FAU Institutional Repository

http://purl.fcla.edu/fau/fauir

This paper was submitted by the faculty of FAU’s Harbor Branch Oceanographic Institute.

Notice: ©1975 Elsevier Inc. This manuscript is an author version and may be cited as: Rice, M. E. (1975). Sipuncula. In A. C. Giese and J. S. Pearse (Eds.), Reproduction of marine invertebrates: Vol. 2. Entoprocts and lesser coelomates (67‐127). New York: Academic Press. Reprinted from: RE:PRODUCTIONREPRODUCTION OF MARINE INVERTEBRATES, VOLVOl. II @1975©1975 ACADEMIC PRESS, INC.INC N*wH_ YwkYork SonSan Fronci>cOFrancisco LoiIcIolILoado«

Chapter 4

SIPUNCULA

Mary E. Rice

4.1 Introduction...... Introduction ...... 67 4.2 Asexual Reproduction ...... 69 K 44..3..3 Sexual Reproduction ...... 71 4.3.1 Sexual Dimorphism 71 ^ 4.3.2 Hermaphroditism 72 4.3.3 Anatomy of the Reproductive System 72 4.3.4 Origin of the Germ Cells 74 4.3.5 CytodifferentiationCytodifierentiation of the Gametes ...... 74 4.3.6 Gametogenic Cycles 85 4.3.7 Factors Influencing Gametogenesis 90 4.3.8 Spawning and Breeding Periods 91 4.4 Development 96 4.4.1 Fertilization 96 4.4.2 Embryonic Development 97 4.4.3 Larvae 112 4.4.4 Summary of Developmental Patterns 123 Acknowledgments 124 4.5 References 125

4.1 Introduction

The phylum SipullculaSipuncula is a simtllsmall group of unsegmentedunsegmcnted coelomcoelomateate marine worms noted for its lack of morphological diversity. Comprised of 17 recognized gClleragenera and approximately 320.320 species, the group is distributed throughout the polar, temperate, and tropical oceans (Stephen and Edmonds, 1972). The worms dwell in benthic habitats, ranging from intertidal shores to abyssal depths. Commonly they burrow into sand, mud, or gravel and, in the tropics, are frequently found inin • burrows within dead coral or other calcareous rock. Some species inhabit discarded mollusc shells and others live under rocks or wedge themselves into crevices. 67 68 MARY E. RICE

:1

FIG. 1.I. Dissected specimen of Themiste lissllm.lUsnm. A, anus; FF,, F"Fj, F"F,,, fixing muscles " of the digestive tract; G, gomld;gonad; 1\,\, nephridium; NC, nerve cord; Rlvl,RM, retractor muscle; T, tentacles; INd, descending intt'stine;intestine; INa,I\'a, ascending intestine.intestine, (Redrawn from Fisher, 1952.)1952. ) I I The adult body consists of an elongated cylindrical trunk with slendcrslender anterior introvert which may be vvithdrawnwithdrawn into thcthe trunk by the contrac­contrac- .I tion of one or more pairs of retractor muscles. The introV<'rtintrovert is usually terminated by tentacles surrounding or dorsal to thcthe mouth. The esscntialessential features of sipunculan anatomy are illustrated in Fig. 1, a dissecteddi.ssected specimen of Themiste lissum. A long narrow csophagus,esophagus, extending the length of the introvert, is continuous with the intestinal spiral in the trunk. Recurved, with descending and ascending limbs typicallytypicallv coiled about oneOne another, the intestine opens to the exterior in a dorsal anusamis commonlycommonlv located on the anterior trunk. Two nephridia, reduced in some species to one, open through ventrolateral pores usually at the .., apprOXimateapproximate level of the anus. ThcThe ncrvousnervous system includes a supraeso­supraeso- phagcal ganglion, circumesophagealcireumesophageal connectives, and a ventral, un­un- f paired and unsegmented n1l'uianmedian nerve cord. The gonad is commonlycommonlv locatedlocatc>d at the base of the ventral retractor muscles. The Sipuncula are classificdclassified as coelomate protostomes. CleavagcCleavage is spiral, the mouth forms in the position of the blastopore, and the coelom l ,

4. SIPUl\'CULASIPU,\CULA 69

is formed by schizocoely. Characteristically there is a trochophore larva which in some species may behe followed by a second larval type, the pelagosphera. A protonephridium,protonephridimn, found in developmental stages of many protostomes, is absent in sipunculans. The complete lack of seg­seg- mentation either during development or in the adult, distinguishes the Sipuncula from the Annelida and, in the opinion of most contemporary zoologists, justifies the recognition of the group as a separate phylum.

4.2 Asexual Reproduction

Asexual reproduction is known to occur in two species of sipunculans. A small, rock-boring species from the Caribbean Sea, Aspidosiphon brocki,hrocki, reproduces asexually, dividing into two unequal parts (Rice, 197D).1970). A constriction separates the smaller posterior daughter portion from the larger parent at a distance approximately one-fifth of the total length of the trunk from the posterior extremity (Fig. 2a). Before the completion of fission, the parent portion and the daughter portion each regencratesregenerates the structures essential to the formation of a new individual. In the case of the daughter the regenerated structures includcinclude thcthe entire anterior body and introvert, anterior gut, retractor muscles, and neph­neph- ridia. Adult structures contributed bybv the parent and incorporated into the daughter are one or two coils of the posterior intestinal spiral, poste­poste- rior parts of the spindle muscle and ventral nerve cord, and coelomocytes (Fig. 3). ThcThe formation of the juvenile is completed at the time of detachment from the adult or shortly.shortly thereafter when the newly regener­regener- ated introvcrtintrovert is everted (Fig. 2c). The parent regenerates only the posterior body wall which is formed as an invagination anterior to the constriction and is everted at the time the daughter individual is de­de- tached. Asexual reproduction is a naturally occurring phenomenon in AspidosiphonAspidosipllOH hrocki and has heenbeen found in 1.5ÍÍ15~t of the specimens col­col- lected in a population in KeyKev Largo, Florida. Sexual reproduction has not been observed in this species, although gonadal tissue is present in most individuals at the basehase of the retractor muscles. A second ascxuallyasexually reproducing species, Si/JUIlClIlllsSipiinciihts rolmsllls,rohusUis, has been ... reported to reproduce both by transversctransverse fission and lateral Imddingbudding (Rajulu and Krishnan, 1969; Hajulu,Rajulu, 1975). The posterior third of the animal may constrict off by transvcrsetransverse fission to form a new individual or the posterior one-half or two-thirds may give rise to as many as 5 incli­indi- viclualsviduals Simultaneously.simultaneously. The newly formed organs and structurcsstructures of the daughter dcvelopdevelop from a blastema composed of coclomocytescoelomocytes and pro------_._------_.. -,

70 MARY~IARY E. RICE

FIG.Fic:. 2. Asexual reproduction in i\sl'idosipllOnAspidosiphon brocki. (a) Animal showing posterior coastriction.L'onstri<:tion. (b)( b ) Recently separated parent amIand daughter.daughter, :-':oleXote everted.everted, newlyne\\'ly regeneratedf(;,gl'nerated posterior end of parent and black anterior cupcap of daughter.daughter, (e)(c) Parent and juvenile, 2 days after separation,separation. Regenerated anterior end of juvenile has \)el'nbeen everted.everted, (d) Posterior end of parent and daughter individual in process of separation,separation. Regenerated posterior of parent haslias heenbeen evertede\erted and daughter remains attached only hyby fragments of the hlaekblack material of the collar. A, anus; AS,.AS, anterior shield; C, collar on constricted animal or cap on daughter individual and juvenil*';juvenile; I, introvert. (From Rice, 1970.)

liferations of epidermal and giltgut cells of the parent. MostMo.st of the internal organs are formed beforcbefore fission, but tentacles, introvert, and anus de­de- velop after clctachmcnl'.detacliment. Lat('ralLateral buds maymav hchv formed illin the region pos­pos- terior to the anus, and at the time of separation from the parent the daughter individual is compkklycompletely formed. Asexual reproduction inin ... S-iJlunculllsSipunculus rolJUstusrohuslus has be'enbeen observed only under conditions of stress such as maintl'nancemaintenance of animals inin stalestal(> sea water in the laboratory.laboratorv. Although asexual reproduction has becnbeen onlyonlv recentlyrecentlv reportedrc-portcd in si­si- puneulans, their regenerative eapahilitic.~capabilities havchave been recognizedrc>cognizcd for some.some time. :n~n cxnerimcntaJlyexnerimentally indncedinduced reg('nerationregeneration of the introvert of severalsevenal species, formation of anterior gut, retractor muscles, and brain have heenbeen 4.4. SI1>UNCUJ.ASIPUNCULA 7171

"

Fifi.l'J(;. 3. Diaj^raniDia~ral11 of dissected posterior end of constricted AspidnsiphonAspidosipholl brockihrocki illustrating intcrnalinternal stl'1lclurcsslniclures of daughter and I'0sl"riorposterior portion of parent. C, collar; E, csopha~lls;esophagus; EP,lOP, epidermal ill\'agination;invagination; C, gonad; INa, ascending intestine; I:--Jcl,INd, cl('secndingdescending intt'stine;intestine; N, nephridium;ncphridiinii; ::\'C,\C, nerve cord; n,H, rectum; R\l,RM, retractor lllllSCIe;muscle; S, internalinternal noncellularnoneelhilar sheet of constriction; S\I,SM, spindle lIlllscle.nuiscle. (From Rice, HJiO.)1970.)

observedob.scrved (Schleip.(Schleip, 1934; 'Wegener,Wegcncr, 193R).193S). EdodcnnalEctodermal clementselements have becnbeen presumed to formform from a strandstrand of regenerative ('('liseells illin thetlie nerve con1cord and mesodermalmesoderm;d elellwntselements fromfrom coelom()('\eoelomocytes.Ics. TheThe literatureliterature on regenerationregeneration inin sipuneulallssipunculans isis reviewedreviewed IJ~!by H~'manHyman (1959).( 1959).

4.34.3 SexualSexual ReproductionReproduction

4.3.14.3.1 SexualSexual DimorphismDimorphism AsAs aa rulerule sipunculanssipunculans ;Hearc dioecious;dioecious; however,however, exlcrnalexternal signssigns ofof sexualsexual dimorphismdimorphism areare t'ntircl~'entirch' lacking.lacking. OnlyOnly inin thosethose sp('eiesspecies inin wlliehwhich thethe bod:'body wallwall isis translllccnttranslucent andand thethe oocytesooc\'tes areare ofof aa distinctivedistinctive pigmentationpigmentation isis itit possihlepossible toto ascertainascertain byby grossgross examinationexamination whetherwhether anan indivi(lualindividual isis malemale oror female.female. ForFor {'xample.example, illin maturemature specimcnsspecimens ofof PlwscoloSUII111Phascolosoma !Jerlllccnsperlucens 72 MARY E. RICE the female will appear bright red in coloration due to the pigmentation .- of the oocytes within thetJie coclom, whereas the males, becausebccause of concen­concen- trations of coelomic sperms, will be palepalc yellow or white. In many sipnnculanssipunculans it is pOSSiblepossible through most of the )'earyear to detcrminedetermine sex with- out permanent injury to the animal by extracting a small samplcsample of coelomic gametes with hypodermic syringcsyringe and needle and examining thcthe sample microscopically for developing oocytcsoocvtes or spcrmatocytes.spermatocytes.

4.3.2 Hermaphroditism HcrmaphroditismHermaphroditism has been documented in only one species of sipnn­sipun- cnlan,culan, GolfingiaColfingia minutamillllta (Akesson, 19,58).1958). In this species, reported to hebe a protandrous hermaphrodite, coelomic ooeytesoocytes and spermatocytesspcrmatoc\'tes may occur simultaneously or only oocytesooc\'tes may be present. Coelomic oocytes of those animals with gamgametesetc's of hothboth SCX('Ssexes arc immature. As the breed-breed­ ing season progresses, the percentage of animals with both male and female coelomic gametes decreases, and the percentage with oocytcsoocytes only increascs.increases. It has been assumed, therefore, that the spermatocytesspermatocvtes require a shorter time than the oocytcsoocvtes to mature and that an animal functions as a male before the definitive maturation of the oocyte'soocytes is completed. The gonad is divided into specialized regions, the more median section near the ventral nerve cord producing onlyonlv oocytcsoocytes and the more lateral parts giving rise to both oocytesoocvtes and spermatocytes. An unexplained prevalence of femalcsfemales has been recorded in popula-popula­ tions of some species. In a collection of 200 specimens of three different species of Go/fingiaGolfingia (G. elongata, G. vulgaris,vitlgaris, and G. minuta), KefersteinKeferstcin (1863) found only females. ClaparedeClaparède (1863)( 186.3) examincdexamined hundreds of specimens of GolfmgiaGolfingia elollgataelongata and found only one or two males. In an Indian population of Themiste signifer, Awati and Pradhan (1936) reported a ratio of I1 male to 60 females. Cole (1952), in a study of the morphology of Golfingia pugettensis,JJugettellsis, found onlyonlv 2 males in 100 specimens. In a later studystudv on the development of this same population, the male to female ratio was 50:50:3737 with 1.3% of undetermined sex (Rice, 1966)1966)..

