Contributions to Zoology, 65 (3) 129-175 (1995)

SPB Academic Publishing bv, Amsterdam

A -eating , revisited (Cirripedia, Pyrgomatidae)

Arnold Ross & William A. Newman

Scripps Institution of Oceanography, La Jolla, California 92093-0202, USA *

Keywords: Cirripedia, Pyrgomatidae, Hoekia, coral-eating , new species, parasitism, functional

morphology

Abstract — est caractérisé par une coquille de forme irrégulière

nichée de manière cryptiqueentre les polypes du Corail hermaty-

The coral-eating barnacle Hoekia monticulariae (Gray, 1831), pique HydnophoraFischer, 1807, à distribution dans la majeure

the internal the Thoracica described l'Océan Indien only parasite among tothis partie de et du Pacifique de l’Ouest. A cause de

is characterized shell nestled des les n’ont day, by an irregularly-shaped cryp- l’aspect protéiforme animaux, cirripédologues pas

tically between the polyps of the hermatypic coral correctement apprécié leur diversité, Hoekia étant considéré

Fischer, 1807, which occurs throughout most of the Indo-West comme genre monotypique. Sont décrites ici 7 nouvelles espèces

Pacific. Para- Because of its protean form, cirripedologistshave failed reparties entre Hoekia et 3 nouveaux genres (Eohoekia,

to the of taxa related to et tribu Hoekiini appreciate diversity Hoekia,, a presumed hoekia Ahoekia), la étant proposée pour ce

monotypic genus. We describe seven new species divided be- complexe. Comme c’est le cas dans d’autres Pyrgomatides, l’en-

tween and three Hoekia new genera, Eohoekia, Parahoekia, and croûtement par le Corail est inhibé le long du bord de la paroi

Ahoekia for which the Tribe Hoekiini is proposed. As in other et au niveau de l’aperture.Mais, dans les Hoekiini,paroi et aper-

calcareous the coral is inhibited le pyrgomatids, overgrowth by ture sont recouverts par un pseudopolype au dépens duquel

around the edgeof the wall and aperture. But in Hoekiini a pseu- Cirripède se nourrit à l’aide de trophi modifiés. De plus, leur

dopolyp, upon which the barnacle feeds with modified trophi, paroi est suspendue dans le tissu du Corail — et non articulée à

coversthe wall and aperture. Furthermore, rather than articulat- une base calcaire; le bord latéral (basal) hypertrophié de cette

ing with acalcareous basis, the wall is suspended in coral tissue. paroi est l’endroit au niveau duquelon suppose que des activités

Its hypertrophied lateral margin (= basal margin), in contact métaboliques se déroulent. Chez Hoekia et Ahoekia on trouve

with the host’s tissue, is the site where metabolic activities are in- dans la paroi des tubes simples ou communiquants, qui aboutis-

In ferred to take place. Hoekia and Ahoekia, the wall develops sent à des orifices dans le bord, fonctionnant comme voies cir-

simple or connecting tubes that lead to openings in the margin, culatoires. Bord hypertrophié et système circulatoire élaboré

which serve A les Hoekiini soient uni- as circulatory pathways. hypertrophied margin suggèrent qu'il est possible que ne pas

and elaborated circulatory system suggests that the Hoekiini quement des consommateurs directs de tissu de l’hôte ou/et de

may not be wholly dependent on feeding directly on host tissue matériel du coelenteron, et qu'ils pourraient être aussi des para- and/or coelenteronic material,but may also be absorptive para- sites absorbtifs. Pour ce quiest d’autres Pyrgomatides (membres

sites. Although other pyrgomatids, in the tribes Pyrgopsellini des tribus Pyrgopsellini nov. et Pyrgomatini nov.), ils restent nov. and Pyrgomatini nov., exercise some control over their planctotrophes, bien qu'exerçant un certain contrôle sur leur hosts frill and discontinuities by an apertural through between hôtes par une collerette aperturaleet par des discontinuités entre the shell and basis, they are still planktotrophic. la coquilleet la base.

Résumé Introduction

Hoekia monticulariae is not Le de Corail Hoekia monticulariae (Gray, 1831) the Cirripède mangeur (Gray, only

1831) - jusqu'à présent unique endoparasite décrit parmi les most unique coral-inhabiting barnaclein the Pyrgo-

* A contribution of the Scripps Institution of Oceanography, new series. 130 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Table I. Species of Hoekiini, host coral and locality.

Hoekiini Host coral Locality

Eohoekia n. gen.

chaos Gulf of n. sp. Hydnophora sp. Aqaba

H. Gulf of Suez nyx n. sp. exesa(Pallas, 1766)

Parahoekia n. gen.

aster n. sp. H. microconos (Lamarck, 1816) New Caledonia

Hoekia Ross & Newman

fornix n. sp. H. exesa Obilatu Isl., Moluccas

monticulariae (Gray) H. exesa Singapore

mortenseni n. sp. H. exesa Mauritius

Ahoekia n. gen.

tanabensis n. sp. H. ? exesa1 , H. bonsai Veron, 1990 Tanabe Bay, Japan

H. of Java chuangi n. sp. rigida (Dana, 1846) Bay Batavia,

1 The host coral as listed on the label accompanying the type specimens was H. exesa. Present evidence indicates that Hydnophora is

H. as as represented in Japanesewaters by bonsai Veron, 1990 well H.¡ exesa, but we have only been able toconfirm that the A. tanabensis we examined occurred on the former (see text for further discussion).

matidae, it is the only known wholly parasitic Two statementsby Darwin (1854: 373) prompted balanomorph barnacle. Because of its peculiar a restudy of H. monticulariae. The first involves shell, Darwin (1854: 373) initially "... did not be- why "... the outer laminaof the shell in this one lieve thatit was a cirripede." He had only the shell species does not nearly reach the circumference of and therefore did not realize that theanimal within the walls ..." And the second, why "Internally ... had cirri fed tissue aberrant and directly on host the walls are ... perforated by many quite irregu- with its trophi (Ross & Newman, 1969: 255). In lar, small orifices ..." that "... admit threads of our earlier study we concentrated on the aberrant corium into certain irregular pores which penetrate cirri and biting trophi of preserved specimens, and the shell." Our answers, presented below, explain failed to sufficient pay attention to the hard parts. not only the functional aspects of these structures

have since discovered that of distant of We the shells but also reveal a previously unknown diversity populations have differentorganization plans, and genera and species. therefore Hoekia includes a number of taxa for A list of taxa, host and localities is provid- which we now propose three new genera and seven ed in Table I. Diagnoses for all of the species are new species of coral-eating barnacles. based wholly on characters of the wall, an approach

Since our earlier studies (Ross & Newman, 1969, necessitated by the rarity of complete specimens,

1973) there has been considerable interest in pyr- especially those preserved in alcohol. Also, the gomatid functional morphology and symbiosis characteristics of the wall laminaeutilizedare limit-

(Moyse, 1971; Newman & Ladd, 1974; Anderson, ed to those readily seen under simple magnifica-

1992; Anderson Characteristics 1978, & Southward, 1987; Cook et tion. of the genera and species are al., 1991). The implication in many of these studies distinguished in the diagnostic key to the Hoekiini is that pyrgomatids are able to manipulate their and the accompanying illustrations of principal

hand hosts, on one limiting deposition of coral anatomical features (Figs. 1-3). This general infor-

certain sclerosepta on areas of the shell surface mation and the accompanying systematics section,

chemical other dis- through interactions, and on the of interest mainly to specialists, is followedby a metabolically stimulating the growth of coral cussion of the functional morphology and host

all coenenchyme that forms or a portion of the bar- relationships of these specialized parasitic bar- nacles' diet (Ross & Newman, 1969: 255). nacles. Contributions to Zoology, 65 (3) - 1995 131

Fig. I. Principal anatomical features of the wall in Hoekiini: a, Schematic cross-section of the wall of a generalized,mature, coral-eating

barnacle viewed from the The outer of central right side. lamina (stippled) consists an elevated portion (peritreme) separated by a sulcus

from a slightly to highly reflexed or ascendant outer margin(left and right ends, respectively). The inner lamina (unstippled) forms a

marginal fringe which is subdivided by radial grooves and channels inferred to serve, or to have served as circulatory passages. These

join each other occasionally along their lengths, and at the circumferential channel beneath the distal margin of the outer lamina. The

inner and outer laminae are separatedby the predecessors of these radial grooves and channels. Those between the marginal fringe and

the circumferential channel in alacunate forms, and between both and the lacunae in lacunate forms, are functional (see Eohoek-

ia/Parahoekia and Hoekia/Ahoekia,respectively). Alternatively, the functionless portionof a radial channel, that central tothe circum-

ferential channel in alacunate forms or the lacunae in lacunate forms, is either filled or simply abandoned. The most obvious abandoned radial channel is that forming the carinal sinus. Ascendant marginalfringes can exceed the heightof the peritreme. The highly vascula- rized marginal and upper surfaces of the marginal fringe are in intimate contact with tissue of the host coral (see Fig. 3). b—e, Schematic cross-sections of the shell wall of Eohoekia, Parahoekia, Hoekia, and Ahoekia, respectively, viewed from the right side.

The and inner laminae and and and channels of the of the outer are stippled unstippled, respectively, grooves marginal fringe rostral

end are omitted. For the these and the distinctions between wall 134 and text, (left) key to genera types depicted here, see p. respectively. - 132 A. Ross & W.A. Newman A coral-eating barnacle, revisited

cut intact. Once the shell was removed, this small block was

againto reveal the relationship of the wall to the basis. We found

that the easiest way to trace pathways or openings on the inner

wall to the marginal fringe was by inserting an eyelash into the

opening, and observing where it exited. All of the specimens

were examined with transmitted light, or "candled", which

proved especially helpfulin elucidating the development of the

pathways within the wall.

Specimens chosen for scanning electron microscopy were

cleaned in a 5.25% solution of sodium hypochlorite, rinsed

several times in tap water and air dried. These were mounted on

with sil- aluminium stubs with double-sided tape, sputter-coated

ver, and viewed in a Cambridge S360 operating at lOkV.

Maximum parsimony analyses, including 50% majority rule

Branch and Bound consensus trees, and bootstrapping with the

search algorithm in the software package PAUP (version 3.1.1)

were developed on a Macintosh Quadra 840 running System 7.

Abbreviations of depositories

MNHN Muséum National d'Histoire Naturelle, Paris

NNM Nationaal Natuurhistorisch Museum, Leiden

SIO Scripps Institution of Oceanography, La Jolla, Ca.

USNM National Museum of Natural History, Washington DC

(formerly United States National Museum) Fig. 2. Principal anatomical features of the opercular plates in ZM Zoologisk Museum, Copenhagen Hoekia monticulariae: a, plan view in situ, with the outlines of ZMA Amsterdam the underlying sheath and orifice of the wall identified by the Zoölogisch Museum,

lightly dashed lines; b, side view of the right plate also viewed

from within. The heavier dashed line labelled “Division T and

S” indicates the approximate division between the fused scutal Systematics

and tergal portions of the plates. When the plates are closed (a),

the arches over the scutal and rostral portions of each plate, the Suborder Pilsbry, 1916 limbus occludens (b), meet to occlude the orifice of the wall as 1825 well as the aperture to the mantle cavity. The two stippled areas Family Pyrgomatidae Gray, represent openings between the plates when closed, the rostral Subfamily Pyrgomatinae Gray, 1825

and carinal fenestrae. As indicated, the insertion of the lateral

depressor is relatively discrete, and that for the tergal depressors

- included in this sub- Remarks. For alist of genera is limited to the region of the carinal tooth. See also Figs. 9a, b. Galkin family, see Newman & Ross (1976: 39),

(1986: 1290), and Anderson (1992: 335). The gener-

ic name Utinomia proposed by Galkin (1986: 1290) Materials and methods is preoccupied by UtinomiaTomlinson (1963: 264),

an acrothoracican barnacle. As areplacement name

The material available either less was dried, or commonly, with indicum An- we propose Galkinia, Pyrgoma preserved in alcohol. A few ofthe specimenswere examined dur- nandale, 1924 as the type species, by original desig- ing our previous studies, but most were acquired over the inter- nation. 25 A well of vening years. relatively preserved sample Hyd- Within the there three nophora exesa(Pallas), containing H. monticulariae from Sin- Pyrgomatinae are major

was especially useful in clarifying the relationships of mode of gapore, taxa, distinguished on shell characters and the well its barnacle toits host as as anatomy. Procedures for the feeding, which are readily separable from one of soft detailed in Newman preparation parts are & Ross (1971: each of these another. For lineages we propose 17). We found a small hand-heldgrinder with either an abrasive tribal names: Pyrgomatini, Pyrgopsellini, and or diamond impregnateddental wheel useful in cuttingthrough

the coral skeleton in order to remove the whole shell and basis Hoekiini, as defined below. Contributions to Zoology, 65 (3) - 1995 133

Tribe Pyrgomatini Gray, 1825 nom. transl.

Definition. — Shell wholly calcareous, wall of 4

plates or concrescent, regular in outline; orifice

large; lacking carinal ridge; wall interlocking with

calcareous basis; cirri and trophi normal, employed

in planktotrophic feeding; living in stony corals.

included 1817 The genera are: Pyrgoma Leach,

(type genus), Creusia Leach, 1817, Nobia Sowerby,

1839, Cantellius Ross & Newman, 1973, Hiroa

Ross & Newman, 1973, Galkinia Ross & Newman,

herein, Paranobia Galkin, 1986, Darwiniella An-

derson, 1992, Wanella Anderson, 1992, Arossella

Anderson, 1992, and Trevathana Anderson, 1992.

Tribe Pyrgopsellini n. tribe

Definition. - Shell only partly calcareous, walls al-

orifice ways concrescent, regular in outline; large;

lacking carinal ridge; wall attached to wholly mem- Fig. 3. Schematic optical cross-section of a lacunate coral-eating branous cirri and in barnacle viewed from the right side, with ashell (black) approxi- basis; trophi normal, employed mately 1 cm in diameter. Note that the outline of the host coral, planktotrophic feeding; living in sponges.

resembles Hydnophora spp. (CC, CT), superficially a simple Type genus: Pyrgopsella Zullo, 1967 (substitute polyp, but whether it is provided with a mouth (CM?) could not name for Pyrgopsis Gruvel, 1907, not Rochebrune, be determined in the material available. The barnacle’s inflated 1884). prosoma and reduced thorax reside in the confines of a close-

fitting mantle cavity (stippled) and the relatively voluminous

formed of the basis is filled Remarks. - is included in this space by growth cup-shaped largely Only Pyrgopsella with fibrous connective tissue (FCT). The oral cone with the tribe. In P. stellula Rosell, 1975, the opercular contained mouth parts (OC), when extended up through the plates are separate, the scutum transversely elon- orifice (O) of the shell, with the contained mouth parts, can cut lacks gated, and the triangular tergum a spur and rasp coral tissue, and the distended stomach (S) is common- (Rosell, 1975: 212), features that ally it with Wanel- ly filled with nematocysts. The shell (black) consists of the wall

la 1992 and tethered by attachment fibers (AF) to the calcareous portion of Anderson, Savignium Leach, 1825, the basis (BC). The margin of the wall is separated from the both of which are in the Pyrgomatini. Pyrgopsella coral the membranous of the basis the by portion (BM). Thus, radiation that has apparently represents an early wall is suspended in the surrounding tissue of the host (CC) by adapted to living in calcareous sponges, which have the hypertrophied inner lamina (LI). Furthermore, circulatory growth characteristics not greatly unlike corals. channels pass through the wall (see Fig. la) from the tissues within to the relatively isolated outer surfaces of the hyper- trophied inner lamina, presumably in connection with an ab- sorptive function. Tribe Hoekiini n. tribe

A?, anus (presence not confirmed by gross inspection); AF, attachment fibers; AM, adductor muscle; BC, calcareous por- Definition. - Shell calcareous, wall concrescent, ir- tion of basis; BM, membranous portion of basis; CC, coral in orifice with not regular outline; remarkably small; coenenchyme (soft tissue); CM?, coral mouth (presence con- firmed by gross inspection); CS, calcareous coral skeleton; CT, coral tentacle; C1, right cirrus I; C6, right cirrus VI; FCT, fi- brous connective tissue; FGA, female genitalaperture at base of ovary; OVD, oviduct; OVD/FGA?, connection ofOVD toFGA

C1; IPA, initial point of attachment; LI, lamina inner; LO, la- not observed; P, penis; S, stomach; T, right testis and seminal mina outer; O, orifice; OC, oral cone; OP, operculum; OV, receptacle. 134 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

inferior of basis carinal ridge; portion calcareous, fringe narrow, not ascendant; carinal and rostral ridge

I superior membranous; lateral margin of wall in prominent Parahoekia aster n. gen., n. sp.

- Outer lamina ovate, reflexed; marginal fringe wide, contact with tissue of host, not with basis; cirri reflexed; carinal ridge obscure, rostral ridge lacking aberrant; trophi modified for feeding on host ...... Eohoekiaa n. gen 7 tissue; living only in species of the stony coral 7. Marginalfringedistally anastomosingor filigreed; foramina

Hydnophora Fischer, 1807. in fringe developby coalescence of proximal elements; lacu-

absent oninner wall Eohoekia chaos n. Type genus: Hoekia Ross & Newman, 1973. nae sp.

