Pacific Science (1984), vol. 38, no. 4 © 1985 by the University of Hawaii Press. All rights reserved

Prey Capture in Lyonsiella formosa (: : Verticordiacea) 1

BRIAN MORTON2

ABSTRACT: A specimen of the bathyal verticordiid Lyonsiella formosa has been obtained from Hawaii at 460 m depth. Assignment of this specimen to L. formosa suggests that this has a much wider range than hitherto believed. Dissection and subsequent histological examination ofthe specimen suggests a mechanism of prey capture completely different from that previously described for this species and resembling that attributed to . Sensory papillae on the siphonal tentacles probably detect the prey. Prey capture is by eversion of an enormous hoodlike cowl of the inhalant . Inversion brings the prey into the cavity. Further distension of the siphon within the mantle cavity is believed to push the prey into the buccal apparatus comprising medially fused labial palps . The unfused tips of the palps or the foot may assist in this. A model of the hydraulic changes that may occur in Lyonsiellaformosa to effect prey capture is described. The similar modes offeeding exhibited by L.formosa () and Poromya granulata () suggest a close affinity.

ON TH E DEEP OCEANFLOOR two bivalve taxa the mechanism of prey capture. Feeding in have adopted a predatory mode of life. Pos- . Poromya was explained by Morton (1981a) sibly all species of the Propeamussiidae are who showed that the inhalant siphon com­ carnivores (Knudsen 1970, 1979), though the prised a huge raptorial cowl beneath which mechanism of prey capture is unknown. prey were trapped before being withdrawn Con versely, the Parilimyidae, Verticordiidae, into the mantle cavity. Morton (1981b) also Poromyidae, and of the ancient explained prey capture in Parilimya , the rap­ and peculiar subclass Anomalodesmata have torial inhalant siphon this time being with­ well-known predatory habits. Each species drawn by elongate siphonal (taenioid) retrac­ has particular morphological adaptations to tor muscles. Allen and Turner (1974) investi­ its predatory life style, but the single most im­ gated the Verticordiidae in great detail but did portant feature of them all is the active mus­ not come to any firm conclusions with regard cular pumping of fluids in and out of the to how prey is captured, stating that (p. 513) mantle cavity, to facilitate prey capture, in­ "the tentacles in L. abyssicola and probably in stead of the ciliary movement of fine sus­ all other species, extend across the surface pended material. The prey is digested by pro­ layer of the abyssal sediment in life and pas­ teolytic enzymes (Reid 1977) within a highly sively wait for organisms to brush against them modified gut (Purchon 1956). Yonge (1928) and adhere to them"; and later (also p. 513) investigated Poromya and but " the tentacles with adhering food contract Reid and Reid (1974) were the first to demon­ and move inwards into the inhalant strate the mechanism of prey capture in Cuspi­ where they are wiped clean by the constriction daria. This was subsequently confirmed by of the inhalant aperture and valve-when the Allen and Morgan (1981) who also investi­ latter is present. " This view was later re­ gated poromyids but could not elaborate on iterated by Allen (1983). Allen and Turner (1974) describe a variety of siphonal types for the verticordiids they in­ 1 Manu script accepted II April 1984. 2 The University of Hong Kong, Depart ment of Zoo­ vestigated, suggesting a greater diversity of logy, Hong Kong. feeding strategies than sticky tentacles. Ac- 283 284 PACIFIC SCIENCE, Volume 38, October 1984 cordingly, Lyonsiella formosa is here investi­ a box dredge operating at a depth of 460 m gated in greater detail, special attention being and with a sea temperature of 9° C. No ob­ paid to the possible mechanism of prey servations on the living were possible, capture. as it was dead by the time the dredge reached the surface. It was preserved in 5 percent neu­ tral formalin. Following dissection, the speci­ men was transversely sectioned at 6 Jlm and alternate sections stained in either Ehrlichs' Lyonsiellaformosa (Jeffreys, 1881)has been haematoxylin and eosin or Masson's tri­ further described by Allen and Turner (1974) chrome. The broken shell has been deposited who also illustrated the structure of the body in the collections of the Bernice P. Bishop as well as the shell. The species has only been Museum (Reg. no. 207491). recorded from the Atlantic (i.e., the Canaries, Azores, Bay of Biscay, and GulfofMexico) at depths ranging between 366 and 3783m (Allen and Turner 1974, Knudsen 1979). I have also FUNCTIONAL MORPHOLOGY examined the type specimen (AMNH Reg. no. The Shell 61238). Lyonsiella elegans (Thiele and Jaeckel, 1931) The shell (Figure lA) is exceedingly fragile, has only been recorded from the type locality , equivalve, and markedly inequilateral. The Station 242 (404 m) of the Valdivia Expedition anterior face is rounded and somewhat re­ (1898- 1899) (6° 34.8' S, 39° 35.5' E) at Dar es duced relative to the posterior which is in­ Salaam. I have not seen the type specimen of flated. The posterior margin is ventrally this species, there being an excellent illus­ angular and has a corrugated outline. tration of it (Thiele and Jaeckel 1931). Whereas the anterior face is smooth, the pos­ Thiele and Jaeckel acknowledge that Lyon ­ terior is radially ridged and concentrically siella elegans resembles the Atlantic L. for­ ringed. From each umbone arises a single rib mosa, and Allen and Turner (1974) have that extends to the ventral margin just pos­ shown that the shells of different -sized speci­ terior to the mid line. Seven or eight further mens ofthe latter species vary considerably in rays extend to the posterior and posterodorsal overall form and in details of the . margin of the shell. The specimen here under consideration was At the junction of each ray with a con­ obtained from a depth of 460 m off Hawaii centric ring a sharp spine is produced, though and bears a very strong resemblance to L. in this specimen most have been broken off. elegans. Bearing in mind the great variability Spines in different positions adorn the shells in shell form of L.formosa, however, and as­ ofLyonsiellaformosa illustrated by Allen and suming that L. elegans is possibly equally vari­ Turner (1974), but a series of spines on the able, it seems very probable that L. elegans is dorsal part of the shell appear more resistant but ajunior synonym of L.formosa. As will be and characterize all individuals, including the seen later, the specimen is morphologically type of L. elegans (Thiele and Jaeckel , 1931). indistinguishable from L. formosa. It accord­ The shell between the posterior rays and the ingly seems likely that L. formosa is not re­ single more median rib is concave. Allen and stricted to the Atlantic, but is a widely dis­ Turner (1974) illustrate exceedingly fine rays tributed bathyal species. on the anterior face, these are barely discern ­ ible in this specimen . The surface of the shell is finely grained. From the dorsal aspect (Fig­ ure 1B) the great expansion of the posterior MATERIALS AND METHODS region of the shell is more clearly seen. This re­ The specimen of Lyonsiellaformosa here in­ sults from , as will be seen later, the location of vestigated was obtained during a short re­ the prey-capturing organs posteriorly. The search visit to Hawaii in December 1981 with shell is widely emarginate posteriorly and less Prey Capture in Lyonsiella-MoRToN 285