4.3.3 Anatomy of the Reproductive System The gonad of most sipunculans extends as a narrow digitate band of tissue along the base of thethc two ventral retractor musclcs,muscles, extending from or ncarnear the lateral edge of oncone muscle, under the ventral nervcnerve cord to the lateral edge of the other muscle (Fig. I).1). In species such.such as Phascolion stromhi,strombi, in which thcthe number of ventral retractors has 4.-\. SIPUNCULASIPU1\CULA 7373 beenbeen reducedreduced toto one,one, thethe gonadgonad maymay bebe asymmetric,asymmetric, extendingextending fromfrom thethe ridgeridge onon thethe líasebase ofof thethe musclemllscle anteriorlyanteriorly forfor aa shortshort distancedistance alongalong oneone sideside ofof thethe ventralventral nerv(>nerve cord.cord. MaleMale andand femalefemale gonadsgonads arearc similarsimilar inin form,form, althoughalthough inin thethe malemale thethe gonadalgonadal digitationsdigitations areare frequentlyfrequently moremarc numerousnumerolls andand thinn(>r.thinner. TheThe lengthlength ofof thethe digitationsdigitations maymay varyvary fromfrom 0.10.1 toto 0.50.5 mmmm andand inin thosethose speciesspecies withwith aa definitedefinite breedinghreeding season,season, thethe gonad may disappear or be considerably reducedredlleed inin sizesize duringdming part of thethe year. Enclosed by a peritoneal sheath andaml suspended by a peritoneal mesen-mesen­ tery,tery, thethe gonad has beenheen presumedpresllmec1 toto originate from peritoneal cells which are transformedtransformed intointo gonia inin thethe pro.ximalproximal regionregion of thethc organ (Andrews, 1889; Hérubel,Herubcl, 1908). Within thethe gonad the gouoc)'tesgonoeytes are foundfound inin a gradient of progressivelyprogressivdy advanced stages fromfrom proximal to distal ends of the digitations. At the distal borderhorder of the gonad thethc gonocytes are released into the coelomcoclom where, as freely floating cells bathed in coclomiccoelomic fluid, thcvth('~! continuecontinm' to grow and differentiate until the time of spawning when thevth('~' arc accumulated from the coelom into the nephridia. The nephridium of a sipunculan is an elongated tubulartuhular structurestrtletnrc with two openings, one external, the nephridiopore, and oneonc internal, often referred to as the nephrostome (Fig. 4). Both openings are located

\ w x^ J)

FI<:.P^ic. 4.4. DiagramDiagram ofof longitudinallongitudinal sectionsection throughthrough neplll'idillmnephridium andand hodvl)ody wallwall ofof Phasco[osomaPhascolo.soma t'lll';ans.variant. (From(From Shipley,Shipley, 1890.)1890.) 74 lvlARYMARY E. RICE at or near the anterior point of attachment of the nephridium to the body wall. The internal opening is ciliated and funnel-shaped and fre­fre- quently bordered by a crescentic lip. It joins the nephridial and coelomic cavities by way of a narrow ciliated canal, and it is through this canal that coelomic gametes are directed into the nephridium before spawning by some still unknown mechanism. After a short period of storage in the nephridium the gametes are discharged to the exterior through the nephridiopore.

4.3.4 Origin of the Germ Cells There is no evidence concerning the origin of the germ cells in sipuncu­sipuncu- lans. The similarity in cytology of peritoneal cells and oogonia has led to the assumption in the past that germ cells arise from the peritoneum surronndingsurrounding the base of the gonads. AndrewsAndi-ews (1889) proposed that ""...... the nuclei of thcthe peritoneum multiplied rapidly to form a mass of germ nuclei." Later he amended his proposal to suggest that germ cells might be retroperitonealretroperitoncal rather than peritoneal in origin (Andrews, 1890).1890 ). Herube1Hérubel (1908),( 1908 ), in an extensive treatise on the biology of sipuncu­sipuncu- lans, stated that the oocytes ""...... sont certaines cellules péritonéalesperitoneales qui se differencient."odifférencient."" Although Gonse (1956a) did not refer to the source of the oogonia, he noted that the cells in the proximal region of the gonad rescmbledresembled those of the peritoneum. A similarity between peritoneal and oogonial cells as well as an apparent elaboration or thick­thick- ening of the peritoneum during the period of gonadal growth has been reported (Rice, 1973). The assumption that germ cells are derived from peritoneum does not take into account the possibility of segregation of the germ lincline in earlyearlv developmental stages. The only ohservationobservation of germ cells during development was made by Gerould (1907). He found what he termed "reproductive cells" at the base of the ventral retractor muscles in larvaclarvae of GolfmgiaGolfingia vulgaris and Phascolo})sisPlwscolo]J-sis gOllldigouldi at 2 to 3 weeks of age, but he did not follow the embryologicalembrvological derivation of these cells. Until more information is available any consideration of the origin of the germ cells in sipuncllianssipunculans remains speculative.

4.3.5 CytodifferentiationCytodlfferentiation of the Gametes

4.3.5.1 DIFFERENTIATION OFOK OVARIAN OOCYTES In the ovary of sipunculans, cytodifferentiationcytodiffcrentiation occurs in sequential stages from proximal to distal ends, beginning with the oogonia or mitoti-

o° Occ)'tesOocytes "are certain peritoneal eellscelts which are differentiated." 4. SIPUNCULA 75 cally dividing cells at the proximal end and progressing distally through the successive nuclear states of the meiotic prophase of the primary oocyte. At the distal border the cells have progressed to the diplotene stage with notable increases in nuclear and cytoplasmic volumes. It is at this stage that the oocytes are liberated into the coelom to undergo the remainder of their growth. Based on the successive nuclear changes, Gonse (1956a) has recognized seven regions in the ovary of Golfingia vulgaris (Fig. 5). In the first and most proximal region, oogoniaoogenia arc characterized by small nuclei similar to peritoneal cells and the divisions are mitotic. Cells of the second region are transitional from oogoniaoogenia to the leptotene stage of the first meiotic prophase. The chromosomes are first condensed into balls in thcthe telophase of the last mitotic division, then despiralized and grouped at the thin end of the nucleus. A nucleolus appears in this stage and the nucleus is increased in size. Region 3 consists of the leptotene stage. The chromosomes are strongly despiral­despiral- ized and the zone is narrow, indicating rapid change. The fourth and fifth regions are, respectively, the zygotene stage of chromosomal synap­synap- sis and the pachytene stage with thickened and elongated chromosomes in the bouquet form. The diplotene stage marks the sixth region; chro­chro- mosomes are spiralized and tetrads are formed. In the seventh and most distal region the prophase is completed. Chromosomes are despiralized, nucleoli appear, and the nuclear and cytoplasmic volumes increase. Other authors have mentioned similar nuclear gradients in ovaries of PlwscolionPJiascoIion stromhi,strom1Ji, GolfingiaGolßngia m;nutaminuta (Akesson, 1958), Phascolosoma agassizi, ColfingiaGolfingia pugeftensispllgettensis (Rice, 1974), and Phascolosoma arcuatwnarcuatum (Green, 1975).1975).' Gonse (1956a) noted, in addition to the oocytes in the ovary of G. vulgaris,Dulgaris, small eellscells which he designated as "annex cells" and assumed to be abortive ooeytes.oocytes. These cells, later to become the follicle cells of the coelomic oocytes, are first distinguishable in region 2 and undergo the same meiotic changes as the developing oocytes. In the ovaries of other species, similar small cells, assumed to become the follicle cells of later stages, have been found among the oocytes; however, meiotic changes have not been observed and they have been interpreted as infolded peritoneal cells rather than as abortive oocytes (Akesson, 1958; Rice, 1974).

4.3.5.2 DIFFERENTIATION OF COELO),[JCCOELOMIC OOCYl"ESOOCYTES After liberation from the ovary,ovarv, oocytes of sipunculans undergo vitello­vitello- genesis and the major portion of their growth as freely floating cells suspended in coelomic fluid and surrounded by coelomocytes. Commonly the oocytes break off from the ovary in clumps of 10-20 cells, interspersed

.J 7676 MARY~,IARY E.E. RICERICE

~[

II

III IV V

VI

VII vi«., ^'

i.V.. u.V.

•4-* .

FIG.Fic;. 5.5. Dial!;ramDiagram of longitudinallongit\iclinal s('etionsection of ovary of GolfingiaGoZ/íiigia cII/garis.vtilf^aris. Ovary isis suspendedsuspended hyl)v pcritonealperitoneal mesentery fromfrom hase of retractorretractor muscle. Seven regionsregions of successivesuccessive nuclear change.schanges arcare designatcd.designated, 1.I, oogonia; II,II, transitionaltransitional fromfrom oogoniaoogenia toto firstfirst meiotic prophase; III,III, leptoteneleptotcne stage;stage; IV,IV, zygotene stage;stage; V, pachytenepacln-tene stage;stage; VI, diplotencdiplotene stage;stage; VII, cOlnpletioncompletion of firstfirst meiotiemeiotic prophase;prophase; chromosomeschromosomes despiralizcd,despiralized, nuclearnuclear andand cytoplasmiceytoplasmic volumesvolumes increased.increased. (From(From Gonse,Gon.se, 1956a.)1956a.) 4. SIPU;\,CULASIPUXCULA 77

with smaller "annex" or peritoneal cells. Once in the coelom thcthe clumps soon disperse into single oocytes, the smaller cells arranging themselvcsthemselves around the periphery of the oocytes to become the follicle cells. An '. cxceptionexception is found in oocytes of Phascolosoma,Phascolosoma., which are usually de­de- tached from the ovary as single cells and lack a covering of folliclcfollicle cells. During the period of coelomic grovvthgrowth the volume of an oocyte increases as much as 200 or more times. Coelomic oogenesisoogénesis has been studied in fomfour spedes:species : Golfingia vulgaris (Gonse, 1956a,b, 1957a,b), GolfillgiaGolfingia ikedaiikedai (Sawada et al.,al, 1968), Gol­Gol- fingia pugettensis, and Phascolosoma agassizi (Hice,(Rice, 1974). Incidental observations on oogenesisoogénesis in additional species are found in various studies on development or reproductive cyclcscycles (Gerould, 1907; Akesson, 1958; Green, 1975). In the most comprehensive of the studies on oogcne­oogéne- sis, Gonse has investigated cytological, cytochemical, and physiological properties of coelomic oocytes of GolfillgiaGolfingia vulgaris. He distinguished six stages of coelomic oocytes, which he referred to as 0, 1, T, 2, 3, and M. Stage 0 is represented by clumps of oocytes and "annex" cells recently detached from the ovary and stage MVI is the final or matmemature stage which precedes nephridial accumulation and spawning. Stages 0, T, and M are considered to be transitory. Each stage is defined cyto­cyto- logically by size, appearance of nucleus, and kind and localization of cytoplasmic inclusions. Characteristics of the stages are summarized in Fig. 6. Similar stages of growth have been demonstrated in coelomic oocytes of other species, although specific variations occmoccur in thcthe persis-persis­ tence of folliclcfollicle cells, size and shape of oocytes,ooevtes, structmestructure of egg envelope, and localization and relative time of appearance of cell inclusions. Follicle cells, when present, are detached during coelomic oogenesis,oogénesis, but the relative time of detachment varies in different species. Detach­Detach- ment occurs in Golfingia vulgaris (stage T) and G. ikedai when the oocytes reach a diameter of 60 ,um./¿m. In G. pugettensis]JlIgettensis it is much later; in this species at the time the oocytesooevtes attain a diameter of 90 ,urn/xm the follicle cells are raised npup from the surface of the oocyte, standing out as blebs over the egg's smface,surface, hutbut they are not detached until the final stage of coelomic oogenesisoogénesis when the oocytes are approximately 1.50,um150 /im in diametcrdiameter (Fig.( Fig. 7a-d). In coelomic oocytes of Themiste pi/roides,pyroides, ranging in diameter from 30.30 to 190 ,ulll,/xm, follicle cells arc elevated from the smfacesurface of oocyte.~oocytes at a diameter of 100 ,um/j,m and are retained to a diameter of 186 ,um/ivn at which stage they are replaced by a jelly layer 50 ,um/¿m in thickess.thiekess. 'VhileWhile in the coelom as freelyfreelv suspended cells, oocytes may manifest their first indication of polarity.polaritv. Oocytes of species of Plwscolosoma,Phascolosoma, released from the ovary as spherical cells usually 20 ,urn/xm in diameter, 78 :--IAHYMARY E. RICE

10,S~1 Volume 5 10 1S 20 so 100 ISO 200 lC

2 $to tI • Hu I.-an .- ' .. a .-, ~--:"t ~ ~.

Q-~.~: l pld. ...,

170jim Diame1erDiameter 150 170l'm Fit;,Fu;. 6. l'ietOlialPielorial summary of differentiationdilFercntiatioii of l:Oelomiecoclomic oocytes of Golfingia lOlIlgaris.v.iiigmis. Six stagesslagcs of coelomiccoclomic dilfcrentiationdillcrcntiation arc (ksignated:designated: O.O, ooeytcsoocytes in dumps,clumps, rceentlyrecently ddacheddetached from¡rom ""aryoiai'v alongalonií with annex cells; 1.1, ('lumpsclumps dispcrsl·ddispersed as single oocytesooeytcs with ~nrroundingsurrounding follicle cells.cells, growth Iwgins;begins; T,'1', transitory stage, follidcfollicle cells lost; 2, main pha~l'phase of growth; 3, growth cleC'l']l'ratcd.decelerated, hurstbmst of R:\'ARXA production,proclul'lion, appcanlllceappearance of yolk platelets;platt'lds; 1\1,M, final or "matm('''"ni;iture" slagestage precedingprl'ceJing lH'phridialncphridial aeclIlIllIlation,accumulation, dissolution of germinal \'l'sicle.vesicle. (From( I'rom Cons("(louse, 11956a,D.5Ga. p. 222.)

are changed during coelomiccoc-lomic oogenesisoogénesis to slightly flattened ellipsoidsellipsoid.s (Fig. 7e-h) which in P. agassi;:,.iaoassizi measure 140 X noIK' X 90 /i,m.p.m. The polar-polar­ ity and bilateral sYlllmctr~'s\'mmetrv exhihitedexhibited at this phasephasc of oogenesisoogenesi.s are re­re- taincdtained in the developing embryo.embrvo. PolarityPolaritv illin coelomiceoelomic oocytcsooevtes of Gol­Gol- fmgiafingia vulgarisvulgariii first hecomcsliecomes apparcntapparent when the nuc:leusnucleii.s undergoes an aS~'l1lmetrieasymmetric Ratteningflattening in stage 3,3. Oocytes of most other species of Gol­Gol- ßnffcijingia and ThcmisleThemisic arc spherical throughout coelomiceoelomic oogenesisoogénesis with littlclittk; or no manifestation of polaritv.polnrit~,. The egg envelope of the fully differentiateddiflercntiatcd coelomiceoelomic oocyte of sipuneulanssipunculans is comprised of several layerslax'ers and perforated bv pore canals. Cytoplasmic cxtensionsextensions or microvilli extend through the poreporc canals and in a fewfevv species the entire egg is covered b\'by an outer jelly coat. ThcThe egg( gg envelope(-nvelope has been characterized cytoehemieallycvtochemicallv as a mucoprotein 4. SIPUNCULA 79

