic features of: supraethmoid with bowl-shaped depression; compressed and medially
expanded kinethmoid; lateral ethmoid inflated posteriorly and containing a tubular olfactory
foramen; elongate lateral occipital fenestra and a pectoral axial fin lobe attached to the
postcleithrum.

Within these two assemblages various sub-groups or lineages can be recognized as follows:

Bariliine assemblage

Barilius represents the lineage which forms the sister group to the other members of the assem-
blage. This taxon possesses all the characters listed on p. 180 as defining the group apart from 3
and 7. Barilius is distinguished from other included taxa by the apomorph characters of: a deep,
strongly curved and narrow ethmoid region; elongate nasals; elongate parietals and truncated
lateral ethmoids.

Leptocypris, Engraulicypris, Opsariichthys, Opsaridium and Raiamas represent the apomorph
sister group to Barilius. These genera are linked by the synapomorphies : pterosphenoid fossa and
anterior trigemino-facialis foramen situated in the lateral face of the prootic.

Leptocypris and Engraulicypris together form the sister group to the other three genera and
possess the synapomorphies: ventrally directed medial maxillary process; elongate supraorbital ;
1st infraorbital partially overlapping the 2nd infraorbital and elongate pelvic scale.

Opsaridium, Opsariichthys and Raiamas possess the synapomorphies: anteriorly extended
maxilla; bulbous vomer; presence of quadrate flange and lateral temporal foramen.

Opsaridium and Raiamas form the apomorph sister group to Opsariichthys linked by the
synapomorphies: dorsally expanded kinethmoid; broad mesocoracoid and extensive posttemporal
foramen.

Luciosomine assemblage

Parluciosoma and Luciosoma form the apomorph sister group to Megarasbora. Apomorphies
characterizing and linking these genera have already been discussed (p. 183).

The relationships of the genera forming the bariliine group are shown as a cladogram in Fig. 46.

When dealing previously with monophyletic groups of cyprinids (Howes, 1978, 1979) I refrained
from assigning them to formal taxonomic categories and presented them as 'informal groups'.
This same policy is followed here and for the same reasons, i.e. lack of congruency at higher
level classification. The nature of the cladistic classification tends to inflate hierarchical levels.
Thus, when the interrelationships of the bariliine group with those of other monophyletic
assemblages are more completely understood it is likely that some of these groups, including the
bariliines, will have to be assigned subfamily status.

Interrelationships of the bariliine group

The search for the sister group of the bariliines has so far proved inconclusive. The reason is that
the majority of cyprinid subfamilies, and even genera, currently recognised are non-monophyletic
in the cladistic sense and it has not been possible to postulate relationships on the basis of
synapomorphic characters. In previous studies I identified some monophyletic lineages (Howes,
1978, 1979) and in dealing with one of these groups, the chelines, I stated that there was no
evidence to suggest close relationship with the bariliines (Howes, 1979). Since then, however,
this hypothesis has been weakened by the identification of presumed synapomorphies associated
with the jaws and otic region of the cranium (work in preparation). Regan (1911) suggested a
relationship between Barilius, Danio and Opsariichthys considering that these genera were related
to Aspius and Leuciscus, and stated ' . . . Hypophthalmichthys is nearly related to the Barilius

BARILIINE CYPRINID FISHES 185

group'. Later, Regan (1922) considered the geographical distribution of the Cyprinidae and
recognized the Bariliinae as the most primitive members of the family. In this work Regan noted
that there were a few Palaearctic genera in the Bariliinae, but he did not name them. Furthermore,
he was quite explicit with regard to their relationships ' . . . the Leuciscine Cyprinidae is derived
from the more primitive and more southern Bariliinae'.

BARILIUS

LEPTOCYPRIS
ENGRAULICYPRIS

OPSARIICHTHYS
OPSARIDIUM
RAIAMAS

PARLUCIOSOMA
LUCIOSOMA

MEGARASBORA

Fig. 46 Cladogram illustrating the relationships of the bariliine genera.

The evidence presented here falsifies Regan's claim for a close association between the leucis-
cines and bariliines. Neither have any synapomorphies been identified which relate Hypophthal-
michthys with any bariliine taxon. As remarked previously (Howes, 1978) the Leuciscinae are a
polyphyletic assemblage including at least three major groups. Current studies indicate that
these three groups can be related at various levels to the aspinines, cultrines and hemicultrines
(including the Xenocyprininae) sensu Howes (1978, 1979). In turn, this higher-level group appears
as the likely sister group to that comprising the bariliines, chelines, rasborines and danioines.
The interrelationships of the various monophyletic subgroups have yet to be deduced.

Matthes (1963) considered that the ' ... "Barbus and Barilius" types share many character-
istics which are indicative of close phylogenetical affinities'. I have found no synapomorph
characters indicating close relationship of the bariliines with any barbine or labeine group.

Gosline (1978) considered Opsariichthys and Zacco to be 'specialized' genera allied with the
'cultrin series of Leuciscinae, but with indications of affinity with the Tribolodon-American
section of the subfamily'. Gosline gives no clue as to what these 'indications of affinity' might be.
The tacit implication of this statement is that the American and Asian 'Leuciscinae' are a mono-
phyletic assemblage. My opinion, based on current investigations of the anatomy of Tribolodon,
is that this genus is indeed related to a group of the American cyprinids (including Ptychocheilus),
but that the sister group to this assemblage is, in all probability, the aspinine group sensu Howes
(1978). There is certainly no indication in terms of shared derived characters of any close relation-
ship between any member of the bariliine group and the 7>/60/06fo/7-American assemblage.

186 G. J. HOWES

Finally, mention must be made of Zacco which in the absence of synapomorphous osteo-
logical characters is excluded from the bariliine group. It must be admitted that as yet I have been
unable to relate Zacco with another monophyletic group. Indications are, however, that derived
characters associated with the ethmoid and otic regions of the cranium are shared with members
of the alburnine assemblage of 'Leuciscinae'.

Biogeography of the bariliines

The bariliine group is represented in Africa, India, Burma, the Amur basin, Korea, Japan (Honshu
Island only) and the south-east Asian archipelago. Both plesiomorph and derived taxa occur
sympatrically in Africa and India. The present distribution and almost complete allopatry of the
derived members of the bariliine (Opsariichthys and Indian Raiamas) and luciosomine (Luciosomd)
assemblages indicates vicariance events occurring after the break-up of Gondwanaland.

The geographical distribution of the bariliines in Africa is as follows:

Opsaridium occurs in the Zambesi drainage and Lake Malawi, extends south to the Limpopo
river and north through Zaire to the Ja (Dja) river, as far west as the Cubango river, and east to
the Rufiji. Opsaridium is the only bariliine genus found in the Zambesi basin and the east-African
drainage systems (see Banister & Clarke, in press, for detailed distribution).

Leptocypris and Raiamas both have similar distribution patterns, ranging through the Nilo-
Sudanian and Zairean provinces. Raiamas extends far to the west in Sierra Leone while its most
southerly distribution is to Lake Tanganyika.

Engraulicypris is confined to Lake Malawi.

The partially sympatric distribution of the bariliine genera in Africa suggest a series of vicariance
events, with the most derived genus, Raiamas, being the most widely distributed. Engraulicypris
is the only bariliine adapted to a planktonophagous niche, this niche in other Rift lakes being
occupied by derivates of a Euro-Asian assemblage of cyprinids (see p. 196).

The pattern of bariliine distribution in India shows that the relatively plesiomorph species
Barilius bendelisis and B. vagra are widely distributed and indeed are the only Barilius species
found in the Indus plain (Mirza, 1974) and Sri Lanka (Silas, 1953). Raiamas extends eastward,
beyond the range of Barilius into Shan and the Malay Peninsula.

Plesiomorph luciosomines (Parluciosoma and Megarasobrd) are also present in India with the
more derived species ofLuciosoma and Parluciosoma extending through the Malay Peninsula and
along the Sunda Island chain.

Opsariichthys occurs through southern China, the Amur basin, Korea and Honshu Island of
Japan. What may prove to be the more plesiomorph species of Opsariichthys inhabit Taiwan and
Hainan (see p. 189).

The affinity of the Indian with the Malaysian ichthyo fauna has been emphasized by Hora (1944)
and Menon (1953, 1955). In Menon's view, Indian ostariophysi had their 'origins' in southern
China and dispersed along the Himalayas and the Burma-Malayan arc. More recently Briggs
(1979) has made a similar assertion that the cyprinoids had their origins in the Oriental Region,
more particularly in South-East Asia. These hypotheses suffer from not having as their base-line
a well-founded theory of phylogenetic relationship. Furthermore, it seems that these ideas depend
upon the ancestral lineages of the present-day ostariophysan groups having been confined to
'evolutionary centres' from which they dispersed (Briggs, 1979). The difficulties inherent in
constructing 'dispersal' hypotheses on the basis of palaeogeography are realized when considering
the varying views concerning the Gondwanic position of the South-East Asian block and its
likely connections with India. Burton's (1970) and Ridd's (1971) proposals are for juxtaposition
of the Malay Peninsula with India, whereas Stauffer (1973) considers the Malayan blocks as once
having African connections.

The system of interrelationships hypothesized here for the bariliines suggests a Gondwanic
(i.e. pre-drift India -(-Africa) distribution (Fig. 47). Banister & Clarke (1977), in accounting
for a Gondwanic distribution of another ostariophysan group, the Clariidae, were puzzled by the
narrower geographical gap separating the Indian and African Barilius compared with that

BARILIINE CYPRINID FISHES

187

separating the Clarias of the two continents. These authors had been misled by the misidentifica-
tion of a middle-Eastern cyprinid as a Barilius (see p. 190). Now it is seen that the distribution
of the bariliines approximates closely to that of the Clariidae and to that of two other cyprinid
groups, the barbines and labeines. However, before too many assertions can be made concerning
the distributional history of the cyprinids, the monophyletic integrity of these two groups must
be tested, as indeed must that of the entire Cyprinoidei.

Fig. 47 Map showing distribution of bariliine genera (dark grey) and dispersal track (broken line).
Blanket coverage is given to Sumatra, Java and Borneo although the actual distribution of bariliines
on these islands is still to be ascertained.

Conclusions

1. Opsariichthys is not a 'primitive' cyprinid - the opinion of previous authors - but a relatively
derived member of its group.

2. Opsariichthys is a member of a monophyletic unit termed the bariliine group comprising two
sister assemblages whose constituent genera are: Luciosoma, Parluciosoma and Megarasbora;
and Barilius, Leptocypris, Engraulicypris, Opsariichthys, Opsaridium and Raiamas.

3. Barilius as previously conceived was a composite genus, it is here restricted to include only
those species inhabiting the Indian subcontinent and parts of South-East Asia. Most of the
species formerly included in Barilius are assigned to Opsaridium, Raiamas and Leptocypris,
others are found to possess apomorph characters which relate them to other cyprinid genera.

4. Engraulicypris as formerly recognized is a composite genus. It is here restricted to include
only the type species, E. sardella. This is a derived member of a lineage within the bariliine
assemblage whose sister taxon is identified as Leptocypris. The other species formerly included
in Engraulicypris are assigned to various non-bariliine genera (see Appendix 2).

5. Rasbora is a composite genus and includes a monophyletic assemblage typified by R. argyro-
taenia here described as a new genus, Parluciosoma. This taxon is the sister group to Luciosoma
and together they form the sister lineage to Megarasbora.

188 G. J. HOWES

6. Transition series of characters correlated with cranial elongation are identified in both the
bariliine and luciosomine clades.

7. The interrelationships of the bariliine group are unresolved although preliminary (unpublished)
evidence suggests that they are the sister group to the chelines (sensu Howes, 1979).