L

2·5mm c

lmm

F IGURE I. LyonsiellaJormosa. The shell as seen from A, the right side, and B, the dorsal aspect; C, an enlarged view of the hinge plate of the left valve. FP, fused pcriostracum; ILL, inner ligament layer; L, lithodesma. so anteriorly. The umbones point slightly tiny pedal retractor muscles-anterior (APR) forward. and posterior (PPR). Other muscles are con­ Hinge structure has similarly been described cerned with pallial retraction (Figure 7A, by Allen and Turner (1974) (Figure IC). PRM). From the exhalant siphon extends a There are no . The primary liga­ thin array of muscles (Figure 2, ESR) which ment is small and internal, comprising inner posteriorly attach to the shell approximately ligament layer (ILL) only, this being calcified where the dorsal margin ofthe ascending lam­ mid-ventrally into an inconspicuous litho­ ella of the outer demibranch unites with the desma (L). The dorsal borders of the shell are mantle. united by a thin layer of fused The inhalant siphon is retracted by a pair of (FP) forming a " secondary" ligament whose muscles within the fused ventral mantle mar­ main function is to assist in the alignment of gin and which have their origins posterior to the valves. the pedal gape. The muscles are attached at the single near-median rib on each valve and extend posteriorly into the inhalant siphon, The Musculature radiating as they do. The two adductor muscles (Figure 2, AA; Since the muscles are probably derived PA) are of approximately equal size (iso­ from the more typical bivalve siphonal retrac­ myarian) and are bordered internally by very tor system, these are called taenioid muscles 286 PACIFIC SCIENCE, Volume 38, October 1984

PPR PA 00 ES ESR APR

H---AA

FLP J....'...'.'..'. Zmrn TM < FO PPG BG T BYG

FIGURE 2. Ly onsiellafo rmosa. The organs of the mantle cavity after removal of the right shell valve and mantle lobe. AA , anterior adductor muscle; APR, anterior pedal retractor muscle; BG, byssal gland ; BYG, byssal groove; DD, digestive diverticula; ES, exhalan t siphon; ES R, exhalant siphonal retractor muscle; FLP, fused labial palps; FO, foot; HE, heart; ID, inner demibranch; K, kidney; LS, lacunal system; 0 , ovary; OD, out er demibranch; P, periostracum; PA, posterior adductor muscle; PE, pericard ium; PPG, posterior limit of pedal gape; PPR, posterior pedal retractor muscle; SC , siphonal cowl; T, testis; TM, taenioid muscle.