FIG.FIO. 7. Photographs of living coelomic oocytes. (a-d) GolfingiaColfinoia pugcttcmis;pugettensis; (e-h)(c-h) PhancolosoniaPhascoloso17la agassizi. Coelomocytes are seen out of focus in background. Scale, 25 t'm.ßm. fe,fc, FollicleFolhcle cells. [From (in part) Rice, 1974.11974.]

which is secreted bvby the oocyteoocvte during the period of coelomic growth (Gonse,(Gonsc, 19.56a;1956a; Hice,Rice', 1974). The envclopeenvelope of oocytesoocvtes of PluiscolosomaPlwscolosoma agassizia<^assizi reaches a thickness of 10 ^m,urn and fully developed envelopes of GolfingiaGoJßn^i^ia imgettensis]ntgettensis and G. ikedai are 5,5 and 2 vm,p.m, respectivel)'.respectively. In these latter species the number of la),erslayers maymav hebe as manyman\' as 14. In P. agassizi, three layers are distin­distin- guishable. As shown in electron micrographs, microvilli pass through pores in the inner and middle layers, branching into fan-shaped structures in the outer la)'crlaver (Fig. 8). Tips of the microvilli arcare marked by an amorphollsamorphous material or fuzz which consolidates in the latest ooc)'tesoocvtes as a homogeneous fringe around thcthe egg and is expanded in the regions of the animal and vegetal poles. The inner layerlaver gives a stronglystronglv positive reaction for mucoprotein and a weak response for protein whereas thcthe

1- 80 MARY E. RICE

., 4

!•

FIG.Fic. 8.B. Electron inicrographmicrograph of a coelomiccoeloniic oocyte of PhascnlosomaPhascolosoma agassizi,agas^izi, showing egg envelope. 0,O, M,/II, I, Outev,Outer, miJdlc,niiclrlle, and inner layers of eggl'gg envelope; f, surrounding fiizz;fnzz; I,1, lipid; mv, microvillus; y, yolk granule. Scale, approximatelyapproximaldy 111m.1 ^m.

reverse is true of the middle la\'er.layer. The outer hiverlayer has beenheen identifiedidentmed as an acid mucopolysaccharide and the surrounding fuzz as a mucoprotein.mucoprotein. On the basis of their electronclectron microscopic studies on0)) coelomic oocytes of GolfitigiaGoTfingia ikedai, SawadaSawadn et al.aT. ((1968) 1968 ) have proposed a schemeschcme of en-en­ velopevelopc formation (Fig. 9). The completed envelope of this species is composed of a fibrous outer layer,laycr, a multilamellate middle laver,layer, and

a diffusedifFuse inner layer.laver.J The earliest oocytesoocvtesJ show no indicationindication of an envelope, but projecting from the surface are numerous microvilli with pinocytotiepinocytotic vesicles at their bases. As the oocyte grows, the diffuse mate-mate­ rial of the innerinner layer, presumed to be precursor substances extruded from thethe cell surface, condenses to form inin succession thethe 14 middle 4. SII'UI\CULASIFUiXCULA 81

1"1(:.Fit;. 9. Schematic representationlepresentation of growth of coelomiccoeloniic oocytcsoocytes of Golfm{!,iaGoljingia ikcdai, bast'dl)ascd on electron microscopic observations. C, chorion; DS,US, diffused substances; EV, endoplasmiccndoplasmic vesicles; F.F, follicle; G, golgi body;bod)'; 1M,IM, inner membrane; .I,[, jellv la)'er;layer; iVI,M, mitochondria;niitoehondria; N, nucleus; 0111,OM, outer membrane; PM, plasma melllbranc;membrane; 1'V,PV, pinocytotiepinocytotic vacuole; R, ribosomeribosonie cluster; RER, rough-surfacedrougli-surfaeed endoplasmic'endoplasmie reticulum;reticnlum; V, microvillus;mierovillus; Y, yolk granule.granule, (From Sawada ctcl at.,al., H)(iS,1968, p. ,37.)37.) layers,laver.s. The origilJorigin of thetlie fibrousfiljrou.s outer laverla~'cr isi.s uncertain, 1)utbut the microvil­microvil- lar tips have been implicat('d.implicated. A jc]]~!jelK' Ja~'erlaver surrounding.surrounding the ooc~·teooc\te in the latest stages m,lymay also hebe a sccretionsecretion of thcthe microvilli."microvilli. The nucleus in earlyearJ~) oocvtesooc~'tes is alwaysalwavs rounded.rounded, occupying a central position in the cell. Later, but at characteristic times in different species, the nucleus is marked by numerous peripheralpcripheral infoltlings.infoldings. In GolfillgiaGolßn^ia vulgaris and PhuscolosomaPlwscolosollla agassi:::.iagassizi this occurs in the final stage of oogenesis;oogénesis; in the formcrformer thcthe folds occur on onlvonly one side of the nucleus. The nucleus of G. pugettcnsis}Jllgettcllsis becomesbecomcs scallopedscall~pecJ much carlin,earlier, whenwjien the diameter of the oocvteoocvtc. is onlvonly. one-half of its mature dimension. The nucleoli appear as numerous fragments in the early oocytes of G.C pugettcnsispllgettellsis and G. l;lIlgaris,vulgaris, later diminishingcJin~inishing in G. pugettensis]JlIgettellsis to two or three discrete nucleoli with few fragments and in G. vulgaris to sevnalseveral discrete spherical strnctmcsstructures arrangcdarranged around the periphery, one of which in the final stage dominates the others in size. The oocytesoocvtes of P. agassizi 82 MARY E. RICE have 2-5 nueleolimicleoli throughout oogenesis,oogénesis, but IIIin younger stages one is larger than the others. Mitochondria, as observed in two species of GolfingiaGoJfingia (ConsC',(Gonse, 19.56a;1956a; Sawada et al., 1968), arc localized in early oocytes near the periphery of the cell, but as their numbers increase in later stages they move throughout the cytoplasm. In electron microscopic studies of G. ikedai the mitochondria are found to be fused with peripheral endoplasmic vesicles in early coelomic oocytes; in later stages they are associated with one of two kinds of yolk granules. Pyroninophilia, interpreted as a probable indication of the presence of ribonucleic acid, has been demonstrated in the cytoplasm of coelomic oocytes of Golfingia vulgaris.mdgaris. Strong perinuclear concentrations occur in the earliest stages and again in the latest stage. Electron microscopy of the oocytes of G. ikedai has similarly revealed an early concentration of rough endoplasmic reticulum.reticuluni. Vitcllogenesis in GolfingiaGoJfingia vulgaris is initiated soon after detachment of oocytes from the ovary when carbohydrate yolk granules, presumed to be galactogen, make their appearance at the periphervperiphery of the cells. As the oocytes develop the granules move toward the nucleus, then are dispersed toward the cortex where, in the final stage (stage M), they form a layer of cortical granules. A second polysaccharide, recog­recog- nized histochemicallyhistochemicallv as glycogen.glvcogen, appears as a diffuse substance in a perinuclear position late in coelomic oogenesis.oogénesis. Lipid does not appear until stage 2 when it is fonnelfovmd in association with\",ith a mitochondrial mass in the perinuclear zone; later it disperses through the cytoplasm. At the time of germinal vesicle breakdown (stage M) the lipid is concen­concen- trated in the peripheral cytoplasm. Prominent elongate bodies or yolk platelets, not defined histochemically,histochemicallv, but presumed to be proteinaceous yolk, occur in well advanced oocytcsoocytes of stage 3 (Fig. 6). From investiga­investiga- tions of physiological properties of the oocytes of this species, two peaks of exogenous respiration have been demonstrated during coelomic oogen­oogén- esis, the earlier corresponding to the beginning of carbohydrate synthesissvnthesis and hothboth corresponding to periods of high concentrations of ribonucleic acid. The peaks of respiration were found to be coupled with metabolismmetaboli.sm of hexoses and pentoses (Conse,(Gonse, 1957a,b). In theth~ elaboration of yolk there is little consistencyconsistenev among different species in the sequence in which carhohydrate',carl)ohvdrate, protein, and lipid make their appearance in the developing oocytes, or in the form and localiza­localiza- tion of the yolk granules within the cells. The carbohydratecarbohvdrate yolk of PhascolosomaPllascolosoma agassizi differs from that of Golfingia vulgaris in that it is associated with protein as a carbohydrate-protein complex in the form of distinctive yolk granules of irregular staining pattern. The gran- 4. SIPUNCULASÎPUNCULA 83 ules are visible first in the perinuclear zone, rather than at the periphery as in G. vulgaris, and after dispersal through the cytoplasm they move in later stages away from the periphery leaving a clear cortical area. The earliest oocytes of G. pugettensisJlugettensis that have been examined show carbohydrate yolk granules midway between nucleus and periphery; no tests for protein have been made. Lipid in this species is present as large droplets at the periphery in the early oocytes, appearing rclativelyrelatively sooner than in G. vulgaris. It is then dispersed and finally, in late oocytes, as in G. vulgaris, it is localized in the peripheral cytoplasm. The large and rather rare lipid droplets of P. agassizi have not been seen until late in oogenesis,oogénesis, but very small granules, suspected to be lipid, arc distrih­distrib- uted through the cytoplasm in the earliest stages of coelomic oocytes. Characteristic of lipid, these small granules appear after centrifugation in the centripetal half of the egg and in electron micrographs thcythey arcare seen as homogenous spheres without surrounding membranes (Fig. 8). Recently spawned, unfertilized eggs of many species of sipunculans have been described in studicsstudies on development. Sizes of the eggs of 17 species can be found in TahleTable II (see Section 4.4.2). All sipunculan eggs are encompassed by a thick characteristic envelope whichv^'hich is com­com- posed of several layers and perforated by pores. There is considerable variation among species in thethc thickness of the envelope and the number of its layers as well as in the size, shape, pigmentation, and yolk content of the eggs (Fig. 10 a-e). The eggs of most sipunculans are spherical, but those of a few species, such as ParaspidosiphonParaspidosilJ1lOn ßscherifisclleri and Golfingia minlltaminuta are oval and those of the genus Phascolosoma are typically flat­flat- tened ellipsoids, frequently with depressed apexes. Sipunculan eggs are commo\llycommonly various shades of red or yellow. Large eggs 'withwith high yolk content, such as eggs of Themiste,Theiniste, may be white or grayish and eggs low in yolk, as found in species of Si}JttnCIIlusSiptincuJus and Siphonosoma, are transparent. At the time of spawning, the gcrminalgerminal vesicle of .thethe eggs has broken down and meiosis is arrested in the first meiotic mctaphasemetaphase (see Sections 4.3.8,4.4.1).4.3.8, 4.4.1).

4.3.5.3 DIFFERENTIATION OF MALE CA1\IETESGAMETES Little is known about cytodifferentiationcytodifFcrentiation of male gametes. TIleThe cytology of the male gonad of Phascolosoma arcuatum has been reported to resem­resem- ble that of the ovary, as described by ConseGonse (1956a) for Golfingia vul­vul- garis, with the cxceptionexception that the cytoplasmic volume of the distal sper­sper- matocytes is less than that of the oocytes in that position (Crcen,(Green, 1975),1975). Spermatocytes break offofF from the testis as loosclyloosely associated clumps of cells which undergo two meiotic divisions and diffcrentiatedifferentiate into sper­sper- matids while floating as morulae in the coelomic fluid. The cells within 84 MARY E. RICE

d

FIG.Fic. 10. Recently spawned gametes of siplinclilans.sipunculans. Photographs of living eggs andaod sperms. Eggs are unfertilized.unfertilized, (a) Egg of Phascolosoma agassizi. (b) Egg of PhaxcoloiomaPlwscolosoma perlucens.pcrltlcens. (c) Egg of Phascolosoma antillarum.antiUarum. (d) Egg of Phasco-Phasco­ liolllion crypl/ls.crijptiis. (e) Egg of ThemisteTheniiste pijroiclcs.pyroidcs. Note thick jelly layer.layer, (f) SpennatozoaSpermatozoa of Thcmi.tcTliemiste pipoidcs.plJroirlcs. SealeScale a-I.',a-e, 252.5 .um.urn, (a and c from Rice, 1967. h,b, c, and d from Rice, 1975). a clump increascincrease in number as divisionsdivi.sions ocemoccur and in each cell the cyto­cyto- plasmic volume decreases and a flagellmllflagellum develops. The clumps of diHercntiatcddiflerentiated spennatidsspermatids breakbrcak lipup into free spermatozoa in the coelomseoeloms of G. lntgettcllsisjm^ettensis and TlwmisteThemiste l'l!Jroidespyroides where they maymav lwbe SCCIlseen through most cf the year. TnIn P. agassi::..iagassizi a few spermatozoa are froefree in the coelom at the time of the breeding season, but the majority remain inin clusters. In P. arclwtum,arcuatum, spermatids and free spermatozoa ocellI'occur in the coelomic fluiclfluid onl),only a short time before spawning. SpermatozoaSpeimatozoa of sipunculans have the typical morphology of the primi­primi- tive sperm as defined bybv FranzenFranzt'n (1956). Similar to those of other species which eJisehargedischarge their gametes fredyfreely into the seawater,scawater, theythev 4. SIPUNCULA 85

are comprised of low acrosomal caps, midpieces each with four mitochon­mitochon- drial spheres, and long f11amentousfilamentous tails. The sperms of Themiste 7l!}l'Oidespyroides are somewhat modificdmodified in that they possess a highly developed acrosomal cap with pointed tip and swollen basal rim (Fig. 10f).lOf). It is probable that this increased complcxitycomplexity is associated with the thick jelly coat of the egg through which the sperms must pass (Rice,( Rice, 1974).