- Marginal fringe distally serrate, not filigreed; foramina in

fringe absent; lacunae rarely present on inner wall Remarks. - We propose this new tribe to include ...... Eohoekia nyx n. sp. Hoekia and the three of new genera below. All the

shells with variable species possess outlines, i.e., no two individuals of the same species have a consis- Eohoekia n. gen. tent shell outline comparable to those found in other and in Pyrgomatidae, balanomorphs general - Definition. Shell thin, delicate in appearance;

(Spivey, 1988: 279). Although there to be appear carinal ridge obscure, rostral ridge lacking; distal in the of which consistent differences soft parts, we margin of outer lamina flat to slightly reflexed; have seen only those in three of the eight species, marginal fringe thin, simple, strongly reflexed, as there to be considerable variation in the appears high as or higher than orifice; foraminapresent in shell characters in the and employed following key, marginal fringe, rarely on inner wall of shell and therefore it remains tentative. then not connected to circumferential channel.

Eohoekia chaos Type species: n. sp.

to the and in the tribe Hoekiini Key genera species - Derived from the Greek Etymology. prefix eos-,

dawn or early, and -Hoekia, relative to the follow-

1. Inner surface with numerous lacunae connected directly to ing genera, especially Hoekia. circumferential channel; shell with thick, complex marginal

fringe 2

— Inner surface occasionally with few foramina occurring in Eohoekia chaos n. sp. marginal fringe, and a few blind "lacunae" on inner wall (Figs. 4, 18b) not connected directly to circumferential channel 6

2. Distal margin of outer lamina slightly ascendant, lower than

Hoekia 3 Material. - Twelve dried from peritreme i specimens Hydnophora sp.,

- Distal margin of outer lamina strongly ascendant, higher Elat, Gulf of Aqaba, Red Sea, Israel, L. Fishelson coll., 1971,

than peritreme Ahoekia n. gen 5 approx. 29°33'N, 34°57'E. Holotype: USNM no. 259978,

3. Orifice apical, round; carinal ridge lower than orifice ... 2 paratypes, same, no. 259979; 1 paratype, Nationaal

H. Natuurhistorisch Tel ...... fornix n. sp. Museum, Leiden, no. C-2537; 1 paratype,

- Orifice subapical, carinal than orifice 4 Aviv no. AR 1 ovate; ridge higher . University, Israel, 27533; paratype, Zoölogisch

4. Shell white, peritreme generally patelliform; carinal ridge Museum Amsterdam, no. ZMA Cirr. 201515; 6 paratypes,

low, broad, commonly obscure Scripps Institution of Oceanography, no. C-5983.

...... H. monticulariae (Gray, 1831)

- Shell pink to rose, peritreme subcircular, carinal ridge high, Diagnosis. - Shell low; marginal fringe thin, sim-

narrow, distinct H.} mortensenii n. sp. ple, exceedingly broad, higher than outer lamina, 5. Shell red, peritreme large, carinal ridge low and broad; commonly higher than orifice; distal elements of border of marginal fringe forming open network concomi-

tant with pronounced gap between outer lamina fringe separate, arborescent or filigreed; elements

A. tanabensis n. sp. of ...... fringe coalesce forming open foramina distal to

- Shell white, peritreme small, carinal ridge low and narrow; outer lamina. border of marginal fringeforming closed network concomi-

with indistinct between tant gap outer lamina

Description. - Shell white, translucent, thin, deli- ...... A. chuangi n. sp. 6. reflexed Outer lamina lobate, not or ascendant; marginal cate in appearance; orifice ovate, subapical on - Contributions to Zoology, 65 (3) 1995 135

4. view Fig. Eohoekia chaos n. sp.: a, apical view of holotype, USNM no. 259978, carinal end at lower right; b, apical of paratype,

USNM carinal end of and closed and some foramina; of no. 259979, at left, showing open filigree fringe some incipient c, portion

fringe at right in b, showing elements forming a foramen and the partially closed radial channel leading to the circumferential channel. 136 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Eohoekia view of USNM carinal end Fig. 5. nyx n. sp., apical holotype, no. 259980, at right.

rostral slope; shell surface verrucose to pustulose Sea; Y. Achituv leg., 20 June 1967, approx. 28°14'N, 33°38'E. without discernible growth ridges; marginal fringe Holotype: USNM no. 259980; 1 paratype, same, no. 259981; 1

paratype, Nationaal Natuurhistorisch Museum, Leiden, no. slightly to greatly reflexed; fringe supradecom- C-2538; 1 paratype, Tel Aviv University, Israel, no. AR 27534; pounded to anastomosing (Fig. 4b); sheath about 1 paratype, Zoölogisch Museum Amsterdam, no. ZMA Cirr. one-half of internal surface height wall, glossy; 201516. slightly roughened; lacunae lacking on inner wall.

Diagnosis. - Shell low; marginal fringe moderately Opercular plates thin, translucent, approximate- thick, distally higher than outer ly 3 times longer than high; growth ridges thin, narrow, serrate, lamina, commonly higher than orifice; distal ele- beaded; long axis strongly arcuate in section; ros-

ments of fringe confluent, or developing foramina tral tooth of scutum slightly elevated and greatly that fill secondarily. produced, inflexed and distally spatulate; tooth of right plate slightly broader than that of left; teeth Description. — Shell white, opaque to semitranslu- about twice as wide and long as tergal end; tergal cent, thick, somewhat delicate in appearance; cari- end distally spatulate. nal ridge prominent in juveniles, obscure in adults;

orifice ovate, subapical on rostral slope; shell sur- Etymology. - Derived from the Greek chaos, an al-

face granulate to pustulose; growth ridges discerni- lusion to utter formlessness, or state of things in of which ble along periphery outer lamina; fringe margin chance is supreme.

weakly branching, not strongly digitiform or

filigreed (Fig. 5); sheath about one-half height of Eohoekia nyx n. sp. wall, glossy; internal surface roughened, chalky; (Figs. 5, 16) blind filled lacunae or secondarily present at or near

periphery of inner lamina. Material. - Four dried specimens and several fragments from

Hydnophoraexesa (Pallas); El-Tur, Egypt, Gulf of Suez, Red Opercular plates thick, opaque, approximately Contributions to Zoology, 65 (3) - 1995 137

4.5 times longer than high; growth ridges thin, 166°30'E. Holotype: Nationaal Natuurhistorisch Museum,

Leiden,no. C-2539. Paratype: preserved specimen from H. beaded; long axis strongly arcuate in section; ros- one

microconos ; Kouare Reef, New Caledonia, 1 m, Georges Bar- tral tooth well elevated and greatly produced, in- gibant coll., 18 May 1994, 22°46.30'S, 166°47.40'E, Muséum flexed, terminally spatulate, slightly wider and National d'Histoire Naturelle, Paris, no. CI-2352. longer than tergal crest; crest thinner and about 1/2

of rostral - length tooth, terminally spatulate. Diagnosis. Shell high; peritreme prominent; mar-

ginal fringe moderately thick, slightly filigreed,

narrow, not reflexed or ascendant, lower than outer Remarks. - Our attempt to restore the dried speci- lamina and orifice. mens in a weak solution of tri-sodium phosphate

was unsuccessful. We did note that the opercular Description. - Shell moderately thick, white to plates, attached by the opercular membraneto the translucent; short carinal ridge as high as or higher sheath, had the adductor muscle bundle attached to than orifice; rostral ridge lower than orifice; orifice the medial edge of the rostral tooth. The adductor of ovate, subapical on rostral slope; distal portion extends from the middleof the fused plate almost outer lamina flat, not reflexed or ascendant (Fig. to the tips of the rostral tooth. Although there were 6b); short low ridges intercalated between carinal portions of the scutal lateral depressors, their and rostral ridges; growth ridges beaded, poorly de- length could not be determined. veloped; marginal fringe exceedingly narrow, wider

at extremities; fringe bifurcates and coalesces to

- Derived from Greek Etymology. mythology, form foraminain extremities; sheath glossy, about nyx, night, the offspring of Chaos. 1/2 height of innerlamina; wide bandbelow sheath

translucent and roughened; lower portion milky

white, similarly roughened.

Opercular plates twice as wide as high, thin, Parahoekia n. gen.

translucent; long axis arcuate in section; growth

ridges slightly beaded; bilaterally symmetrical and Definition. - Shell moderately thick, somewhat equal in size; scutal rostral tooth greatly produced, delicate in appearance; surface ornamented with comprising about 1/3 length of plate, slender, but carinal, rostral, and subsidiary lateral ridges; distal twice as wide as tergal end; tip of tooth inflexed; margin of outer lamina irregular, narrow, flat, not tergal end acicular, about 1/4 length of plate. reflexed; sulcus lacking between margin and peri- Crest of labrum deeply notched, devoid of setae treme; orifice higher than surrounding surface; and teeth; oral surface clothed with downward marginal fringe moderately thick, flat, foraminate; projecting ctenoid scales (Fig. 7a). Palps fused lacunae absent on inner wall of shell. basally to lateral margins of labrum; clothed with Type species: Parahoekia aster n. sp. long, simple setae. Mandible with leading edge

straight; divisible into three parts; superior 1/4 de-

Etymology. — Derived from the Greek prefixpara-, void of teeth, but armed with a few short, slender beside close and to denote their or by, -Hoekia, inferior divisible into two setae; 3/4 parts, upper relationship. armed with wide-spaced, strong, stout, simple or

bicuspid teeth; inferior angle armed with close-

spaced multicuspid teeth and short, slender setae

along inferior margin; cutting edge of whole Parahoekia aster n. sp. ap- pendage clothed with ctenoid scales. Maxilla I with (Figs. 6, 7) moderately short, uniform, simple, equally spaced

spines; lateral surface of appendage clothed with Material. - One dried specimen from Hydnophora microconos ctenoid scales. Maxilla II bilobed; lobe (Lamarck, 1816); Passe Dumbea, outside barrier reef, New upper

20-35 21/22 reduced to knob-like witheither few Caledonia, m, June 1976; approx. 22°10'S, protuberance a 138 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Fig. 6. Parahoekia aster n. sp.: a, apical view of holotype,NNM no. C-2539, carinal end at right; b, lateral view of same showing carinal, rostral, and subsidiary lateral ridges; c, enlarged view of fringefrom below, illustrating convolute channels leading to circumferential

channel; minute in border of outer lamina for that the shell. pits are passages setae traverse - Contributions to Zoology, 65 (3) 1995 139

Fig. 7. Parahoekia aster n. sp., mouthparts and appendages: a, labrum and palps; b, mandible; c, enlarged view of inferior angle of

maxilla mandible; d, maxilla I; e, enlarged view of apical portion of maxilla I; f, median languette and reduced right II; g, cirrus III; h, cirri IV—VI; i, distal end of penis.

or several short, slender, apical setae. Median lan- Cirri I-V distinctly aberrant; coxopodite of pro-

20 guette exceedingly large, laterally compressed; topodite of cirrus I broad, about times larger

of rami peripheral margin with minute, slender, simple se- than basipodite; segments distinct, armed tae; anterior edge straight; superior edge sulcate; with a few long, simple setae. Distinct hiatus be- posterior of appendage strongly protuberant (Fig. tween cirrus I and II; cirrus II biramous, segments

7f). indistinguishable; terminal setae short, slender. 140 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Cirrus IV biramous; rami equal in length, and equal the Cirripedia collected by expeditions of H.M.S. in length to basipodite of protopodite; about 3/4 "Challenger" and the Dutch Man-of-War "Sibo- length of coxopodite of cirrus V. Cirrus V shorter ga". For a recent expose of his early career, and than protopodite of cirrus VI; outer ramus shorter correspondence with Darwin and others, which led than first 3 segments of inner ramus; inner ramus to his work on the cirripeds of the "Challenger"

articulation Pieters Wintha- with short, simple setae at each along and "Siboga" expeditions, see &

Cirrus VI with rami shorter than who do mention greater curvature. gen (1990), not his incomparable protopodite; outer ramus shorter than inner, with studies on the cirripeds collected by the "Siboga".

13 segments, inner with 15 segments; intermediate

with of lesser segments one pair apical setae on cur- Hoekia fornix n. sp. vature; at articulation on greater curvature 2 long (Fig. 8) and 1 shorter, single setae. Intromittent organ an- nulaied, more than twice length of cirrus VI.

Material. - One dried specimen lacking opercular plates from Basidorsal point low, slender, apically acuminate. Hydnophora exesa, on reef of Poeloe Toesa east of Obilatu Isl.,

Moluccas, Indonesia; approx. 1°47.5'S, 126°59.5'E, Snellius

Remarks. - The shell the of preserved paratype, Exped. sta. 253a, 23-27 April 1930. Holotype: Nationaal

Natuurhistorisch Museum, Leiden, no. C-2540. which agrees in all particulars with the holotype, was chipped when removed from the host coral.

Diagnosis. - Shell suffused with pink; peritreme The outer lamina has a crystalline calcareous cover-

carinal ing, apparently laid down by the coral, that masks large; orifice apical, circular; ridge narrow,

lower than orifice. the growth lines. The subsidiary radial, but not the carinal and rostral ridges are poorly defined owing

- Shell delicate, somewhat to this overgrowth. Description. thin, translucent; peritreme highly domed, conical,

orifice Etymology. - Derived from the Latin aster, a star, growth ridges indistinct; apical, circular;

carinal ridge narrow, lower than orifice; distal edge in reference to the appearance of the outline of the shell. of outer lamina slightly thickened, slightly ascen-

dent, with distinct gap between it and fringe (Fig.

than 8a); marginal fringe higher outer lamina, nar-

but wider sheath less than 1/2 Hoekia Ross & Newman, 1973 row at extremities;

height of wall, hyaline, glossy; inner wall opaque,

lacunae circular to of - roughened; elongate-oval, Definition. Shell relatively thin, delicate in ap-

various sizes. pearance; carinal crest narrow, prominent; orifice apical or subapical; peritreme small to large; distal

Remarks. - The shell surface is encrusted cal- margin of outer lamina ascendent, low; marginal by

which have been fringe moderately thick, complex, ascendant deli- careous growths may or may not

the result of coral overgrowth (Fig. 8b). quescence lower than orifice; lacunae of innerwall separate, ovate to circular in outline, slightly ir-

Etymology. - From the Latin, fornix, arched or regularly arranged, connecting directly to circum- ferential channel. vaulted, in reference to the pronounced domeshape

of the Type species: Pyrgoma monticulariae Gray, peritreme.

1831, by original designation (Ross & Newman,

1973: 162). Hoekia monticulariae (Gray, 1831)

(Figs. 9, 10, 15, 17, 18a, 19)

Remarks. - We proposed Hoekia to honor the late

Dr. P.P.C. Hoek, a contemporaryof Charles Dar- Pyrgoma (Daracia) monticulariae Gray, 1831: 6; Darwin, 1854: in of his win, recognition monumental studies on 372, pi. 12 fig. 5A-E; Kolosvàry, 1947: 427; Ross & Newman, Contributions to Zoology, 65 (3) - 1995 141

Fig. 8. Hoekia fornix n. sp.: a, apical view of holotype, NNM no. C-2540, carinal end at left; b, oblique view of same, carinal end at right; the slightly ascendant marginalfringe is distinctly separated from the outer lamina. Also note disorganized calcareous overgrowth

on shell.

delicatein well elevated 1969, in part, fig. 1A only; Ross & Newman, 1973: 162, in part, thin, appearance; peritreme fig. 18 right, fig. 19 top. in small but lower in large specimens; carinal ridge

prominent in large specimens, less pronounced in

Material. - One dried specimen lacking opercular plates from small; ridge not extending to margin of outer lami- Hydnophora exesa, Singapore, John Hendley Barnhart coll.,

orifice rostral approx. 1°20'N, 103°50'E, American Museum ofNatural His- na; ovate, subapical on slope; growth

coral 1883. no. Twelve specimens from Hydnophora tory, exesa, ridges faint, beaded, widely spaced on peritreme,

Singapore; S.H. Chuang coll., 3 Dec. 1975. Ten topotypes, but proximate along distal portion of outer lamina; Scripps Institution of Oceanography, no. C-8375; 1 topotype, distal edge of outer lamina thin, slightly ascendant; Zoölogisch Museum Amsterdam, no. ZMA Cirr. 201517; 1

same as or than out- topotype, Nationaal Natuurhistorisch Museum, Leiden, no. marginal fringe slightly higher

C-2541. er lamina, narrow in small specimens, wider in

large, widest at extremities in large specimens; in- Diagnosis. - Shell white, peritreme patelliform, terior surface dull, translucent, smooth, becoming large; carinal ridge broad, higher than orifice. wrinkled near margin; sheath 1/2 or more height of

Description. - Shell semitranslucent, moderately wall, glossy; lacunae of different sizes, ovate, elon- 142 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Fig. 9. Hoekia monticulariae (Gray, 1831): a, external surface of left opercular plate, limbus occludens ( = occludent edge) at top (pad- like flangealong basal marginto the left of center is the attachment surface for lateral depressor muscle); b, internal surface ofbroken right opercular plate; c, apical view of young specimen with marginalfringenot fully developed; noteessentially ovate shape, low carinal ridge at left, and granularto slightly beaded growth ridges. Topotypes, SIO no. C-8375. Contributions to Zoology, 65 (3) - 1995 143

Fig. 10. Hoekia monticulariae (Gray, 1831), mouthparts and appendages: a, labrum and left palp; b, mandible; c, enlarged view of inferior angle of mandible; d, maxilla I; e, enlarged view of apical portion of maxilla I; f, posterior portion of median languette and reduced maxillae cirrus cirrus cirrus cirri V and cirri II; g, I; h, II; i, III; j, VI; k, V, VI, and penis. 144 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

but of for the gateor irregularly shaped, irregularly arranged H. monticulariaeexcept type locality are

essentially in a circle. questionable (see under A. tanabensis).