(Figure 2, TM). and can Pll compared with outer hemocoel separating epithelium from those of Lyonsiella fragilis (Allen and Turner the central tissues, and I believe the structure 1974) and Parilimya fragilis (Morton 198Ib). to be basically the same as tha t described for Poromya by Morton (198Ia). Thus, the en­ closing epithelium is formed into numerous The Siphons small proj ections or papillae that are sensory The siphons of Lyonsiellaformosa are pale (Figure 4A, SP). Internally there is a large cream and surrounded by a ring of approxi, nerve (N) surrounded proximally by but a few mately 48 tentacles arranged in two cycles; longitudinal muscle fibers, and from between those of the outer cycle are larger and longer it and a large hemocoelic space (H) arise an than those ofthe inner. The appearance ofthe array of radial muscle fibers (RM) that extend siphonal apparatus as seen from the posterior to the outer epithelium and are separated aspect is shown in Figure 3. Allen and Turner from each other by numerous smaller (1974) record a maximum number of45 ten­ hemocoelomic spaces. At higher resolution tacles. Each tentacle is profusely papillate . (Figure 4B), each papilla comprises conical Allen and Turner have described the structure cells surrounding a flask-shaped chamber of a tent acle as comprising a central core of (CH). Arising from the base ofthis chamber is tissue containing a subdivided hemocoel and a single, long (8,um) flagellum (F). It may be surrounded but separate from the outer that there is more than one flagellum, butif so epithelium by yet a further hemocoel. I believe I cannot differentiate them, as has been pos­ this to be an artifact and tha t in life there is no sible by electron microscopy for Prey Capture in Lyonsiella-MoRToN 287

trally , not dorsally as suggested for this species by Allen and Turner (1974). At the tip and in transverse section (Figure 5A), the hood comprises an inner (IE) and outer (DE), much folded, squamous epithelium, cross­ connected by fine transverse muscle fibers (TF). Internally; there is an enormous hemocoel (H). Further toward the posterior base of the siphon (Figure 5B), the hood becomes penetrated by the radiating longi­ tudinal fibers of the taenioid muscles (LM). There is still an extensive hemocoel (H) and the enclosing, much less folded epithelia are cross-connected by thicker transverse muscle fibers (TF). Itis suggested that the hood ofthe inhalant siphon can be everted in much the same way as that of Poromya to ensnare the prey beneath it: This is illustrated in Figure 6, and the mechanism of prey capture will be dis­ cussed later.

The Mantle The inner and outer pallial epithelia are widely separate as in Cuspidaria except that in the latter, wide separation only occurs dorsal to the point of union of the with the FIGUR E 3. Lyonsie/la formo sa. The siphons and si­ phonal tentacles as seen posteriorly with the shell buried mantle (Reid and Reid 1974). In Lyonsiella in sand. The exhalant siphon is dorsal. the mantle epithelia are everywhere , except laterally, widely separate (Figure 7A) and and Cardiomya (Adal and Morton 1973, Reid cross-united by strands of muscle fibers (TF). and Crosby 1980) in which somewhat similar Underlying the inner epithelium is an extensive butmuch larger organs occur at the eosinophilic unicellular gland that also stains of each tentacle. The tentacles of Lyonsiella dark green in Masson's trichrome (Figure 7A, have a structure almost identical to those of MG; Figure 7B). Anteriorly the gland is Poromya (Morton 1981a), including most im­ sparse , but increases in size posteriorly so that portantly Hie ciliated sense organs on the it virtually completely lines both supra- and papillae. infrabranchial chambers (Figures 9 and 10; Allen afid Turner (1974) describe Lyonsiella MG) . A similar gland is seen in Poromya but formosa, and other verticordiids, as possess­ there it only lines the supraseptal chamber ing a large valve to the inhalant siphon. Yonge (Morton 1981a). The epithelium of the (1928) and Bernard (1974) also thought Poro­ general mantle surface is unciliated. my a granulata and P. tenuiconcha possessed The pedal gape is long, extending from a similar structure. Morton (1981a), how­ beneath the anterior adductor muscle to a ever, showed for P. granulata that the "valve" point on the shell margin where the single was in fact the inverted inhalant siphon, with near-median rib demarcates a change in shell the tentacles surrounding the base of this form (Figure 2, PPG). The mantle margin of structure. This is also true of L. formosa. The the pedal gape (Figure 7) is exceedingly "valve" comprises a deeply folded siphonal simple, comprising unspecialized inner cowl (Figure 2, SC), or hood, that is open ven- (IMF), middle (MMF), and outer (OMF) 288 PACIFIC SCIENCE, Volume 38, October 1984