Acknowledgements

I am most grateful to all my colleagues in the Fish Section of the British Museum (Natural
History) for their assistance during this work. In particular Drs Humphry Greenwood and Keith
Banister have aided by means of their stimulating and fruitful discussions and their criticims of
the manuscript. In this regard I must thank Dr Gordon Reid who also permitted me to use some
of his unpublished findings concerning the use of cyprinid tubercles.

My sincere thanks are due to Dr Tyson Roberts for providing me with information on Indian
and African bariliines and to Drs Max Poll and C. Lavett-Smith for allowing me to borrow the
type specimens then under their care of, respectively, Engraulicypris katangae and E. congicus.

Finally, special thanks to Gina Sandford for typing the manuscript and to Jeff Jacobs for
obtaining for me a live (which subsequently became a dead) specimen of Luciosoma trinema.

Appendix 1

Annotated list of bariliine genera and species

The following notes concerning the taxonomy of the bariliine taxa are in no way intended as a
formal revision but serve to indicate the assignment of species to the genera as now defined. In
the course of this study it has been necessary to examine type specimens of many of the species
included in the group and some discrepancies in earlier descriptions have been corrected in the
light of these examinations.

OPSARIICHTHYS Bleeker, 1863
TYPE SPECIES: Leuciscus uncirostris Temminck & Schlegel, 1844.

DISTRIBUTION: Northern, central and southern China, Southern tributaries of the Amur, Korea,
Japan (Honshu), Hainan and Taiwan.

Banarescu (1968) appears to be the only author to have revised the genus. This revision was in
effect a note on the so-called subspecies of Opsariichthys. Banarescu (1968) made no attempt to
define the genus and failed to recognize that two species placed in the genus Zacco in fact belong
with Opsariichthys. These are the species pachycephalus Gunther, 1868 and barbatus Regan, 1908
(see below). I have the following remarks to make concerning the three species now assigned to
the genus :

Opsariichthys uncirostris (Temminck & Schlegal, 1844)

Banarescu (1968) follows Nichols (1943), Berg (1949) and Wu (1964) in recognizing other nominal
species of the genus as being subspecies of O. uncirostris. The separation of the subspecies appears
to have been made only on the basis of the number of lateral line scales, a character in which
there is complete overlap throughout the range of the species. Other characters such as the
varying morphology of the jaw, cranial osteology and buccal epidermis have been completely
disregarded by previous authors.

BARILIINE CYPRINID FISHES 189

Opsariichthys pachycephalus Gunther, 1 868

This species has been placed in Zacco by various authors but is, in my opinion, a member of
Opsariichthys. It shares with other species of the genus the notched lower jaw; a metapterygoid
process; separated hyomandibular condyles; a hyomandibular flange; pterosphenoid-para-
sphenoid connection; subtemporal fossa connected with posttemporal fossa; deep infraorbitals;
maxilla narrowly separated from the premaxilla. It does not share with Zacco the deep, narrow
ethmoid block, enlarged carotid foramen nor the produced anterior cleithral lamina.

Opsariichthys pachycephalus differs from other species in its low number of vertebrae (40 cf.
44 in uncirostris and 42 in barbatus} there being fewer caudal vertebrae (17-18) than in all other
bariliine taxa except Luciosoma and Parluciosoma (16-17).

Banarescu (1968) notes that specimens recorded from Taiwan as Zacco temmincki or Zacco
platypus are most probably misidentified and should be referred to Opsariichthys pachycephalus.

Opsariichthys barbatus Regan, 1908

This species is recorded only from Lake Candidus on Taiwan and was placed by Jordan &
Richardson (1909) in a separate genus Candida on the basis of possessing a pair of posterior
maxillary barbels. Banarescu (1968) included the species in Zacco, considering the presence of
barbels insufficient evidence to warrant generic status. However, Banarescu had overlooked the
form of the lower jaw which is quite different from that of Zacco. I disagree with Banarescu's
(1968) assignment of the species barbatus to Zacco and believe that it should be placed in Opsari-
ichthys. Having only the type specimens available I have been unable to check that all the osteo-
logical characters necessary for inclusion in this genus are present. However, from radiographs
it is clear that the cranium and pectoral girdle have an overall opsariichthine-morphology (the
ethmoid block is broad and shallow and a metapterygoid process is present).

The lower jaw of O. barbatus is shorter than that in any other Opsariichthys species and is
completely overlapped by the upper jaw when the mouth is closed (Fig. 48). The jaw is also
inclined at a more oblique angle so that its articulation with the quadrate comes to lie on a line
passing through the centre of the orbit whereas in other Opsariichthys it is below the posterior
third of the orbit. The lateral line scales number 54-56, is comparable only to that in O. pachy-
cephalus (53) and O. uncirostris uncirostris (50-59).

Jaw morphology and low vertebral numbers suggest that Opsariichthys pachycephalus and O.
barbatus are relatively plesiomorph species as compared with O. uncirostris. However, the
interrelationships of the three species can only be resolved when osteological material of O.
pachycephalus and O. barbatus is available and synapomorphies are identified. Meanwhile, I
would hypothesize that the two Taiwan species form the plesiomorph sister group to the Opsari-
ichthys uncirostris species complex of China, Japan and the Amur basin.

BARILIUS Hamilton, 1822
TYPE SPECIES: Cyprinus barila Hamilton, 1822.

The genus as now restricted includes only those species occurring in India, Sri Lanka, Burma and
Thailand.

Hamilton (1822) first used Barilius as a 'Division' of Cyprinus under which he included B.
barila, noting that it lacked barbels. Heckel (1842) apparently overlooked or disregarded
Hamilton's use of Barilius and proposed Pachystomus in which he included species with 4 barbels
and those without barbels.

Bleeker (1849) introduced Bendelisis as a subgenus of Barilius to include species with 2 barbels.
Later, in the Atlas (1863) he designated the species bendelisis as the type of the subgenus, stating
'cirri 2'. However, Bleeker was in error as B. bendelisis (Hamilton) has 4 barbels.

Day (1877) used Pachystomus, Bendelisis and Barilius as divisions of Barilius. Species with 4
barbels he included in Pachystomus, species with 2 barbels in Bendelisis and those without

190 G. J. HOWES

barbels in Barilius. Day (1877) included only a single species in Bendelisis, namely barila. But,
Day too was in error as B. barila also has 4 barbels. (Day's mistake was pointed out by Hora,
1921.) At the same time, Day (1877) noticed Bleeker's previous oversight with respect to the
number of barbels in B. bendelisis (see above) and placed the species in Pachystomus. Finally,
Jordan (1919) designated barila as the type species of Barilius by logotypy.
Thus, it would appear that Pachystomus and Bendelisis are synonyms of Barilius.

5mm
Fig. 48 Opsariichthys barbatus, lateral view of the head to show form of the lower jaw.

As noted above, p. 180, two groups of Barilius can be recognized, but it is possible that these
'groups' form part of a morphocline. Included species: group (i): B. barila (Hamilton, 1822),
B. bendelisis (Buchanan, 1807), B. radiolatus Gunther, 1864, B. vagra (Hamilton, 1822), B. shacra
(Hamilton, 1822); group (ii): B. baker i(Guniher, 1868), B. barna (Hamilton, 1922), B. canarensis
(Jerdon, 1849), B. evezardi Day, 1872, B. gatensis (Val., 1844), B. modestus (McCl., 1842), B.
ornatus Sauvage, 1883, B. tilheo (Hamilton, 1822).

The following species are ones which I have not examined and am unable to allocate with
certainty to Barilius: B. bernatziki Koumans, 1937, B. dogasinghi Hora, 1921, B. huahinensis
Fowler, 1934, B. infrafasciatus Fowler, 1934, B. karatensis Smith, 1931, B.pulchellus Smith, 1931,
B. ponticulus (Smith, 1945), B. nanensis Smith, 1945.

Two other species formerly placed in Barilius are found not to belong to this genus; they are:

Barilius auropurpureus Annandale, 1918 from Inle Lake, Burma, possesses characters which
relate it to the cheline group.

Barilius mesopotamicus Berg, 1933 possesses apomorph characters shared with some species
presently assigned to Leucaspius.

LEPTOCYPRIS Boulenger, 1900
TYPE SPECIES: L. modestus Boulenger, 1900.
DISTRIBUTION: Africa; Nile, Niger and Zaire.

Two other species are included in the genus, L. niloticus (Joannis, 1835) and L. weynsii (Blgr.,
1899). The characters distinguishing the three species are as follows:

Gill-rakers on 1st ceratobranchial

Lateral line scales

Anal fin rays, branched

Maxillary valve

Pelvic scale, % of pelvic fin length

Extent of maxilla

L. niloticus

3

36-40

11-12

non-papillate

25

L. modestus

40-42

8-9

papillate

75

to centre of eye anterior

L. weynsii

2-3

44-45

14-15

non-papillate

75

centre

BARILIINE CYPRINID FISHES 191

Barilius guineensis Daget, 1962 appears from the description and figure also to belong to Lepto-
cypris. Characteristic are the inferior position of the mouth and rather prominent snout. Daget
(1962) notes the close similarity between B. guineensis and B. weynsii.

ENGRAULICYPRIS Boutenger, 1911
TYPE AND ONLY SPECIES: Barilius sardella Giinther, 1868.
DISTRIBUTION: Lake Malawi and upper Shire river.

OPSAR1DWM Peters, 1852

Pelotrophus Giinther, 1864.

Barilius (non Hamilton), Boulenger, 1899 (part).

TYPE SPECIES: Leuciscus zambesensis Peters, 1852.

DISTRIBUTION: Africa; Zambesi, Zaire, Quanza and Lower Guinea provinces (Roberts, 1975).

Opsaridium is the plesiomorph sister taxon to Raiamas and is characterized by an extended anal
fin in the males, larger and granular tubercles, large and well-patterned dorsal fin (13-15 branched
rays) and fleshy or lobate pectoral axial scales. Osteological characters are listed on p. 180.

Included species: O. zambesense (Peters, 1852), O. christyi (Blgr, 1920), O. loveridgi (Norman,
1922), O. microcephalus (Gnthr, 1864), O. microlepis (Gnthr, 1864), O. ubangense (Pellegrin,
1901).

Opsaridium zambesense (Peters) and O. ubangense (Pell.)

On the basis of their external morphology and osteology these two species appear to be closely
related. Indeed, in Boulenger's (1911) key there is a complete overlap of diagnostic characters.
Jubb (1967) placed Barilius neavii Blgr, 1907 and B. peringueyi Gilchrist, 1913 into the synonymy
of B. zambesensis, stating that these 'species' represented various growth stages.

I have examined all the specimens in the BM(NH) collections identified as B. zambesensis and
B. ubangensis and have the following comments to make: In Opsaridium zambesense (as repre-
sented by the types and the types of Barilius neavii} the maxilla extends as far as the suture of the
2nd and 3rd infraorbitals, the operculum is long and rather attenuated, the dorsal fin is patterned
by dark bands of pigment extending between the rays, except the 6th to 8th rays where the
pigment is confined to the distal margin of the fin membrane. (This pattern is admirably illustrated
in Poll, 1967, fig. 89.) The pectoral axial scale is small with a fleshy lower margin, the dark bars
along the flanks are almost always split anteriorly and number 6-12, lateral line scales number
41-43. The pectoral axial lobe is small; the dark bars along the flanks are almost always split
anteriorly and number 6-12, lateral line scales number 41-43.

Specimens from the Cubango river, Angola (BMNH 1965.3.15 : 34-39) previously determined
as B. zambesensis differ from the above description on having somewhat longer jaws, longer
operculum and a clear submarginal band on the dorsal fin.

In Opsaridium ubangense (specimens from Ubanghi, Luala and Luluaburg), the maxilla
extends to just beyond the anterior border of the eye, the operculum is short, the dorsal fin is
similarly patterned to that in O. zambesense except that the dorsal anterior margin is without
pigment and the tips of the membrane between the last two rays is very dark. The pectoral axial
scale is fleshy and short, the dark bars along the flanks are split and number 12-13, lateral line
scales number 39-42.