Opercular valves thin, translucent; strongly arcu- Unconfirmed records: Annandale (1924: 67)

ate to U-shaped in section; growth ridges slightly identified H. monticulariaesolely on the configura-

beaded; plates bilaterally asymmetrical, right plate tion of a basis in H. exesa, from the coast of Arra-

longer and not as arcuate as left; rostral depressor can, and another from an undisclosed locality in the

tooth of scutal portion elevated; tooth about 1/4 Indian Ocean.

length of plate; tergal crest about 1/3 width of ros-

tral tooth, and about 1/5 length of plate (Fig. 9a,

b). Hoekia mortenseni n. sp.

Crest of labrum slightly sulcate, devoid of setae (Figs. 11, 12)

and teeth; anterior surface lacking ctenoid scales

fused to lateral of (Fig. 10a). Palps basally margins Pyrgoma monticulariae; Ross & Newman, 1969, in part, figs.

2A-C labrum, lowerthan labral crest, clothed with long IB, only.

Hoekia monticulariae; Ross & Newman, 1973, in part, fig. 18, simple setae. Cutting edge of mandible almost left, only. straight, divisibleinto three parts; superior half de-

void of teeth, but armed with a few slender setae;

Material. - Three preserved specimens from Hydnophora exe- inferior half divisible into two parts, upper armed sa, Th. Mortensen Java-South Africa Expedition, sta. 44, be- with strong, stout teeth with subsidiary cusps; in- tween Gunner's Quoin Isl. and Flat Isl., north of Mauritius, 46

ferior armed with teeth and angle more simple long, m, Sigsbee trawl, tugboat "Maurice", 15 Oct. 1929; approx.

stout setae along inferior margin; leading edge 19°54'S, 57°37'E. Holotype: Zoologisk Museum, Copenha-

CRU 2 gen, no. 701; paratypes, same, no. CRU 702. Two dried of whole appendage clothed with long ctenae. Max-

specimens lacking opercular plates, paratypes, Pointe au Sable illa I with long, uniform, simple, equally spaced (south of Port Louis), Mauritius, outer reef slope, 22 m, July spines; lateral surface of appendage clothed with 1971, approx. 20°18'S, 57°31'E, Gerard Faure coll., Muséum

ctenae. Median of languette same size as adjoining National d'Histoire Naturelle, Paris, no. CI-2354 and 2355.

trophi; peripheral margin with minute, slender,

of - Shell to subcircu- simple setae; posterior part appendage protuber- Diagnosis. pink rose; peritreme

ant Maxilla (Fig. lOf). II fused basally to languette; lar, small; carinal ridge narrow, higher thanorifice.

with bilobed, a few moderately long apical setae.

Coxopodite of protopodite of cirrus I broad, Description. - Shell thin, partly translucent; peri-

about 25 than of times larger basipodite; segments treme low, conic (Fig. lib); orifice subapical on

rami fused, malformed. Cirri II and III aberrant, rostral slope; carinal ridge prominent; growth

medially situated and parallel with but partly hid- ridges closely spaced, pustulose or beaded; distal

den by coxopodite of cirrus I. Cirrus IV vestigial. margin of outer lamina thin, ascendant; marginal

V rami Cirrus reduced, distinctly biramous, un- fringe narrow, low, slightly higher than outer lami-

equal; medial to, and of same height as coxopodite na, only slightly wider on extremities; sheath less

of protopodite of cirrus VI. Cirrus VI not mal- than 1/2 height of wall, glossy; area below sheath

formed, rami nearly equal in length, exopodite with rough to slightly pustulose, translucent with radi-

14 with segments, endopodite 13; greater curvature ating alternate light and dark rays emanating from

with few short a simple setae at articulations, lesser near sheath; lacunae approximately same size,

curvature essentially devoid of setae. Intromittent small, ovate.

organ distinctly annulated,long, slightly more than Crest of labrum sulcate, lacking teeth (Fig. 12a);

twice of cirrus length VI; sparsely clothed with a anterior (outer) surface densely clothed with cte-

few scattered short, irregularly setae. Basidorsal nae. Palps slender; lower than crest of labrum;

rounded. point large, evenly fused basally to lateral margins of labrum; sparsely

setose, setae simple. Mandible with cutting edge

— All records Remarks. previous for the occurrence straight; superior margin covered with single row of - Contributions to Zoology, 65 (3) 1995 145

Fig. 11. Hoekia mortenseni n. sp.: a, apical view of holotype,ZM no. CRU 701, carinal end at top; b, oblique view of same; c, apical view of same, in situ on Hydnophora exesa, carinal end at top, x 1.85 (image reversed).

closely spaced, simple, short setae; edge divisible reduced, fused basally to languette; apically clothed into three parts; superior half devoidof teeth or se- with short, simple setae. tae; inferior half armed with short, multicuspid Coxopodite of protopodite of Cirrus I broad,

inferior armed with multi- 20 times than of teeth; angle similarly about larger basipodite; segments cuspid but shorter teeth; inferior margin with row rami fused, malformed. Cirri II—IV malformed. of simple, stout setae. Maxilla I with straight cut- Cirrus V questionably biramous, segments indistin- ting edge, armed with simple, thin, uniformly guishable. Cirrus VI with rami essentially equal in spaced spines diminishing in length basally; inner length; exopodite with 13 segments, endopodite lateral face with Median covered ctenae. languette with 10; greater curvature with short simple setae at approximately of same size as adjoining trophi; articulation, lesser curvature essentially devoid of

clothed with bris- annulated; periphery minute, slender, simple setae. Intromittentorgan sparsely cov- tles; anterior edge slightly sulcate; posterior portion ered with irregularly scattered simple setae. Basi- strongly protuberant (Fig. 12f). Maxilla II greatly dorsal point large, evenly rounded. 146 A. Ross & W.A. Newman A coral-eating barnacle, revisited

Fig. 12. Hoekia mortenseni n. sp., mouthparts and appendages: a, labrum and palp; b, mandible; c, enlarged view of inferior angle of mandible; d, maxilla I; e, enlarged view ofapical portion ofmaxilla I; f, median languetteand reduced maxillae II; g, cirrus VI and penis; h, distal end of penis.

- find tribute Etymology. We can no more fitting ascendent, high; marginal fringe thick, complexly than to name this species for that intrepid explorer deliquescent, as high as or higher thanorifice; lacu- and renowned zoologist, the late Dr. Th. Morten- nae of inner wall separate to confluent, ovate to sen of Denmark, who collected the present elongate, irregularly arranged on inner wall, com- specimens. monly laterally concentrated; lacunae connected

directly to circumferential channel.

Type species: Ahoekia tanabensis n. sp.

Ahoekia n. gen.

Definition. - Shell thick, massive in appearance;

carinal - Derived from the a peritreme small to minute; orifice subapical; Etymology. Greek, an-, nega- ridge low, broad; distal margin of outer lamina tive prefix meaning not, and -Hoekia. Contributions to Zoology, 65 (3) - 1995 147

relate observations on the TanabeBay Ahoekia tanabensis n. sp. 405) likely to

(Fig. 13) form, which we consider a new species, Ahoekia

tanabensis.

Pyrgoma monticulariae; Hiro, 1935: 18, fig. 11; Hiro, 1938: Several specimens of this species were discovered

405. in the Seto Marine recently in a display case Biologi-

corals from the cal Laboratory, Japan, on several

Material. - Four dried specimens lacking opercular plates from same region as the types (Harada, pers. comm., Hydnophora?exesa, Tanabe Bay, Honshu Isl., Japan, H. Uti-

Based on Harada's the corals no. 1995). photographs, nomi leg., approx. 33°43'N, 135°22'E. Holotype: USNM identified Veron 259982; 1 paratype, same, no. 259983; 2 paratypes, Scripps were by (pers. comm., 1995) as

Institution of Oceanography, no. C-7980; 5 paratypes, Seto H. bonsai Veron, 1990. In Japan, H. exesa occurs

Marine Biological Laboratory, no. 383, date and collector sympatrically with H. bonsai (Nishihira & Veron, unknown, on H. bonsai. hoekiine 1995), but whether this or another occurs

on the former is presently equivocal. Diagnosis. - Shell rose to pink violaceous, opaque; peritreme relatively small; carinal ridge low, broad; Etymology. - The Latin suffix, -ensis, denotes lo- periphery of marginal fringe openly perforate, with cality, in referenceto Tanabe Bay where this species it and lamina. pronounced gap between outer occurs.

Description. - Shell dull, opaque, massive; orifice

ratio of ovate, subapical on rostral slope; rostro- Ahoekia chuangi n. sp. carinaldiameterto orifice approximately 25:1 ; cari- (Fig. 14) nal ridge broad, short; distal edge of outer lamina ascendent, high; marginal fringe narrow to wide, Material. - Three dried specimens from Hydnophora rigida

width laminaand greatest at extremities; outer mar- (Dana, 1846) "Eiland Onrust-oost, Noordelijk deel", Bay of

Batavia, Java Sea, Indonesia, J. Verwey coll., 28 Mar. 1931, ginal fringe separated by distinct gap; ancillary ap-

prox. 6°08'S, 106°45'E. Holotype: Nationaal Natuurhistorisch openings in periphery of marginal fringe lead to Museum, Leiden, no. C-2542; 2 paratypes, same, no. C-2542 lacunae on inner wall; internal surface roughened (PI, P2). or wrinkled; sheath less than 1/2 height of wall,

lacunae - glossy; irregularly arranged, irregularly Diagnosis. Shell white, translucent; peritreme of different sizes. shaped, small; carinal ridge low, narrow; periphery of mar-

ginal fringe compact, apparently poorly perforate,

Remarks. - The four dried shells from Tanabe with indistinct gap between outer and innerlamina. identified labelled Bay, Japan were originally and as monticulariae the late Huzio Utino- Pyrgoma by Description. - Shell vitreous, translucent, thick, Hiro mi (= Fujio Hiro; see Newman, 1981). (1935: massive (Fig. 14b); peritreme only slightly elevated,

illustrated two different fig. 11) specimens purport- orifice ovate, subapical on rostral slope; ratio of of this from Tanabe He omitted edly species Bay. rostro-carinal diameter to orifice approximately mentioning lacunae on the internal wall, although 18:1; carinalridge narrow; growth ridges not visible he illustrated them in the but not left right speci- due to coral overgrowth; distal edge of outer lamina men. However, the external view of the opercular ascendant, high; marginal fringe narrow to wide, he illustrated is different from that plate distinctly decidedly wider at extremities; outer lamina and

Darwin fig. figured by (1854: pi. 13, 5F; repro- marginal fringe not separated by distinct gap; duced Ross & 1973: In by Newman, fig. 19, top). sheath about 1/2 height of wall, semi-glossy, with his 1937 on the barnacles in the vicinity of report translucent radial lines; inner wall opaque to the Seto Marine mention Biological Laboratory no chalky, roughened, irregular; lacunae irregular, is made of H. monticulariae. in his 1938 study But, ovate to elongate, irregularly arranged (Fig. 14c). of the pyrgomatids of Palau he discussed the host specificity of H. monticulariae, which was not Etymology. - Named in honor of Prof. S.H. Chu-

1938: National of who known there. Therefore, his remarks (Hiro, ang, University Singapore, kindly 148 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

of monticular- under brought us preserved specimens H. high magnification (Fig. 15c).

circumference of the wall in most balano- iae on Hydnophora exesa that he collected. These The

oval and formed the basis for our anatomical work. morphs approximates an or circle, this

shape is essentially characteristic for each species

(Spivey, 1988: 279). There are variations, but in Structure of the wall and basis arise from in the good measure these irregularities

substratum and individuals. The general development of the balanid shell was crowding by adjoining described by Darwin (1854), Costlow (1956), New- In Hoekiini, dueto thearrangement of the hydnons man et al. (1967), Ross & Newman (1967), Newman of the host, there is no fixed form to the circumfer-

Their best & Ross (1976) and Bourget (1977), and the crystal- ence (Figs. 4a, 8a, 13a). outline is

described and this them from line microstructure by Newman et al. (1967), New- as protean, sets apart

all known The fea- man & Ross (1971), and Bourget (1977, 1987). Ex- balanomorphs. only constant ternally, the surface of the shell is covered com- tures of the adult wall are the presence of a raised pletely by a thin cuticle, the epicuticle. Where the peritremal area, minute aperture, and the hyper-

lateral opercular plates are suspended in the orifice of the trophied margin.

and and in sessile barna- shell, the cuticle is uncalcified and coriaceous, In some pyrgomatids, most is termed the arthrodial membrane. Internally, the cles, the conical wall has an apical and basal desig- entire shell is lined by the hypodermis which se- nation. However, in Hoekiini, as in many pyr-

wall lie cretes the calcareous layers as well as the cuticle ex- gomatids, the basal portion of the comes to

horizontal posed to the exterior, including thatof the body and essentially in a plane (Figs, lib, 14b).

well that of the uncalcified the basal in such forms is appendages as as por- Therefore, margin tion of the basis. referred to herein as the lateral margin.

Ross & Newman (1973: 139) reviewed what was In most balanomorphs the wall consists of an in-

laminaoften tubes and known aboutthe growth pattern of the pyrgomatid ner and outer separated by shell, from the time of settlement to the early adult related structures (Darwin, 1854; Costlow, 1956;

& stage. The following discussion covers how the Newman et al., 1967; Newman Ross, 1971, 1976).

Hoekiini depart from it. In Hoekiini, the shell in cross section is reminiscent

The shell shape in Hoekiini is conical in early on- of the Greek letter Omega (Fig. la). The outer (up-

in Hoekiini does extend togeny, with the orifice opening apically or on the per) laminaof the wall not

far the inner rostral slope. All of the specimens have a concres- laterally as as (lower) lamina(Figs, la,

wall with of the line of The distal of the lamina is thin and cent no indication juncture 4a). edge outer between individual wall plates, but the earliest on- comes to lie in essentially a horizontal plane, or it togenetic stages were not available. may turn upward and reflex, or even become ascen-

External ornamentationis limited to granular or dant (Fig. 14b), to degrees varying witheach group. beaded The lateral be of such as to ex- growth ridges, and where the ridges are ob- margin may height scure the surface remains granular or beaded. Each ceed the height of the peritreme and/or the orifice. bead represents the site of a seta that originates The inner lamina in plan view, is usually widest internally (Fig. 15a, b) and traverses the wall at the extremitiesof the shell (Fig. 6a). In some spe-

almost the (Bocquet-Védrine, 1966: 339; Newman & Ross, cies it is narrow and turns upward to

1977: the but in others it is 1971: 146; Bourget, 293; Clare et al., 1994: same height as outer lamina,

The holes which the A feature of the inner 30). through setae grow range significantly higher. unique in deli- diameterfrom 2.73 to 3.09 nm, and may be seen laminae is the deliquescent periphery. This

Fig. 13. Ahoekia tanabensis n. sp.: a, apical view of holotype, USNM no. 259982, carinal end at right; b, oblique view of same, carinal end of of at left; c, internal view paratype, USNM no. 259983, carinal end at bottom; note lateral arrangement lacunae; d, enlarged view of right extremity of specimen shown in a, showing opening of ancillary connection from periphery of fringe to circumferential channel. - 149 Contributions to Zoology, 65 (3) 1995 150 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Table II. Measurements of the shell, orifice and opercular plates of the type and figured specimens of Hoekiini. Abbreviations: RCD,

rostral-carinal diameter; LD, lateral diameter; RCOL, rostro-carinal diameter of outer lamina; L, length; H, height; ?, unknown or

not determined. All measurements in mm.