A N CH SP

RM 2jJm H NU SO,um

FIGURE 4. Lyonsiellaformosa. A, a transverse section through a single siphonal tentacle; B, detail of a single sensory papilla of a siphonal tentacle. CH, chamber; F, flagellum; H, hemocoel; N, nerve; NU, nucleus; RM, radial muscle fibers; SP, sensory papilla.

IE DE IE

TF

10,AJm 20,AJm A TF B H

FIGURE 5. Lyonsiellaformosa. Transverse sections through the wall of the siphonal cowl; A, at its tip (note the absence of longitudinal muscles); and B, at its base. CM, circular muscle fibers; H, hemocoe l; IE, inner epithelium; LM, longit udina l muscles; DE, outer epithelium; TF, transverse muscle fibers. PreyCapture in Lyonsiella-s-bscwtcm 289

F IGURE 6. Lyonsiel/a fo rmosa. The animal in a natu ral position in the sediment and with the inhalant cowl of the siphon fully extended. The animal is here illustrated capturing a bottom-dwelling . Th is has not been observed, but is the interpreted feeding mechani sm. folds. A pallial retractor muscle (PRM) ex­ torial species, in an anterior direction to the tends into the margin. There are no glands and pedal gape. In Cuspidaria, however , pseudo­ there is an extensive hemocoel (H). Postero­ feces are still ejected from the inhalant siphon ventrally (Figure 8) left and right union ofthe (Reid and Reid 1974). mantle lobe margins is by fusion of the inner and middle folds (Type B) (Yonge 1982), these The Ctenidia and Labial Palps forming a small ridge (FMMF) extending be­ tween greatly enlarged outer folds (OMF). A The ctenidia (Figure 2) are of the typical thin periostracum (P) arises from the anomalodesmatan plan comprising an entire periostracal groove (PG). The mantle margins inner demibranch (ID) and the ascending lam­ also have an enormous hemocoel, the two ella only of the outer demibranch (OD) . An epithelia being joined laterally and centrally important difference is that wherea s in the by transverse muscle fibers (TF). The central typical anomalodesmatan , the ctenidia are zone of fusion is characterized by the pair of vertically aligned , in Lyonsiella they are hori­ large siphonal retractor muscles (taenioid zontal (Figures 9 and 10). The ctenidia are re­ muscles) (TM) earlier described. The inner duced and the horizontal orientation largely surface ventrolaterally comprises two rows of concerns itself with the posterior regions of ciliated columnar cells. These longitudinal the body where the shell is inflated (Figure 9, tracts (RT) presumably collect fine material CF) . Horizontal positioning is also permitted that might inad vertently enter the mantle cav­ by the fact that, as in all anomalodesmatans ity during feeding, and transport it either to hitherto studied (Morton 1981c), the ctenidial the inhalant siphon (as in conventional axis (CA) separates from the visceral mass bivalves) for disposal, or possibly, in this rap- posteriorly, coincidentally forming a capaci - 290 PACIFIC SCIENCE, Volume 38, October 1984

8

100,vm

30,(Jm

IMF

MMF P

FIGURE 7. Lyonsie/laformosa. A, a transverse section through the right mantle lobe margin; B, detail of the mucous glands in the mantle. H , hemocoel; IMF, inner mantle fold; MG, mucous gland; MMF, middle mantle fold; OMF, outer mantle fold; P, periostracum; PRM, pallial retractor muscle; TF, transverse muscle fibers. " EEI! ~