Specimens identified as O. ubangense from the Kribi and Ja (Dja) rivers (Cameroon) differ
from 'typical' O. ubangense in that the 3rd and 4th infraorbitals are broad, almost covering the
cheek, the pectoral axial lobes are longer and the dorsal fin lacks the intensely dark posterior
distal markings.

192 G. J. HOWES

Specimens from the Upper Mwanza and Upper Shire rivers are also identified as belonging
to the O. ubangense group.

At present the limits of the various taxa included within the zambesense-ubangense complex
cannot be defined and the problem may only be resolved when further, larger samples from a
wider geographical area become available.

Opsaridium loveridgi (Norman, 1922)

Described from a single specimen from the Rufiji river, the fish has a well-developed pectoral
axial lobe and colour pattern characteristic of the genus Opsaridium. Norman (1922) noted that
the species was ' . . . Near B. ubangensis and B. peringueyC . Certainly O. loveridgi belongs to the
O. ubangense-zambesense group but appears to be a distinct species.

Opsaridium microcephalus (Gnthr, 1864) and O. microlepis (Gnthr, 1864)

These two species both occur in Lake Malawi. Giinther (1864) referred them to the genus Pelo-
trophus on account of their extended anal fins.

Opsaridium microcephalus differs from other Opsaridium species in possessing a ventral frontal
foramen and an extensive lateral temporal foramen (p. 147, Fig. 17). It shares with O. microlepis
a high vertebral number (47-48) and a high number of dorsal fin rays (15-17).

The relationships of the Lake Malawi species have not yet been explored in great depth but
indications are that the closest riverine relative is O. ubangense.

Three species formerly included in Barilius are recorded from Lake Tanganyika; moori,
salmolucius and tanganicae (Poll, 1953; Worthington & Ricardo, 1936). The first two species are
now included in Raiamas (see below) but tanganicae has an Opsaridium-\ike morphology.
Barilius tanganicae Blgr, 1900 is known only from the holotype, and despite subsequent collecting
no specimens have been recorded again. Poll (1953) makes no mention of the species; Worthington
& Ricardo (1936) note only the type. The holotype conforms in almost every respect to the descrip-
tion of Barilius microlepis given by Boulenger (1911). Although Boulenger (1911) states for
tanganicae ' ... 16 or 17 blackish vertical bars on each side of the body' and for microlepis
' ... no markings', there are in fact similar markings present in microlepis (see Poll, 1953).

The holotype of B. tanganicae was collected by J. E. S. Moore and bears the BMNH register
number 1906.9.6 : 26. The locality given in the accessions register is 'N end of L. Tanganyika'.
Regrettably there is no collection number or label which can help to identify the locality. In view
of the fact that Moore's locality labels have proved to be unreliable in the past (see Trewavas,
1946 : 244) I suspect that tanganicae is a specimen of microlepis and that it was collected from
Lake Nyasa (L. Malawi). Thus, until there is evidence to the contrary I place Barilius tanganicae
into the synonymy of Opsaridium microlepis. This means that the only bariliine species known
from Lake Tanganyika are of the genus Raiamas.

Species which I have not examined but which appear from the literature to belong to Opsari-
dium are: Barilius maculicauda Pellegrin, 1926. Described from 3 specimens collected at Tshikapa,
Congo. The description and Pellegrin's observation that the species resembles O. zambesense
indicate its inclusion in Opsaridium.

Barilius leleupi Matthes, 1965. The prolongation of the anal fin, body markings, shape of the
mouth and distribution of the tubercles indicate that this species belongs to Opsaridium. Matthes
(1965) discussed the likely affinity of B. leleupi with O. zambesense and B. boweni Fowler, 1930;
the latter species he considered might be synonymous with O. zambesense.

Barilius engrauloides Nichols, 1923. Described from a single specimen collected from the
Ubangui river. In his description Nichols states ' ... It seems to be more or less intermediate
between species of that genus (Barilius) and of Engraulicypris\ The figure accompanying the
description is crudely drawn and of no help in determining the generic allocation of this species.

BARILIINE CYPRINID FISHES 193

RAIAMA S Jordan, 1919

Barilius (non Hamilton), Boulenger, 1 899 (part).
Bola Gunther, 1868.
Sagittabarilius Fowler, 1936.

TYPE SPECIES: Cyprinus bola Hamilton, 1822.

DISTRIBUTION: Africa, Nilotic, Guinean and Zairean ichthyofaunal provinces (see Roberts, 1975
for definition of these provinces); northern India; Burma and the Malay Peninsula.

Gunther (1868) first recognized a generic separation for Barilius bola and used the generic
name Bola. However, this name was preoccupied and the name Raiamas was proposed as a
replacement by Jordan (1919).

Hora & Mukerji (1936) accepted Jordan's substitute name but later, Hora (1937) discussed the
nomenclatural history of Barilius bola and decided that Raiamas had been proposed in error
because the name Opsarius McClelland, 1839 was available. Hora's (1937) contention was that
Day (1869) had restricted the use of Opsarius McClelland to Barilius bola. Day had, however,
simply placed B. bola in the genus Opsarius without designating it as the type of that genus, and
later (1889) he included Opsarius in the synonymy of Barilius. McClelland (1839) had not desig-
nated a type species for the genus Opsarius, an action subsequently taken by Jordan (1919) who
selected the species maculatus on the grounds of logotypy. Opsarius maculatus is a synonym of
Barilius tilheo (Hamilton, 1822).

Smith (1945) used Jordan's name Raiamas for bola although his reasons for justifying its generic
allocation, i.e. on the basis of its pharyngeal dentition, are not those used here. Smith (1945)
stated that the two rows of pharyngeal teeth in R. bola distinguish it from all other Barilius,
which have three rows. In all specimens of R. bola I have examined there are three rows, a feature
also noted and illustrated by Hora (1937).

Included African species: R. ansorgii (Blgr, 1910), R. batesii (Blgr, 1914), R. buchholzi (Peters,
1876), see comments below, R. kingsleyae (Blgr, 1899), see comments below, R. loati (Blgr, 1909),
see comments below, R. nigeriensis (Daget, 1951), R. salmolucius (Nichols & Griscom, 1917),
R. silax (Schultz, 1942), R. senegalensis (Steind., 1870), R. steindachneri(Pe\l, 1908), R. longirostris
(Blgr, 1902), see comments below, R. macrostoma (Blgr, 1913), R. moorii (Blgr, 1900), R. weeksi
(Blgr, 1899), R. shariensis (Fowler, 1949).

Raiamas buchholzi (Peters)

Of this species, Boulenger (191 1) stated that it was separable from Barilius kingsleyae only on the
number of lateral line scales (46 cf. 49-51 according to Boulenger) and ' . . . may ultimately have
to be regarded as not separable from it'. Although I have not had the opportunity to see the type
specimen of R. buchholzi it is clear from Peters' (1876) illustration that it should be included in
Raiamas. In R. kingsleyae the number of lateral line scales ranges from 46 to 49 (51 only if
counting onto the caudal fin). I find no other differences between R. kingsleyae and Peters'
description of R. buchholzi apart from the apparent lack of body markings in the latter (faint
vertical bars are present in R. kingsleyae but are lacking in some specimens, no doubt due to the
effects of preservation). I therefore propose that Raiamas kingsleyae be regarded as a junior
synonym of Raiamas buchholzi.

Raiamas longirostris (Blgr)

Known only from the holotype collected at Ubanghi, this species was placed by Fowler (1936)
in the subgenus Sagittabarilius along with R. salmolucius. The principal features Fowler used to
characterize Sagittabarilius were the slender body, long maxilla and wide infraorbitals - characters
shared by all Raiamas species.

194 G. J. HOWES

Dr K. E. Banister has examined the holotype in Tervuren Museum and reports that the
characteristic elongate snout is not an artefact and in this regard R. longirostris closely resembles
R. bola (see below).

Raiamas loati (Blgr) group

The complex of species represented by R. loati in the east, R. salmolucius in the west and by R.
ansorgii in the south-west appears to be a morphoclinal series of taxa. If the descriptions of the
west African species given by Boulenger (1911, 1916) are compared closely, the only differences
to be found are in lateral line scale counts and some body proportions.

Raiamas loati is readily distinguished from the west and south-western species complex by its
long maxilla (extending to the posterior third of the orbit), elongate pectoral and pelvic axial
fin lobes (half the length of their respective fins), length and shape of the 1st infraorbital bone
(equal to eye diameter and with the upper posterior edge slightly extended along the 2nd infra-
orbital).

In contrast R. senegalensis, R. steindachneri, R. buchholzi and R. ansorgii all have a relatively
short maxilla, short pectoral and pelvic axial fin lobes, and a short 1st infraorbital bone -the
upper border of which extends posteriorly as a narrow triangle.

Regarding R. macrostoma, Boulenger (1916) noted ' . . . Intermediate between B. loati and B.
senegalensis'. Its elongate axial fin lobes and shape of the 1st infraorbital bone suggest that it is
more closely allied to R. loati.

Even within the relatively small number of specimens of the west African species examined,
'intermediates' between all species can be found.

Two Asian species are included in Raiamas, R. bola (Hamilton, 1822) and R. guttatus (Day,
1869). These species are almost identical in cranial morphology but differ in certain other
characters. The lateral line scales in R. bola number 88-94 but 44-48 in R. guttatus. Barbels are
absent in R. bola but in R. guttatus there are two minute posterior maxillary barbels. Smith (1945)
says of R. bola that its 'scales are finer than any known species of Barilius\ However, they
compare in shape and striation with those of other Raiamas, Opsaridium and Barilius species.
According to Hora & Mukerji (1936) R. bola is confined to the hilly areas of the Northwest
Provinces of India, Assam, Bengal, Orissa and Burma. Fowler (1934) records the species from
Thailand but appears to have mistaken R. guttatus for R. bola (see Smith, 1945). Raiamas guttatus
occurs in the Shan States, Burma, Thailand, Cambodia (Khmer Republic) and the Malay
Peninsula.

LUCIOSOMA Bleeker, 1855
TYPE SPECIES: Barbus setigerus Valenciennes, 1842.
DISTRIBUTION : Thailand, Java, Sumatra, Borneo and Laos.

The following species are included in the genus: L. setigerum (Val., 1842); L. spilopleura Bleeker,
1865; L. bleekeri Steind., 1879; L. trinema Bleeker, 1852; L. pellegrini Popta, 1905.

Luciosoma fasciata Yang & Hwang, 1964 appears from the figure not to belong to this genus.
The shape of the head and mouth, the position of the dorsal fin, length of the pectoral fin, and the
high number of lateral line scales all indicate a species of Raiamas, possibly R. guttatus.

PARLUCIOSOMA Howes, 1980
TYPE SPECIES: Leuciscus argyrotaenia Bleeker, 1850.
DISTRIBUTION: India, Thailand, Malay peninsula and Sunda Islands.

Included species: P. argyrotaenia (Bleeker, 1850), P. dusonensis (Bleeker, 1851), P. daniconius
(Hamilton, 1822), P. cephalotaenia (Bleeker, 1859), P. volzi (Popta, 1905).

BARILIINE CYPRINID FISHES 195

Species descriptions and details of their distribution are given by Brittan (1954). Species which
probably belong to this genus, but which I have not had the opportunity of examining, are:
Rasbora steineri Nichols & Pope, 1927, R. hubbsi Brittan, 1954, R. myersi Brittan, 1954,
R. tawarensis Weber & de Beaufort, 1916.