Species Shell Orifice Opercular plates

RCD-LD-RCOL RCD-LD L-H

Eohoekia

chaos holotype; Fig. 4a 6.0-7.0-3.4 0.3-0.3 1.3-0.3

chaos paratype; Fig. 4b 5.8-5.0-3.7 0.3-0.25 7

nyx holotype; Fig. 5 6.4-5.1-4.0 0.4-0.3 1.4-0.3

16 4?-3.0 nyx paratype; Fig. 0.4-0.3 1.25-0.4

Parahoekia

aster holotype; Fig. 6 4.5-3.8-3.2 0.35-0.25 1.3-0.3

aster paratype 4.4-3.9-3.1 0.35-0.30 1.4-0.4

Hoekia

fornix holotype; Fig. 8 7.3-5.7-5.5 0.4-0.4 ?

monticulariae 7.5-5.3-6.2 0.4-0.3 1.7-1.2

monticulariae;Fig. 9 5.8-3.4-5.1 0.35-0.3 ?

monticulariae ; Fig. 18a 9.1-8.2-? 0.5-0.4 7

monticulariae;Fig. 19 4.8-3.4-? 0.25-0.22 7

mortenseni holotype; Fig. 11a 6.8-4.0-5.7 0.4-0.4 7

mortenseni paratype 6.1-6.1-5.7 0.4-0.3 7

mortenseni paratype 5.1-3.5-4.2 0.35-0.2 1.0-0.3

A hoekia

tanabensis holotype; Fig. 13a 10.1-9.0-7.5 0.4-0.3 ?

tanabensis paratype; Fig. 13c 8.6-7.8-8.1 0.4-0.35 ?

chuangi holotype; Fig. 14a 8.9-9.0-6.3 0.5-0.4 7

chuangi paratype; Fig. 14c 8.0-5.8-5.2 0.3-0.3 1.3-0.35

quescence is verrucose to papillose, and in those fused compartments it is due mainly to abrasion by

where it translucent reminiscent of species is it is a water-borne particles and corrosion (Darwin, 1854: thick sheet of icicles. the to orifice 324). Obviously, only way enlarge the

The minute size of the orifice relative to the size in barnacles that have a concrescent shell, is of the shell is unparalleled in other balanomorphs. through abrasion and/or corrosion. In most Hoeki-

Rostro-carinal diameters of the orifice range from ini the aperture is neatly "worn" so that the wall of

0.2 0.5 and 0.2 0.4 the to mm widths from to mm orifice is perpendicular to the adjoining shell

This would limitations size increases (Table II). seemingly pose to surface (Fig. 14a). Orifice evidently as normal biological activities owing to the otherwise needed, but how might this be accomplished?

normal size essentially of the body. In the absence Neither the appendages nor the trophi appear cap- of functional cirral a net, the presence of a minute able of being rotated 180°, and hence they are in- orifice is due largely to a wholly unique method of capable of abrading theorifice. The likely explana- feeding unknown in other balanomorphs (see the tion is chemical corrosion by the opercular mem- sections Cirri, and Feeding and nutrition). brane (tergo-scutal flaps), but no significant

Aperture enlargement is not uncommon in inter- membrane or frill was evident in the preserved tidal balanomorphs. In balanids it is due largely to materials available. diametric growth, whereas in tetraclitids that have The basis in balanomorphs is attached to the sub-

14. Ahoekia view of Fig. chuangin. sp.: a, apical holotype, NNM no. C-2542, carinal end at upper right, note perpendicular edge of orifice view of carinal end internal view of carinal appears uniformlyworn; b, oblique same, at left; c, paratype, NNM no. C-2542(P1), end at left, note incompletely formed lacunal opening and the channel toward the lower right. - 151 Contributions to Zoology, 65 (3) 1995 152 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

15. Hoekia monticulariae view of sheath surrounded for that the Fig. (Gray, 1831):a, enlarged adpressed by pores setae traverse wall;

b, view of left and view of lamina external for that enlarged sheath, upper corner, pores; c, enlarged outer showing pores setae traverse

the from wall; pores range 2.73 to 3.09 µm in diameter. Topotype, SIO no. C-8375.

stratum by secretions from the cement glands. growth may also occur along the sutures between

Growth occurs in a narrow zone at the edge of the the plates.

basis and the lower lateral margin of the wall. In Utinomi (1943: 19) noted that the basis in the

those forms in which the wall earliest plates are separate, stages of pyrgomatids is wholly mem- Contributions to Zoology, 65 (3) - 1995 153

behind and membranous branous and persists until the rostro-carinal dia- ly lags growth, thus a por-

is which tion of basis is continuous of the later- meter of the shell nearly 1 mm, is generally the with that

the case in balanomorphs having a calcareous basis. al and superior margins of the hypertrophied inner

Among the pyrgomatines, only Pyrgopsella Zullo, lamina. Consequently, this membranous covering

in 1967, addition to an incompletely calcified wall, of the wall broadly contacts coral tissue rather than

retains basis life a wholly membranous throughout the calcareous basis.

The bottomof the and evidently as an adaptation to living in calcareous basis has a higher opacity,

cementation and calcification is than twice thick its vertical walls. sponges. Following more as as

of the basis to the skeleton of the coral, lateral There are mammillated deposits at the bottom of growth of the shell and basis proceeds rapidly "... the basis that are unknown in other pyrgomatines.

so that the shell reaches essentially maximum di- As notedabove, the peripheral margin of the wall

in ameter early life" (Ross & Newman, 1973: 139). is filled with tissue covered by a cuticle that is in

This effectively establishes the area that the barna- contact with coral tissue rather than the calcareous cle will occupy on the coral until overgrown. Subse- basis of the barnacle or the surrounding calcareous

to accommodate the ovarian skeleton of quent growth, body, the coral. By applying light pressure to and mantle is tissue, largely accomplished by alcohol-preserved specimens the shell can be "wig- lengthening the basis. This phenomenon was illus- gled" both vertically as well as horizontally, as is trated by Hiro (1938: 409), and is readily seen in undoubtably the case for living . In effect,

fossils that have been leachedfrom corals (Newman the shell wall and surrounding portion of the basis

& Ladd, 1974: pi. 2, fig. H). We subscribe to the ob- are "floating" in soft coral tissue, and this is unique servation of Baluk & Radwanski that the (1967: 488) among pyrgomatids. growth of the basis is largely synchronous with that From the foregoing it is apparent that the barna- of the host but with cle is surrounded coral skeleton, some exceptions effectively a cyst by skele-

1976: Ross & 1973: for the and the exterior (Newman et al., 69; Newman, ton, except marginal fringe

140). The coral will eventually overgrow the barna- surface of the wall which is surrounded by coral tis- cle, and not infrequently with the opercular plates sue. A plausible explanation of why the top portion cemented in life position by the coral. of the basis is membranous and the shell does not

Lateral growth of the basis, and concomitantly rigidly interlock with it, as in other pyrgomatines, the wall in Hoekiini, is determined by the configu- is that there is metabolicactivity betweenthe barna- ration of the coral's hydnons. Most subsequent cle and host coral tissues. This activity could be sim-

is additions to the distal of the that which coral skeleton from growth by margin ply keeps overgrow- basis, in keeping pace with the growth of the host. ing the barnacle, but considering the extent of the

Shell profile in pyrgomatines is conical to essen- area, and that it is highly vascularized, it could tially flat. When flat, the lateral margin parallels serve as a pathway for the exchange of nutrients. the vertical surface of the basis. Thus, the peripher- al outline is a reversed image of the outline of the basis. Although the basis may continue to grow Shell development slightly wider after the initial period of rapid lateral expansion, it and lateralgrowth of the shell wall are In Hoekia, as with pyrgomatids in general, the wall essentially determinate.On theother hand, vertical is solid, but it is apically thickened by the sheath, a growth of the basis is virtually indeterminate. secondarily deposited innermost layer (Ross &

The unusual aspect of the basis in Hoekiini is that Newman, 1973: 143; Bourget, 1977: 289). In primi-

is not only the circumference irregular in outline, tive Pyrgomatini, such as Cantellius Ross & New- relative to the configuration of the surrounding man, 1973, the basal margin of the sheath is not at- hydnons, but above the calcareous portion there is tached to the wall, and is said to depend freely

membranous a zone. Calcium deposition apparent- (Newman & Ross, 1976: 27). An extension of the 154 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

sheath forms an inner wall or lamina that develops tantly, beyond the circumferential channel, the parallel to, and is fundamentally shorter than, the margin of the inner lamina begins to digitate. The

inner the form outer wall or lamina. Although it is called an spaces between digitations passages ra- lamina in higher pyrgomatids, it is not homologous diating from the inner surface through the inner with that of higher balanids (Costlow, 1956: 370; lamina to the circumferentialchannel. In morpho-

Ross & Newman, 1973: 143). In pyrgomatines hav- logically simple forms, such as Eohoekia, the digi-

tubiferous ing a wall, the wall and sheath are sepa- tations deliquesce as they extend peripherally. In

vertical form elements between rated by partitions (ribs) to one or some cases, radiating déliques- more rows of tubes. cences coalesce to form a filigree (Fig. 4b) chan-

In the ontogenyof Hoekiini, the innerlaminabe- nelizing the passages to the circumferential chan-

but and the of gins simply as an adpressed sheath, quickly de- nel, to outer surface the marginal fringe

the it but the in this velops a dependent lower margin as indicated by as expands, openings filigree may position of the proximal portion of the abandoned fill secondarily. carinal sinus (see below). The lower margin of the Parahoekiais similar to Eohoekiain the develop-

of dependent sheath becomes weakly cemented or ment the marginal fringe, but there are several fused, by radial ribs or bridges (Fig. 16), to the out- differences. The elements of the fringe are thicker, er lamina. As growth progresses, thebasal margins narrower and not reflexed (cf. Figs. 4c and 6c). At of both laminae come to liein essentially a horizon- the digitations or extremities, adjacent elements tal plane, after which growth of the outer lamina eithertouch or coalesce to form one or two terminal slows relative to the inner. When beyond the outer apico-basal foramina. Internally, at the extremi- lamina, the lateral margin of the inner lamina ties, there is a shallow sulcus that deepens as it leads hypertrophies to form the marginal fringe which, in to the base of these foramina. The sulcus may be

H. monticulariaeaccording to Darwin (1854: 373), obscure on some extremities, but obvious on resembles a "mass of moss". others.

Between the outer and inner laminae, before de- In Eohoekiaand Parahoekiagrowth of the outer velopment of the marginal fringe, an intervening lamina and marginal fringe of the inner lamina is but minor and temporary lamina develops, the accomplished by the hypodermis that extends from basal of which is edge lobate. At the carinal end, be- within the shell up and between the comb-like teeth low basal of the edge the outer lamina, one lobe of and foraminaof the innerlamina. The irregular up- this minorlamina forms of inner a complex tongue-like ex- per surface the lamina, broken into ridges tensionthat keeps pace with the growth of the outer and valleys, provides space for thin threads or laminabefore coalescing with the wall (Fig. 16). In sheets of tissue contacting those marginal surfaces

it becomes doing so, the floor for the carinal sinus, abutting against coral tissue.

abandoned an channel, which may fill with calcare- In morphologically more complex forms, such as ous material to varying degrees. Hoekia and Ahoekia, the undersides of the radiat-

The space between the inner lamina and the ing deliquescences close off as they grow laterally. growing edge of the outer lamina forms a peripher- In so doing, they leave behind lacunae that lead to al circumferential channel, which we infer provides radial channels that connect with the circumferen-

for inferior = circumferential space the ( basal) ves- tial channel and its tributaries from the marginal sel found in balanomorphs (Burnett, 1977: 301). fringe. The fringe is thick and massive and embod-

The circumferential channel moves laterally with ies the lacuno-circumferential channel. We infer lateral growth of the outer lamina, and in doing so that this elaboration of the marginal fringe and the much of the tissue between the laminae withdraws included lacuno-circumferential portion of the cir- and the vacated channels for in may become partly or culatory system allows a substantial increase wholly filled. the area of contact between tissue of the barnacle

At an early stage, before the inner lamina begins and that of the host in which it is suspended (see to extend beyond the outer lamina and, concomi- Figs. la-e). Contributions to Zoology, 65 (3) - 1995 155

Fig. 16. Eohoekia nyx n. sp.: internal view of paratype, USNM no. 259981, carinal end of shell and basal edge of inner lamina with tongue-like extension that forms the floor of the carinal sinus; to the right bridge-like structures unite the inner and outer lamina; the adpressed sheath is visible at top.

Eo- Peripheral shell shapes hoekia in being conical and ovate, but the overall

outline is stellate (Fig. 6a). In this case settlement

occurred between All of the specimens of E. chaos available are ap- probably essentially equally

of of the parently of the same age class because compara- spaced hydnons. Subsequent growth margi- ble size. The outer lamina is primarily ovate in out- nal fringe appears to be uniform in all directions.

limitedobservations of line, but the marginal fringe varies in shape. The Based on our Hydnophora fringe is strongly reflexed further suggesting matu- microconos, we infer that any other species found

in of which have this coral will have stellate outline similar rity. Shellshape early juveniles, we on a to

that of seen only one specimen (.E. nyx:) approximates that P. aster. of other pyrgomatids; a conical wall with a circular Hoekia exhibits the most variable shell pattern, to ovate outline (Hiro, 1938: 409; Moyse, 1971: but all three species develop a marginal fringe that

The Hoekiini take and close the 135). on protean shapes as they parallels remains to outer lamina ex-

of condi- at the monticulariaethe shell grow, but they retain part this primitive cept digitations. In H. tion throughout life in the form of the peritreme. assumes a more digitiform outline than in H. mor-

Their growth is constrainedby thearrangementand tenseni, which tends to be trapezoidal (Fig. 11c). growth of coral hydnons. Bourget & Crisp (1975a: The shell in Ahoekia is much like that of Hoekia.

245) note that in ordinary barnacles there must be However, it is more trapezoidal, and does not as- some mechanism controlling the rate at which in- sume theprotean outline characteristicof the latter. dividual parietal plates grow when hinderedor un- The shape suggests that whereas the two species of hindered by obstacles, but what the mechanism is Ahoekia may be limited in their ability to grow has not been determined. laterally, they have essentially unlimited vertical

The wall of Parahoekia is much like that of growth. 156 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Fig. 17. Hoekia monticulariae (Gray, 1831): a, fixation fibers with anastomoses attached to membrane; b, fixation fibers removed from wall of shell; note disc-like surfaces at termination of fibers that enhance attachment to wall. Topotype, SIO no. C-8375.

Wall-basis interface lamina and insert on the calcareous portion of the

basis.

Observations on the interface between the wall and Gutmann (1960: 13) first discovered "Fix- basis are limited because of the paucity of preserved ierungsfasern", or fixation fibers, in Semibalanus material available. However, by applying light balanoides (Linnaeus, 1766), which anchor the wall

to the shell of to the also do in modestus pressure an alcohol-preserved speci- basis, as they Elminius

it be noted earlier. This is 1854 1975: & men, can "wiggled", as Darwin, (cf. Bubel, 288). Bourget puzzling when one considers that the wall of all Crisp (1975b: 457) suggested that the fixation fibers other for the allow the limited movement relative pyrgomatids, except Pyrgopsellini, .. parietes mechanically interlock with the basis. In Hoekiini, to the base". Each fiber consists of a long fine ten- the lower portion of the basis is calcareous, whereas don, a short section of striated muscle, and a short- the upper is membranous.This membrane is white, er basal tendon. The fibers, which are remarkedly flexible, tough, but relatively thin. The distal mar- robust in H. monticulariae, lack the short striated gin of the membranepasses up and over the margi- muscle section and, therefore, presumably do nal fringe to attach around the outer margin of the not contract. At the origins and insertions they

lamina and hold the wall in which broad of outer helps place (Fig. anastomose, provides a area attach-

3). However, the wall is also connected to the cal- ment (Fig. 17). When the wall is pulled free, the careous basis by fixationfibers, as describedbelow. fibers generally remain attached to the basis rather

If wall is free the pulled from the coral, the oper- than to the wall. cular plates and commonly the body remain at- tached to it, whereas the lining of the membranous basis, the accompanying tissue and sometimes the Coral overgrowths body remain with the coral.The separation general- ly occurs along a horizontal plane where the wall is All pyrgomatids, except for the sponge-dwelling attached to the fixation fibers (Fig. 17) which Pyrgopsella, live on stony corals and are subject to

beneath the of the inner originate basal margin overgrowth by the host coral. There are two types Contributions to Zoology, 65 (3) - 1995 157

of overgrowth, coenenchymal or calcareous. The approximates a classic text-book scleractinian

cover of coenenchyme keeps the barnacle shell polyp. This "polyp", or pseudopolyp, includes two

"... clean and free from epizoic growths" (Moyse, layers of tissue separated by radial mesenteries or

1971: 137), and it also tends to make the barnacle septa that extend from the periphery of the shell

less conspicuous. This blanket of tissue extends to almost to the aperture, and covering the lateral

the margin of the orifice, and Anderson (1992: 305) margin of the shell there are one or two whorls of

hypothesized that it is kept from overgrowing the en echelon tentacles (Fig. 3). In the preserved

modified materials available oral mouth orifice by tergo-scutal flaps, the "apertur- no disc, or pharynx,

al frill", which he stated is unique to pyrgomatids. situated over the aperture of thebarnacle, were ob-

The frill itself may be the site of inhibitory secre- served. If a mouth were present, it is possible that

food to tions, although in some species, e.g., Pyrgoma the coral's is going directly the barnacle, as

elongatum Hiro, 1931, Wanella milleporum (Dar- well as the barnacle acquiring food indirectly by

win, 1854), and Pyrgoma cancellata (Leach, 1818), eating coral tissue (Ross & Newman, 1969: 255).

there are large subopercular glands that may be the

frill secretory source (Anderson, 1992: 337). A as

such probably does not exist in Hoekia because the Opercular plates

orifice and underlining aperture are completely

overgrown by coral tissue, and we failed to detect Closure of the orifice by paired opercular plates,

in the of Hoekia formed the articulated and charac- subopercular glands specimens by scuta terga,

and Parahoekia that we examined. terizes most balanomorphs. The opercular plates

In many pyrgomatids the shell is partly or almost are attached to the sheath by a flexible arthrodial

covered coral skeleton. These form wholly by over- membrane. The sheath may or may not a dis- growths extend from the coral on to the shell proper. tinct structural strengthening element of the wall

But they are either broken at the wall/basis inter- (Newman & Ross, 1971: 146; Bourget, 1977: 289;

face or contiguous growth at this junction is inhib- 1987: 275). Commonly, a narrow band of this ited the it is around the membrane is the environment in interti- by barnacle, as aperture exposed to

(Moyse, 1971: 139; Ross & Newman, 1973: 139; dal forms. Because of the size of the orifice in

Anderson, 1992: 292). In the sulcus between the Hoekiini, only the limbus occludens (see Fig. 2) of peritreme and ascendent margins of the shell, both each opercular plate occludes theorifice and, there-

A. tanabensis and A. chuangi have crustose over- fore, the opercular membrane is completely con- growths which extend on to the lower portion of the cealed.