Prey Capture in Lyonsiella~MoRToN 291

TM RT RT

MG

OMF PG P FMMF 200pm

FIGURE 8. Lyonsiella fo rmosa. A transverse section through the fused mantle margins posterior to the pedal gape. FMMF, fused middle mantle folds; H , hemocoel; MG, mucous gland; OMF, outer mantle fold; P. periostracum; PG, periostracal groove; RT, rejectory tract ; 1'P, transverse muscle fibers; TM, taenioid muscle. ous suprabranchial chamber and with impor­ threads. Within the visceral mass there is an tant implications for the evolution of the ovary (0) and a testis (T) (i.e., Lyonsiellafor­ poromyid and cuspidariid septum. As in other mosa is a simultaneous hermaphrodite as in anomalodesmatans the ascending lamella of most anomalodesmatans; Morton 1981 c). the inner demibranch is but weakly attached 'The heart (HE) lies posterior to the umbones to the visceral mass. and comprises a central ventricle, penetrated The labial palps of Lyonsiella formosa by the rectum (R) , and lateral auricles. (Figure 2, FLP) are complex and have been Beneath it lies the paired kidney (K), the distal described by Allen and Turner (1974). Essen­ limbs being enormous and the cells occupied tially, inner and outer palps are fused into a by spherical excretory granules. Surrounding pouch that opens laterally in a pair (left and each kidney, however, is a large, complex, sys­ right) of fluted funnel-shaped tubes which cor­ tem of lacunae (LS), that in life are white and respond to the separated palp tips. The cteni­ could superficially be mistaken for the kidney, dia terminate in these orifices. The complex which they obscure. Much of the posterior in­ structure of the palps means that they cannot flation of the shell accommodates this system. be dramatically extended as in Poromya to Allen and Turner (1974)described this lacunal seize prey in the mantle cavity (Morton system for L. abyssicola and showed that the 1981a). Rather, prey must be taken to them , situation in L.formosa is essentially the same. as will be discussed. The lacunal system in fact occupies the mantle (Figures 9 and 10), proliferating around the The Foot, Visceral Mass, and Pericardium kidney and extending down into the mantle enclosing the suprabranchial chamber. The The foot (Figure 2, FO) is long and clearly lacunal system, on first inspection, appears to functions as a digging tool. There is a distinct be a gland (Figure lOB), but close inspection byssal groove (BYG) mid-ventrally and a reveals the absence of any secretory surfaces functional byssal gland (BG) but no byssal and the thin-walled lacunae (LW) contain 292 PACI FI C SCIENCE, Volume 38, October 1984 R

PA

CF MG

LS SC

250,.um

FMMF

FIGURE 9. Lyonsiellaf ormosa. A transverse section thro ugh the whole body at the posterior add uctor muscle and point of ventral mantle fusion posterior to the pedal gape. CA, ctenidial axis; CF, ctcnidial filament; FMMF, fused middle man tle folds; H , hemocoel; L S, lacunal system; MG, mucous gland ; PA, posterior add ucto r muscle; R, rectum; SC, siphonal cowl. Prey Capture in Lyonsiella-MoRToN 293

R K

CA CA

VOG MG

SC

FIGURE 10. Lyonsie//a formosa. A, a transverse section thro ugh the whole body in the region of the kidney and lacunal system; B, a detail of the lacunal system. CA , ctenidial axis; IE, inner epithelium; K, kidney; LS, lacunal system; LW, lacunal wall; MG, mucous gland; OE, outer epithelium; R, rectum; SC, siphonal cowl; VOG, ventral oral groove. blood cells. The significance of this lacun al The method of prey capture in Lyonsiella system will be discussed. formosa has not been described , but based on this study it seems clear the not ion that tentacles around the siphonal orifices are DISCUSSION "sticky" to entrap prey (Allen and Turner As currently recognized, there are four 1974) is not true . The tent acles possess no families ofanomalodesmatan predators- the gland s. No r, importa ntly, do the tentacles sur­ Parilimyidae, Verticordiidae, Poromyidae, round the apex of the siphons as suggested by and Cuspidariidae- each, save the former, Allen and Turner (1974) for this species and presently located in its own superfamily (Mo r­ by Yonge (1928) for Poromya. Rather, as in ton 1981 c). The Parilimyidae for a number of Poromya (Morton 1981a), the tentacles sur­ reasons are regarded,as belonging to the an­ round their bases. cient Pholadomyacea, and Morton (l 98l b) Moreover, the tentacles of both Poromya has suggested that these bivalves are primi­ and Lyonsiella possess flagellate sense cells tive, with representatives later giving rise to that presumably detect the presence ofprey as the more advanced families of predatory postulated for Cuspidaria and Cardiomya by anomalodesmatans. Reid and Crosby (1980). Also, like Poromya , Me