4.3.6 Gametogenic Cycles Annual reproductive cyelescycles arcare known for four species of sipunculans:sipunculans; Golfingia vulgaris (Gonse, 1956b), G. pugettensispllgettensis (Rice, 1966), and two populations of PhascolosomaPhascolosonw agassizi (Rice, 1966; TowlcTowle and Giese, 1967), all from tempcratetemperate waters, and one tropical specics,species, PhascolosonutPlwscolosonUl arCtlatumarcuatum (Green, 1975). Cycles have been dcflneddefined in these stndiesstudies by estimations of brecdingbreeding seasons from observations on spawning, cyto­cyto- logical examination of gonads, and measurements of coelomic oocytes throughout the year. Oocyte measurements have been expresscdexpressed in rela­rela- tive frequencies and no absolute estimations arcare available of the various sizes of cells. Golfingia vulgaris has a limited breedingbrccding season lasting from June to September in RoscoH,Roscoff, France. The gonad, howcver,however, is continuously active, releasing small oocytes into the coelom throughout the year, with reduced aetivitvactivitv, in the winter. MonthlvMonthly, measurements of coelomic oocytes over a year show that although oocytes are alwaysalwavs piesentpresent in the coelom their growth is arrested during part of the year (Figs. 6 and 11). In the winter all of the oocytesooc\tes are small, i.c.,i.e., less than 61 J-lm,)Lim, and the size-frcquencvsize-frequency curvecurv<> is unimodal, indicating arrested growth. Growth commences in the spring and with the appearance of a slightly larger group of cells the size-frequcncysize-frequencv curve becomes bimodal, the second mode at 66 I,m,/mi, close to thcthe first. In thcthe summer the latter mode disappears and all sizes of oocytesoocvtes are fonnd:found: the smallest oocytcsoocytes persist as a prominent mode, intermediatcintermediate ooeytesooe\'tes arcare present but in low freqnencies,frequencies, and a population of large oocytesoocvtes makes its appearance, forming a second mode at 1.54154 /¿m.p.m. In the au­au- tumn, at the conconclusionelusion of the breeding season, growth of small oocytes is again arrested and oocytes of intermediate and large size graduallygraduallv disappear. Low frequencies are interprctedinterpreted as indicative of a rapid rate of growth and high frequencicsfrequencies as slow growth. Thus, those popu-popu­ lations that form prominent modes, that is, the small aneland large cells, grow morcmore slowlyslowlv than the ccllscells of intermediate size (Fig. 6). In the annual cycle of Golfingia pugettensis,pugettel1sis, endemic to the Northwest Pacific Coast of the state of \Vashington,Washington, the l)rcedingbreeding season occurs 86 MARYr-IARY E. RICE

I Frequency

FIG. II.11. Three-dimensionalThrcc-dimen.sional diagram of size-frc(!uencysize-fie(iuciicy curves of coelomic ooeyte.~oocytes of GolfillgiaGoJfinnUi Gu/garisvult^aria over a period of 1 year. The absolutealxsolute value of the fre(|iiencyfrequency of the small.small oocytesoocyte.s is arbitrarily kept constantcon,stant except in the autumn when the ovary is less active and the production of coelomic oocytesoocvtc.s isi.s decreased. (From( From Gon.se,Conse, 1956b, p. 232.)

in October, November, and December. During this period size-frequency eurvescurves show, as for G. vulgaris, two major populations of coelomic oocyt(·s,oocytes, one of small oocytes and one of large, with only a small propor­propor- tion of oocytes of intermediate size (Fig. 12). In January, after the breeding season, large oocytes are no longer present but the population of small oocytes persists and a ncwnew population of slightlyslightlv larger ccllscells becomes evident. This exists through the spring but disappears in the summer at the same time several small populations of cells of intc'nnecli­intermedi- ate size arcare formed. Although the hreedingbreeding periods of GolfingiaGolßngia pllgettensispugettensis and G. vulgariSvtilgaris arcare at different seasons, the latter spawning during the summer months, there is, nevertheless, the same sequcncesequence of oocyte growth with similar phases of apparent arrest and acceleration. However, the relative time of appearance of the various phases differs. In G. ]>ugettellsispugettensis there is an apparent arrest of growth of small oocytes during the breeding season, and after the large eggs are discharged by spawning,spawning, the smallsmall

l- _ -I

4.4. SIPUNCULASIPUNCULA 8787

10

JANUARYJANUARV 15

10

20

I Y

[".BEA 15

10

5 -

52 787 8 104104 130130 156 DIAMETERDIAMETER OFOF COELOMICCOELOMIC OOCYTESOOCYTES lum)lum)

Fic.FIG. 12.12. Size-frequencySize-frequency cun'escurves ofof coelomiccoelomic oocytesoocytes ofof GolfingiaGolfingia pugettensispugettensis (]962-63).(1962-63). InIn JanuaryJanuary 450450 eggseggs werewere measured,measured, 5050 fromfrom eacheach ofof 99 animals.animals. InIn bothboth JulyJuly andand November,l'\overnher, 300300 eggseggs werewerc measured,measured, 100100 from[rom eacheach ofof 3.3 animals.animals. (From(From Rice,Rice, 1966.)1966.) eggseggs beginbegin toto grow.grow. InIn G.G. vulgarisvtllgaris thethe frequencyfrequency ofof smallsmall eggseggs passingpassing intointo thethe intermediateintermediate stagestage isis muchmuch higher;higher; growthgrowth ofof smallsmall oocytesoocytes continuescontinues throughoutthroughout thethe breedingbreeding seasonseason andand isis arrestedarrested onlyonly afterafter spawning,spawning, thethe periodperiod ofof arrestarrest lastinglasting forfor severalseveral months.months. ThusThus duringduring thethe periodperiod ofof spawningspawning inin G.G. vulgarisvulgaris intermediateintermediatc oocytesoocytes areare prevalent.prcvalent. ThoseThose remainingremaining afterafter spawningspawning areare presumablypresumably resorbed.resorbed. InIn G.G. puget-puget­ tensistensis byby thethe timetime ofof spawningspawning mostmost ofof thethe intermediateintermediate cellscells havehave alreadyalready givengiven riserise toto maturemature cellseeUs andand thethe growthgrowth ofof thethe oocytesoocytes hashas ceasedceased oror beenbeen considerablvconsiderably reduced.reduced. 88 MARY~IARY E. RICE

The oogenic cycle of a population of PhascolosomaPlwscolosoma agassiziagmmzi min the San Juan Archipelago off the Northwest Pacific Coast of \Vashington,Washington, differs from that of hothboth GolfillgiaGolfingia pugetienmJillgcttCIlSis and G. Dulgarisvulgaris in that oocytcs of all sizes, small, intermediate,intcnnodiate, and large, arc present throughout the year (Rice, 1966). An examination of the frequency polygons reveals that there are always two principal size groups of cells, the largest oocytcsoocytes usuall:vusually predominating exeeptexcept immediately after spawning (Fig. 13). This population of P. agassizi spawns from early June through August.

21) -- LfNGrHLE NCI I H -_ ••• WIDTH

l~

10 /' ."', /"-. A:/~-~ ., i /~/ j _/_-\...._.- ''-__r_ __----1>._

<.:l .. 10 z /\::...... \:., ..., / ~\ .. ~

~, .0 . , 1U . .•

52 78 104 156 51ZESIZE Of COELOMIC OOCYTES 'l,n1)uinii

[o'IC;.FK;. 1.3.13. Size-frequency curves of coelomic oocvtesoocyte.s of PhaRcolosomGVhascolosoma ngíLmzia{!/L~sizi from the(he San Juan Arehip"!ago,Archipelago, vVa,hingtonWashington (1964-H)6.3).(1964-196.5). \leasmementsMea.snrements each month represent a total of .500500 oocyte',oocytcs, 100 from each of 5 animals. (From Rice, 1966.) 4.4. SlPUNCULASIPUNCULA 8989

I -L NGT 75 •• L: H ~ ~50 II! 2

,,'

2 " 34.5JJ34.5« 69.0)J69.0 il I035JJ103 5u 138.0JJ138.0^. 1715)J1715íJ MONTHLY CHANGES ININ LENGTH AND WIDTH OF DEVELOPING OVA

FIG. 14. i\'lonthlyMonthly changeschan~c~ in size~ize of coe!omi<.:coelomic m:ytcsncvtos of P/ifisroio-soDi«rllO~coloso))l(l ar^aasiziag(l~sl;:;1 from 1I1olltereyMonterey Bay, California (1960-1961).(1960--1961). Frequency distribuliondi~lribution each month representsrepre~t'nts nieasurementsmeasnrempnts of 500,500 ocieytcs,ooc~'tcs, 20~() from each of 25 females.ft'lllal,'~. [From[1'rIIlll Towle and Gicse,Giese, 1967, fig. 1 (in part), p. 232.]

A population of thethe samosame .speciesspecies at Vfontcrcv,Ivlontcrey, California spawns threethree months earlier, inin March, and for a 3-month3-montl1 period followingfollowing spawning both male and femalefemale gametes arearc absent inin thethe majority of animals (Towlc(Towle and Giese, 1967). BvBy' thethc fourthfourth month, July,Julv, all sizessizes of coelomic oocytesoocytes areare found,found, includingincluding high proportionsproportions ofof both smallsmall andand largelarge oocytesoocytes (Fig.(Fig. 14). During thethe fallhll andand winterwinter thethe frequencyfrequency ofof thethe smallsmall oocvtesooc)'tes diminishesdiminishcs whilewhile thethe largelarge oocvtesoocytes proportionatelyproportionately increaseincreasc untiluntil aa monthmonth beforebefore thethe March\'farch spawning,spawning, whenwhen appro.\imatclvapproximately 90%90% ofof thethe coelomiccoelomic oocytesoocytcs areare large.large. JîvBy contra.st,contrast, thethe monthmonth beforehefore thethe brec-dingbreeding seasonseason inin thethe populationpopulation fromfrom thethe SanSan JuanJuan ArchipelagoArchipelago onlyonly 55155% ofof thethe coelomiccoelomic oocytesoocytes areare inin thethc largelarge sizesize group.group. AnAn importantimportant differencedifferencc inin thethe oogcnicoogenic cvclescycles ofof thethe twotwo populationspopulations seemsseems toto bebe thethe degreedegree andand durationduration ofof reductionreduction inin gonadalgonadal activity.activity. TheThe gonadgonad ofof thethe MontereyMonterey animalsanimals isis active

4.3.7 F:actorsF-actors Influencing Gametogenesis

Little is known of the factors controlling the growth of gametes in sipunculans. It has been suggested (Akesson,(Äkesson, 1961a) that nelll"osecretoryneurosecretory products mavma~J play S0111esome roleroh; in the reproductive cycle since Carlisle (1959), in studies of Sipl/Ilcllil/sSipunculus Iltllills,midus, fonndfound a greater amount of neuro­neuro- secretory material in animals before the breeding season than following it. Ncv.rosccretoryNet:.rosceretory material has been demonstrated in other sipunculans (Cabe,(Gabe, 19.53;1953; Akesson,Äkesson, 1961a), hutbut its significance in reproduction has not been investigated. The possible influcnceinfluence of cxogenousexogenous factors such as tcmperahlretcmpcratin-e I.Jnon gametogencsis of sipunculans is suggested by thcthe difference in the dura­dura- tion of breeding scasonsseasons of some tropical and temperate species. Observa­Observa- tions on spawning in certain tropical or subtropical species, e.g., Phascoh-Phnscol~­ soma perlucens, TlwmisteThemiste alutacea, and T. lagenifonnis, indicate that breeding may occur throughout the year (Section 4.3.8). However, no information is available on gametogenesisgametogencsis in these species, and in other tropical species, such as P. arcuatllm,arcuatum, the breedingbrceding season is known to be restricted, as it is in all species from temperate waters. In species with well-definedwell-d(!fined breeding seasons differences may be found in the time of initiation of breeding in the samcsame species at diffcrentdifferent latitudes. At Mon­Mon- terey, California, Phascolosoma agassizi(wassizi breeds 3 months earlier than at II11°0 farther north in the San Juan Archipelago, Washington (Rice, 1966; Towle and Giese, 1967). However, in two populations of P. arctUI-arcua- 4. SIPUNCULA 91

tumturn in Australia, separated by 8° in latitude, correlation between temper­temper- ature and spawning is tIlethe reverse, that is, the population at the higher latitude spawns first (Green, 1975).