Mention may be made at this point of Rasbora zanzibarensis Gunther, 1866 reportedly collected
from the Rovuma river, East Africa, and to date the only known representative of the genus in
Africa. Day (1877) placed the species in the synonymy of Rasbora daniconius, an action followed
by Brittan (1954). Neither author made particular mention of the African locality of the species.
I have examined the unique type of R. zanzibarensis and find that it is a Rasbora, but I am
doubtful that it belongs to the species daniconius. I also doubt the provenance of the fish is
Africa and I suspect that there has been an error concerning the locality citation. For the present
R. zanzibarensis must remain a species inquirendum.

MEGARASBORA Gunther, 1868

TYPE AND ONLY KNOWN SPECIES: Cyprinus elanga Hamilton, 1822.
DISTRIBUTION : Bengal, Assam and western Burma.

The species has been described by Brittan (1954) and an osteological description is given above,
p. 183.

Appendix 2
A note on the taxa formerly included in Engraulicypris

As Engraulicypris is now restricted to the type species, E. sardella (see p. 182), the species pre-
viously included must be referred to another genus. A preliminary review of the anatomy of these
species has indicated on the basis of synapomorphic characters that they belong to three genera,
namely Neobola, Rastrineobola and Chelaethiops. The apomorph characters defining the genera
and their contained species are as follows :

NEOBOLA Vinciguerra, 1895
TYPE SPECIES: Neobola bottegi Vinciguerra, 1895.

APOMORPH CHARACTERS: Deep, narrow ethmoid block, the dorsal surface (supraethmoid)
channelled, elongate nasal, raised frontal sensory canal, narrow infraorbitals, separation of the
dermo- and autopterotic by a posttemporal fossa, long anterior premaxillary processes, attenuated
opercular border, rounded anterior articular face of 1st vertebra.

INCLUDED SPECIES: TV. bottegi Vine., 1895, N.fluviatilis (Whitehead, 1962), N. brevianalis (Blgr,
1908), N. bredoi (Poll, 1945), N. spinifer (Bailey & Matthes, 1971), N. stellae (Worthington, 1934),
N. moeurensis (Blgr., 1915).

RASTRINEOBOLA Fowler, 1934
TYPE SPECIES: Neobola argentea Pellegrin, 1904.

APOMORPH CHARACTERS: Long, ventrally directed medial maxillary processes, long post-coronoid
dentary border.

INCLUDED SPECIES: R. argentea (Pellegrin, 1904).

CHELAETHIOPS Boulenger, 1899
TYPE SPECIES: Chelaethiops elongatus Blgr, 1899.
APOMORPH CHARACTERS : Supraethmoid narrow, channelled and overlapped by the frontals, raised

196 G. J. HOWES

frontal sensory canal, elongate jaws, maxillaries contacting one another medially, separation of
the dermo- and autopterotic by a posttemporal fossa, anterior articular face of 1st vertebra
rounded with caudally directed lateral processes, elongate pectoral axial scale, numerous olfactory
lamellae. In addition the jaw musculature shows derived features.

INCLUDED SPECIES: C. elongatus Blgr, 1899, C. bibie (Joannis, 1835), C. congicus (Nichols &
Griscom, 1917), C. minutus (Blgr, 1906).

I have examined the type specimens of both Chelaethiops congicus and C. katangae Poll, 1948,
and it is my opinion that the latter is a synonym of the former. Specimens catalogued in the
BM(NH) collections as C. congicus are not this species and represent an undescribed taxon (or
taxa). Ricardo (1939) described from Lake Rukwa a subspecies of C. congicus. However, her
description was based on the examination of the misidentified comparative material noted above.
Although rukwaensis is indeed a discrete taxon it is not closely related to C. congicus, although it is
related to the undescribed taxa from Lake Tanganyika and the Malagarasi river. Thus, Ricardo's
subspecies is here recognized as a species, Chelaethiops rukwaensis Ricardo, 1939.

The relationships of Neobola, Rastrineobola and Chelaethiops remain to be investigated, but
my preliminary observations indicate that all three genera form a monophyletic assemblage
related to middle Asian and European groups as represented by Phoxinellus (of which there are
two African species, P. chaignoni and P. callensis). There is thus no close relationship with the
bariliine group.

References

Banarescu, P. 1968. Revision of the genera Zacco & Opsariichthys (Pisces, Cyprinidae). Vest. csl. zool.

Spol 32 (4) : 305-210.
Banister, K. E. & Clarke, M. A. 1977. The freshwater fishes of the Arabian Peninsula. In: The scientific

results of the Oman flora and fauna survey 1975. //. Oman Stud. 2 : 111-154.

- In press. A revision of the large Barbus of Lake Malawi with a reconstruction of the history

of the southern Rift Valley lakes. /. nat. Hist.
Berg, L. S. 1964. Freshwater fishes of the USSR and adjacent countries. 2. (English translation of the

fourth ed. of 1949.) 496 pp.
Bleeker, P. 1859. Conspectus systematis Cyprinorum. Natuurk Tijdschr. Ned.-Indie'2Q : 421-441.

- 1863. Atlas ichthologique des Indes Orientales Needlandaises. Cyprins 3. Amsterdam : 1-150.
Boulenger, G. A. 1911. Catalogue of African freshwater fishes in the British Museum (Natural History)

2 : 529 pp.

- 1916. Catalogue of African freshwater fishes in the British Museum (Natural History) 4 : 392 pp.
Briggs, J. C. 1979. Ostariophysan Zoogeography: an alternative hypothesis. Copeia 1979 (1) : 111-118.
Brittan, M. R. 1954. A revision of the Indo-Malayan fresh-water fish genus Rasbora. Institute of Science &

Technology, Manila. 224 pp.
Brousseau, R. A. 1976. The pectoral anatomy of selected Ostariophysi. II. The Cypriniformes and siluri-

formes. /. Morph. 150 : 79-115.
Burton, C. K. 1970. The palaeotectonic status of the Malay Peninsula. Palaeogeogr. Palaeoclimat.

Palaeoecol. 1 : 51-60.
Chu, Y. T. 1935. Comparative studies on the scales and on the pharyngeals and their teeth in Chinese

cyprinids, with particular reference to taxonomy and evolution. Biol. Bull. St. John's Univ., Shanghai

2 : 1-225.

Daget, J. 1962. Les poissons du Fouta Dialon et de las basse Guinee. Mem. Inst.fr. Afr. noire 65 : 1-210.
Day, F. 1869. On the fishes of Orissa. Proc. zool. Soc. Lond. : 369-387.

- 1877. The fishes of India Pt. Ill : 369-552. London.

- 1889. The fauna of British India, including Ceylon and Burma. Fishes 1. London, 548 pp.
Evans, H. M. 1940. Brain and body offish. London. 164 pp.

Forey, P. L. 1975. A fossil clupeomorph fish from the Albian of the Northwest Territories of Canada,
with notes on cladistic relationships of clupeomorphs. /. Zool., Lond. 175 : 151-177.

Fowler, H. W. 1934. Zoological results of the third de Schaunensee Siamese Expedition Part 1. Fishes.
Proc. Acad. nat. Sci. Phi lad. 86 : 67-163.

- 1936. Zoological results of the George Vanderbilt Africa expedition of 1934. Part III. The fresh-
water fishes. Proc. Acad. nat. Sci. Philad. 88 : 243-335.

BARILIINE CYPRINID FISHES 197

Gilbert, C. R. & Bailey, R. M. 1972. Systematics and zoogeograply of the American cyprinid fish Notropis

(Opsopoedus) emiliae. Occ. Pap. Mus. Zool. Univ. Mich. 664 : 1-35.
Goren, M., Fishelson, L. & Trewavas, E. 1973. The cyprinid fishes of Acanthobrama Heckel and related

genera. Bull. Brit. Mus. nat. Hist. (Zool.) 24 (6) : 293-315.
Gosline, W. A. 1978. Unbranched dorsal-fin rays and subfamily classification in the fish family Cyprinidae.

Occ. Pap. Mus. Zool. Univ. Mich. 684 : 1-21.
Gosse, J.-P. 1966. Remarques systematiques sur quelques especes de la faune ichthyologique congolaise.

Revue Zool. Bot. afr. 73 (1-2) : 186-200.
Greenwood, P. H. 1976. A review of the family Centropomidae (Pisces, Perciformes). Bull. Brit. Mus. nat.

Hist. (Zool) 29 (I) : 1-81.

, Rosen, D. E., Weitzman, S. H. & Myers, G. S. 1966. Phyletic studies of teleostean fishes, with a

provisional classification of living forms. Bull. Am. Mus. nat. Hist. 131 : 339-456.
Giinther, A. 1 864. Report on a collection of fishes made by Dr Kirk in the Zambesi and Nyassa Regions.

Proc. zool. Soc. Land. : 303-314.

1868. Catalogue of fishes in the British Museum. 7 : 1-512.

Hamilton, F. 1822. An account of the fishes found in the River Ganges and its branches. Edinburgh, 405 pp.
Heckel, J. J. 1842. Ichthyologie (von Syrien). In: Russeggar, J. V. Reisen in Europa, Asien und Africa etc.

Pt. 1 : 993-1099.
Hora, S. L. 1921. Fish and fisheries of Manipur with some observations on those of the Naga Hills.

Rec. Ind. Mus. 22 : 165-214.

- 1937. The game fishes of India. /. Bombay Nat. Hist. Soc. 39 : 199-210.

& Mukerji, U. D. 1936. Fish of the Eastern Doons, United Provinces. Rec. Ind. Mus. 38 (2) : 133-

146.
Howes, G. J. 1978. The anatomy and relationships of the cyprinid fish Luciobrama macrocephalus

(Lacepede). Bull. Br. Mus. nat. Hist. (Zool.) 34 (1) : 1-64.

- 1979. Notes on the anatomy of the cyprinid fish Macrochirichthys macrochirus with a review of the
subfamily Cultrinae. Bull. Br. Mus. nat. Hist. (Zool.) 36 (3) : 147-200.

Jordan, D. S. 1919. The genera of fishes, Part 3. Leland Stanford Junior Univ. Pubs, University Series :
285-410 +i-xv.

- & Richardson, R. E. 1909. A catalog of the fishes of the island of Formosa, or Taiwan, based on the
collections of Dr Hans Sauter. Mem. Carnegie Mus. 4 (4) : 1 59-204.

Jubb, R. A. 1967. Freshwater fishes of Southern Africa. Cape Town. 24 pp.
McClelland, J. 1839. Indian Cyprinidae Asiat. Reschs 19 (2) : 217-471.

Matthes, H. 1963. A comparative study of the feeding mechanism of some African Cyprinidae (Pisces,
Cypriniformes). Bijdr. Dierk. 33 : 3-35.

- 1965. Quelques poissons nouveaux du Congo. Revue Zool. Bot. afr. 71 : 177-193.

Menon, A. G. K. 1951. Further studies regarding Hora's Satpura hypothesis. Proc. natn. Inst. Sci. India
17 (6) : 475-497.

- 1953. Age of transgression of the bay of Bengal and its significance in the evolution of the freshwater
fish fauna of India. Bull. natn. Inst. Sci. India 1 : 240-247.

1955. The external relationships of the Indian freshwater fishes, with special reference to the

countries bordering the Indian ocean. /. Asiatic Soc. Calcutta 21 (1) : 31-38.
Mirza, M. R. 1974. Freshwater fishes and ichthyogeography of Baluchistan and adjoining areas of the

Indus plain, Pakistan. Biologia 20 (1) : 67-82.
Nakamura, M. 1969. Cyprinid fishes of Japan. Tokyo. 451 pp.
Nichols, J. T. 1923. New African fishes. Amer. Mus. Novit. 57 : 1-3.
1943. Freshwater fishes of China. Nat. Hist, of Central Asia 9. Amer. Mus. of Nat. Hist., N.Y.

322 pp.