294 PACIFIC SqENCE, Volume 38, October 1984

the inhalant siphon is deeply inverted into the therefore forced into the pallial hemocoels, infrabranchial chamber. Both Yonge (1928) filling them. Overfilling is prevented by the for Poromya and Allen and Turner (1974) for transverse muscle fibers cross-connecting the the Verticordiidae thought it formed an inter­ two epithelia. From here blood is then forced nal "valve" but, in fact, as in Poromya (Morton into the siphonal hood causing it to evert and 1981a), the "valve" can be everted far beyond to entrap the prey (Figure 11C). The siphon the ring of basal tentacles and formed into a can be everted from between the shell valves hood, beneath which, it is postulated, the prey even though they are closed, because of their is trapped. Retraction of'the siphon will bring wide emargination posteriorly. The siphon, the prey into the infrabranehial chamber, as with prey, is returned to the mantle cavity by in Poromya. A major difference between L. relaxation ofthe adductor muscles so that the formosa and P. granulata, however, lies in valves part and internal pressures are neutra­ the fact that the labial palps of the former are lized. This causes blood to flow back from the formed into a fused, medial sac, only the tips siphon into the lacunal system so that the forming narrow fluted funnels laterally. In the siphons can now be retracted by contraction latter, the anterior palps are capable of great of the taenioid muscles in the ventral mantle extension and , reaching back into the mantle margin (Figure lID). The prey has now to be cavity, can grasp the prey and stuff it into the transferred to the mouth, and as noted earlier mouth. This cannotbe so in LJormosa and it this cannot be, as in Poromya, by great back­ is suggested that in this anima}, the inhalant ward distension of the anterior palps to grasp siphon may extend forward and push food the prey from the siphon, because in Lyon­ into either one of the lateral funnels leading siella formosa, the palps are medially fused to the mouth. Possibly the palp tips are capa­ into a buccal sac. It seems more likely that the ble of some distension to assist in this or, as siphon itself is extended anteriorly and passes suggested by Bernard (1974) for Poromya the prey to the free tips ofthe palps which may tenuiconcha , the highly active foot has some in turn assist in stuffing the food into the role in this process . mouth. Anterior extension ofthe siphon must Figure 11 attempts to interpret the mechan­ be essentially the same as for eversion. Thus, ism of prey capture in Lyonsiella formosa. closure of the valves again forces blood from Three hemocolic spaces are important: (1) the the lacunal system reservoir into the siphon lacunal system, (2) the pallial hemocoel, and which expands within the mantle cavity (ever­ (3) the hemocoel ofthe inhalant siphon. In the sion being prevented, possibly by the sus­ resting condition, the bivalve presumably tained contraction of the taenioid muscles) functions like any other, valve movements pushing the food toward the palps (Figure and the eulaterofrontal cilia of the ctenidial lIE). Following ingestion the animal returns, filaments creating a respiratory flow (Figure by opening the valves and equalizing pres­ 11A). The locomotory movements of the prey sures throughout the body, to a balanced state are presumably detected by the sense organs (Figure 11 F). on the tentacle papillae, as in Cuspidaria , The foregoing mechanism of prey capture Cardiomya, and Poromya (Reid and Reid fits interpreted structure. Certainly with no 1974, Reid and Crosby 1980, Morton 1981a) glands in the tentacles, prey capture is not by (Figure 11 B). "sticky" tentacles as proposed by Allen and Once the prey is detected, the prey­ Turner (1974) and Allen (1983). Rather, as in capturing process is initiated (Figure 11 C). all predatory, deep water, anomalodesmatans Since this involves eversion of the inhalant hitherto studied, a raptorial inhalant siphon is siphonal hood by hydraulic pressure changes everted as a result ofinduced pressure changes in the body, the following process, based upon translocating blood from a reservoir into the the structure of the various hemocoelic siphon and its subsequent return by muscular spaces, is postulated. retraction. Adduction compresses mantle fluids and It is important that attempts be made to in­ the blood within the lacunal system. Blood is vestigate living members of these predatory 'DUmM...... !I!E

Prey Capture in Lyonsiella-MoRToN 295

A BALANCED STATE Slow movement of valves co-ordinated with opening and closing of siphons effects respiratory exchange

PREY DETECTED • B FEEDING (l) Shell valves open (2) Taenioid muscles relaxed

PREY CAPTURED (3) Shell valves tightly clo sed (except at posterior gape) (4) Blood forced from blood reservoir in lacunal system into pallial hemocoe l (5) Blood flows from pallial hemocoel into inhalant siphon (6) Inhalant siphon everts