The of is peritreme. overgrowth A. tanabensis Although we have not seen opercular plates for white, opaque, and compactly nodular or verru- all Hoekiini, the ones examined differ in relative cose. On A. chuangi it is semitranslucent and more proportions, flexure, symmetry and other details.

in crystalline appearance. These overgrowths are They are thin, translucent, and exceedingly fragile.

of normal coral On All not suggestive sclerosepta. one are compound or concrescent plates that lack a specimen of H. mortenseni the shell is partly cov- distinct juncture between the scutum and tergum. ered by a coral hydnon, and thebalanceof the outer The plates are wider than high with the scutal longer lamina covered completely by drusy calcite. than the tergal portion. External ornamentation

Hydnophora differs from most scleractinians in consists of closely spaced simple or slightly beaded that the mouths occur in the "valley" between hyd- growth ridges (Fig. 9a) covered with a hirsute cutic- nons, and the tentacles encircle hydnons rather ular membrane. The cuticle covering the plates is than mouths. The shell morphologies of the Hoeki- denseand darkly pigmented in Parahoekia, but not ini and the coral tissue overgrowing them are sug- in those species of Hoekia that we examined. gestive of these tuberosities. Although it appears In most balanomorphs growth increments to the that the barnacle/coral basal mimics a hydnon because opercular plates are added at the margins the tentacles encircle the barnacle, the combination (Darwin, 1854: 54). Each increment increases the 158 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

of its and width. the tetraclitines where orifice is thickness the plate as well as height enlargement by at-

is addition to the occludent trition 1854: There rarely any or corrosion (Darwin, 324; Ross,

effect margins. 1969: 238). But this has very little on growth

and form of the added When discussing opercular morphology and opercular plates, which are growth in some pyrgomatids, Darwin (1854: 54) in- to at the basal margins, because growth of the wall troduced the term limbus occludens (= occludent remains conical. The "reverse" growth pattern of

ledge or edge; Hiro 1938: 403; Soong & Chang, the valves seen in Hoekiini is necessitated by the

for of the from horizontal 1983:247) growth along the apical margin change conical to growth, and con- plate. He noted that this edge formed an addition crescence of the wall plates. With growth of the

along the occludent margins of the scuta and terga wall, and enlargement of the orifice, enlargement

(Fig. 9a, b). This is readily seen on theexternal sur- of the superior portion of the opercular plates for

occlusion orifice is if low face, and represents "unimportant exceptions to of the required a profile the rule" that growth increments are added only at is to be maintained. This growth forms the occlu- the basal margin. He also noted that this edge dent edge, and it is anticipated in other pyrgoma- reaches its maximum development in H. mon- tines such as Trevathana dentatum(Darwin, 1854)

ticulariae, and that the balance of the plate, and Savignium crenatum (Sowerby, 1823) where

and is less than half of the show representing the original tergum scutum, one plates apical growth,

"... reduced to a mere basal edge or border" (Dar- as opposed to more than one half in Hoekiini win, 1854: 356). However, his description of H. (Soong & Chang, 1983: 247).

inter- monticulariaeis ambiguous, insofar as orientation The occludent margins of the scutal plates and manner of growth are concerned, because he lock to varying degrees in nearly all balanomorphs. stated the itself is with theocclusal surfaces rela- "... scutum curved, a slip, In primitive barnacles are along the true occludent margin (best seen at the tively smooth, as they are in Hoekiini. Interlocking

rostral end), lying in a differentplane from the rest closure of the occlusal surfaces has obvious adap-

of the valve ..." (Darwin, 1854: 373; emphasis ad- tive value in barnacles exposed to predators.

Hoekiini is ded). This "slip" is depicted in his figure below the However, for there apparently no adap-

"basal edge or border" (pi. 13, fig. 5f). From this tive value in possessing such a mechanism because

one might be led to believe that there are two occlu- of the minute orifice which is covered by a cloak of

dent margins, but the "true" margin is simply the coral tissue.

"original" margin to which growth increments The unexposed upper margins of the opercular producing the occludent edge have been added. plates in Hoekiini do not occlude completely. Only

As Darwin noted, opercular growth in Hoekiini the medialoccludent edges are brought into apposi-

the of in other tion. On the rostral and carinal ends the contravenes general concepts growth edges balanomorphs in that major growth increments are diverge laterally leaving distinct, somewhat ellipti- made along the occludent rather than the basal cal openings or fenestra, with the rostral larger than margins. Furthermore, the modified scutal adduc- the carinal (Fig. 2). A similar situation exists at the tor and muscle carinal end of the in Nobia ridge tergal crest (see below) wrap operculum grandis around both sides ofthebasal margin of each plate, Sowerby, 1839 (see Ross & Newman, 1973: fig. and where they contact theexterior surface there is 12C). The fenestraeare simply the free spaces adja- a narrow, shallow sulcus (Fig. 9a). It is along this cent to the hinge points formed by the tergal and juncture, effectively the original basal margin of scutal extensions of the valves. the plate, that the arthrodial membrane attaches In those balanomorphs having scuta with inter- the plates to the sheath. locking occludent margins, the plates are slightly

Why have Hoekiini reversed the manner by bilaterally asymmetrical. This results from inter- which increments growth are added to the opercu- digitation of basal growth increments along the oc- lar valves? It has to do with the manner of growth cludentmargin. Although the occludent margins do of the wall. Monometric shell growth characterizes not interlock in Hoekiini, the plates are asymmetri- - Contributions to Zoology, 65 (3) 1995 159

cal. The apico-medial portion of the plate, the oc- mate. The opposed spatulate distal extremities in

for cludent edge, is semicircular in outline, and when A. chuangi may serve as pivotal or hinge points

form the but how the in the plates are in apposition they a hemisphere. plates, inequality length is com-

But, one side is usually more inflated than the pensated for is unknown. other, and the disparity in length between the two

of plates may arise from asymmetry the wall.

In H. monticulariaethe opercular plates average Body shape and size four to five times longer than the rostro-carinal di-

in whether the ameter of the orifice. One specimen, with a rostro- Shell shape pyrgomatids dictates

diameterof 0.35 has that will be housed within the confines the carinal orifice mm, plates body of wall, are 1.7 mm and 1.2 mm long, respectively (Table partly within the walland basis, or essentially with-

within II). Thus, a minute orifice has not precluded the in the basis. In Hoekiini the space available need for proportionately large opercular plates. the confines of the conical portion of the adult wall

for that sufficient exists for of This is because they are required body and mus- suggests only room part cle attachment. the oral cone and terminal cirri. The balance of the

in the basis The scutal portion of the plate lacks depressions prosoma and thorax reside (Fig. 3). for the adductor and the lateral and rostral depres- According to Anderson (1992: 290) the same rela- sor muscles. However, in H. monticulariae, along tionship exists in Cantellius euspinulosum (Broch, the basal margin there is a medially projecting 1931).

is attachment site for the the is than flange (Fig. 9a) that the In many pyrgomatids prosoma larger lateral depressor. The basal portion of the scutal the thorax (Anderson, 1992: 299), and in many it is segment is greatly thickened and constitutes the ad- globose or spheroidal (Ross & Newman, 1969:253).

correlation:the ductor ridge. It serves as the site of attachment for Thus, there appears to be a reverse the scutal adductor muscle well the and the the smaller as as prosoma, larger more globose prosoma, and laterally extends well beyond the rostral angle, and more slender the thorax. According to Ander-

of the thorax thus forming a rostral point or tooth. The distal end son, concomitant with diminution of is in there in and and the tooth simple and spatulate H. mon- is reduction cirral size length, ticulariaeand E. chaos, but in A. chuangi it is fold- hence perhaps functional dependency on the cirral ed and inflexed medially. The terminationis the site net as a captorial feeding mechanism. of attachment for the rostral depressor muscle. The prosoma in large specimens of H. morten-

of lacks The the thorax in there is no clear dis- The tergal portion each plate a spur. seni eclipses size; basal margin at the tergal end is produced, but not tinction between these two regions because what is to the same degree as at the rostral end. The tergal leftof the thorax is engulfed by inflationof the pro-

thoracic end is a modified crest, or fusion of several crests, soma. And, the appendages are vestigial for insertion of the tergal depressor muscles (Figs. except for theaberrant cirrus VI (see section Cirri).

2, 9a, b). The distal portion of the tergal extension There is a distinct thorax in H. monticulariae, but may be simple and spatulate or complex and in- it too is small relative to the prosoma, and in large flexed medially. However, unlike the rostral specimens it is rotated anterioventrally. In this

depressor muscle, the tergal depressor muscle bun- orientation, the tips of the posterior cirri approach

beneath dles in H. monticulariae and P. aster are arrayed the apex of the oral cone directly theorifice along the length of the tergal crest. (Fig. 3). The thorax in Parahoekia is also diminu-

Darwin (1854: 373) stated that the "U-shaped" tive and positioned anterioventrally with the inflat- opercular plates of H. monticulariae, when viewed ed prosoma extending posteriorly beyond it.

of in section along the major axis "... form a bow Because of the low profile the wall and size of with the two ends curled rather abruptly inwards the body in Hoekiini, thebulk of the body must be

...". Because of the length disparity in large H. housed in the basis. When raised to the feeding po- monticulariae the "two ends" may not be proxi- sition, the oral cone or alternatively, the intromit- 160 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

be extended tent organ and/or cirrus VI, can out The tergal depressors originate on the ridge along through the orifice. thebasal margin of the tergal portion of the opercu-

lar plate, and they also insert on the wall. They are

thin and shorter than the lateral scutal depressors,

and dividedinto four Opercular musculature are three or distinctly separate

bundles in H. monticulariae. In P. aster they are

In balanomorphs there are four groups of opercu- divided into six or seven distinct bundles, and each

muscles lar and their size, shape and length vary appears to have a greater cross-sectional area than

1992: 1994: that found in (Anderson, 334; 74). Although opercu- H. monticulariae.The depressor mus- lar musculature than culature in Hoekiini follows may serve more one function a similar pattern as

(Anderson, 1981: 447), the primary role is in that depicted by Anderson (1992: 331) for type 3

the regulating aperture between the opercular pyrgomatids. plates. The three peripheral groups include the Because the shell in Hoekiini is covered by coral

and and the aberrant cirri paired rostral, lateral, tergal depressor mus- tissue, no longer serve a cap- cles. The fourth is the unpaired scutal adductor. torial function, the activities normally associated

muscles in The opercular balanomorphs may com- with opercular and cirral movements are largely

of the soft of . prise a large portion parts the precluded. However, in concert with the prosomal

But, they form an exceedingly minute percentage of muscles, the depressor muscles may create the the tissue mass in Hoekiini. pumping action that ventilates the mantle cavity

Inorder for the opercular plates to close, the scu- and aids in expelling larvae in ordinary balano- tal adductor, the most conspicuous muscle in morphs (H. Barnes, 1955: 114).

Hoekiini, must be more apical than the lateral

depressors; that is, above the pivot or hinge points.

It pulls the opercular plates medially, thus closing Trophic structures the aperture. Unlike that of other balanomorphs and this is pyrgomatids, muscle bundle broad, flat The oral cone, thus far known only in Parahoekia and thin rather than ovate in and holds being essentially cross Hoekia, a prominent position on the pro- section. It is attached to the rostral tooth the and its size overshadow the thorax and along soma, can basal margin of the plate, but does not extend all its appendages (Fig. 3). It was this difference that the to way the end of the tooth (Fig. 2). originally led us to the fact that Hoekia feeds on

The muscles the depressor open opercular plates, host tissue rather than capturing foodwith cirri ex- and can generally draw theentire operculum down- tended into the water (Ross & Newman, 1967: 255). ward. The lateral depressors in Hoekiini, although The labrum in both H. mortenseni and H. mon- moderately slender, are exceedingly long. In speci- ticulariae is only slightly sulcate rather than deeply

of mens H. monticulariaeand H. mortenseni they notched as it is in P. aster (Fig. 7a). In all three spe-

than three times the of the are more length opercu- cies the crest lacks setules and teeth. The mandibu- lar plates. The proximal half is striated muscle, lar palps, fused laterally to the labrum, are small, whereas the distal half is tendonous. The latter por- clothedwith setae of various lengths, and do not ex- tion anastamoses and inserts on the basis rather tend to the top of the labrum; that is, they tend to

the shell than on wall. In Parahoekiathis muscle is close the sides rather than the top of the oral cone

it is shorter smooth, not striated, and greater in (cf. Figs. 7a, 10a, 12a). cross section than in and it lacks the The of the mandibles that Hoekia, ten- structure suggests they dinous distal portion. serve more than one function. The cutting edge of

The rostral depressor muscles are more slender the superior portion is smooth, but parallel to the than the lateral depressors. Their origin is on the edge there are exceedingly long, fine ctenae that apical surface of the rostral tooth, and they extend extend well beyond the cutting edge. The inferior somewhat before wall. of the is divisible into laterally inserting on the portion cutting edge two Contributions to Zoology, 65 (3) - 1995 161

functionalregions armed with multicuspid spines, veloped in these biting parasitic is likely

functional it is in the upper of which projects beyond the inferior for the same reasons as primitive

angle. ascothoracidans.

Of the three regions of the mandible, the func-

tion of the superior portion is equivocable, but in

light of the lateral position of the palps, it likely Cirri

strains or brushes food particles back toward the

mouth. The medial portion, acting much like a Balanomorphs, except those noted below, are

coarse-toothed saw, shreds coral tissue, maintains characterized by six pairs of biramous thoracic ap-

access to the "outside", and triturates large food pendages which fall into two functional groups. In

particles. The inferior angle, armed with smaller balanoids the anterior three pairs are maxillipeds

spines, transfers food to the mouth. that clean the posterior three pairs of food and

cirri Maxillae I have an unbroken array of equal transfer it to the trophi. Furthermore, the

length spines in H. mortenseni and P. aster but respond to mechanical stimuli, such as the direction

there are two discontinuities in thearray in H. mon- of water currents. They also serve some role in the

ticulariae, suggestive of functional groups (Fig. reproductive process (Ross, 1969: 248; M. Barnes,

1992: and function in lOd). The arrangement of spines, suggestive of a 520) may respiratory ex-

comb more than a saw, probably clean the mandi- change (Crisp & Southward, 1961: 290).

bles as well as assist in moving fine food particles The Hoekiini are unusual in that the cirri are

to the mouth. reduced and aberrant. Those of H. mortenseni

Maxillae II in Parahoekia, densely and darkly represent the most extreme case: the first three pairs

pigmented, are reduced to a mere knob-like promi- are wholly vestigial; cirrus IV is malformed and

nence surmounted by a few short setae (Fig. 7f). tipped with a few short setae; cirrus V is question-

They are fused to a large, unpaired, laterally com- ably biramous, unsegmented, essentially devoid of pressed, median languette (hypostome), similar in setae, and only slightly longer than the coxopodite appearance to that found in the ascothoracidan of cirrus VI, theramiof which are unequal in length

and the Synagoga Norman, 1913, and related genera (Nor- shorter than protopodite. man, 1913: 163; Moyse, 1983: 170; Grygier, 1984: The cirri of H. monticulariaeand P. aster (Fig.