Nieuwenhuys, R. 1962. Trends in the evolution of the Actinopterygian forebrain. J Morph. Ill : 69-88.
Norman, J. R. 1922. Four new fishes from Tanganyika Territory. Ann. Mag. nat. Hist. (9) 9 : 686-688.
Okada, Y. 1960. Studies on the freshwater fishes of Japan. II. Special Part. /. Fac. Fish. pref. Univ. Mie-

Tsu 4 (2) : 267-588.
Patterson, C. 1975. The braincase of pholidophorid and leptolepid fishes, with a review of the actino-

pterygian braincase. Phil. Trans. Roy. Soc. B 269 (899) : 275-579.
Pellegrin, J. 1926. Description d'un Cyprinide nouveau recolte au Congo Beige par le Dr Schouteden.

Revue Zool. Bot. afr. 14 : 40-41.
Peters, W. C. H. 1876. Uber die von Professor Dr. Reinhold Buchholz in Westafrica gesammelten Fische.

Mber. dt. Akad. Wiss. Berl. : 244-252.
Poll, M. 1953. Poissons non Cichlidae. Result, sclent. Explor. hydrobiol. Lac Tanganyika (1946-1947)

3 (5 A) : 1-251.

198 G. J. HOWES

- 1967. Contribution a la faune ichthyologique de 1'Angola. Publicoes cult. Co. Diam. Angola no. 75 :
1-381.

& Gosse, J.-P. 1963. Contribution a I'etude systematique de la faune ichthyologique du Congo

Central. Ann. Mus. roy. Afr. Cent. Zool. 116 : 43-1 11.
Ramaswami, L. S. 1955. Skeleton of cyprinoid fishes in relation to phylogenetic studies: 7. The skull and

Weberian apparatus of the Cyprininae (Cyprinidae). Acta Zool. 36 (3) : 199-242.

Regan, C. T. 1908. Description of new fishes from Lake Candidus, Formosa, collected by Dr A. Moltrecth.
Ann. Mag. not. Hist. (8) 2 : 358-360.

- 1911. The classification of the teleostean fishes of the order Ostariophysi. 1. Cyprinoidei. Ann.
Mag. not. Hist. (8)8 : 13-32.

1922. The distribution of the fishes of the order Ostariophysi. Bijdr. Dierk. 22 : 203-207.

Ricardo, C. K. 1939. The fishes of Lake Rukwa. /. Linn. Soc. (Zool.) 40 (275) : 625-657.

Ridd, M. F. 1971. South-East Asia as a part of Gondwanaland. Nature, Lond. 230 : 531-533.

Roberts, T. R. 1973. Interrelationships of ostariophysans. In: Interrelationships of fishes. Greenwood,

P. H., Miles, R. S. & Patterson, C. (Eds) : 373-395. London and New York.

- 1975. Geographical distribution of African freshwater fishes. Zool. J. Linn. Soc. 57(4) : 249-319.
Rognes, K. 1973. Head skeleton and jaw mechanism in Labrinae (Teleostei : Labridae) from Norwegian

waters. Arbok Univ. Bergen Mat.-Naturv. Ser. no. 4 : 1-149.

Schnitzlein, H. N. 1964. Correlation of habit and structure in the fish brain. Amer. Zool. 4 : 21-32.
Seggar, J. 1961. Telencephalon and behaviour in Gasterosteus aculeatus. Behaviour 18 : 256-287.
Silas, E. G. 1955. Speciation among the freshwater fishes of Ceylon. Bull. natn. Inst. Sci. India 7 : 248-259.
Smith, H. W. 1945. The fresh-water fishes of Siam, or Thailand. Bull. U.S. natn Mus. 188 : 1-622.
Stauffer, P. H. 1974. Malaya and southeast Asia in the pattern of continental drift. Bull. geol. Soc.

Malaysia 7 : 89-138.

Tretiakov, D. K. 1946. Systematic groups of Cyprinidae. Zool. Zh. 25 (2) : 149-156.
Vandewalle, P. 1974. On the anatomy and function of the head region in Gobio gobio (L.) (Pisces, Cyprin-
idae). 2. The trigemino-facialis chamber and some adjacent structures. Forma Functio 7 : 119-124.
Verraes, W. 1974. Discussion on some functional-morphological relations between some parts of the

chondrocranium and the osteocranium in the skull base and the skull roof, and of some soft head parts

during postembryonic development of Salmo gairdneri Richardson 1836 (Teleostei : Salmonidae).

Forma Functio 7 : 281-292.
Yang, Y-J. & Hwang, H.-C. 1964. Leuciscinae. In: Wu, H.-W. Cyprinid fishes of China 1. Shanghai.

228 pp.
\\ eit/nian, S. 1962. The osteology of Brycon meeki, a generalized characin fish, with an osteological

definition of the family. Stanford ichthyol. Bull. 8(1) : 3-77.
Wiley, M. L. & Collette, B. B. 1970. Breeding tubercles and contact organs in fishes: their occurrence,

structure and significance. Bull. Am. Mus. nat. Hist. 143 art. 3 : 143-216.
Worthington, E. B. & Ricardo, C. K. 1936. The fish of Lake Tanganyika (other than Cichlidae). Proc.

zool. Soc. Lond. (4) : 1061-1112.
Wu, Hsien-Wen, 1964. Cyprinid fishes of China 1. Shanghai. 228 pp.

Manuscript accepted for publication 28 March 1979

British Museum (Natural History)
Publication on fishes

The Cichlid fishes of Lake Victoria, East Africa. The biology and

evolution of a species flock. P. H. Greenwood.

Bull. British Museum (Natural History) Zool. Suppl. No. 6, 1974

vi+ 134 pp, 1 coloured plate, 77 text figures, 4to, (paper) 6.00
(boards) 8.25

The feeding mechanisms of a deep sea fish, Chauliodus sloani

Schneider. V. V. Tchernavin

1953, viii+ 101 pp, 10 plates, 36 text figures, 4to paper, 3.85

Forty drawings of fishes, made by the artists who accompanied
Captain James Cook. P. J. P. Whitehead

1968, xxvii pp. 36 collotype colour plates, demy folio boards, 55,00

The Clupeoid fishes described by Lacepede, Cuvier and
Valenciennes. P. J. P. Whitehead

Bull. British Museum (Natural History) Zool, Suppl. No. 2, 1967,
180 pp, 11 plates, 15 text figures, 4to paper, 7.00

The Clupeoid fishes of the Guianas. P. J. P. Whitehead

Bull. British Museum (Natural History) Zool. Suppl. No. 5, 1973,

227 pp, 72 text figures, 6 tables, 4to paper, 13.50

Titles to be published in Volume 37

Miscellanea

The echinoderms of Aldabra and their habitats. By N. A. Sloan,
Ailsa M. Clark & J. D. Taylor

The anatomy, phylogeny and classification of bariliine cyprinid fishes.
By G. J. Howes

The Fellodistomidae (Digenea) of fishes from the northeast Atlantic.

By Rodney A. Bray & David L. Gibson

Printed by Henry Ling Ltd, Dorchester

Bulletin of the

British Museum (Natural History)

The Fellodistomidae (Digenea) of fishes
from the northeast Atlantic

Rodney A. Bray & David I. Gibson

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and are not
necessarily completed within one calendar year. Subscriptions may be placed for one or more
series. Subscriptions vary according to the contents of the Volume and are based on a forecast
list of titles. As each Volume nears completion, subscribers are informed of the cost of the
next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to:

Publications Sales,

British Museum (Natural History),
Cromwell Road,

London SW7 5BD,
England

World List abbreviation: Bull, Br. Mus. nat. Hist. (Zool.)

Trustees of the British Museum (Natural History), 1980

This number completes Volume 37

ISSN 0007-1498 Zoology series

Vol 37 No 4 pp 199-293
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 28 February 1980

The Fellodistomidae (Digenea) of fishes from the
northeast Atlantic

Rodney A. Bray & David I. Gibson

Jk

Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD

Contents

Introduction 199

Materials and methods ........... 199

Systematic section ............ 200

Family Fellodistomidae 200

A taxonomic history of the group ........ 200

Some comments on forms previously treated as fellodistomids . . . 201

Aspects of biology ........... 202

Subfamily Fellodistominae ......... 202

Fellodistomum fellis .......... 205

Steringophorus furciger . . . . . . . . . .212

Steringophorus agnotus . . . . . . . . . .217

Steringophorus blackeri .......... 220

Steringophorus pritchardae ......... 223

Steringophorus thulini .......... 226

Olssonium turneri ........... 229

Steringotrema pagelli .......... 234

Steringotrema divergens .......... 239

Steringotrema ovacutum .......... 242

Prudhoeus nicholsi ........... 245

Subfamily Baccigerinae .......... 248

Bacciger bacciger ........... 250

Pronoprymna ventricosa .......... 255

Subfamily Xenoperinae .......... 259

Proctoeces maculatus .......... 262

Subfamily Monascinae .......... 268

Monascus filiformis .......... 268

Subfamily Tergestiinae .......... 274

Tergestia laticollis ........... 276

Host-parasite list . . . . . . . . . . . .281

Acknowledgements ..........

References ............. 283

Introduction

This is the third in a projected series of papers on the helminth-fauna of fishes from the northeast
Atlantic region. The family discussed is the Fellodistomidae, a group restricted mainly to marine
fishes, but occurring occasionally in freshwater fishes and as adults in marine invertebrates.

Materials and methods

The materials and methods used in this investigation are outlined in the first two papers in this
series (Bray & Gibson, 1977; Gibson & Bray, 1977). In addition to the sources of material
mentioned in these two papers, we have collected further material at Aberdeen in March-April
1977, and aboard MAFF RV Cirolana in NE Atlantic in 1978. In this report we have used the
nomenclature and classification of fishes presented by Hureau & Monod (1973).

GENERAL
29 FEB1<

Bull. Br. Mus. nat. Hist. (Zool.) 37 (4) : 199-293 Issued 28 February 1980

199

200 R. A. BRAY & D. I. GIBSON

Systematic section
Family FELLODISTOMIDAE Nicoll, 1909

Steringophoridae Odhner, 1911.
Xenoperidae Poche, 1926.
Monascidae Dollfus, 1947.

DIAGNOSTIC FEATURES. Body large to small; globular to elongate. Body-surface usually smooth;
occasionally bearing spines or muscular ornamentation. Oral sucker subterminal; globular.
Ventral sucker globular; large to small; in middle or anterior half of body. Prepharynx short to
apparently absent. Pharynx well developed; globular to elongate. Oesophagus absent to long.
Gut-caeca normally two; narrow to wide; reaching to testes or beyond; occasionally with single
caecum (? opening into excretory vesicle). Testes two; oval to globular; entire or deeply lobed;
symmetrical, oblique or tandem; in anterior or posterior hindbody; normally post-ovarian.
Cirrus-sac usually well developed, occasionally absent; oval to claviform; containing seminal
vesicle, pars prostatica, prostatic cells and ejaculatory duct. Seminal vesicle usually bipartite;
occasionally globular or convoluted and tubular. Pars prostatica usually wide with filamentous
lining; occasionally greatly reduced. Spermatophores may be produced. Ejaculatory duct usually
wide and convoluted; occasionally small and narrow. Genital atrium short to long. Genital pore
ventral; usually sinistral; in middle or posterior half of forebody. Ovary entire to multilobate;
in hindbody; usually anterior to testes. Laurer's canal present. Uterine seminal receptacle usually
present; canalicular seminal receptacle (see Gibson & Bray, 1979) occasionally present. Uterus
normally reaching posteriorly to testes; bulk of uterus usually post-testicular. Eggs numerous;
small; variable; operculate; egg-shell occasionally ornamented. Vitellarium follicular; typically
in two lateral fields; occasionally amalgamated to form two oval masses; occasionally confluent
medially. Excretory pore terminal; vesicle 'Y'- or 'V'-shaped, with anterior arms reaching to
about level of pharynx. Parasitic in intestine, pyloric caeca, bile duct and gall-bladder of marine,
and occasionally freshwater, teleosts; occasionally occurring as adults in molluscs (bivalves and
gastropods).