4.3.8 Spawning and Breeding Periods GametcsGametes are spawned fromnom the nephridia by forceful ejection through thcthe nephridiopores into the seawatcr where fertilization ocems.occurs. V/henWhen spawning is imminent the nephridiopores are often swollen and the ani­ani- mals may becomebecomc quite active, frequently extending and retracting the introverts. Phascolosoma agassizi, when maintained in an aquarium with sand and gravel, has been observed to extendcxtend the anterior end above the surface of the substratum so that the nephridia are well exposed, and at the time gametes are releascdreleased thcthe body is cxtendedextended and turgid (Rice, 1966). Although usually in sipunculans all of the gametes from both nephridia are spawned at once, variations of this process have been observed. A male specimen of ParaspidosiphonParaspidosipllOn ßscherifischeri spawned 7 times in the laboratory over a period of 40 minutes, first several times from the right nephridium,nephridiuin, then from the left. Phascolion cryptus,cnjptus, with only a single nephridium, was observed to spawn short intermittent spurts of sperm over a period of 15 minutes (Rice, 1975). Since sipunculans, along with the majoritymajoritv of marine invertebrates, spawn their gametes into the seawater, synchronization of spawning is essential to assure fertilization and the survival of the species. Synchro­Synchro- nization in sipunculans is dependent on the regulation of two separate events: (1) uptake of coelomic gametes by nephridia; (2) release of gametes from the nephridia into the seawater. Factors controlling thescthese events have not been analyzed, but hypotheses have been proposed for nephridial uptake and data have been accumulated on such aspects of spawning as sequence of male and female activity, time of day that spawning occurs, and intervals between spawnings within a breeding season. Before spawning, gametes are accumulated into the nephridium from the coelom by way of thcthe internal funnels or nephrostomes. The process of nephridial selectivity by which the most mature gametes, either oocytesoocytcs or sperm, are taken into the nephridium, while the immature gametes and coelomocytescoelomoeytes are rejected, is poorly understood. Since Docytesoocytes are known to enter the nephridium soon after breakdown of the germinal vesicle (Gerould, 1907; Akesson, 1958; Riel',Rice, 1966) it can be assumed that nephridial selectivity is related to the initiation of matu­matu- ration of the gametes. (It should be noted, however, that maturation divisions do not occur until after sperm entry; see Section 4.4.1.) Gerould (1907) noted in Golfingia vulgaris that before spawning the nephridia 92 MARYiMARY E. RICE

becamebecamc greatly distended with fluid, which he presumed to be seawaterscawater taken in through the nephridiopores. Assuming that the eggs selcctedselected by the nephridium were hydrotropic because of their lower specific grav­grav- ity, GerouldGc-rould proposed that these eggs were accumulated in the region of the nephrostome and directed by ciliary currents into the nephridia where theythev resorbed water. AkessonÂkesson (1958)( 1958 ) further suggested that a chemical change occurring in the oocyteooc\'te at breakdown of the gcrminalgerminal vesicle might alter the direction of nephridial cilia which wOllldwould then move ,hesediese eggs into the nephridium. In explaining nephridial uptake of mature spcrm,sperm, Akesson noted that seawater induced motility of coelomic sperm. Helie then hypothesized that as the nephridia filled with seawater, some of the water passed into the coelom through the nephro­nephro- stomes and the mature sperm were activated to swim toward the incrcas­increas- ing concentration of seawater to the nephridium. Breakdown of the germinal vesicle in the coelomcoclom and dispersal of coelomiccoelon~jc sperm clusters arcare apparent prerecjuisitesprerequisites for nephridial uptake. In many species oocyte'soocytes of maximnmma.vimum size with intact llllcleinuclei arcare present in the coelom through much of the year (Section 4.3.6),4..3.6), hutbut onlyonlv dlll'ingduring the breeding seaSOnseiison shortly before spawning does dissolution of the germ­germ- inal vcsiclevesicle occur. Few studies have heenbeen made of the mechanisms regu­regu- lating the maturation of gametes. PasteelsPasteéis (1935) carried out in vitro studies on the breakdown of the germinal vesicle in coelomic oocytes of Phascolion strombi. He found that the addition of increasing quantities of calcium ions induced maturation of the eggs. In preliminary studies (Rice, 1966) it was found that crude extractse-xtracts of coelomic sperm, coelomic oocytc~,oocytes, coelomocytes, brain, and mnsclemuscle all induced germinal vesicle breakdown in vitro of coelomic oocytes of PhascolosomaPhllSCO!OSOlllll agassbi.agassizi. The majority of coelomic sperm in P. agassizi remain in the form of clusters until a short time before nephridial uptake when the clusters break down into free spermatozoa. Experimental dispersal of spermatid clusters "vaswas accomplished in vitro bvby exposure to hypertonichvpertonie seawater. The resultant free spermatozoa became motile within 24 hours afteralter transfer to seawater. After nephridial uptake gametes remain in the nephridium onl~'onlv a brief period before spawning OCClll'S.occurs. Gerould (1907) noted in GolfingiaGolftngia vulgaris and PhascolopsisPhascoloJ)sis gouldigOllleli that the nephridia filled with gametes a few hours bdorebefore spawning. Akesson (19.58)(1958) reporteJreported that oocytcsoocytes of Phascolion strombi underwent gcrminalgerminal vesicle breakdown 12 hours before spawning. In small transparent specimens of ParaspidosiphonPar{/spidosipholl fischerifischer! in which nephridia were distended with gametes at the time of collection, spawning occurredocciu'red within periodspcriods varying from 1 to 24 hours. 4. SIPU0:CULASIPUN'CULA 93

The sequencesecjuence of male and femalelemale spawning is variablevarial)lc in differentdiffeient species.s pecics. At least two species of sipunculans,sipunciilans, GolfillgiaGolfmnia vulgarisüiilßiiris and Vhas-P!ws­ colion strombistromlJi (Gerould, 1907, Akesson, 1958), follow Thorson's rule of cpidemicepidemic spawning in which the males spawn first, stimulating the spawning of the females (Thorson, 1946). In populationspopnlations of Phascolo-PJwscolo­ soinasoma agassizi from ~'[onterc~T,Monterey, CalCaliforniaifornia (Towle and Giese, 1967) and from the San Juan ArchipelagoArchipi-lago (Rice, 1967), spawning of females has been observed to precede that of males. In the latter population the female spawning may be followed by the male or cithcreither sex may spawn independently.independentl\'. OnlyOnl\' a small proportion of the animals placed in one container spawned at the same time. It waswa.s found that spawning can often be triggered in the laboratory, if gametes arc present in the neph­neph- ridia, by a changcchange to freshlyfreshlv aerated waterwatei- or a sudden change in tem­tem- perature. No observations have heenbeen mackmade on the spawning of sipunculans under naturalnattiral conditions in the field. GeronldG(MOuld (1907), in his studies of GolfingiaGolfingid vulgaris and PhascoJopsisPlwscolopsis goiildi,gouldi, reported that spawning is confined to homshours of darkness between 2000 and 0400 or 0500. He stated further that when animals wercwei^c main­main- tained continuollslyeontinuouslv in a dark aquariumaquariiun the rhythmrhvthm was interrupted and spawning sometimes occurred during the day.dav. His explanation for spawning at night was that the sipunculan's bodvbody relaxes in darkness, resulting in distensiondistt-nsion of the nephridia with consequentconsecjuent nephridial up­up- take of surrounding.surrounding seawater and aecumnlationaccumulation of mature h~'drotropichvdrotropic eggs from thcthe coelom. Spawning at night has also heenbeen rcportedreported for PhascolionPlwscolioll sfroinI)istrol1lhi and GolfingiaGolßngia elongata (Akesson, 1958, 1961n).1961a). How­How- ever, inill PhascolosomaPlwscnlosnma agassi;:,iagassizi spawning did not oecuroccur exelusivelyexclusively at night. Of a total of 89 spawnings, 9% spawned during da;'lightdavlight hours, 38% at nightiiight but illin the artificial ilhlll1inationillumination of the laboratory,laboratorv, and 53%53f in darkness. Thus Gerotlld'sGerould's ('xplanationexplanation for night spawning would not I",be applicable in 47%47'l of thcsethese cases. Again, in ThemisteTheiniste l>!Jroides,pijroides, of the spawnings of 20 animals recorded over a 3-year period, 5.5%55% oc­oc- curred overnight, the remainder during daylight hours (Rice, 1967). The onlyonlv snggcstionsuggestion for periodieity'periodicitv in spawning in sipunctllallssipunculans is found in records for PliascolosoiiiaPlwscolosOll1ll agassizi (Rice, 1967). A total of 99 spawnings were recorded in the laboratorylaboratorv over a period of 3 years.)ears. The data, summarized in Fig. 15, reveal a peak of maximum spawning each year\-ear as well as a certain periodicityperiodicitv throughout the breeding season. In 1962 and 1963 the periodicit~,periodicitv between spawning peaks ranged from 21 to 29 days. There was no obvious correlation of spawning periods with either tides or phases of thcthe moon. Since thescthese observations were made under uncontrolled conditions in the laboratorv it cannot hebe as­as- sumed that the same periodicity exists in thcthe natural habitat. 94 MARY E. RICE I 2 1

ale , 4 I mOle

f' 1.1 ""tu.--r"----.---y-~-r__._--_r_-.r--r- 10 ~ 10 ~ "0 7~ I 5 JUt., AUGUst 'SEPT

FIG.FK;. 15. Spawning dates and number of spawnings of PhascolosomaPlwsco!osoma agasslzlagOKsizi from the San Juan Archipelago, Washington.\Vashingtnn. Spawnings were recordedrecord~d in tIlt'the lahora­labora- tory from 1962 to 1964.]964. (From( From Rice, ]967,J967, p. ]47.)J47.)

Although there may be a periodicity in spawning during tliethe breeding season there is not a complete gametogeniegametogenic cycle between each spawning in sipunculans. A single animal may spawn repeatedly during a breeding season as evidenced in the six recorded spawnings of a single female of Themiste pyroiclespyroides during an 8-weck period (Rice, 1967). Itft seems probable that each spawning is preceded by an malmationmaturation of only a certain percentage of the largest oocytesoocytcs in the coelom which are then collected by the nephridium. Estimations of breedinghrceding periods have been made for 10 species of sipunculans byhy one or more of the following methods of study: observa-observa­ tions on spawning in the laboratory over a period of a year,y('ar, annual studies of gametogeniegametogenic cycles with peiiodicperioclic e.xaminationexamination of coelomic oocytes and gonads, or from studiesstudics concerned primarily with develop-develop­ ment. Additional observations on spawning of 9 species have been made at sporadic intervals, butbnt not carried out over a period of a year. It is important to note, therefore,therefore, that in these instances the times of re-re­ corded spawning are not indicative of thethc range or duration of breeding seasons, but show only thethe months in which spawning was observed -- and can be expected to occur. The information is included for its possible value toto persons who may wish to carry out further investigationsinvestigations inin thethe localities named. The data arearc summarized in Table I. TABLE I BRKEDINGI3R~;ED[NG PERIODSPKRIODS OF SIPUNCULA:-ISSIPUNCULA.VS

BrecdingBreeding period of months Species Locality HefercnceHeference of observcdobserved spawning

Golfingia c10ngataclonijala Hoscoff,I'osc-off, FranccFrance Akesson, 1961a July, Aug. GoljingiaGolfingia minutaminula Kristinebcrg,Kristincbcrg, SwcdenSweden Akcsson,Akesson, 19.58195S Sept.-Nov.Sept.-Nov.»a GoljingiaGolfingia pcUucidapel/ucida Fort Pierce, Florida Rice,Kice, unpublished Feb.-.t\Iay,Feb.-May, Aug., Sept. Nov. GoljingiaGolfing i a ppllge[{('1lsis ugctic iiis is Han•San .JuanJuan Islands, Washington Hice,Pice, 1967 Oct.-Jan."Oct.-Jan.» GoljingiaGolfingia vulgaris Hoseoff,Koscoff, FranccFrance Gcrould, 1907; Gonse, I956b1956b June-Sept.June-Sept."a ParaspidosiphopParaspido.nphori slecnslrupisteclIstrllpi British Honduras Hice,Kice, unpublished June Parasp idosiphoh fischcrijischeri Isla ?lIargarita,Margarita, \"enczuela\'enezuelá Hicc,Hice, 1975 Oct.-Dec. Isla PcPerico,rico, Panama Wcc,Rice, unpublished July Galeta,üalcta, Panama Hicc,Pice, unpublishcdunpublished Junc,June, July KcyKey Largo, Florida Hicc,Kice, unpublished Oct. Phascolion cryptuscryplus \"irginiaN'irginia Key,Ke\', Florida Hicc,Kice, 1975; Riec,Kice, unpublished July-Nov. CO l'hascolionPhascolion strombisirombi Kristineberg, SwedcnSweden Akcsson,Akesson, 19.18195S Sept., Oct., Nov.Nov.«a I'hascolopsisI'ha.5colopsis gouLdigouldi Newport,Ncwport, HhodeJihode Island Gcrould, 1907 Junc-Aug.June-Aug. § Woods Hole, ?lIassachusettsMassachusetts Gcrould,Gerould, 1907 Aug.-Sept. PhMcolosomaPha.5colosoma agassizi :\Iontcrcy,Monterey, California Towle and Giesc,Giese, 1967 l\Iar.-l\IayaMar.-May° San Juan Islands, Washington Hicc,Kice, H1671967 June-Sept." > l'hascolosomaPhascolosoma antillant/IIanlillantm Key Biseayne,Biscayne, Florida Hicc,Rice, 1975; Hicc,Kice, unpublishcdunpublished July-Scpt.Julv-Scpt. Phascolosomaj'hascolosoma arcnatumarcualum Hossl{oss Hiver, Qucensland,Queensland, Australia Green, 1!l7i)1975 Dec.-Feb.IJec-Feb."a l'hascolosomaPhancolosoma pcrtiiccn^peri,/cellS La Par!!;ucra,l'arguera, PucrtoPuerto HicoKico Hice,Kice, 1975 April Barbados Hice,Kice, 1975 Feb., l\lar.Mar. Isla :\Iargarita,Margarita, \'\'onezuelaenezucla Hice,Rice, 1975 Oct.-Feb. KcyKey Biscaync,Biscayne, Florid~i.Florida Hicc,Rice, 1975, unpublished Nov.-Apr., June-Scpt.June-Sept."a Themi",tcThomiste alutaceaahxtacea Isla :\Iargarita,Margarita, \'cnezuclaVenezuela Hiec,Kice, 1975 Oct.-Dcc.Oct.-Dec. Fort Pierce, Florida Hicc,Rice, unpublished l\Iar.-Aug.,Oct.Mar.-Aug., Oct."a 1'hemistcThemislc lagcnifonnislagcniform is Fort Pierce, Florida Hice,Kico, unpublishcdunpublished :.\Iar.-Aug.,Mar.-Aug., Oct., Dec.Dec."a ThrmIisteThf.misle pyroidcspyroides San Juan Islands, Washington Hicc,Rice, l!)671967 l\Iarch-August"March-August" \"ancoU\'crN'ancouver Island, British Columbia Siphonosoma CllllIanellSCciiniancnse Tampa Bay, Florida Hiec,Kice, unpublished June Sipuncll{usSipuiKulus nuduanut/lIs Tampa Jhy,Bav, Florida Ricc,Rice, unpublished June Naples, Italy Hatschck,Hatschek, 18831S83 July

" Indicates duration of breedingbrecding season; population (losen'edobserved during all scasonsseasons of thethc year. CO 96 MARY~[ARY E. RICE

Of the 10 species for which the brcedingbreeding season has been determincu.determined with some certainty, 6 are temperate and 4 tropical or suhtropical.subtropical. All of the temperate species and one of the tropical species (P7wscolosoma(Phascolosoma arcllaturrt)arcuatum) havchave well-defined hreedingbreeding seasons. Three of the tropiealtropical or subtropicalsiibti'opical spccies,species, ThcmistcThemiste lagcniformis,lageniformis, T. all/tacca,alutacea, and PhascoloaomaPlwscolosmna perhicens,perlucens, appear to spawn throughout much of the year.