168). Although Hoekiini and ascothoracidans have 7h) are more readily identified, but they are also an attenuatedoral cone and mandibles and maxil- aberrant. The location of cirri II and III is normal lae with saw-like edges, those of S. mira (Norman, in P. aster whereas in H. monticulariaethey attach

1913:161) are terminally accumulate, and therefore medially, just posterior to the mouth field, and they better adapted for piercing as well as rasping the are somewhat reminiscent of the mouth cirri (Fig. host's tissue. Flanked by the maxillules, in Hoekiini 10g, h) in acrothoracicans (Tomlinson, 1963: 268). the languette forms a seal to the posterior border of In this position they are flanked and "hidden" by theoral cavity. The apical border is higher than the the protopodite of cirrus I, the coxopodite of labrum times and adjoining appendages, and the periph- which, some 20-25 larger than the basipodite ery is armed with a single row of short, slender, sim- (Fig. 10g), likely retains its "normal" size because ple setae. In H. monticulariae (Fig. lOf) and H. of the presence of the oviducal gland and female mortenseni(Fig. 12f) the languette is comparable in genital aperture (Walley, 1965: 118). All three of size to the adjoining trophi and does not broadly the maxillipeds have discernible protopodites, but close the back of the mouth field as it does in P. the remaining segments are variously fused or in- aster. The presence of a languette that equals or discernible; they are distinguished only by indenta- dwarfs the tions remaining appendages is apparently a in the cuticlerather than by true articulations. synapomorphy that unites the Hoekiini, because Cirrus IV in P. aster, only slightly separated from they are the only balanomorphs known to possess the posterior two pairs, is aberrant, has a poorly such a structure. Why a hypostome is so highly de- segmented protopodite, and equal-sized rami that 162 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

are shorter than the protopodite (Fig. 7h). In H. Darwin (1854: 373) described the inner walls of H. monticulariaecirrus V is essentially fused to cirrus monticulariaeas being smooth, but noted that "...

one-fourth its VI (Fig. lOj), and about length, but they are perforated by many quite irregular, small the rami of cirrus which have the of been VI are equal to or slightly longer orifices, appearance having than the protopodite. All of the posterior cirri formed by some boring animal ..." His illustra- have setae, but these, as in H. mortenseni, occur tion (1854: pi. 13 fig. 5b) shows a highly pock-

of almost wholly on the greater curvature the marked surface, which we have not observed. The rami. lacunae may be circular, ovate or irregular in out-

On the basis of the foregoing it is evident thatthe line(Fig. 19b, c). Some may be confluent, but more maxillipeds no longer function in the transfer of commonly they are discrete, and in Ahoekia they

the the three food to trophi. Similarly, remaining are concentrated along the lateral portions of the pairs cannot function in a captorial manner. Conse- shell (Fig. 13c). "More or less irregular" best

quently, they must serve other functions. Severe describes their spatial distributionbecause there is size reduction inall but cirri VI, which flank the in- no fixed pattern from one specimen to another of tromittent the Each lacuna leads organ, suggests that the latter serve a same species. to a lumen,

function in these which in into reproductive species. turn opens a narrow, partially en-

closed, circumferential channel that lies between

the marginal fringe and the distal margin of the

Lacuno-circumferential system outer lamina. In Ahoekia there are ancillary con-

nections from the circumferential channel to the

Among the unique and obvious structures in Hoeki- periphery of the marginal fringe (Fig. 13a, d). ini is theexternal carinal ridge. It is short, common- Internally, access to the abandoned carinal sinus ly does not extend to the edge of the outer lamina, is generally through a carinally situated lacuna in and is a seemingly anomalous protuberance (Figs. these species. The same lacuna also leads distally to

6b, 8b) that bears little resemblance to the ridges or the circumferentialchannel. Eohoekialacks the en- ornamentationof other barnacles. closed lacunal interconnections, but access to the

Under transmitted light, the ridge in Hoekia ap- circumferentialchannel, and in turn to the sinuses,

house hollow within pears to a cavity or sinus the is via shallow, basal and distal sulci that occur be-

and revealed shell, grinding a cavity. In Eohoekia, tween the digitations of the marginal fringe. The a rostral as well as a carinal sinus can be observed same situation prevails in Parahoekia, but the sulci when specimens are candled. are deeper and more pronounced, and in one in-

A prominent rostral and a carinal ridge occur in stance a sulcus has been closed by secondary calcifi-

Parahoekia, and there are indications of lateral cation. All of these connections and interconnec-

between them. These tions transmitted ridges interspersed ridges ap- are traceable by light, grinding

to be hollow under transmitted but the the and/or lash pear light, shell, by inserting a probe (an eye opacity of the shell precludes further study. The is excellent) into the passage and observing where it broad, low carinal ridge in A. tanabensis also emerges around the fringe (Fig. 18a). houses a sinus, but grinding revealed that it was On the basis of the foregoing it is evident that the

These circula- and the lacunae but filled secondarily. sinuses represent carinal ridge (sinus) are two tory channels that were abandoned as the circum- aspects of the same system, but which system? We ferential circulatory channel moved peripherally inferthatthis extensive and elaborate interconnect-

those during ontogeny. In forms with lacunae, ing system accommodates part of the circulatory

is in there generally a well-developed carinal lacuna system, connection with pronounced elaboration connecting with the carinal sinus (see below). of the lateral margin of the shell which is in direct

The of development lacunae is a unique feature contact with host tissue. Both Eohoekia and Para- in Hoekia Ahoekia. and These occur below or distal hoekia have only the circumferential and carinal to the basal margin of the sheath (Figs. 18a, 19a). portion of the network, and thus the circumferen- Contributions to Zoology, 65 (3) - 1995 163

Fig. 18. a, Hoekia monticulariae (Gray, 1831),internal view ofshell; note essentially symmetrical arrangementoflacunae, with eyelash traversing shell from inner to outer surface; topotype, SIO no. C-8375.

Eohoekia is the b, chaos n. sp., fragment of paratype, USNM no. 259979, view of circumferential channel; flat portion outer lamina with the reflexed marginal fringe to the left and rear. 164 A. Ross & W.A. Newman A coral-eating barnacle, revisited

tial channel must be vessels that al for it supplied by pass system respiratory exchange. However, ap-

frombeneath the shell through the marginal fringe. parently has been elaborated beyond all needs of

this for The lacunate system shortens route the the barnacle as far as respiration is concerned.

the cir- Goreau al. the hemolymph, between the body proper to According to et (1970: 169) zoo- cumferential channel, but it potentially separates xanthellaewithincolonial corals produce more oxy-

afferent circulation the and efferent which would gen in daylight than can be consumed by the host,

for elaboration of the allow greater marginal and metabolic waste products are taken up by the

fringe. zooxanthellae.Although, by virtueof their zooxan-

The only recent study on the balanomorph hemo- thellae, reef corals behave much like plants, the lymph system (Burnett, 1977) provides a back- zooxanthellae "function as intracellular lungs and

which the be ground against present situation can kidneys", thereby reducing the host's work involv- interpreted. Although Burnett's study was on a ing "respiratory gas exchange and transport of higher balanid, Megabalanus californiens (Pilsbry, waste products". Elaboration of a marginal fringe

it is 1916), likely that barnacles having a solid and and its proximity to the host coral tissue fosters our

wall differ in detail from that wholly concrescent only hypothesis it functions as a metabolicpathway this general plan. Although we could not trace the to and from the host, and hence, the development vessels themselves, from the preserved material it and elaboration of a lacuno-circumferential cir-

that appeared branches from thebasal (= inferior) culatory system to better serve these functions. circumferential that in vessels, may be observed

Megabalanus, formed the lacuno-circumferential channel network in the lacunate Hoekiini. It is also Reproduction evident that there is a major carinal vessel, previ-

in ously unrecognized balanomorphs, that supplies Balanomorphs are fundamentally cross-fertilizing the Based reexami- tergal depressor muscles. upon hermaphrodites. Although there is no elaborate nation of Burnett's in the collections of specimens courtship behavior, individuals acting as receptive

Institution Scripps of Oceanography, this vessel females induce surrounding individuals to act as evidently not only supplies these muscles, but also males at the time of mating (H. Barnes et al., 1977: the branchiae. 197; Crisp & Southward, 1961: 288). The three spe-

In cirral cies general, movements associated with ac- we examined have a well-developed intromit- tivities affect circulation of fluids & do all body (Crisp tent organ as presumably Hoekiini, and

Southward, 1961: 292). Obviously, the presence of therefore it appears that cross-fertilization remains

reduced cirri greatly necessitates some other integral to their reproductive biology. The pedicel mechanism, rhythmic or otherwise, to assure circu- of the penis arises between cirri VI, extends to an in- lation. In the absence of cirral movements, Blatch- flated basidorsal point from where the probosci- ford (1970: 914) suggested that "... rhythmic con- form portion turns abruptly anteriorly and comes tractions of the body wall in theregion between the to lie along either side of the prosoma. The pedicel adductor scutorum muscle and the oral cone may is almost the same height as the protopodite of cir-

in the serve to produce regular pressure pulses body rus VI in H. monticulariae (Fig. 10k), but it is half fluids which aid circulationof the and the of the may blood", height coxal segment in H. mortenseni this is likely the case here. (Fig. 12g). In all of the dissected specimens of H.

Balanomorph respiration takes place by diffu- monticulariaethe paired seminal vesicles were swol- sion the the through body surfaces, mantle, and ex- len and readily visible through the body wall. They

thereof such the branchiae. of tensions as Because were whitish, vermiform, large, and traceable to the reduced opercular area for vascularization, cou- where they unite in the pedicel of the penis. pled with limited exposure to the external milieux The length of the penis may vary with the via theorifice noted and the absence of branchiae, Hoeki- reproductive season, thus the differences ini may depend on the highly vascularized peripher- may not be significant (M. Barnes, 1992: 518). In Contributions to Zoology, 65 (3) - 1995 165

Fig. 19. Hoekia monticulariae (Gray, 1831): a, topotype, SIO no. C-8375, internal view of shell, carinal end at left; b, enlarged view of paired carinal lacunae shown in a, above, at left center; c, enlarged view of single lacuna shown in a, above, top right; note that wall of lacuna is perforated with pores that accommodate setae that exit on outer wall of shell.

H. mortenseni it is less than twice the length of cir- Munn et al. (1974: 96) to the setae that clothe the

but it is than twice in in balanoides rus VI, more the length H. penis Semibalanus (Linnaeus, 1766). monticulariae and P. aster. There are simple, Only cirrus VI appears truly functional in Hoekia minute setae scattered sparsely over the surface, ex- and Parahoekia, and then perhaps only during

in cept P. aster there are more setae, and at the dis- mating (see also section Feeding and nutrition). tal end there are four equally spaced clusters of four Distention of the penis in balanids does not take or five short, slender setae in each (Fig. 7i). A place until the cirri have been fully unfurled, and

function chemosensory has been attributed by consequently their presence may be requisite for 166 A. Ross & W.A. Newman A coral-eating barnacle, revisited

copulation (Crisp & Southward, 1961: 288). the orifice, the oral cone, now in a protruded state,

In tropical or warm-water barnacles breeding is could enter the host's gastrovascular canal system.

and it not uniform throughout the year, may be If it did, it could divert gastrovascular fluids from

asynchronous in a population (Achituv & Barnes, the coral to itself.

1978: 316; M. Barnes, 1989: 143; 1992: 493). Our It is instructive to note a somewhat parallel situa-

samples were too few and too small to add to this tion involving the mytilid Fungiacava eilatensis generalization, but we noted the following. One Soot-Ryen, 1969 and various monostomatous

specimen of H. monticulariae collected in Decem- corals of the family Fungiidae. It resides in a cavity

mid-summerin the southern within of ber, hemisphere, was the skeleton the host coral, and its en- brooding eggs. They were somewhat ovate and larged inhalant siphon opens directly into the coel- average 0.3 mm in diameter. The specimens of H. enteron of the host. Its diet consists of phytoplank- mortenseni taken in October, early spring in the ton, zooxanthellae, and other particles foundin the southern hemisphere, were barren. coelenteron (Goreau et al., 1970: 164; Patton,

We recovered from the sole specimen of P. aster, 1976: 11). The footof this mytilid is involved in the

in collected May, or late fall in the southern food selection process as well as serving some sen- hemisphere, some pre-stage I nauplii that were eyed sory role. This suggests that retention of cirri VI in and filledwith and oil The yolk droplets. specimen Hoekia may also have a sensory role in feeding as was received detached from the host coral and well as in reproduction. therefore estimate it we cannot brood size, but did In balanomorphs, the anus is situated at the base

that effort of the last thoracic above the not appear reproductive was particularly segment, immediately

If this is curious high. not, because parasites nor- pedicel of the intromittent organ (Darwin, 1854:

' mally have a high reproductive output. 87). There appears to be an "anal depression' in H.

limitedobservations Our suggest that population monticulariaeand H. mortenseni, but no anal flaps densitiesin Hoekiiniare low. However, considering or opening could be detected. The same also applies the in difficulty obtaining preserved material, and in P. aster, except that we did not detect an anal the likelihood of their being washed off in the depression. preparation of coral skeletons, they could be rela- Acrothoracicans (Apygophora), which likewise

feed fine lack anal tively common. apparently on particles, an open-

ing, andTomlinson(1969: 19) summarized the vari-

ous aspects of how metabolic wastes are voided in

and nutrition Feeding such forms. Therefore, if an anus is lacking in

Hoekiini, it is the bar- not unprecedented among

As noted earlier, the absence of normal cirri, the nacles. nature of the trophi, and the presence of nemato-

in the cysts stomach of Hoekia led to our inference thatHoekia feeds on coral tissue (Ross & Newman, Host specificity and exploitation

1969: 255). A somewhat similar scenario is evident in the ascothoracidan Isidascus 1983. This Moyse, The faviinidHydnophora Fischer, 1807, which was minute, piercing and sucking, gall-forming parasite widely distributed in the Cretaceous, is presently of octocorals feeds on host tissues which are prob- limited to the Indo-West Pacific region (Veron,

"... ably continually proliferating to replace any 1993b: 428). This coral is unusual in that corallite used the by parasite" (Moyse, 1983:175). However, centers are arranged around protuberant collines or we now believe that in the Hoekiini ingestion of hydnons, theresult of circummural budding. Short coral tissue may not be the sole source of nutriment. tentacles surround the base of each hydnon, with

In the preserved specimens we have examined, one tentacle between each pair of sclerosepta the is wall totally covered by living coral tissue (Fig. (Veron, 1993b: 428). Ogawa & Matsuzaki (1992:

in fact it to mimic 3); appears a hydnon. After rasp- 97) reported the apparently facultative occurrence ing and consuming the tissue immediately covering of Creusia indicaAnnandale, 1924 ( = Galkinia in- - 167 Contributions to Zoology, 65 (3) 1995

but hoekiines from the cleaner water of the reefs". Similarly, in dica) on Hydnophora spp., usually

this coral. Tanabe Japan, Hiro found coral- are the only pyrgomatids encountered on Bay, (1935: 21)

abundant where What it is about the unique morphology of this inhabiting barnacles to be rather

silt-laden Further- coral that favors hoekiines, but tends to exclude there was an influx of waters.

thickness of the the Hiroa Ross & Newman, 1973 other pyrgomatids, may be the more, pyrgomatid

zone of the coral tissue in the "valleys" between the hydnons in was first discovered in the high energy which the wall of the barnaclebecomes suspended, reef front (Ross & Newman, 1973: 153) where the

to the and to which its highly vascularized marginal fringe corals are subject to heavy surf as compared

would be is applied. rest of the reef front. Thus it not sur-

is if the larvae of barnacles Newman et al. (1976: 73) note that there a good prising coral-inhabiting

to excellent correlationbetween being morphologi- capitalize on localized or micro-disturbances.

limitednumber of Veron cally specialized and exploiting a According to (1993b: 428) Hydnophora

case here exesa is more common than H. rigida, which is genera of host corals, and this is the (see

and microconos which is section Biogeography). Furthermore, based on our sometimes common, H.

reef and limiteddata for the Hoekiini, species in a particular least common on protected slopes especial-

in Because occur in area settle on one host species, and wide-ranging ly lagoons. they essentially pro-

than low environments, hosts are generally exploited by more one spe- tected, or energy hydnophorans

3 4 not be readily subject to wave perturbation. cies. But so far it appears that only or possibly may

such freshwater of the 22 species of Hydnophora (Veron, 1993a:77; But other forms of stress, as run

table I) have been identified as hosts of this off and/or siltation, or perhaps prédation, may

Hoekiini established barnacle. play some role in becoming on

Success in becoming established on a host may them.

and settle on the coenen- depend on several factors. Coral-inhabiting Pyrgomatid cyprids directly

other symbiotic barnacles apparently have patchy chyme of the host (Duerden, 1902: 38; Gravier,

1971: Utinomi, 1943: distributions and settlement may be due large- 1911: 211; Moyse, 127; 16),

of otherwise but what involved or what time of ly to some sort perturbation, they cues are day

been would be far more common. For example, cyprids cyprids settle remains unknown. It has sug-

of the acrothoracican Cryptophiallus melampygos gested that symbionts achieve some immunity to

of of substances Berndt, 1907 preferentially settle on those areas of the defenses the host by absorption

and if these aid the the host mollusk where the periostracum has been secreted from them, so may

eroded (Batham & Tomlinson, 1965: 143). Gomez settling symbiont (Patton, 1976: 29; Ogawa &

(1973: 163) found that cyprids of Conopea galeatus Matsuzaki, 1992: 91).

nocturnal (Linnaeus, 1767) preferentially settle on those areas Hydnophorans, except H. exesa, are

and it that of the host gorgonian that have been striped or feeders (Veron, 1993b: 428), seems likely

Hoekiini when the host is browsed by an ophistobranch mollusk. In the clas- cyprids settle not actively

rhi- Unlike where each sic study by Ritchie & Hoeg (1981: 343) on the feeding. most corals, polyp oc-

in there dis- zocephalan Lernaeodiscus porcellanae F. Müller, cupies a corallite, hydnophorans are no

tinct corallites. Mouths occur where several 1861, host crabs living among rocks, as compared may

incidence of tentacle-encircled but to those among mussels, had a higher hydnons come together,

of there be mouth in similar parasitism presumably because more frequent may no visible a array.