EXHALAN T SI PHON (7) Shell valves open OPEN (B) Blood flows back from siphons into pallial hemocoel and thence back to lacunal system (9) Taenio id muscles contract (10) Inhalant siphon inverts (11) Prey taken into mantle cavity

EXHALA§TSIPH ON E INGESTION CLOSEO (12) Shell valves close , .- (13) Blood forced from blood reservoir in lacunal system intopallial ' .... -e- • hemocoel (14) Blood flows from pallial hemocoel into inhalant siphon (15) Siphon extends, pushing food towards labial pulps and buccal apparatus

(16) Valves open (17) Taenioid muscles contract (18) Hydraulic system stab ilized RETURN TO BALANCED STATE

FIGURE I I. Lyonsiella fo rmosa . The proposed mechanism of prey capture (for an explanation see text). families to determine the precise mechanism ancestor. This viewwas also held by Allen and of prey capture. Turner (1974). Runnegar (1974) placed the The relationships existing between the vari­ Cuspidariidae in the Palaeotaxodonta-a ous families of predatory deep water bivalves now discredited notion (Yonge and Morton is far from clear. Runnegar (1974) placed the 1980) originally formul ated by Purchon Verticordiidae in the Pandoracea, believing (1956)-leaving only the Poromyidae in the them to have evolved from a Iyonsiid-like Poromyacea. Bernard (1974) believed the fM

296 PACIFIC SCIENCE, Volume 38, October 1984

group to comprise two superfamilies, the various species but, also, of their origins and Verticordiacea (Verticordiidae) and the relationships with the other predatory anorna­ Poromyacea (Poromyidae and Cuspidariidae). lodesmatan bivalves. Subsequently, however, Bernard (1979) has altered his views somewhat and now considers all three families to have superfamily status, ACKNO WLEDGMENTS the Verticordiacea (Verticordiidae and Lyon­ siellidae) belonging to the order Pho la­ I am especially grateful to Tom Burch of domyoida and the Poromyacea and Cuspi­ Kailua, Hawaii, for a most pleasant research dariacea belonging to the order Septibranch­ trip on his vessel Janthina VII and for obtain­ oida. Morton (198Ic), reviewing such diverse ing this specimen of Lyonsiella. Thanks are opinions, preferred to maintain each family also due to J. B. and Peggy Burch for their in its own superfamily but all within the single warm hospitality. I am grateful to Jorgen order Pholadomyoida. Allen and Morgan Knudsen, Zoologisk Museum, University of (1981) placed the Verticordiidae and Copenhagen, for bringing L. elegans to my at­ Poromyidae in the Poromyacea and the Cus­ tention and for translating into English the pidariidae in its own superfamily. Mo st re­ original description ofthis species. Thanks are cently Morton (1984) has reviewed statocyst finally due to H. C. Leung, University of structure in representatives ofall, as currently Hong Kong, for histological assistance. recognized, anomalodesmatan familial line­ ages and has come to a similar conclusion. Thus, the statocysts ofall species of Cuspidaria LITERATURE CITED are different from those of representatives of the Verticordiidae and Poromyidae, arguing ADAL, M. N., and B. S. MORTON. 1973. The that the latter families are allied and distinct fine structure ofthe pallia l eyes ofLaternula from the Cuspidariidae. truncata (Bivalvia: Anomalodesmata: Pan­ This study ofprey capture in Lyonsiellafor­ doracea). J. Zool., Lond. 171 :533-556. mosa further substantiates this view. The feed­ ALLEN, J. A. 1983. The ecolog y of deep sea ing mechanisms of Poromya and L. formosa molluscs. Pages 29- 75 in W. D . Ru ssell­ are almost exactly the same, and both are Hunter, ed. The , vol. 6. Ecology. different from that described for Cuspidaria Academic Press , New York. and Cardiomya (Reid and Reid 1974, Reid ALLEN, J. A., and R. Morgan. 1981 . The func­ and Crosby 1980). Though Allen and Turner tional morphology of the Atlantic deep (1974) described the siphonal apparatus of water species of the families Cuspidariidae many verticordiids, evaluations ofhow differ­ and Poromyidae (Bivalvia): An analysis of ent species might feed are not given; all are the evolution of the septibranch condition. presumed to have sticky tentacles. Morton Phil. Trans. R. Soc. Ser. B 294:413-546. (1981b) has pointed out the great similarity ALLEN, J. A., and J. F . Turner. 1974. On the existing between the siphonal apparatus of funct ional morphology of the family Ver­ Lyonsiella fragilis (Allen and Turner 1974, ticordiidae (Bivalvia) with descriptions of fig. 50) and that of Parilimya fragilis (Pari­ new species from the abyssal Atlantic. Phil. limyidae: Pholadomyacea) (Morton 1981b, Trans. R. Soc. Ser. B. 268 :401-536. figs. 8 and 9). Such a similarity argues for a BERNARD , F . R. 1974. Septibranchs of the close link between the Parilimyidae and the Eastern Pacific (Bivalvia: Anornalodes­ possibly least specialized Verticordiidae. It mata). Allan Hancock Monographs in is clear that the Verticordiidae, as currently Marine Biology 8. Univ. of Southern defined, possess a great variety of prey cap­ California, Los Angele s. ture techniques. Each species will have to be - --. 1979. New species of Cuspidaria from examined in closer detail , and interpreted, in the Northeastern Pacific (Bivalvia: Anoma­ order to obtain a better understanding of the lodesmata), with a proposed classification relat ionships existing not only between the of septibranchs. Venus 38: 14-24. Prey Capture in Lyonsiella-MoRToN 297