4.4 Development

4.4.1 Fertilization

Sipunculan eggs when spawned into the seawater arcare arrested in the first meiotic metaphase. The processproccss of fertilization, 11eginningbeginning wiwithth the entrance of the sperm into thethc egg, includes the extl'llsionextrusion of the first and second polar ]:)odiesbodies and enlargement of the sperm nucleus to form the male pronucleus, and, finally,finall\', the fusion of the female and male pronuclei,pronuclei. Gerould (1907) described fertilization in the eggs of Golfingia vulgarisvu]garis and PJwscolopsisPlwscolopsis gouldi, hutbut he was unable to observe sperm penetration. Both Gerould (1907) and Akesson (1958)(195S) supposed the sperm to cnterenter through a pore of the egg membrane. However, the width of the sperm head excecdsexceeds that of the pore several times and, in more recent studies (Rice, 1966), it has been demonstrated that sperm penctrationpenetration is effectedeliected bybv thcthe formation of a hole in thcthe egg cnvelope.envelope. Penetrating sperm or the resulresultanttant sperm cntryentry holes have heenbeen observed in living and/and/oror sectioned eggs of the following species: PlUlSColoso1/w.PJuiscolosonm agassiz.i,agassizi, P. per-per­ lucens,hicens, P. varians, Golfil1giaGoJjxngia pugeftensis,pugcttellsis, G. pellucida,pel/ueir/a, and TllemisteThemiste pyroicles.pyroides. In addition to the sperm entry hole in the eggcgg cnvelope,envelope, the penetrating sperm of TllCmisteThemiste pi/roidespyroidcs also Icavesleaves behind it a clearly discernible traektrack in thcthe thiekthick jelivjelly coat of thcthe egg. Before penetration this sperm extends an a<:rosomalacrosomal filament 50 p'm|U.m in length from the edge of the jellyjellv to attach to the egg envelope. The sperm entry hole may persist for 2 to 3 days in thetlie developing embryo.embrvo. FollovvingFollowing sperm pcnetration,penetration, maturation of the egg is completed with the formation of the two polar bodies. In the elliptical cggseggs of Phascolos-Phascolos­ OlHlloma the cytoplasm mundsrounds up, pulling away from the egg envelope at the two poles, and the polar bodies arcare formed at the animal polc.pole. In spherical eggs such as those of most Golfingia and Thcmiste,Themiste, a space between egg envelope and cytoplasm forms at the animal pole at thcthe time or shortly before extrusion of the polar bodies. 4.4, SIPU~CULASIPUNCULA 97

Details of sperm migration and zygote formation have been studied in GolfingiaGolftngia vulgarisvtllgaris and Plw8colo)JsisPhascolopsis gouldigouleli (Gerould, 1907). In these species, the sperm, after entrance into thetlie egg, rotates until its long axis is parallel to the surface of the egg. An aster with centrosome appears at the base of the nucleus and within about 10 minutes the sperm, led by the aster, migrates to the center of the egg where it increases in size and the chromatin becomes separated into a loose nct­net- work. At the same time the astrosphere enlarges and from it promi-promi­ nent fibers radiate throughout the cytoplasm except in the direction of thetlie animal pole. During this period of sperm enlargement the two polar bodies arcare given off by the egg. According to Geroulcl'sGerould's account, the reduction or transverse division occurs at the first polar body division and the equational or longitudinal division takes place at the second division, the revcrsereverse of the usual process. During maturation divisions thcthe number of chromosomes is rcdueedreduced from 20 to 10, the haploid number. After extru­extru- sion of the second polar body the nucleus of the egg is brokenbrokcn up into 10 chromatic vesicles which soon fuse to form the female pronucleus. A centrosome, from which astral fihersfibers radiate, lies in a fold of the pionucleuspronucleus toward the center of thcthe egg. As thcthe two pronuclei move toward cacheach other, the astral rays of the female pronucleus become less prominent and the centrosome moves to oncone side, about 9090°0 from that point which will first contact the spcrmsperm nucleus. As the sperm nucleus mOvesmoves toward the animal pole, its astrosphere moves to one side and decreases in size. At the time the two pronuclei unite, their astrosphcres arearc of approximately the samcsame size, hutbut the astral fibcrsfibers of the male pronucleus,pronuclcus, are much more prominent. As the first cleavage approaches, the two asters are equal in size.

4.4.2 Embryonic Development

4.4.2.1 CLEAVAGE Cleavage of the sipullculansipunculan egg is spiral, unequal, and holoblastic (Fig. 16a,b). Cell lineage of one species, Go/fingiaGolfingia vulgaris, was studied through the 48-cell stap;estage bybv CerouldGerould in 1907. Startinp;Starting with a dexiotropic cleavage at the third division, the cleavage planes alternate in direction to the 48-ce114(S-cell stage, after which the spiral pattern continues onlyonlv in certain areas of the egg. A characteristic feature of the cleavage of Gol­Gol- fillgiafingia vulgaris is the relativelyrelativelv large size of the micromeres at the S-cell8-cell stage. At the 16-cellIG-cell stage, the 8 micromeres of the animal hemisphere, all of approximatelyapproximatelv equalecjual volume, exceed illin size all of the cells of 9898 MARY E. RICE -t. 99

lb't l

FIG. 17. Forty eight-cell stage of GolfillgiaGolfingia vulgaris,viilgaris, anterior hemisphere, showing the molluscan cross and intermediate cells. Rosette cells are dotted, cross cells are dashed, and intermediate cells barred. Prototroch cells are around the periphery. (Redrawn from Gerould, 1907, fig. D, p. 99.) the vegetal hemisphere except those of the D quadrant; the 2d cell or somatoblast is the largest cell at this stage and the second largest is the 2D. The relative size of the micromeres in early cleavage is reflected in the extraordinarily large size of the prototroch cells in the embryo. At the 48-cell stage the animal hemisphere consists of the 32 cells whichv^'hich make up the apical plal:eplate and the prototroch. The spiral pattern is inter­inter- rupted at the 48-cell stage by the radial division of the lql",Iq'-, the result of which is the formation of the apical cross. The cross cells are in the position of the molluscan cross, that is, the arms extend out from the rosette cells along the future sagittal and frontal planes of the embryo in a radial direction rather than in the interradial direction characteristic of the anneIidanannelidan cross (Fig. 17). This feature has been misunderstood by manymanv authors probably because of Gerould's adoption of terminology

FIG.Fic. 16. Diagrammatic representation of developmental stages of Phascolosoma perlucens,perlllcens, illustrating events in the metamorphosis of the trochophore. (a) Two-cell stage.stage, (b) Eight-cellFight-cell stage.stage, (c) Early trochophore. (d) Late trochophore. (e-f) Premetamorphosis stages, (g) Planktotrophic pelagosphera !;lrva,larva, immediately after trochophoral metamorphosis, 2~~32^,-2-3 days of age. (h) Pelagosphera larva, about 1 week old.old, As the trochophore elongates dUringduring metamorphosis (e,O(c,f) the egg envelope loses its porositypOl'Osity and lamellation and is transformed into the cuticle of the pelago­pelago- sphera larva (g,h). Immediately before metamorphosis (f) the gut cavity and coelomic cavity are formed and metatrochalmotatroehal cilia appear.appear, At."^t the time of metamorpho­metamorpho- sis the anus and mout]lmouth are opened to the exterior, the prototroch reduced and terminal organ everted (g). n,a, anus; at, apical tuft; cu, cuticle; en, egg envelope; es,as, esophagus; in, intestine; 111,m, metatroch; 1110,mo, mouth; p, prototroch; s, stomach; st, stomodeum; to, temlinalterminal organ. 100 MARY E,E. RICE

previously used for polychaetes (Clark, 1969);1969): however,howcver, when\-vhen using these terms he places them within quotation marks. Thus he labelslauel.> the cross cells "intermediate" cells, apparently referring to the intermediate cells of annelids, not of sipuneulans.sipunculan.>. In his figure of the apical region >!Í of the 48-ceH48-cell stage, reproduced here (Fig. 17), the prominent cross i cell arearc clearlyckarly seenseCll to be inin thethc radial position. The 4 cells of the rosetterosettc and the 16 cells of the prototroehprotntroeh arearc notable for theirthcir large size. The prototroeh, derived from the trochoblasts, Iq-1q" cells, consists of a girdle of 16 primary cellscclls at the 48-eell48-e('11 stage; later threethrec secondarysccondary cells, probablyprobahly derived from intermediate cells,ceJJs, are added to complete the prototrochal girdle of the trochophore. A gap persists in the proto-proto­ troehtroch between cellsceJJs of the C and D quadrants, joining apical and somatic plates, and has been designated as the dorsal cord. The division of 3D gives rise to 4d which immediatelyiml11eeliatel~T divides to form two equal cells which sink below thethc surface to form thethc mesodermal teloblasts.tcloblasts. Additional information availableavailablc on early cleavage of otherothcr species indicates that the relatively large micromeres arearc a feature of volkyyolky eggs and lecithotrophic development, whereas inill eggs with relatively small amounts of yolk,volk, micromeres arearc smaller than, or in some cases equal to, the macromeres.macromcres. For example, in the 8-cell stages of Phascolo-Phascolo­ SOI/Wsoina agassi;:,iagassizi and P. perhtcens,perlllcells. species with eggsc-ggs low in yolk content, the micromercsmicromeres and macromeresmacromercs in the A, ll,13, and C quadrants are all approximately the same size; Id is slightly larger and ID is by far th(~the largest of the 8 cells. The largest eellcell of the 16-c('1116-cell stage of P. agassizi is the somatoblast 2d whiehwhich is nearlv twice thcthe size of 2D. The relative size of micromcresmicromeres and macromeresmacroIl1ere,~ at the 8-cellS-cell stage of 10 species of ,5ipullculanssipuneulans is found in TallIeTable II.11.

4.4.2.2 GENEHAI.GêNERAI, DEVELOl'~IE;o.;TALDEVELOI'MEXTAL PATTERN.'>PATTERNS Several patterns of development arcare known to occur in the Sipul1cula.Sipuncula. OncOne is direct development, defined here as nonpelagic clevelopmentdevelopment in which ciliated stages are ahsentabsent and the emllryoembryo transforms directly into a small, crawling worm, usually hatching from aileone or more egg coats. The remaining patternspattc'rns are classified as indirect development and include either one or both of the following swimming larval stages: ((1)1) trochophore, characterized bybv ciliated prototrochprototroeh and apical tuft, and (2) pelagosphera with prominent ciliated metatroehalmetatrochal banel,band, lacking a prototrochprototroeh or with considerably reduced pl'ototroch.prototroeh. Information is now available on the development of 17 species of sipunculans,sipuneulans, including 5 species of Golfingia, 4 speeiesspecies of Plwscolosoma,PhascoJosoma, 3 of Themiste, 2 of Phascolion.Phascolion, and aileone each of PhascolopsisPhascolopsis,.. Paraspidosi­Paraspidosi- phon, and SiPIiIlCllllls.Sipunculus. Table II lists the species according to devel- 4. SIPUNCULA 101

opmcntalopmental pattel'l1s.patterns. Four common catcgoricscategories are designated: I, direct development; II, indirect development with one larval stage, thcthe tro­tro- chophore, which transforms into the vermiform stage; III, indirect development with two larval stages, thcthe trochophorctroehophore and lecithotrophic pelagosphera; and IV, indircctindirect development with two larval stages, the trochophoretroehophore and planktotrophic pelagosphera. The species listed range in distribution from the cold waters of the Northeastern Pacific and North Atlantic to the tropiCilltropical waters of the Atlantic. ThrceThree speciesspecie.~ arcare known to unuergoundergo direct development: Golfingia minl/taminuta from Kristineberg, Sweden (Akesson, 19.58),1958), PIU/scolionPlia^colion cryptllscrijptus from Virginia Key, Florida (Riel',(Rice, 1975), and ThemisteThemiate pijroidespyroitles from San Juan Archipelago, Washington (Hice,(Rice, 1967). The eggs are rich in yolk, blastulae are solid, and gastrulation occurs bvby epiboly.epibolv. In two of thcthe species in which 8-celI stages have becnbeen CXilminedexamined (Themiste(Tliemiste pijroidespyroides and PhiiscolionPhascolion cryptus),cnjptus), the macromeres in the A, B, and C quadrants are smaller than their respective micromercs.micromeres. DerivativcsDerivatives of the large micromercs,micromeres, the prototroch celIs,cells, form a prominent band of noncilatednoneilated ceJlscells of extraordinary size. At this stage, comparable to thcthe trochophoretroehophore although lacking cilia, the cytoplasm is retracted from the egg envelope in the pretrochal area around the apex of the cmbryo,embryo, forming an apical groove which delimits the ceJlscells of the nonciliated rosette.rosettc. Joining prctrochalpretrochal and posttrochal hemispheres a dorsaldor.sal groove interrupts the prototrochal band and immediately posterior to the proto-proto­ troch a space between cytoplasm and egg envelope marks the p(jsitionposition of the stomodeum. At the age of 2 da:'sdavs (2.5°C),(25°C), Phascolion cryptcnjptusliS hatches.hatches as a crawling worm from the anterior portion of the egg envelope and the surroundingsurromiding jelly coat (Fig.( Fig. 18).IS ). The emhr:'oembrvo first elongates posttrochallv,posttrochalIy, rupturing the jellyjeJly coat postt'riorlypost(>riorlv and then the pretrochal region, folIowedfollowed by the prototrochal, separates from the egg envelope resulting in a large anterior cavity.cavitv. As the retractor musclcsmuscles hecomebecome functional the head is extendedext(-nded and withdrawn and through this activityactivit\' the antcrioranterior egg envelope is severed from its posttrochal connection to the embryo. Thus the jellvjelly eoatcoat and anterior egg envelope are cJiscarded,discarded, a new cuticle is formeu.formed to cover the pretrochal and prototrochal I'l'gions,regions, and the post-post­ trochal egg ellvelopeenvelope is transformed into the posterior cuticle of thetlie vermiform stage. In TlwlIlisteThemiste pi/roicles,pyroides, the entire egg cnvelopeenvelope appears to be retained as the cuticle of the V('rmiformvermiform stage. TIl('The cmbryoembryo develops within the envelope and surrounding jelly coat until the age of 8S to 9 daysdavs (12°-13°C)(12°-13-C) when, hyby a seriesscries of movements extencJingextending and contracting the body, it hatches out of the jelh'jelIy coat as a smalI,small, crawling worm. TABLETABLK II o bo A SUMMARY OF'OF 1)EVl·:LOp,'.n::-rr.\LDI;VI;LOI>íII:.\TAL CIUR.\CTERISTH;SCHARACTI-;BISTICS 0,'o¡r TH,:THK BIPCCi"CCL.\SIPUN'CULA