There sites for the settle damage to their cleaning appendages by shifting are two possible cyprid to

tissue between eitherin mouthof the rocks. The observation of Duerden (1902: 38) on a on hydnons, a

West-Indian coral, Siderastrea radians (Pallas, host or in the intervening areas. Although the bar-

nacle to mimic a rather than a 1766), indicates that, among others, pyrgomatids grows hydnon polyp,

mouth of "... are particularly numerous on specimens ob- it could have settled in the a preexisting

tained from shores with those well the between them. muddy as compared polyp as as open space 168 A. Ross & W.A. Newman A coral-eating barnacle, revisited

Biogeography might not reflect the same diversity trends seen in

the other its setose-feeding allies. Hiroa, on hand,

the excluded because it known from the In a study of diversity trends in Pyrgomatidae was was only

of involv- & 1973: and (Newman et al., 1976) a number indices type locality (Ross Newman, 153), one

ing 10 genera were identified, quantified, and would expect such a moderately generalized form

much ranked. The principal ones were: (1) structural to be relatively wide-ranging. We now know

about both of these and ask diversity (SDI), (2) host exploitation (HEI), (3) ge- more genera one can

ographical diversity (GDI), and (4) generic species what bearing this new knowledge has on these

richness (GSR). Six rank order comparisons were hypotheses.

made and rank-difference coefficients r calculated As for the new concerns its d Hoekia, knowledge

The results and and (cf. Tate & Clelland, 1957). were sum- structure, systematics, distribution, the

marized in matrix which is and first of these bear SDI. Hoekia a updated repro- two parameters on

duced here (Table III; certain recently proposed is now not only divisibleinto four genera compris-

generic level taxa have yet to be evaluated, and ing the Hoekiini, but the basic wall and opercular

therefore these have been omitted in the tables and valves in this group are now recognized as the most

discussion that follows). modified of the pyrgomatines. Therefore, the

It can be seen in Table III that the three original Hoekiini exchange places with Pyrgopsella (Pyr-

principal rank order comparisons (HEI, GDI, and gopsellini, the only pyrgomatine occurring on

than GSR with SDI) support to a greater or lesser extent sponges rather corals) because they have the

three general hypotheses, viz., that the relatively most advanced walls and operculae among the Pyr-

unspecialized genera tend (1) to exploit a greater gomatidae (Table IV). Furthermore, several of the

variety of hosts (HEI), (2) to have broader new genera appear to have relatively restricted

geographical distributions (GDI), and (3) to be ranges despite the wide geographical range of the

richer in than tribe. Based the of species (GSR), specialized genera. on proximity some species

it will observed that in achieve of However, be order to (Gulfs Suez and Aqaba; Fig. 20) we predict that

reasonable coefficients and concomitant probabili- newly discovered populations will more likely re-

ties, it was necessary to exclude the monotypic present new species than range extensions.

from for feeder with genus Hoekia the second, and the monotypic As Hiroa, a normal setose a

modified it first discovered genusHiroa from all three of these rank-order com- moderately shell, was

parisons with SDI. entombed in fragments of coral blasted from the

Hoekia was removed because it was the only reeffront on Truk (Chuuk), Caroline Islands, more

member of the entire family known to be wholly than 20 years ago. Since then it has been found to

parasitic (Ross & Newman, 1969), and therefore be widely distributed: Taiwan (Soong & Chang,

Table III.Summary of rank difference coefficients r and p from Newman et al. with calculated r and ( d) values, (1976: fig. 2a) newly d

values forrank-order comparisons of GDI and GSR with with rather than without, Hiroa and Hoekia in parentheses p HEI, SDI, , given below the original values. For further explanation, see text.

HEI GDI GSR

SDI World-///roo World-Hiroa + Hoekia World-Wiroa

r = .485, < .20 r = .037, p < .00 = .443, p < .00 d p d

(r -725, p< .01) (r =.585,p<.05) p<. 10) d~ d

HEI - Worldwide Worldwide

r = .77, < .05 r = .83, p < .01 d p d

GDI - Worldwide

r = .60, p < .10 d - Contributions to Zoology, 65 (3) 1995 169

Table IV. (A) Correlation between structural modifications of Table V. Rank order comparisons between structural diversity

walls (W) and valves (V) based on the rankings of Newman et (SDI), host exploitation (HEI), geographical diversity (GDI),

al. emended reflect the diversification of the Hoekiini and species richness The and new- (1976) to generic (GSR). r d p < values,

herein. The rank difference coefficient is ly calculated and from Newman et 1976 (inparentheses), are reported (rd) now al.,

.896, p < .01, considerably more significant than that of .732, discussed in the text.

p < .02 in the previous study. (B) Structural diversity indices,

SDI, formed by multiplicationof W x V and ranked according- SDI HEI GDI GSR

ly. The SDI ranking can then be used to test certain hypotheses

Ceratoconcha 13.0 6 6 9.0 taken in Table V; the r values are summarized in Table III. up d Cantellius 12.0 13 11 13.0

A B Hiroa 11.0 7 10 2.0

Creusia 9.5 10 7 12.0

W r V r SDI r Megatrema 9.5 12 13 7.5

Savignium 7.5 11 12 10.0

Ceratoconcha 1 2.5 1 1.0 1 13.0 Nobia 7.5 9 8 11.0

Cantellius 2.0 2 2.5 2 12.0 I Pyrgoma 6.0 8 5 2.0

Hiroa I 2.5 3 4.0 3 11.0 Pyrgopsella 5.0 3 4 5.0

Creusia 1 2.5 4 6.5 4 9.5 Eohoekia 3.5 3 3 5.0

Megatrema 2 5.0 2 2.5 4 9.5 Parahoekia 3.5 3 3 2.0

Savignium 3 7.0 4 6.5 12 7.5 Hoekia 1.5 3 9 7.5

Nobia 3 7.0 4 6.5 12 7.5 Ahoekia 1.5 3 3 2.0

6.0 3 7.0 5 9.0 15 = Pyrgoma r & < .725, .01 .585, .05 .497, .10 d p

4 9.0 4 6.5 5.0 Pyrgopsella 16 & = .20 .00 (rd p < .485, .037, .443, .00)

Eohoekia 5 10.5 6 10.5 30 3.5

Parahoekia 5 10.5 6 10.5 30 3.5

Hoekia 6 12.5 7 12.5 42 1.5 and GSR rank-order comparisons with SDI this ap- Ahoekia 6 12.5 7 12.5 42 1.5

pears largely due to adjustments made to the SDI

stemming from the discovery of the complexity of

1983: 1990: its four 248), Japan (Ogawa& Matsuzaki, 75), the hoekiine shell and divisioninto genera.

and other localities in the western Pacific as far west On the other hand, considering the robustness of

as the Marshall Islands and south as the Gilbert rank-difference correlations, it is reassuring to see

Islands and New Caledonia, and into the Indian that despite the mismatch due to Hiroa being

Ocean as far as Mauritius (Achituv & Newman, in monotypic, including it as well as Hoekia in the

prep.). Thus, there is no longer an excuse to exclude analyses, produces r and p values that support the d it from the comparisons. However, in spite of being hypotheses that (1) generalized genera inhabit a

it is considered the wide-ranging, same species, greater variety of hosts, (2) have greater geographic

Hiroastubbingsi Ross & Newman, 1973 (Achituv & ranges, and (3) include more species than special-

in 2 Newman, prep.). ized (Table V; r provides an approximate genera d

for the record the Except of astrocoeniid Stylo- percentage of the variation of either variable ex- phora cited by Ogawa & Matsuzaki (1992: 101; plained by its correlate, 52%, 34%, and 25%,

Ogawa, pers. comm.), Hiroa is known only on the respectively).

astrocoeniid of Astreopora (reexamination a frag- ment of the host skeleton from the type locality in- dicated that it also belongs to this genus; S. Cairns, Affinities and relationships pers. comm.). Whether Hiroa turns up on hosts other than and remains of the Astreopora Stylophora to One unique characters that unites the genera be but in consideration of the and them from seen, its moderately comprising Hoekiini, separates

this specialized valves appears unlikely. other Pyrgomatinae, is the development of a margi-

Including Hoekia and Hiroa in the rank-order nal fringe and concomitantly elaborationof the cir-

of and GSR with SDI cumferential of comparisons HEI, GDI, part the circulatory system. Each

a result In the HEI the Hoekiini has this in produces significant (Table V). genus within developed a 170 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

Fig. 20. Distribution of the species of Hoekiini. All of the species appear to be limited in their distribution,i.e., only a single species

occurs at each location, even though the host, Hydnophora, ranges throughout the Indo-West Pacific.

but differentmanner (Fig. lb-e), and there is a clear thicker, slightly narrower more complex margi-

evolutionary trend. In the more generalized taxa, nal fringe places E. nyx at a higher level of organi-

Eohoekia and Parahoekia, foramina, distal to the zation. Deep incisions in the fringe of E. nyx seem-

outer lamina, appear to develop haphazardly in ingly coalesce, but theborders are not fused. In one

where foramina various positions and are ephemeral, and access to fragment of a marginal fringe, are

the around the inner of circumferential system must come from apparent only on surface, coalescence

beneath the shell. In the advanced taxa, Hoekia and adjoining fringe elements occurred, but there are

Ahoekia, haphazardly to more or less uniformly no visible or traceableinterconnections to the exter-

circumferential channel. these spaced permanent lacunae and passages in the wall nal surface or Why

not only provide direct access to the circumferential foramina have been abandoned is not clear, but it

sealed system from the interior, but make it possible to is obvious that they have been partially or

afferent and efferent vessels. filled separate secondarily.

ovate Eohoekiachaos possesses the greatest numberof Parahoekia shares a conical peritreme and

shell with but the shell of the former is plesiomorphic characters. It has an essentially coni- Eohoekia,

cal shell coupled with an ovate outer lamina, the set apart by a prominent rostral as well as carinal

retention of which reflects the condition seen in and subsidiary ridges inaddition to a flat, marginal

both more primitive Pyrgomatini such as Cantellius. The fringe. Under transmitted light Eohoekia has

within the same morphology prevails in the somewhat more a rostral and a carinal sinus shell,

specialized E. nyx, where ontogenetically, an ovate presumably heralding the development of the ros-

and conical shell persists untiltherostro-carinal di- tral and carinal ridges in Parahoekia. Another

the ameter is some 4.0 mm. Although E. chaos and E. difference between these genera concerns how

circumferential In nyx undoubtedly share a common ancestry, the ridges relate to the system. both, Contributions to Zoology, 65 (3) - 1995 171

Table VI. Character states for analysis of relationships in the Hoekiini and outgroups.

1. Shell configuration: conic - 4 plates (0); modified conic - concrescent (1); flat - concrescent (2).

2. Shell to basis interface: interlocking (0); not interlocking (1).

3. Length of shell to orifice ratio: <1 : 10 (0); > 1 : 15 (1).

4. Carinal ridge: absent (0); present (1).

5. Carinal ridge development: absent (0); obscure (1); prominent (2).

6. Outer lamina outline: subcircular (0); elongate-oval (1); irregularly ovate (2); lobate (3).

7. Outer lamina configuration: flat (0); reflexed or ascendant (1).

8. Outer lamina elevation to orifice: same (0); lower (1); higher (2).

9. Marginal fringe: absent (0); present (1).

10. Marginal fringe type: absent (0); filigreed (1); lacunate (2).

11. Marginal fringe elevation: absent (0); flat (1); reflexed (2); ascendant (3).

12. Marginal fringe width: absent (0); broad (1); narrow (2).

13. Lacuno-circumferential system: absent (0); diffuse (1); concentrated (2).

with with without 14. Opercular plate configuration: separate, tergum spur (0); concrescent, tergum spur (1); concrescent, tergum spur

v4/*

15. Limbus occludens: absent (0); poorly developed (1); well developed (2).

16. Feeding method: setose (0); parasitic (1).

17. Labral crest: deeply incised with teeth (0); deeply notched without teeth (1); shallow-sulcate without teeth (2).

the sinuses were evidently functional channels dur- development of the peritreme in Ahoekia appears

with earlier which ing early ontogenetic development, access to to be arrested at an stage, suggests

the circumferential system and marginal fringe that ontogenetic development of the marginal

from beneath the shell, but for some reason the ros- fringe begins earlier than in Hoekia. tral and carinal ridges have taken on an unusual im- It is from a relatively narrow and essentially flat,

in Parahoekia. but thick such that of Parahoek- portance marginal fringe, as

Although somewhat comparable in size to Eo- ia, that the ascendant, or vertically elaboratedand hoekia, the shell of Parahoekia has a lobate outline complex fringe of Hoekia was derived. Within and a well-developed, narrow, marginal fringe that Hoekia there appears to be a general trend towards is lower than the orifice. Coalescence of proximal a low or reduced peritreme: inH. fornix a high peri- elements in the marginal fringe of P. aster results in treme and narrow marginal fringe occur in parallel foramina, but their final outcome is unknown. and the progression through H. monticulariae to

The lacunate evolved and H. mortenseni results in low generaare more highly ultimately a peritreme the lacuno-circumferential system in Ahoekia and narrow fringe. Although the soft parts are represents a significant advance over Hoekia. In unknown in H. fornix, those of H. mortensenisug-

Hoekia this is of cirral in system more uniformly arranged gest a greater degree vestigiality compar- than in Ahoekia where the lacunae are commonly ison with H. monticulariae. Trophic structures in confluent and concentrated along the lateral mar- these latter two species are obviously similar, but

of is this gins the shell, and the marginal fringe higher may relate to comparable diets rather than in- and more compact or dense. In both, the peritreme nate similarities. remains essentially conical, but in Hoekia the cari-

nal sinus - ridge apparently is not subsequently filled Cladistics. There is a consensus, first promulgat- or sealed as it is in Ahoekia. Moreover, in Ahoekia ed by Ross & Newman (1973: 147), that Cantellius lateral expansion of the marginal fringe is arrested has the greatest number of plesiomorphies (four when it begins to ascend vertically, which provides wall plates and generalized opercular plates; see

directed toward and in Table and Therefore it is a greater surface area con- VI Fig. 21). from the level tact with the adjacent coral tissue. This represents of organization found in Cantellius that all of the a significant departure from Hoekia. Furthermore, genera withinthe Pyrgomatinae have been derived, 172 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

five of the species and therefore, for two genera we

have scored this character with a (?), but when a

member of character a genus has that state we con-

clude a priori that all of the species have it.

Among the 17 characters selected, there are

numerous multistate characters. To obviate prob-

lems inherent with continuous variables we estab-

lished character polarity using outgroup compari-

sons based on studies by Ross & Newman (1969,

1973), Newman & Ross (1976), Soong & Chang

(1983), and Anderson (1992).

Our character matrix, cladogram, consensus tree,

consistency index (CI), homoplasy index (HI), re-

tention index (RI), and rescaled consistency index

(RC) are presented in Fig. 21, and the characters

and character states are summarized in Table VI. It

will be observed that the results corroborate the

conclusions regarding affinities reached above.

Fig. 21. Character matrix, cladogram, consensus tree and in- dices (CI, consistency index; HI, homoplasy index; RI, retention index; RC, rescaled consistency index) for the Hoekiini. See Ta- Parallels with other parasitic thecostracans ble VI for the summary of characters and character states used in developing the above.

Hoekiini have adapted to living within tissue of the and therefore we selected this taxon as our first out- host coral, and thus they are internal parasites. group. Selection of a second outgroup, interme- Growth of coral tissue directly over the barnacle diate between Cantellius and the Hoekiini, posed conveniently places it where it can be eaten. The

that feeds oral problems stem from recent taxonomie changes barnacle by extending the cone and as- that the of this sociated the are beyond scope study. Con- mouthparts up through aperture, so

refer the second that and the basal tissue of the sequently, we to outgroup as they can rasp cut

Pyrgoma elongatum Hiro, 1931 (: 154), which also coral and carry it to the mouth. Moreover, not only has a concrescent shell and opercular plates. is theoral cone in a position to feed directly on coral

This preliminary cladistic analysis is based on tissue, it is likewise in a position to penetrate into relatively few characters only because of inadequa- the gastrovascular canal of the coral. Therefore, cies in the materials available. Moreover, it is based Hoekiini may also be able to feed directly on food almost wholly on shell characters, because it is the being collected by the host. shell that is most commonly encountered (Table This remarkable group of barnacles also appears

All of the characters VI). were unweighted because to be in a position to draw upon the host's resources

contain sufficient information to in additional they resolve rela- an way. This is accomplished by tionships among the taxa without biasing the data elaborationof the distal margin of the inner lamina set. and its loss of contact with the calcareous basis, ap-

In three instances (characters 14, 15, and 17) we parently to accommodateincreased vascularization

coded character of have a as present inall members and contact with the tissue of the host coral. We

when have observed its in direct evidence a taxon we presence only have no that there is metabolic ex- one or two of its members. Also, on the basis of change across the membraneseparating the tissues shell and opercular morphologies we have conclud- of the coral from those of the barnacle, but the ed a priori that none of the Hoekiini are setose feed- pronounced modifications of the shell makes the

have the labrum in ers (character 16). We not seen likelihood worth exploring. Contributions to Zoology, 65 (3) - 1995 173

is The lateral margin of the shell or carapace un- Acknowledgements usual not only in being hypertrophied and highly

but in the wall of the bar- vascularized, suspending For providing us with specimens, access to specimens in their nacle in the host tissue rather than locking it into a charge or other courtesies, we thank Torben Wolff, Universitets

Zoologiske Museum, Copenhagen; the late Huzio Utinomi, Seto calcareous basis, as in other pyrgomatids. In fact, Marine Biological Laboratory, Japan; S.H. Chuang, Depart- development of the calcareous basis in the region of ment of Zoology, National University ofSingapore; William K. potential contact has been completely curtailed. Emerson, American Museum of Natural History, New York;

and vascularized surface is stra- Thus an extensive Lev Fishelson and Yosi Loya, Tel Aviv University, Israel; D.T.