A taxonomic history of the group

Nicoll (1909a) erected the subfamily Fellodistominae in the suborder Prosostomata Odhner, 1905,
to include the two genera Fellodistomum Stafford, 1904, and Steringophorus Odhner, 1905.
Two years later Odhner (191 la), in erecting the new family Steringophoridae, provided the first
detailed concept of this group. In it he included Steringophorus, Fellodistomum, Tergestia Stossich,
1899, and four new genera, Rhodotrema, Steringotrema, Haplocladus and Proctoeces. The name
Steringophorinae was used by Odhner to replace the name Fellodistominae, as he felt that it was
more appropriate and, in addition to Steringophorus and Fellodistomum, he included Rhodotrema
and Steringotrema. For the remaining genera, Haplocladus, Proctoeces and Tergestia, he erected the
subfamily Haplocladinae. Woodcock (1912) first used the name Fellodistomidae, and Nicoll
(1913) pointed out its priority to the appellation Steringophoridae.

The two subfamilies Steringophorinae [ = Fellodistominae] and Haplocladinae were recognized
by Fuhrmann (1928) and Issaitschikov (1928), the Steringophorinae having grown to include six
genera and the Haplocladinae to include four genera. The next important developments were
made by Dollfus (1952), who recognized four subfamilies, the Fellodistomatinae [sic], containing
four genera with a pre-testicular uterus, the Steringophorinae, containing 14 genera with the
uterus mainly post-testicular, the Monascinae Dollfus, 1947, containing five genera with the
uterus distributed both anteriorly and posteriorly to the testes, and the Discogasteroidinae
Srivastava, 1939, containing two genera with a number of characteristics including short caeca
and a seminal receptacle. Cable (1953), using evidence from life-history studies, included the
Fellodistominae, Haplocladinae, Tandanicolinae Johnston, 1927, and Gymnophallinae Odhner,
1905, in the Fellodistomatidae [sic]; but Yamaguti (19530), although including 20 genera in the
family, did not recognize any of the subfamilies. Skrjabin & Koval (1957), however, further sub-
divided the family into 1 1 subfamilies, the Fellodistomatinae [sic], the Discogasteroidinae, the

THE FELLODISTOMIDAE 201

Haplocladinae, and eight new subfamilies, the Ancylocoeliinae, the Antorchiinae, the Lissolo-
matinae, the Markevitschiellinae, the Pyriformiinae, the Proctoecinae, the Tergestiinae and the
Yamagutiinae. Yamaguti (1958) recognized 10 subfamilies, the Fellodistominae, the Monascinae,
the Discogasteroidinae, the Antorchiinae, the Pyriforminae, the Tergestiinae, the Lissolomatinae,
the Heterorchiinae Dollfus, 1950, and two new subfamilies, the Pentagramminae and the Sym-
metrovesiculinae [being unaware of Skrjabin & Koval's (1957) work, Yamaguti also considered
that four of the other subfamilies were new]. Baer & Joyeux (1961), however, retained only four
subfamilies, the Fellodistomatinae [sic], the Antorchiinae, the Gymnophallinae and the Heter-
orchiinae. Mehra (1963) recognized 9 subfamilies, synonymizing the Discogasteroidinae with the
Fellodistominae and the Pentagramminae with the Antorchiinae. He also included the Bacci-
gerinae Yamaguti, 1958, transferring it from the Cryptogonimidae Ward, 1917. Overstreet (1969)
and Angel (1971) considered the Monodhelminthinae Dollfus, 1937, as a subfamily of the
Fellodistomidae and, finally, Yamaguti (1971) recognized 15 subfamilies, the Baccigerinae,
Proctoecinae, Parantorchiinae Yamaguti, 1958, Lintoniinae Yamaguti, 1970, Stenakrinae
Yamaguti, 1970, and two new subfamilies, the Trigonocryptinae and the Infundibulostominae,
being added to his earlier (1958) list. Yamaguti (1971) followed Travassos, Teixeira de Freitas &
Biihrnheim (1965) in giving family-status to the Monascinae and considered the Pentagramminae
to be synonymous with the Baccigerinae.

It can safely be said that not since the work of Odhner (191 la) has a revision of the family been
solidly based on a study of specimens of the genera involved. As can be seen from the brief
summary presented above, much juggling of the subfamilies has occurred during the history of
this family. This is not the place to attempt a complete reclassification of the group, being beyond
the scope of this series. A detailed study of those forms from the northeast Atlantic region has,
however, been undertaken. The following five subfamilies are, therefore, considered here: the
Fellodistominae, the Baccigerinae, the Xenoperinae Poche, 1926, the Tergestiinae and the
Monascinae.

Some comments on forms previously treated as fellodistomids

Certain genera, which at one time or another have been considered as fellodistomids, have been
omitted from this study. Ancylocoelium Nicoll, 1912, was originally included in the Haplocladinae;
but Nicoll indicated that its relationships are not straightforward and that its designation was
provisional. In our opinion, its most satisfactory position is in the Monorchiidae Odhner, 1911.
A close comparison of this genus with Chrisomon tropicus (Manter, 1940) as described by
Kovaleva (1970#) from the same host may reveal that Nicoll (1912) misinterpreted certain
features.

The Stenakrinae was included in the Fellodistomidae by Yamaguti (1970, 1971); but an
examination of members of this subfamily has led us to believe that it is best placed in the
Opecoelidae Ozaki, 1925. Stenakron Stafford, 1904, was not well known until studied by the
Russian workers Strelkov (1960) and Mamaev, Parukhin & Baeva (1963). Prior to the latter
work this genus was always considered an allocreadiid, but Mamaev et al. placed it in the
Fellodistomidae. It has been confused with Rhodotrema Odhner, 1911, particularly by Russian
workers: this confusion is discussed below (p. 245). Having studied Stenakron vetustum Stafford,
1904, S. kerguelense Prudhoe & Bray, 1973, Anisorchis opisthorchis Polyanski, 1955, and Caudo-
testis nicolli Issaitschikov, 1928, Bray (1979) was able to point out that the members of this
subfamily differ from the Fellodistomidae in the following characters: I-shaped excretory vesicle;
long claviform cirrus-sac containing single large sac-like seminal vesicle; small narrow pars
prostatica; and long ejaculatory duct. Other normal characters of the Stenakrinae are unusual
in the Fellodistomidae: these include large eggs, a pre-testicular uterus and vitelline fields con-
fluent in the median line. It is also possible that the cercariae of Stenakron are microcercous and
occur in gastropods, as in the case of the opecoelids; but the cercariae apparently lack a stylet
(Chubrik, 1966). Gaevskaja & Kovaleva (1977) follow Prudhoe & Bray (1973) in including
Stenakron in the Opecoelidae.

202 R- A. BRAY & D. I. GIBSON

Another putative member of the Fellodistomidae is Yamagutia anarhichae Brinkmann, 1956,
from the gall-bladder of Anarhichas minor from Iceland. This taxon is based upon a single well-
flattened specimen, which, through the kindness of Dr F. Gudmundsson of the Museum of
Natural History, Reykjavik, we have been able to examine. It appears to us to be a teratological
specimen of the zoogonid Dere trema pycnorganum (Rees, 1953), which occurs commonly in the
gall-bladder of this host off Iceland. A comparison with specimens of D. pycnorganum in the
collections of the British Museum (Natural History) indicates that there is a great similarity in the
alimentary system and the terminal genitalia. In addition, some of our specimens, especially the
smaller ones, exhibit no tegumental spines, which, according to Brinkmann (1956), are also
missing in Y. anarhichae. The 'opening of Laurer's canal', as described by Brinkmann, appears
to be, in fact, the excretory pore which has been displaced onto the dorsal surface by pressure
during fixation. We have no doubt, therefore, in considering Yamagutia anarhichae (the type-
species of the genus Prolateroporus Yamaguti, 1971) to be a synonym of D. pycnorganum.

Aspects of biology

The fellodistomids have small eggs and, according to Cable (1974), miracidia with cilia confined
to patches. The sporocysts develop in bivalve molluscs and the daughters normally give rise to
motile cercariae. These cercariae are basically furcocercous, the excretory vesicle opening
terminally on the furcae, if the latter are present. The tail, however, may be trichocercous and
long, or much reduced and deciduous (see Cable, 1954). The metacercariae occur in amphipods,
ctenophores and echinoderms, but in some instances the life-cycle may be telescoped, with the
adults occurring in a gastropod or with the complete cycle occurring in the lamellibranch. Whilst
this telescoping may occur as a general pattern in some species, other species occasionally occur as
progenetic metacercariae in the second intermediate host. Apart from these examples, adult
fellodistomids are found in the intestine or neighbouring organs of marine, and occasionally
freshwater, teleosts. Some species exhibit a fairly rigid host-specificity, whilst others appear to be
specific in part of their range and less specific in other parts. A common pattern of specificity is for
the parasite to show a strong preference for a particular host species or group of species, but to
occur occasionally, sometimes in an immature condition, in other hosts. The group is widespread
in sub-polar, temperate and tropical waters and in both the shallows and the deeps.

Copulation is apparently reciprocal and involves the juxtaposition of the genital atria. In some
species at least it involves the transfer of spermatozoa encased in fibrous spermatophores.

Key to the subfamilies of the Fellodistomidae in the northeast Atlantic

1 Caecum single (? unites with excretory vesicle) .... MONASCINAE (p. 268)
Caeca double ...........

2 Ring of conical papillae surrounding oral sucker; ridges on lateral surface of forebody; cirrus-sac

bipartite; pharynx elongate TERGESTHNAE (p. 274)

Body-surface lacking ornamentation; cirrus-sac oval to globular; pharynx oval . . .3

3 Genital atrium long, narrow; seminal vesicle coiled, tubular; cirrus-sac surmounted by muscular

papilla XENOPERINAE (p. 259)

Genital atrium short to long; seminal vesicle bipartite ... . . . 4

4 Seminal vesicle almost fills cirrus-sac; genital atrium long, narrow; vitellarium in 2, or a few,

compact bunches of follicles; Laurer's canal opening at or near posterior extremity; canalicular
seminal receptacle present; ovary post- or inter-testicular . . BACCIGERINAE (p. 248)
Seminal vesicle small relative to size of cirrus-sac; genital atrium short, wide; vitellarium 2
(occasionally 4) lateral follicular fields; Laurer's canal opening dorsal to gonads; uterine
seminal receptacle present; ovary pre-testicular . . . FELLODISTOMINAE (p. 202)

Subfamily FELLODISTOMINAE Nicoll, 1909

Steringophorinae Odhner, 1911.
Lissolomatinae Skrjabin & Koval, 1957.
Markevitschiellinae Skrjabin & Koval, 1957.

THE FELLODISTOMIDAE 203

DIAGNOSTIC FEATURES. Body large and robust to small. Body-surface smooth. Ventral sucker
usually larger than oral sucker, occasionally smaller; in middle of body or in anterior half.
Pharynx globular to oval. Oesophagus absent, short or long. Caeca narrow to wide; reaching to
testes or beyond and almost to posterior extremity. Testes in anterior hindbody; post-ovarian.
Cirrus-sac well developed; oval to claviform; containing many prostatic gland-cells. Pars
prostatica wide; straight or gently curved. Ejaculatory duct wide; pocketed. Spermatophores
occasionally seen. Genital atrium small. Genital pore in mid-forebody; sinistral to median line.
Ovary entire to multilobate; usually anterior to right testis. Laurer's canal and uterine seminal
receptacle present. Uterus usually reaching posteriorly to testes, but does not extend into post-
testicular region in type-genus. Metraterm joins genital atrium from left. Eggs numerous;
operculate; small; often variable; egg-shell may be ornamented. Vitellarium follicular; in two
(occasionally four) lateral fields; in fore- and/or hindbody. Excretory pore terminal; vesicle 'V-
or 'Y'-shaped; with arms reaching to level of pharynx or oral sucker. Parasitic in intestine,
pyloric caeca, bile duct and gall-bladder of marine teleosts.