Length of pelagicpelag;ic staf!;estage X-Cell8-Cell stage Egg size relative ~izesize of Pelagosphera diameter or micro- and length X width macromeresmaeromeres in Lecitho­Lecitho- Plankto­Plankto- Species" (I'm)C/jm) quadrants A, B, C Gastrulation TrochophoreTroohophore trophic trophic

Category I GolfingiaGotfingia ntminuta'imlla I 260-280 X 215-23021~-230 ? EpibolyJCpiboly 0 0 0 1'hemisteThemiste pyroides^pyroides' 190 2\1Micromeresicromeres > Epiboly 0 0 0 s:: :\1Maeromeresacromeres >-> ::0 PhascolionP hascolion eryptuscryplun'^G 136 2\licromeresMicromeres > Epiboly 0 0 0 >< :'lacromeresMaeromeres t"l Category II ::d 1')~ Phascolion slrombi'strombi1 12.-)_.) Micromeres > EpibolyP^piboly 8S DaysDa3-s 0 0 n :'lacromeresMaeromeres t=:I Phascolopsis gouldi^gouldi' lS0-1801.50-180 ;\]Micromeresicromeres > Epiboly 3;~ Days 0 0 2\JMaeromeresacromeres Category III Golfingia vulgaris'vulgarise 1,jO-I80150-180 :'IicromeresMicromeres > Epiboly 3 DaysDaj-s 2 DaysDavs 0 :'IacromeresMaeromeres Golfingia elongalaelongala'2 12;j125 ? Epiboly + 2 Days 4 Days 0 invaginatiollinvagination Golfingia pugettensis^pugetlensis· 160 :'IicromeresMicromeres = EpibolyICpiboly S8 DaysDnvs 13 DaysDavs 0 :'lacromcresMaeromeres ThemThemisteisle alulaceaahUacea'6 138 ? Epiboly 2 Days 60 DaysI>ays 0 Themiste lagenljonnis¡ageniformis'S 145 l\IicromeresMicromeres > Epiboly 0 8-12 Days 0 :'IacromeresMaeromeres Category IV GoljingiaGotfingia •peUucida'pellucida7 16;;16Ö ? ? 3 Days 0 1 :\IonthMonth ParaspidosipkonParaspidosiphon fischeri^fLScheri· 103 X 94 ? ? 2 Days 0 1 ?IonthMonth PhascoloaomaPhascolosoma aga.ssizi·ayassisi^ 140X110140 X 110 :\IicromeresMicromeres = Epiboly + 8-10 Days 0 1 2\IonthMonth :\1Macromeresacromeres invagination PhascolosomaPha~colosoma antiliarwII·anlillarum^ 127 X 97 ?•? ? 3 Days 0 1 Month PkascolosomaPhascolosoma perhtcens''perlucens· 112 X 91 ?llicromeresMicromeres = Epiboly + 3 Days 0 1 l\lonthMonth l\lacromeresMacromeres invagination Phascolosoma varians·varians" 104 X 90 ? ? :33 Days 0 1 ?IonthMonth SipuncllilisSipuncutun nudlls'niidus'' 120 l\IicromeresMicromeres < Invagination 3 Days 0 1 :\IonthMonth :\lacromeresMacromeres

a» References: 1. Akcsson, 19;;/);19Ó8; 2. Akesson, 1961a; 3. Gerould, 1907; 4. Hatschek, 1883; .'i. nice,Hice, 1967; 6. nice,Rice, 1975; 7. Rice, unpublished; S.8. Williams, 1972. f'- ....Ul '"tl C Zz (")n CG t-< >->

o 104 MARY E. RICE

a c d

FIG. 18. Den,lopmentalDevelopmental stages of l'hascolionPhascoliol1 cryptcnjptiis,LIS, depidingdepicting dir('ctdirect development and partial hatching from egg envelope. Photographs of living embryos. Scale, 2.525 I'm./im. (a)( a ) Two-cell stage. Pres('ncePresent»; of adhesive jelly layerla\ er is indicated by attached debris.debris, (b-c) Egg envelope has detaeheddetached in pretrochalpretroehal and prototrochal regions. Anterior space develops between envelope and embryo. Anterior end can be retracted as in (b) or extended as in (c), and by such movements thcthe l'I1\'c1opeenvelope is e\'entuallyeventually ruptured at the postlrochalpostlrocha! junction.junction, (c1)(d) Hatched vermiform stage. Head is retraded.retracted. Anterior egg envelope is cOlllpletelycompletely detached and posterior egg envelope has hccnbeen .: transformed into cuticle. [From Rice, 1975 (in part).]part ).]

The sticky jelK-jelly coats of the eggs of ThemisteThemisfe pyroides and PhascolionPlwscolion cryptllscryptus cause adhesion to anyanv surface contacted; thus these embryos develop attached to the substratum in the area of spawning. The dircctdirect development of GolfillgillGolßngia minuta, exemplifyingcxempHfying a form of primitive hroodbrood protection, takes place over a period of 3.3 to 4 weeks within the burrows occupied by the £t'males.females. The emhryosembryos do not clevPlopdevelop prototrochal cilia, but cytoplasmiccvtoplasmic blebs, similar to those noted in other species before formation of cilia, have been interpreted hyby Akesson to indicate that lack of prototrochal cilia in this species is secondary rather than primitive. An extremely weak and temporary ciliation with no loco­loco- motory function forms on marginal and rosette cells. The egg en­en- velope transforms into the cuticle of the juvenile, butbnt beneath it a definitive cutic-Iecuticle is apparent. Development is slow in this species; ~'Ollllgvoung wormsworm.s begin to migrate from the hurrowburrow at 1 month, but heads are not retractable until the agea~c of 6 weeks. The development of both PhascolionPlwsco!ioll stromhistrolllbi from Kristincherg,Kristineberg, Sweden (Akesson, 1958)195S) and VhascolopsiaP1IIIscololJsis gOMldigO/lldi from Newport,Ncwport, Hhoo<.,Rhode Island and '\loadsWoods 1I0k,Hole, },IassnchusettsMassachusetts (Cel'Onld,(Geronld, 1907) is marked by a single pelagicpelagiC larval stage, a kcithotrophiclecithotrophic trochophore (Category II, Table II). Eggs of the two species are macrolecithal with micromcresmieromeres exceeding macromeres in size in the A, 13,B, and C quadrantsc(iiadrants at the 8-cdl8-cell stage, the blastulae are solid, and gastrulation occurs bybv epiboly.epiboK-. Trocho­Trocho- phoresphares are characterized by rosette cells with long cilia and a handband of ciliated prototrochprototroeh ceils (Fig. I9a).19a). },IarginalMarginal cells on eithereitlier sidcside of 4. SIPUNCULA5lPUNCULA 105

b. a.

c. d.

1"1(:.I'n:. I19.D. Troehophorcs'I'locliophorcs of Sirllllculan~.sipiinculans. (a) Tnlt'hophorc'I'rochopliore of J'Il/1sco/"I'"i,Pli(iscolo}KÍs gOI/'di.U.^niUli. SmfaceSurface view; 20 hours. (Hedra\\"ll( Ucclrawu from CC'fIlllld,Clcrould, HJ07,1907, plat,·plate 8, figfig... 59.)SU.) (b) TrochophorcTrocliopiiore ofoF C;o/fil/giaGolfin^iu ('[ouga/a.elon^iila. J)ia~ramDiai^rani of sa~iltal.sagittal sct'lion.sec'ti{)n. (I1.l'drawl1(Redrawn fromfroui AkcsSOII,Äkesson, 1961a, p. 51516.)B.) (c) TroehophorcTrochoi^hore of Golfingil/Golfingia GII/garis.tyulgaris. Ventral vie\\',view, indicating l'cllcell outlincs;outlines; aboulabout -1040 hours. (Redrawll(Redrawn frolllfrom Cerould,C'.erould, I1907,D07, plateplatte 7, ligfig.. .sO.)50.) (d) Trol'hophoreTroehophore of 5ip"ICII[II,I'Siptincidiis Tllldll'.ntichis. Latent!Lateral view. Eggligg ellvelopeenvelope an underlying eeJlcell laycrlayer (serosa) have split at posterior end, prcecdingpreceding shedding of enw!ope.envelope.'(Redrawn(Redrawn from IJatsclwk,Ilatsehek, 11183,1883, plate 3, figlII'C:'figure 45.) a, anlls;anns; ag, apical groovc"groove; atat, aricalapical tllft·tuft; !lobo, huccalbuccal org,lfj'organ; Clllcm, cocloI11'coeloni; C'e, ,'Ve"eye; cnelend, cndodenll'endoderin; 19,Ig, lip ~land;gland; m, I11etatro~h;metatroch; 1;'p, prototroch; s(':se, scro~a;serosa; st, StOI;lO(iclll~""stomodenm. ,

the prototroch of Phaseo/ionPhascoHon slrollllJistroinl)i are¿we weakly ciliated and anterior to the prototroch of PlwscolopsisPhmcolopsis gouldi isis a band of well-developedwell-dewlopcd preoralprcoral cilia. At metamorphosismctamorphosi.s the prototroch cellscell.s degenerate and the 106 MARY E. RICE trochophore is transformed into an elongate worm, capable of extending and retracting the anterior portion of the body. The cggegg envelope of Phascolo'psisPlwscolopsis gouldigOHldi is shed at metamorphosis, but in PhascolionPlJascolion strombi it is rctainedretained as the cuticlc.cuticle. A developmental pattern consisting of two pelagic larval stages, a leci­leci- thotrophic trochophore and a lecithotrophic pelagosphera, is found in the following species: Golfingia elongata, G. pugettensis, G. vulgaris,mtlgaris, and Themiste alutacea (Category III, Table II). All are temperate or cold-water species except T. alutacea which ranges from the temperate water off North Carolina and Brazil to the tropical waters of the Carib­Carib- bean Sea. In these species a pelagic lecithotrophic trochophore metamor­metamor- phoses into a lecithotrophic pclagospherapelagosphera which swims for a short timc,time, often near the bottom, undergoes a gradual loss of cilia, and transforms into the veriform stage. The 4 species within this pattern show a gradient in development from least to most highly modified from that of the pre-pre­ ceding pattern. The development of Golfingia vulgaris most nearly re­re- sembles that previously described, whereas that of G. pugettensis is probably the most highly modified. In G. vulgaris the egg is rich in yolk and gastrulation is epibolic (Fig. 20a).2()a). There is no abrupt formation of metatrochal cilia since these are developed in the trochophore stagcstage (Fig. 19c). Metamorphosis of the trochophore occurs at 2 days when the egg envelope is ruptured and eastcast off and the anterior end of the body becomes retractable (Fig. 21c,d). Prototrochal cilia appear often to be destroyed, but metatrochal cilia slip through the envelope without injury. The larva continues to twirl, usually near the bottom, by means of the metatroch until the cilia arcare lost at 5 days of age and the young vermiform stage assumes an elongate shape and begins to creep along the bottom (Gerould, 19(7).1907). Development of Golfingia elongata is similar, exeeptexcept that during gas­gas- trulation an archenteron is formed. TlwThe trochophore is spherical with a long apical tuft, short prototrochal cilia, and long metatrochal cilia (Fig. 19b). Metamorphosis consists of elongation, beginning at 48 hours ((17°C),l7a C), and disintegration of the prototrochal cells. The egg envelope is transformed into the cuticle of the vermiform stage,stage. The trochophorctrochophore of TllemisteThemiste alutacea changes from the spherical shape of the blástulablastula to an oval shape with a prominent equatorial band of prototrochal cilia and a long apical tuft. Metamorphosis,VIetamorphosis, beginning at 28 hours (25(25°C)a C) with elongation of the trochophore, results at 32 hours in a leeithotrophielecithotrophic pelagosphera. During metamorphosis a well­well- developed circlet of metatrochal cilia and a postmetatrochalpostmctatrochal sphincter appear, the stomodeum opens through the egg cnvelopeenvelope to form the mouth and ventral ciliated surface of the head, and the cggegg envelope 4,4. SIPUNCULA 107

de

mes

a,

mes

c.