L.B. tegically located where it can absorb nutrients from Anderson, University of Sydney, Australia; Holthuis, Na- tissues of the host. tionaal Natuurhistorisch Museum, Leiden; J.E.N. Veron, Aus-

tralian Institute of Marine Science, Townsville; Toshiyuki The Hoekiini are in some ways paralleling the Yamaguchi and Kiyotaka Asami, Chiba University, Japan; Ber- evolutionary histories of the Ascothoracida and trand Richer de Forges and Georges Barbigant, Noumea, New histories perhaps the Rhizocephala. These appar- Caledonia; Alain Crosnier, Muséum National d'Histoire ently involved biting parasites becoming absorptive Naturelle, Paris; Mark J. Grygier, Fisheries Laboratory, Hiro-

but there One that shima University, Japan; Kazunari Ogawa, Z. Nakai Laborato- ones, are some differences. is

ry, Tokyo; Robert J. Van Syoc, California Academy of while has to setose feeding clearly given way a biting Sciences, San Francisco; George Wilson, Australian Museum, parasitism in Hoekiini, the piercing and biting as- Sydney; Diana S. Jones, Western Australian Museum, Perth; cothoracids have a non-setose feeding ancestry and Charles Graham, SEM Laboratory, Scripps Institution of

1969: The ascothoracids (Newman et al., R253). Oceanography. We have enjoyed and benefitted from many dis-

cussions with Yair Achituv, Bar-Ilan University, Israel, while on and Hoekiini apparently parallel one another not sabbatical leave in California. We especially thank him for al- only in the similarity between their mouthparts but

lowing us to examine specimens from the Nationaal Natuur- in the elaborationof the carapace as an absorptive historisch Museum, Leiden in his care, and for forwardingother (see Norman, 1913: 161, covering mouth- organ specimens pertinentto this study. Financial support for his visit

of and Bresciani & parts Synagoga, Jespersen, to Leiden was through a scientific exchange agreement between

Israel and the Netherlands, and wehave benefitted from the 1985: 148, concerning the carapace of the as- sup-

port for biological research by these nations. We especially cothoracidan Ulophysema Brattström, 1936). thank Vicki Ross, who typed innumerable drafts of this "note", It has long been considered that the rhizocepha- and ask her forgiveness when attention to detail became exceed- lans evolved from an epizoic, setose-feeding barna- ingly tedious. cle that developed roots at the base of the peduncle that penetrate the host (Newman et al., 1969:

R259). However, a study of the metamorphosis of Literature cited the rhizocephalan cyprid that infects its host crab through the gill membranes (Hoeg, 1985) suggests Abele, L.G., T. Spears, W. Kim & M. Applegate, 1992. an alternative a hypothesis involving free-living, Phylogeny of selected maxillopodan and other

A non-setose feeding, biting ancestor (Newman, taxa based on 18S ribosomal nucleotide sequences: prelimi-

1982: 201; 1987: 37). Recent studies, involving nary analysis. Acta Zool. (Stockholm), 73(5): 373-382. Achituv, Y. & H. Barnes, 1978. Some observations in Tetraclita molecularas well as morphological evidence (Abele rufotincta J. Biol. squamosa Pilsbry. exp. mar. Ecol., 31: et al., 1992: 377; Hoeg, 1992: 325; Jensen et al., 315-324.

1994: 141; Spears et al., 1994), tend to favor the Anderson, D.T., 1978. Cirral activity and feedingin the coral-

If indeed this is original setose-feeding hypothesis. inhabiting barnacle Boscia anglicum (Cirripedia). J. mar.

biol. Ass. 58: 607-626. the case, then the evolution of the Hoekiini, in giv- U.K.,

1981. Cirral and in the barna- and and Anderson, D.T., activity feeding ing up setose-feeding becoming a biting ap- cle Balanus perforatus Bruguière (Balanidae), with comments parently absorptive parasite, parallels theevolution on the evolution of feeding mechanisms in thoracican cirri- of the rhizocephalans, albeit on a scleractinian pedes. Phil. Trans. R. Soc. London, 291(1053): 411-449. rather than a crustacean host. Anderson, D.T., 1992. Structure, function and phylogeny of 174 A. Ross & W.A. Newman - A coral-eating barnacle, revisited

coral-inhabitingbarnacles (Cirripedia, Balanoidea). Zool. J. and the barnacle Savigniummilleporum.Hydrobiologia,216:

Linn. Soc., 106: 277—339. 285-290.

Anderson, D.T., 1994. Barnacles: structure, function, develop- Costlow, J.D., 1956. Shell developmentin Balanus improvisus

ment and evolution: 1-357 (Chapman and Hall, London). Darwin. J. Morphol., 99(2): 359-415.

D.T. A.J. Anderson, & Southward, 1987. Cirral activity of bar- Crisp, D.J. & A.J. Southward, 1961. Different types ofcirral ac-

nacles. In: A.J. Southward (ed.), Crustacean Issues, 5. Bar- tivity of barnacles. Phil. Trans. R. Soc. London, (B) 243

nacle biology: 135-174 (A.A. Balkema, Rotterdam). (705): 271-308.

A Annandale, N., 1924. Cirripedes associated with Indian corals Darwin, C.R., 1854. monograph on the sub-class Cirripedia,

of the families Astraeidae and Fungiidae. Mem. ind. Mus., 8: with figures of all the species. The Balanidae,the Verrucidae,

61-68. etc.: 1-684 (Ray Society, London).

Baluk, W. & A. Radwanski, 1967. Miocene cirripeds domiciled Duerden, J.E., 1902. The coral Siderastrea radians and its post-

in corals. Acta paleont. pol., 12(4): 457-521. larval development. Carnegie Inst. Publ., 20: 1-130.

1955. The in barnacles. Oi- 1986. The of barnacles Barnes, H., hatching process some Galkin, S.V., system coral-inhabiting

kos, 6(2): 114-123. (Cirripedia, Balanomorpha). Zool. Zh., 65(9): 1285-1295.

& W. Barnes, H., M. Barnes Klepal, 1977. Studies on the Gomez, E.D., 1973. Observations onfeeding and prey specifici-

reproduction of cirripedes, 1. Introduction: copulation, ty of Tritonia festiva (Stearns) with comments on other

release of oocytes, and formation of the egg lamellae. J. exp. tritoniids (Mollusca: Opisthobranchia). Veliger, 16(2): 163—

165. mar. Biol. Ecol., 27: 195-218.

Barnes, M., 1989. Eggproduction in cirripedes. Oceanogr. mar. Goreau, T.F., N.I. Goreau, C.M. Yonge & Y. Neumann, 1970.

Biol. ann. Rev., 27: 91-166. On feeding and nutrition in Fungiacava eilatensis (Bivalvia,

Barnes, M., 1992. The reproductive periods and condition of the Mytilidae), a commensal living in fungiid corals. J. Zool.,

penis in several species of common cirripedes. Oceanogr. London, 160: 159-172.

mar. Biol. ann. Rev., 30: 483-525. Gravier, C., 1911. Sur quelques animaux parasites ou commen-

E.J. & J.T. 1965. On des du Galaxea C.r. hebd. Batham, Tomlinson, Cryptophialus saux Madreporaires genre (Oken).

melampygos Berndt, a small boring barnacle of the order Séanc. Acad. Sci., Paris, 152(4): 210—212.

Acrothoracica abundant in some New Zealand molluscs. Gray, J.E., 1831. Description of two new species of Serpula, and

Trans. R. Soc. New Zealand, 7(9): 141-154. an undescribed coral barnacle, discovered by Samuel Pearce

1970. In: Blatchford, J.G., Possible circulatory mechanism in an Pratt, Esq. The zoologicalmiscellany. To be continued oc-

operculatecirripede. Comp. Biochem. Physiol., 34: 911-915. casionally: 1-86 (Treuttel, Wurtz & Co., London).

Bocquet-Védrine,J., 1966. Les soies et les expansions épineuses Grygier, M. J., 1984. Ascothoracida (Crustacea: )

du test calcaire chez le Cirripède operculé Acasta spongites parasitic on Chrysogorgia (Gorgonacea) in the Pacific and

(Poli). Archs. Zool. exp. gén., 107(2): 337-348. western Atlantic. Bull. mar. Sci., 34(1): 141-169.

Bourget, E., 1977. Shell structure in sessile barnacles. Natura- Gutmann, W.F., 1960. Funktionelle Morphologievon Balanus

liste Canadien, 104(4): 281-323. balanoides. Abh. senckenb. naturforsch. Ges., 500: 1-43.

Bourget, E., 1987. Barnacle shells: composition, structure and Hiro, F., 1931. Note onsome new Cirripediafrom Japan.Mem.

growth. In: A.J. Southward (ed)., Crustacean Issues, 5. Bar- College Sci., Kyoto Imp. Univ., (B) 7(3): 143-158.

nacle biology: 267-285 (A.A. Balkema, Rotterdam). Hiro, F., 1935. A study of cirripeds associated with corals occur-

E. & D.J. 1975a. of Bourget, Crisp, Factors affecting deposition ring in Tanabe Bay. Rec. oceanogr. Works Japan, 7(1): 1-28.

the shell in Balanus balanoides (L.). J. mar. biol. Ass. U.K., Hiro, F., 1938. Studies on the animals inhabitingreef corals, II.

55: 231-249. Cirripeds of the genera Creusia and Pyrgoma. Palao trop,

Bourget, E. & D.J. Crisp, 1975b. An analysis of the growth biol. Stat. Stud., 3: 391-416.

bands and ridges of barnacle shell plates. J. mar. biol. Ass. Heeg, J.T., 1985. Cypris settlement, kentrogon formation and

U.K., 55: 439-461. host invasion in the parasitic barnacle Lernaeodiscus porcel-

Bresciani, J. & A. Jespersen, 1985. The ultrastructure ofthe in- lanae (Müller) (Crustacea: Cirripedia: Rhizocephala). Acta

tegument of Ulophysema oeresundense Brattström, 1936 zool. (Stockholm), 66(1): 1-45.

(Crustacea, Ascothoracica). J. crust. Biol., 5(1): 146-159. H0eg, J.T., 1992. The phylogenetic position of the Rhizocepha-

Bubel, A., 1975. An ultrastructural study of the mantle of the la: Are they truly barnacles? Acta zool. (Stockholm), 73(5):

barnacle, Elminius modestus Darwin in relation to shell for- 323-326.

mation. J. exp. mar. Biol. Ecol., 20: 287-324. Jensen, P.G., J. Moyse, J.T. Hoeg & H. Al-Yahya, 1994. Com-

circulation Burnett, B., 1977. Blood in the balanomorph barna- parative SEM studies of lattice organs: putative sensory struc-

californicus cle, Megabalanus (Pilsbry). J. Morph., 153(2): tures on the carapace of larvae from Ascothoracida and Cir-

299- 306. ripedia (Crustacea Maxillopoda Thecostraca). Acta zool.

Clare, A.S., S.C. Ward, D. Rittschof & K.M. Wilbur, 1994. (Stockholm), 75(2): 125-142.

increments of the barnacle Balanus Growth amphitrite am- Kolosvary, G., 1947. A study of cirripedes associated with corals

phitrite Darwin (Cirripedia). J. crust. Biol., 14(1): 27-35. in the collection of the Hungarian National Museum,

Cook, P.A., B.A. Stewart & Y. Achituv, 1991. The symbiotic Budapest. Proc. zool. Soc. London, 117(2—3): 425-428.

between the relationship hydrocoral Millepora dichotoma Moyse, J., 1971. Settlement and growth pattern of the parasitic Contributions to Zoology, 65 (3) - 1995 175

barnacle Pyrgoma anglicum. In: D.J. Crisp (ed.), Fourth Eu- unpublished letters by Charles Darwin. Bijdr. Dierk.,

239-247. ropean Marine Biology Symposium: 125-141 (Cambridge 60(3/4):

Univ. Press, London). Ritchie, L.E. & J.T. Haeg, 1981. The life history ofLernaeodis-

1983. Isidascus bassindalei cus and co-evolution Moyse, J., gen. nov., sp. nov. (As- porcellanae(Cirripedia: Rhizocephala)

cothoracida: Crustacea) from north-east Atlantic with a note with its porcellanid host. J. crust. Biol., 1(3): 334-347.

J. 63: on the origin of barnacles. mar. Biol. Ass. U.K., Rosell, N.C., 1975. On two noteworthy balanids (Cirripedia

161-180. Thoracica) from the Sulu Archipelago, Philippines.

Munn, E.A., W. Klepal& H. Barnes, 1974. The fine structure Crustaceana, 29(2): 206—214.

Thora- and possible function ofthe sensory setae of the penis of Bala- Ross, A., 1969. Studies on the Tetraclitidae (Cirripedia:

J. Biol. Revision of Tetraclita. Trans. San Soc. nat. nus balanoides (L.). exp. mar. Ecol., 14: 89-98. cica): Diego

Newman, W.A., 1981. Huzio Utinomi. Crustaceana, 40(1): Hist., 15(15): 237-251.

100-108. Ross, A. & W.A. Newman, 1967. Eocene Balanidae of Florida,

Newman, W.A., 1982. Cirripedia. In: L.G. Abele (éd.), The bi- including a new genus and species with a unique plan of

ology ofCrustacea, 1:197-221 (Academic Press, New York). "turtle-barnacle" organization. Amer. Mus. Novitates, 2288:

Newman, W.A., 1987. Evolution of cirripedes and their major 1-21.

In: A.J. Bar- groups. Southward (ed.). Crustacean Issues, 5. Ross, A. & W.A. Newman, 1969. A coral-eating barnacle. Pa-

nacle biology: 3-42 (A.A. Balkema, Rotterdam). cific Sci., 23(2): 252-256.

Newman, W.A. & H.S. Ladd, 1974. Origin of coral-inhabiting Ross, A. & W.A. Newman, 1973. Revision of the coral-

balanids (Cirripedia, Thoracica). Verhandl. naturf. Ges. inhabitingbarnacles (Cirripedia: Balanidae). Trans. San Die-

Soc. 137-174. Basel, 84(1): 381-396. go nat. Hist., 17(12):

Newman, W.A. & A. Ross, 1971. Antarctic Cirripedia. Mono- Soong, K.Y. & K.H. Chang, 1983. The coral-inhabitingbarna-

graphic account based onspecimens collected chiefly under cles (Crustacea: Thoracica: Pyrgomatidae) from southern-

the United States Antarctic Research Program, 1962-1965. most coast ofTaiwan. Bull. Inst. Zool., Acad. Sinica, 22(2):

Antarctic Res. Ser., 14:1-257 (American Geophysical Union, 243- 253.

Washington). Spears, T., L.G. Abele & M.A. Applegate, 1994. A phylogenetic

Newman, W.A. & A. Ross, 1976. Revision of the balanomorph study of cirripeds and selected relatives (Crustacea: Theco-

barnacles;including a catalog of the species. Mem. San Diego straca) based on 18s rRNA sequence analysis. J. crust. Biol.,

Soc. nat. Hist., 9: 1-108. 14(4): 641-656.

Newman, W.A., P.A. Jumars & A. Ross, 1976. Diversity trends Spivey, H.R., 1988. Shell morphometry in barnacles: quantifi-

in coral-inhabiting barnacles (Cirripedia, Pyrgomatinae). cation of shape and shape changes in Balanus. J. Zool., Lon-

Micronesica, 12(1): 69-82. don, 216: 265-294.

Newman, W.A., V.A. Zullo & W.A. Wainwright, 1967. A cri- Tate, M.W. & R.C. Clelland, 1957. Nonparametric and shortcut

tique on recent concepts ofgrowth in Balanomorpha(Cirripe- statistics: 1-171 (Interstate Printers and Publishers, Illinois).

dia, Thoracica). Crustaceana, 12(2): 167-178. Tomlinson, J.T., 1963. Two new acrothoracican cirripeds from

& Newman, W.A., V.A. Zullo T.H. Withers, 1969. In: R.C. Japan. Pubis. Seto mar. biol. Lab., 11(2): 263 -280.

Moore (ed.)., Treatise on invertebrate paleontology, part R. Tomlinson, J.T., 1969. The burrowing barnacles (Cirripedia:

Arthropoda, 4(1): R206-R296 (Geol. Soc. Amer., Univ. Order Acrothoracica). Bull. U.S. natn. Mus., 296: 1 — 162.

Kansas, Lawrence). Utinomi, H., 1943. The larval stages of Creusia, the barnacle in-

Nishihira, M. & J.E.N. Veron, 1995. Hermatypic corals of habiting reef corals. Annotnes. zool. jap., 22(1): 15-22.

Japan: 311-312 (Kaiyusha Pubs. Co. Ltd., Tokyo). Veron, J.E.N., 1993a. A biogeographic database of herma-

of Norman, A.M., 1913. Synagoga mira, a crustacean of the order typic corals; species the central Indo-Pacific and genera of

Ascothoracica. Trans. Linn. Soc. London, (2) 11: 161-166. the world. Australian Inst. mar. Sci., Monogr. Ser., 10:

Ogawa, K. & K. Matsuzaki, 1990. Revision of the coral- 1-433.

inhabiting barnacles in Japan - preliminary note. Nanki- Veron, J.E.N., 1993b. Corals of Australia and the Indo-Pacific:

seibutu: Nanki Biol. Soc., 32(2): 73-79. 1-644 (Univ. Hawaii Press, Honolulu).

& An A Ogawa, K. K. Matsuzaki, 1992. essay on host specificity, Walker, G., 1983. study of the ovigerous fraena of barnacles.

systematic , and evolution of the coral-barnacles. Proc. R. Soc. London, (B) 218: 425-442.

Bull, biogeogr. Soc. Japan, 47(10): 87-101. Walley, L. J., 1965. The developmentand function ofthe ovidu-

Patton, W.K., 1976. Animal associates of living reef corals. In: cal gland in Balanus balanoides. J. mar. biol. Ass. U.K., 45:

O.A. Jones & R. Endean (eds.), Biology and geology of coral 115-128.

reefs, 3(2): 1-36 (Academic Press, New York).

Pieters, F.F.J.M.& D. Winthagen, 1990. Discovery ofscientific

correspondence of P.P.C. Hoek (1851-1914),includingthree Received: 16 January 1995