JEFFREYS, J. G. 1881. On the Mollusca pro­ deposit and suspension-feeding bivalves. cured during the "Lightning" and " Porcu­ Compo Biochem. Physio!. 56A: 573-575. pine" Expeditions, 1868-70. (Part IV). REID, R. G. B., and S. P. CROSBY. 1980. The Proc. Zoo!' Soc., London 1881 :922-952. raptorial siphonal apparatus of the car­ KNUDSEN, J. 1970. The systematics and bio­ nivorous septibranch Cardiomya planetica logy of abyssal and hadal Bivalvia . Gal­ Dall (Mollusca: Bivalvia), with notes on athea Rept. 11: 1-241. feeding and digestion. Can. J. Zool. ---. 1979. Deep-sea bivalves. Pages 195­ 58: 670-679. 224 in S. V. d. Spoel, A. C. v. Bruggen, and REID, R. G. B., and A. M. REID. 1974. The J. Lever, eds. Pathways in malacology. carnivorous habit of members of the sep­ Bohn , Schellema, and Holkema, Utrecht. tibranch Cuspidaria (Mollusca: 295 pp. Bivalvia). Sarsia 56: 47-56. MORTON, B. S. 1981a. Prey capture in the car­ RUNNEGAR, B. 1974. Evolutionary history of nivorous septibranch Poromya granulata the bivalve subclass Anomalodesmata. J. (Bivalvia: Anomalodesmata: Poromyacea). Paleonto!' 48 :904-939. Sarsia 66: 241- 256. THIELE, J., and S. JAECK EL. 1931. Muscheln ---. 1981 b. The functional morphology of der Deutschen Tiefsee-Expedition. Wissen­ Parilimya fragilis (Bivalvia: Parilimyidae schaftliche Ergebnisse der Deutschen Tief­ nov. fam.) with a discussion on the origin see Expedition auf dem Dampfer "Val­ and evolution of the carnivorous septi­ divia": 1898-1899. 21: 159-268 (1-110), branchs and a reclassification ofthe Anorn­ pI. VI-X (I-V). alodesmata. Trans. Zoo!' Soc., London YONGE,C. M. 1928. Structure and function of 36: 153-216. the organs of feeding and digestion in the - -- . 1981 c. The Anomalodesmata. Mala­ septibranchs, Cuspidaria and Poromya. cologia .21:35-60. Phil. Trans. R. Soc. Ser. B. 216:221-263. - - -. 1984. Statocyst structure in the ---. 1982. Mantle margins with a revision Anomalodesmata. J. Zool., London. of siphonal types in the Bivalvia. J. Mol!. PURCHON , R. D. 1956. The stomach in the Stud. 48: 102-103. Protobranchia and Septibranchia (Lamel­ YONG E, C. M., and B. S. MORTON. 1980. Liga­ libranchia). Proc. Zoo!. Soc., London ment and lithodesma in the Pandoracea and 127:511 -525. Poromyacea with a discussion on evo­ REID, R. G. B. 1977. Gastric protein digestion lutionary history in the Anomalodesmata in the carnivorous septibranch Cardiomya (Mollusca: Bivalvia). J. Zoo!., London planetica Dall , with comparative notes on 191 :263-292.