COMMENT. The Steringophorinae is retained as a distinct subfamily by Dollfus (1952) on the
basis of the distribution of the uterus; but the present study indicates that the type-species of
Fellodistomum Stafford, 1904, and Steringophorus Odhner, 1905, are similar, and that there is a
form, Steringophorus agnotus (Nicoll, 1909), which is in some ways intermediate. We, therefore,
consider it unnecessary to distinguish the Steringophorinae. The other synonyms listed, the
Lissolomatinae and the Markevitschiellinae, are not well known and are not represented in our
region ; but it appears that Lissoloma Manter, 1934, is close to Steringophorus and Markevitschiella
Skrjabin & Koval, 1957, is close to Steringotrema Odhner, 1911.

This subfamily, as represented below, is a fairly homogeneous group, and it is noticeable, for
example, how similar the cirrus-sac and contents are in virtually all of the species described. The
typical arrangement is well-figured and described by Nicoll (1909a, p. 466) and Odhner (191 la,
p. 101). The terminology of the distal part of the male-duct is rather confusing, for as the ejacula-
tory duct is muscular with irregular, shallow diverticula and muscular lobations, it has been
considered to be the genital atrium or sinus by some authors. These diverticula have also been
termed the 'atrial diverticle' by Yamaguti (1940) in Pseudosteringophorus and 'atrial sac' by
Manter (1947) and Armstrong (1974) in Megalomyzon and a number of fellodistomines, respec-
tively. Nicoll (1909#), for example, states, 'at first sight the genital sinus appears to be of great size,
but this is due to a wide expansion of the ductus ejaculatorius. In reality the genital sinus is
comparatively small'. Our observations agree with those of Nicoll: the ejaculatory duct in
relaxed specimens, we found, is always included within the cirrus-sac. The pars prostatica is a
wide, straight or slightly curved vesicular structure and is lined with numerous narrow filaments
which often extend into the lumen of the ejaculatory duct. These filaments appear to be involved
in the formation of a fibrous spermatophore. The seminal vesicle is normally bipartite, but one
or both parts may be reduced when the spermatozoa have been ejected.

The ovary in this subfamily appears in two basic forms. Fellodistomum and Steringophorus
each has a distinctive acinous, multilobate ovary, while in the other genera the ovary has a
smooth surface, having either a more or less globular or a trilobed shape.

A comment should be made on the unreliability of egg-measurements as taxonomic criteria in
this group. As can be seen from Steringophorus furciger and Steringotrema pagelli (Tables 2 & 7)
in particular, the egg-length can vary considerably, and it is apparent that in these two species at
least the literature suggests that there may be an overall bimodal distribution of egg-sizes. The
latter phenomenon may, however, be the result of authors measuring eggs at different stages of
development.

Key to the genera of the Fellodistominae from the northeast Atlantic

1 Uterus reaching posteriorly to testes; hindbody at least as long as forebody
Uterus not reaching posteriorly to testes; body almost globular; ventral sucker large

FELLODISTOMUM Stafford, 1904 (p. 204)

2 Excretory vesicle 'Y'-shaped .......

Excretory vesicle 'V'-shaped ........- 4

204 R. A. BRAY & D. I. GIBSON

3 Ovary multilobate; caeca reach to level of testes or, more usually, beyond; posterior limit of

vitelline follicles lies posterior to anterior margin of ventral sucker; prepharynx small,

indistinct STERINGOPHORUS Odhner, 1905 (p. 210)

Ovary trilobed; caeca reach to about level of anterior margin of testes; vitelline follicles entirely
in forebody; prepharynx small, distinct . . . OLSSONIUM gen. nov. (p. 228)

4 Body stout; vitellarium extending into forebody and posteriorly as far as testes, mainly in lateral

fields; caeca reaching to region of testes . " . STERINGOTREMA Odhner, 191 1 (p. 231)

Body elongate; vitellarium confined to hindbody, reaching well posteriorly to testes; caeca extend
almost to posterior extremity PRUDHOEUS gen. nov. (p. 245)

Genus FELLOD1STOMUM Stafford, 1904

DIAGNOSTIC FEATURES. Body large; robust. Ventral sucker large; larger than oral sucker; globular;
at middle of body or just posterior to it. Prepharynx short. Pharynx well developed; globular.
Oesophagus absent. Caeca wide; reaching close to posterior margin of testes. Testes two;
symmetrical; oval; close to posterior extremity. Cirrus-sac claviform. Seminal vesicle bipartite.
Pars prostatica wide; surrounded by gland-cells. Ejaculatory duct wide; diverticulate. Genital
atrium small. Genital pore close to ventral sucker; sinistrally submedian. Spermatophores may be
present. Ovary multilobate; just anterior to right testis. Uterus not extending posteriorly to testes
or small portion only posterior to testes [especially in flattened specimens]; bulk of uterus dorsal
and lateral to ventral sucker. Eggs numerous; small; with smooth shells. Vitelline follicles in two
lateral fields; lateral to and reaching just anteriorly to ventral sucker. Excretory vesicle 'Y'-shaped;
stem reaching forward to anterior margin of testes; arms reaching to pharynx. Parasitic in gall-
bladder of marine teleosts (Anarhichadidae).

TYPE-SPECIES. Fellodistomumfellis (Olsson, 1868) (by subsequent designation: Nicoll, 19090: 471).

COMMENT. The present concept of the genus Fellodistomum includes only the large, almost
globular type-species in which little or none of the uterus lies posteriorly to the testes. Our
reasons for distinguishing Steringophorus are discussed below (p. 210). Eighteen species have, at
one time or another, been assigned to this genus. These are:

(1) F. agnotum Nicoll, 1909, which is herein considered to be a species of Steringophorus (see
p. 217).

(2) F. anarhichaelupi (Rathke, 1799) Dollfus, 1968; Dollfus (1968) used this name as a senior
synonym of F. incisum (Rudolphi, 1809). We do not consider that either Rathke's (1799) or
Rudolphi's (1809) descriptions are recognizable as species of Fellodistomum (see p. 209).

(3) F. breve Ching, 1960 (emend. Yamaguti, 1971), herein considered a species of Steringophorus
(see p. 211).

(4) F.fellis (Olsson, 1868) Nicoll, 1909; type and only valid member of the genus.

(5) F.furcigerum (Olsson, 1868) Yamaguti, 1953; type-species of the genus Steringophorus (see
p. 212).

(6) F. incisum (Rudolphi, 1809) Stafford, 1904; unrecognizable (see p. 209).

(7) F. lethrini (Gupta, 1956) Yamaguti, 1971; recorded from Lethrinus sp. from the Gulf of
Manaar off India, this species was originally placed in Steringophorus. Yamaguti (1971),
in addition to including it in Fellodistomum (p. 70), made it the type-species of a new genus,
Guptatrema, which he placed in the Callodistomidae Odhner, 1910 (p. 118). Although
certain marine callodistomids, e.g. Callodistomoides Yamaguti, 1970 (see p. 211), are
difficult to distinguish from fellodistomids, Guptatrema appears to have characters not
usually associated with the latter group, such as a long, narrow cirrus-sac, a large seminal
receptacle and caeca reaching to the posterior extremity. Its position is, therefore, uncertain ;
but it resembles, superficially at least, Paracryptogonimus ovatus Yamaguti, 1952, from an
unknown marine fish from off the Celebes (Yamaguti, 1952) and Pornadasys hasta from the
Bay of Bengal (Madhavi, 1976) and Paracryptogonimus rostratus Nagaty & Abdel Aal, 1961,
from Lethrinus rostratus in the Red Sea, species which are considered synonyms by Manter
(1963).

THE FELLODISTOMIDAE

205

(8) F. magnum (Manter, 1934) Yamaguti, 1953, herein considered a species of Steringophorus
(seep. 211).

(9) F. mendezi Sogandares-Bernal, 1955, from the intestine of a freshwater fish, Brachyrhaphis
episcopi in Gatun Lake, Panama. The uterus of this worm contains an average of only 10
eggs, and it is figured as possessing a considerable proportion of its uterus in the post-
testicular field. As it is described mainly from a single specimen, its status is in some doubt.

(10) F. melanostigmum Noble & Orias, 1975, herein considered a species of the genus Steringo-
phorus (see p. 211).

(11) F. ovatum (Price, 1934) Yamaguti, 1971; this is possibly a species of the genus Bacciger
(see p. 249).

(12) F. phrissovum Aldrich, 1961, herein considered a species of the genus Steringotrema (see
p. 232).

(13) F. preovaricum Caballero y C., Bravo Hollis & Grocott, 1952, from the intestine of Galeich-
thys seemanni off the Pacific coast of Panama. This species was made the type of a new
genus, Allofellodistomum, by Yamaguti (1971). The location of the ovary in the forebody
appears to justify this action, although the spinose body-surface suggests that it may not be
a fellodistomid.

(14) F. profundum (Manter, 1934) Yamaguti, 1953, herein considered to be a species of the genus
Steringophorus (see p. 211).

(15) F. rotundum (Manter, 1954) Yamaguti, 1971, herein considered to be a member of the
genus Steringotrema (see p. 232).

(16) F. sebastodis Yamaguti & Matumura, 1942, herein considered to be a species of the genus
Steringophorus (see p. 211).

(17) F. saviniense Dyk & Dykova, 1964, from the gall-bladder of Salmo trutta and Thymallus
thymallus in the River Savinja, Yugoslavia. It seems likely that this species belongs to the
zoogonid genus Pseudochetosoma Dollfus, 1951, which contains species from the gall-bladder
of freshwater fishes in central Europe (Kakacheva-Avramova, 1966a, 19666).

(18) F. thapari Srivastava & Ghosh, 1968; a synonym of F.fellis.

Fellodistomum fellis (Olsson, 1868) Nicoll, 1909

Distoma fellis Olsson, 1868.

Distoma incisum Rudolph! of van Beneden (1871).

Fellodistomum incisum (Rudolph!) of Stafford (1904) in part (?).

(?) Cercaria limae Nicoll & Small, 1909.

(?) Metacercaria limae (Nicoll & Small) James, Sannia & Bowers, 1977.

Adolescaria ophiurae Tauson, 1917.

Fellodistomum thapari Srivastava & Ghosh, 1968.

TYPE-HOST AND LOCALITY. Anarhichas lupus, Varberg, Sweden.

RECORDS

(i) Material studied

(a) From the NE Atlantic

Anarhichas lupus [gall-bladder] Iceland (see Rees, 1953: 15). BM(NH) 1976.4.8.179.

- [gall-bladder] Herdla, Norway (July, 1971). Material of J. Thulin.

- [gall-bladder] Eldey Bank, off Reykjanes, Iceland (64N, 24W; depth 148-152 m; May,
1974). BM(NH) 1977.3.1.1-20.

- [gall-bladder] Orkney Islands, Scotland (59N, 04W; depth 164-172 m; July, 1976).
BM(NH) 1977.3.1.21-40.

- [gall-bladder] Moray Firth, Scotland (58 N, 02W; depth 164m; July, 1976). BM(NH)
1977.3.1.21-40.

(b) From elsewhere

Anarhichas denticulatus [gall-bladder] Newfoundland (see Bray, 1979 : 419). BM(NH) 1977.2.15.
202-203.

206 R. A. BRAY & D. I. GIBSON

Anarhichas lupus [gall-bladder] Locality unknown; donated by Zoological Museum, Amsterdam.
BM(NH) 1946.12.20.1-20.

- [gall-bladder] Newfoundland (see Bray, 1979 : 419). BM(NH) 1977.2.15.186-200.
Anarhichas minor [gall-bladder] Newfoundland (see Bray, 1979 : 419). BM(NH) 1977.2.15.201.

(ii) NE Atlantic records from the literature

Anarhichas lupus [gall-bladder & intestine] Varberg, Sweden (July, Aug.). Olsson (1868 : 44;
as Distoma fell is).

- [gall-bladder & intestine] Belgian coast, van Beneden (1871 : 48; as Distoma incisum);

- [gall-bladder] North Sea (Feb., 1893). Jacoby (1899 : 12; as Distomum fellis).

- [gall-bladder] Northumberland, England. Lebour (19080 : 36; as Distomum fellis).

- [gall-bladder] St Andrews, Fife, Scotland. Nicoll (19090 : 458).

- [gall-bladder] East coast of Iceland (66N, 13 W; depth 119m; Aug. 1948). Rees( 1953 : 15).

- [gall-bladder] Reykjavik, Neskaupstadur & Hiisavik, Iceland (June, July 1955). Brinkmann
(1956 : 20).

Anarhichas minor [gall-bladder] Neskaupstadur & Husavik, Iceland (July, 1955). Brinkmann
(1956 : 20).

- [gall-bladder] Umivik, East Greenland (July, 1959). Brinkmann (1975 : 51).

ASPECTS OF BIOLOGY. Records of intermediate hosts in the NE Atlantic:

(a) First intermediate host
Nil.

(b) Second intermediate host

Ophiura albida [stomach] Kristineberg, Sweden. Mortensen (1920 : 67; identification by T.
Odhner).

[1]Lima hians [mantle] Millport, Firth of Clyde, Scotland. Nicoll & Small (1909 :241; as
Cercaria limae).

[?] Macoma baltica [?] Burry Inlet, South Wales. James, Sannia & Bowers (1977 : 13; as Mela-
cere aria limae).

The life-history of this species has been studied by Chubrik (1952, 1966) in the Barents Sea. This
author states that the sporocysts develop in the bivalve Nucula tennis and the daughters contain
7-10 cercariae. These cercariae are non-oculate, brevifurcate and distomatous, possessing an
almost 'V'-shaped excretory vesicle with its base forming a reservoir. The tail is broad and flat.
The metacercariae are found in the digestive tract of echinoderms of the genus Ophiura and
progenesis occasionally occurs. Adolescaria ophiurae, described by Tauson (1917) from Ophiura
sarsi in the Barents Sea, is so similar to the Fellodistomum metacercariae described by Chubrik
that there can be little doubt of its identity with F. fellis. A similar worm, Cercaria limae, was
described from the gaping file-shell bivalve, Lima hians, by Nicoll & Small (1909). Single encysted
specimens were found loosely attached to the underside of the mantle-edge, projecting into the
mantle-cavity of two file-shells. Nicoll & Small considered that the evidence strongly suggested
that they were F. fellis; but that they could also belong to a species of Steringophorus, particularly
as Anarchichas lupus was rare in their area. This same worm has also been found recently by
James et al. (1977) in the Baltic tellin Macoma baltica from south Wales.

This species normally occurs in the gall-bladder of anarhichadid fishes, sometimes in great
numbers (over 4000, according to Polyanski, 1955); but it is occasionally found in the intestine
or urinary bladder (Olsson, 1868; van Beneden, 1871; Polyanski, 1955). Inaddition, Polyanski
(1955) found immature specimens in the intestine and gall-bladder of the plaice Pleuronectes
platessa, and Srivastava & Ghosh (1968) record their Fellodistomum thapari from the gall-bladder
of the anglerfish Lophius piscatorius and from the intestine of Anarhichas lupus. A study of the
literature indicates that F. fellis is restricted to the northern Atlantic Ocean and the Barents Sea.

We found two specimens which are apparently in the act of copulation (Fig. Id). The genital
pores are juxtaposed with the forebodies of each worm at an acute angle to each other. When
separated, structures interpreted as being spermatophores were seen, apparently having been
expelled from the genital pore of both worms, lying close to these pores. In sections, the fibrous

THE FELLODISTOMIDAE

207

coat of the spermatophore can be traced back to the cells lining the pars prostatica (Fig. 2). The
bipartite seminal vesicle in one of this pair which was sectioned contained very few spermatozoa.
It seems likely, therefore, that sperm-transfer takes place via the metraterm and not via Laurer's
canal.

PREVIOUS DESCRIPTIONS. Olsson (1868 :44; as Di stoma fellis) ; Jacoby (1899 : 12; as Distomum
fellis); Nicoll( 19090 : 458); Miller (1941 : 43); Rees(1953 : 15); Polyanski(1955 : 20); Srivastava
& Ghosh (1968 : 46; as F. thapari).

Fig. 1 Fellodistomum fellis: (a) slightly flattened whole mount; (b) cirrus-sac; (c) diagram of female
proximal genitalia; (d) two worms in copulation (free-hand sketch); (e) lateral view of fixed but
unflattened worm (free-hand sketch). Bar scales: a= 1 mm; b = 0-25 mm.

DESCRIPTION (Figs 1 & 2). This description is based upon 38 whole-mounted and three serially
sectioned specimens which represent a wide range of sizes and states of maturity. The dimensions
of these worms, in addition to some measurements from the literature, are included in Table 1.

The adult worms are stout (Fig. le) and are almost subglobular with an anterior protuberance
carrying the oral sucker in addition to a small posterior protuberance. The body-wall is thin and
unarmed. The subglobular oral sucker opens subterminally, and the large, globular ventral sucker
is deeply embedded in the middle of the body (Fig. la). The ventral sucker is invariably much
broader than the oral sucker, in the ratio of 1 : 1-5-3-0, with a distinct tendency for the ratio in

208 R. A. BRAY & D. I. GIBSON

Table 1 Dimensions of Fellodistomum fellis from the present material and from the literature

Authority

Jacoby ( 1 899)

Nicoll (1909)

Rees (1953)

Present material

Name used

Distomum

Fellodistomum

Fellodistomum

Fellodistomum fellis

fellis

fellis

fellis

Hosts

Anarhichas

Anarhichas

Anarhichas

see text

lupus

lupus

lupus

Locality

North Sea

Scotland

Iceland

see text

Length (mm)

2-2-5

2-5-3-3

2-5-3-3*

0-97-3-7*

Breadth (mm)

1

1-1-1-6

1-1-1-8

0-38-2-44

Length : forebody

(1 :0-41)

-

-

1 : 0-23-0-48

ratio

Oral sucker (mm)

0-36 dia.

0-40-0-45 dia.

0-40-0-45

0-1 8-0-56x0-22-0-67

Ventral sucker (mm)

0-88 dia.

0-9-1-0 dia.

0-90-1-50

0-34-1 -62x0-33-1 -84

Sucker-ratio

(1 : 1-9)

-

1 : 2-25-3

1 : 1-5-3-0

Pharynx (mm)

0-22 dia.

0-16-0-21 long

0-18-0-21

0-1 3-0-27x0- 13-0-28

Cirrus-sac (mm)

0-29x0-23

-

-

0-92xO-39f

Testes (mm)

-

0-45x0-19

-

0-1 0-0-80x0-09-0-60

Ovary (mm)

-

-

-

0-42xO-22f

Eggs (urn)

40x20

42x23

42-43x23-24

37-48x19-29

* Some measurements from flattened specimens.
t Measured from sections.

the larger worms to be greater than in the smaller. The oral sucker is connected by a short
prepharynx (visible only in sections) to a globular pharynx. There is no oesophagus as the intestine
divides immediately on leaving the pharynx and forms two wide, but thin-walled, caeca, which are
almost invariably filled with the semi-digested remains of blood-cells. This opaque mass often
obscures much of the internal details of the worm in whole-mounts. The caeca, which are appar-
ently lined with squamous epithelium, run in the dorso-lateral fields and terminate blindly at the
level of the testes.

The excretory pore lies terminally and leads via a short, narrow duct into the wide stem of the
excretory vesicle, which reaches to the anterior margin of the testes. At this point narrow arms
arise which pass dorso-laterally to the ventral sucker and expand slightly in the forebody. They
terminate laterally at the level of the pharynx.

The irregular but more or less oval testes lie symmetrically close to the posterior extremity.
The two vasa efferentia unite to form the seminal vesicle immediately after passing through the
base of the thin-walled cirrus-sac. The latter structure lies just anteriorly to the ventral sucker.
As it lies more or less perpendicularly to the ventral surface, the measurement for it given in
Table 1 is taken from longitudinal sections. It is claviform in shape, narrowing proximally, and
contains, in addition to a bipartite seminal vesicle, a wide pars prostatica and a wide, irregular
ejaculatory duct (Fig. Ib). The proximal part of the seminal vesicle may be slightly coiled, and
occasionally one or both parts may be reduced to a narrow duct, especially when a spermatophore
is present. The wider distal part of the cirrus-sac contains the dilate pars prostatica which is
surrounded by a dense mass of gland-cells and lined by numerous filamentous projections. These
projections extend forward into the ejaculatory duct through a narrow opening and appear to be
involved in the formation of the wall of the spermatophore (Fig. 2). The ejaculatory duct is large
and may be mistaken for a genital atrium. Its wall is strongly muscular and forms numerous
irregular pockets and lobes. The genital atrium itself is small and opens, via the genital pore, in
the posterior region of the forebody just to the left of the median line.

The multilobate ovary lies close to the anterior margin of the right testis. From it a narrow
oviduct passes dorsally, leading to a diffuse Mehlis' gland. Laurer's canal, which opens dorsally
at about the level of the ovary, and a minute common vitelline duct open into the oviduct just
prior to its entry into Mehlis' gland (Fig. Ic). The initial coils of the uterus form a uterine seminal
receptacle, there being no other seminal storage organ in the female part of the reproductive

THE FELLODISTOMIDAE

209

4,

Fig. 2 Photomicrograph of section showing spermatophore production in Fellodistomum fellis.

Bar scale: =0-1 mm.

system. The uterine coils extend posteriorly between, or fractionally beyond, the testes; but the
great bulk of the uterus occurs between the testes and the cirrus-sac, especially around the ventral
sucker. The distal extremity of the uterus, which is thin-walled and wide, enters the genital atrium
posteriorly through a muscular sphincter (Fig. Ib). The numerous eggs which fill the uterus are
operculate and possess a smooth shell. The vitellarium consists of a number of irregular follicles
of varying sizes, which are packed tightly in lateral fields between the anterior margins of the
testes and the anterior margin of the ventral sucker or the cirrus-sac. There are two main lateral
vitelline ducts at the level of the ovary which pass medially and unite to form the short common
vitelline duct immediately prior to uniting with the oviduct.

DISCUSSION. Distomum fellis of Olsson (1868) is the earliest name that can unequivocally be used
for this species. Older names, such as Distoma anarhichaelupi Rathke, 1799, and D. incisum
Rudolphi, 1809, from the stomach of Anarhichas lupus, cannot be compared as the descriptions
are totally inadequate. In addition, these names have fallen out of use since Nicoll (1909), who
discussed their validity at some length, came to the same conclusion. The figure of Distoma
incisum produced by van Beneden (1871) strongly suggests that he was dealing with F. fellis: we
have, therefore, included this as a synonym. The genus Fellodistomum was initially erected by
Stafford (1904) based on his very brief description of 'F. incisum'; but his specimens were re-
described by Miller (1941), who came to the conclusion that they were F. fellis.

Srivastava & Ghosh (1968) described a new species, F. thapari, from the gall-bladder ofLophius
piscatorius and the' intestine of Anarhichas lupus from an unknown locality [neither of these two
fishes occur outside the North Atlantic region or, in the case of A. lupus, the Arctic Oce