FIG. 20. GastrulationGa.strulation and mesodermme.soderm formation. PrototrochPrototioch cells are marked by dashes,dashe,s, other ectoderm cells arcare clear; endoderm is dotted; mesoderm is harred.barred. ((a)a) Sagittal section of embryoeml^ryo of GolfillgiaGolfingiu vulgarisvnh¿arís showing hlastopore,hlastopore. GastrulationGastnilation is epibolic; 14~~141/4 hours. (Rl'dnl\vn(Redrawn from Gerould, 1907, plate 6, fig,fig. 40b.) (h)(b) Cross section of embryo off)f GolfingiaGolfingiti vtdgaris;viilgaris-, 24 homs.hours. Mesodermal bands have split into splanchnic.splanchnic and somatic layers. (Redrawn( Redrawn from Gerould, 1907, plate 6, fig. 45. Prototrochal cilia, althoughaltliough described in text, are not shown in Gerould's ngures.)figures.) (c) Optical median section of embryoembr\'o of SipullculusSiptmcuhi.': mululinudus showing forma­forma- tion of ectodermalectodemial somatic plate and embolicembolie gastrulation. (Redrawn( Redrawn from Hatschek, 1883, plate 2, ng.fig. 15.) (d) Optical median seetionsection of embryo of SipllllculusSipunculiis nut/us,Jitidus, later than (c). Prototroch cells have surrounded embryO.embryo. (Redl'awn(Redrawn from Hatschek, 1883, plate 2, fig. 23.)23. ) ag, apical groove; at, apical t;lft;tuft; bp, blastopore; dc, dorsal cord; end, endoderm; mcs,mes, mewderm;mesoderm; p, prototroch; pc, posterior caVity;cavity; r, rosette cells; sp, somatic plate; st, stomodeum. is transformed into the larval cuticle. Transformation into the juvenile may begin as early as 7 to 8 claysdays with loss of metatrochal cilia, and by 2 weeks the coelomic yolk hasha.s 'beenbeen absorbed and the gutbrut completed. Within 1 month it has assumed the shape of the juvenile with elongated introvert and 4 ciliated tentacular lobes. The trochophore of GolfingiaGolftngia pugettensispllgettensis resembles that of ThemisteTheiniste alutacea, but a characteristic feature in this species is the presence 108108 ~lAHYMAKY E.!•:. nlCERICE

-01

c

c.

FIG. 21. Metamorphosis of Phascolopsis ^ouldigouldi and GolfingiaGo/fillgia v.ul^iiria.lOu/garis. (a) Tiocho-Tl'Ol'ho­ pliorophore of P.1'. gOllldi,¡ioi/lili, 4H41> houp's,hours, sliowini;showing circlt'lcirclet of preorulpreoral tiliacilia andanu developingde\Tlllping eiUielel'lllid,· beneath!>cneath eggt'!~g envelope.cnn·lopl'. Lateral view.vic\\'. (Redrawn fromfrolll Cierould,Ccwllld, 1907,1~)07, platepIal<' cS,S, fig.fig. 62.) (I))(h) Metamorphosing\I<'lamorphosing troehophoretrol'hophore of /'.P. goííWígouleli in ))roeessproL'ess of siic-ddingslw,lding egg envelope;envelupe; 434:3 hours. JJorsalDorsal \'ie\v.vielV. (Redrawn(Hedl'awn fromfrom Cerould,Cl'l'Ould, 1907, plateplat(· 9,n, fig.B.c:. 68.) (e) Metamorphosing troehophorelroehophorL' ofor G. viilf^aris.ladgaris. Egg envelopeel1\'l.'lopL' ha.shas beenl>eel1 easteasl off, prototroehalprototwehal eiHaeilia lost,lost, and prototroehalpmtotren)[i]d,Gewukl. f907,] 907, plate 7, fig.fig. .52.)52.) (d) LeeithotrophicL<'L'ilhotwphie pelago-pelago­ sphera larvalarva ofof G. i;i//g«n'.v;t:1I1gllTis; 60 hours.hours. LateralLall'ral view.\'icw. GutCui isis incomplete. ((Hcdrawn Redrawn fromfrom Gerould,Ccwuld, 1907, platepial" 7, fig.fig . 54.).5-1.) at,at, apical tuft;tllft; c, cnticle;elltic1e; e,e, eye;c)'l'; on,Cll, eggegg envelope;cnvclope; in,in, intestine;intestillE'; ni,m, mctatroch;Illdatrol'h; mo,mo, month;mouth; p,p, prototroch;prototroeh; pr,pr, preoralpn'()r:ll cilia;cilia; st, stomodemii.stomodeum. 4. SIPUNCULA 109 of conspicuous lipid droplets in the prototroch cells. MetamorphosisVIetamoiphosis of the trochophore occurs at 8S days (Bo-goe).(8°-9°C). Differing from others with this developmental pattern, the lecithotrophic pelagosphera of Golfingia rnJgettensisfyugettensis has a well-developed terminal organ by which it attaches to the substratum. A species with an apparently highly modified development, similar to that described above but not readily classified in any of the develop­develop- mental categories commonly found in sipunculans, is TlzemisteThemiste lageni-Iageni­ formis. Its early development is essentially the same as that of directly developing species in that it lacks a pelagic trochophore stage. However, thcthe nonpelagic stage, comparable to a trochophore, metamorphoses to a lecithotrophic pelagosphera \vithwith metatroch and terminal attachment organ (Williams, 1972). The final developmental pattern known to occur in the phylum in­in- cludes two larval stages: a pelagic lecithotrophic trochophore and a planktotrophic pelagosphera (Category IV, Table II). The develop­develop- ment of 7 specicsspecies falls \vithinwithin this classification. Four belong to the genus Phascolosoma:PhllScolosoma: P. agassizi, P. antantillamm,illaru m, P. perhicens,perll/cens, and P. varians. Other species are Golfingia pellucida, ParaspidosiphonParaspidnsiphon fischeri, and Sipul1cullisSipunculus 1111dus.nudus. PhmcolosomaPhascolosollla agassizi is a cold- to temperate-water species; others are either tropical or range from temperatctemperate to tropical waters. ViTithWith the exception of Siplll1cllIlisSipunculus 1l1lelliSnudus in which thcthe trochophore is uniquely modified, the dcvelopmentdevelopment is very similar in all species. Some of the characteristic features of this developmental pattern are summarized in Fig. 16, a diagrammatic representation of development through'metamorphosisthrough metamorphosis to the pelagosphera larva in PluiscolosomaPlwscolosol1la per-per­ hicens.Il/cens. The egg of this species is relatively low in yolk content, the blástulablastula has a small blastocoel in the anterior hemisphere, and gastrula­gastrula- tion, although mostly epibolic, is in part achievcdachieved by invagination with the formation of a small archenteron. During the trochophore stage the gut and mesodermal handsbands are differentiated and posttrochal elonga­elonga- tion heginsl)egins (Fig. 16c-e). Long prototrochal cilia, extending through the pores in the egg envelope function to propel the larva in a spiral swimming movement while the cilia of the apical tuft arc directed for­for- ward. Metamorphosis~1etamorphosis of the trochophore into tlwthe planktotrophicphlllktotrophic pelago-pelago­ sphera takes place at 212\ to 3 days of age (2.5°C) and lasts over a period of several hours. Metamorphosis~1etamorphosis is marked by formation of a new ciliary band, the mctatroch,metatroch, reduction of the prototroch, loss of the apical tuft, posttrochal elongation, expansion of coelom, rupture of the egg envelope in the )"('gionregion of the stomodeum to give rise to the ventral ciliated surface of the head and mouth, opening of the anus to complete thcthe gut, and formation of thcthe adhesive terminal organ (Fig. 16f-h). Pelago- nolio MARY E. RICE sphera larvae of P. perillcellsperlitcens have been maintained in the laboratory as long as 6 months, but a second metamorphosis into the juvenile form has not been observed. Other species included within this pattern, except SipllnculusSipiinculus nudus,/ludus, show essentially the same developmental features. The early development of S. l1udllSnudus is unique, particularly in the elaboration of prototroch and embryonic cavities. BlastomeresElastomeres are nearly equal in size, but slightly larger at the vegetal pole. In the blástulablastula a nearly central blastocoel is displaced slightly toward the animal pole. Gastrulation occurs by invagination of the endoderm eellscells which pull away from the egg enve­enve- lope at the vegetal pole, leaving a posterior cavity (Fig. 20e,d).20c,d). At the same time the cells of the apiealapical plate surrounding the rosette cells sink from the envelope forming an apical groove or cavity of the head around the central rosette. The posterior cavity and the apical groove are connected by a narrow dorsal canal "vhichwhich results from the sinking of a double row of small ectoderm cells. Ciliated cells, assumed to be homologs of trochoblasts (Gerould, 1903), then spread out to surround the embryo and form in conjunction with the egg envelope a structure which has been termed a "serosa" (Hatschck, 1883). The trochophore develops within the serosa and at the time of metamorphosis to the planktotrophic pelagosphera the entire serosa is shed (Figs. 19d and 22a,b). The pelagospherapclagosphcm has a ciliated metatrochal band and is similar to other such larvae in this developmental category. After 1 month in the laboratory a second metamorphosis occurs into the juvenile form (Fig. 22c).

4.4.2.3 GASTnULATIONGASTHULATION AI\'l>ANH) FORMATIONFOR},fATION OF MEsouEm\,[MESODERM In species with macrolecithal eggs, as in the first three developmental categories, gastrulation occurs entirely by epibolic movements with the exception of Golfingia elongata in which a small archenteron is formed. This mode of gastrulation, best described for Golfingia vulgaris and Phascolopsis gouldi (Gerould, 1907), occurs bvby an overgrowth of the ceJIscells of the dorsal somatic plate ventrally and laterally to enclose the solid mass of large endoderm cells. The blastopore is representedrepi-esented by the narrow ends of the club-shaped endoderm cells which maintain contact with the egg envelope until closed over by the somatic plate (Fig. 20a,b). Descendants of the stomatoblasts divide rapidly and in­in- vaginate in the region of the trochoblasts at the anterior end of the blastopore to form the stomodeum. In species with less yolky eggs and planktotrophieplanktotrophic larvae, as found in the fourth developmental category, gastrulation is accomplished in part by invagination, but the major role is usually played by epiboly. 4. SIPUNCULA III111

vnc

rm-

Fic.Fl(:. 22. Mctnmoiphosis;\Ietamorphosis of SiptinciihnSipllllCUlrtg iiiitlux.IllteillS. (a) TmchophorcTrochophore hatching fromfrom "serosa," i.e.,Lc., oggcgg cnvflopcenvelope plus prolotrochprototroch cells.cclls. Serosa is still attached in head region.region. (Redrawn from Hatschek,Hatsehek, 1883, plate 5, fig. 49.) (h) Pelagosphera larva, 2 days after shedding of serosa. (Redrawn from Hatschek, 1883, plate .5,.5, fig. .51.)51.) (c) Juvenile, after metamorphosismetamorphosi, of pelagosphera.pelagosplwra. (Redrawn(Rcdrawn from Hatschek,llatschek, 1883,[883, plat(^platt· ß,0, fig.fig. 71.) a, anus; h, hrain; bo, buccalhuccal organ; e,c, eye; in, intestine; Ig, lip gland; ni,m, metatroch;melatroch; mo,IIlO, mouth; n, nephridiuni;l1t'phridium; rm,I'm, retractor muscle; s, stomach; se,Sl" serosa; \nc,\'I1C, ventralventra] nerve cord.

Endoderm cells in bla.stulaeblastulae of .speciesspeeies of PbascoLosomu,Phascolosoma, such as P. agas-agas­ sizi and P. perlucens, arearc separated from thethe egg envelope bvby a large space around thethe vegetal pole. During gastrulation, cells of the somatic plate grow down over thethe entomerescntomcres filling thethe space inin an epibolic 112 MARY E. RICE

gastrulation. A narrow slit in the region of the blastopore marks an invagination leading into a narrow archenteron. Endodcrmal invagination is the primary process of gastrulation in SipunculusSipuncubis maIllsniidus (Fig. 20c,d). Closure of the hlastopore in this species is achieved by a fonvardforward and ventral growth of ectoderm of the dorsal lip; this ectodcrmalcctodermal growth at the same time gives rise to thcthe median somatic plate.platc. Mesoderm~desoderm is derived from a pair of mesodermal telolllaststelohlasts located on either side of the endoderm cells. Proliferation of these cells, observed in GoJpngiaGolfingia vidgaris,vlllgaris, PhascolopsisPlwscolopsis gouldi<^ouldi ((;erould,(Gcrould, 1907), Si}JunwlllsSipunculus nudusniidus (Ifatschek,(Hatsehek, 1883), and PhawolosomaPlwscolosoma agassiziagaasizi (Rice, 1967), results iIIin the formation of the lateral bands of mcsodermmesoderm (Fig.( Fig. 20b).

4.4.34.4.2 larvae ¿r

4.4.3.1 TROClIOPIlORETROCIIOPIIORE LAHVALAUVA St\ A trochophore stngestage is a chnracteristiccharacteristic phase in the developmentdevelopmcnt of all sipuncllians,sipunculans, although in some species it maymav be highly modified. Trochophores of several specicsspecies nreare illustrated in Fig. 19. The apical plate of the pretrochal hemisphnehemisphere consists of numerous smnllsmall ectouermectoderm cells which later give riscrise to the headh(nid ectoderm and brain and of a central circle of large rosette cells, usually bearing the long cilia of the apical tuft. The apical groove around the rosette cells may persist in thcthe early trochophore,trochophore. but later it is fillcdfilled bvby a growth of ectoderm cells. EmhedekdEmbedded in the apical platcplate in a dorsal position just anterior to the prototroch is a pair of small eye'spots.evcspots. The somatic plate covers the posterior hemispherchemisphere and isi-s the source of the ectoderm of the trunk. The stomodcumstomodeum is in a medioventral position, extending anteriorlyantcriorlv into the region of the prototroch. The large cells of the prototroch (19 in Golfingia vulgaris)üulgarín) form a prominent equatorial handband which spreads out over a large proportion of the surface of the larva. Prototroch cells are usually ciliated and always concentrated with yolk granules. The early trochophore retains the shape of the egg, but then an oval shape is assumed and in later stages the length increases as posttrochal elonga­elonga- tion begins. Enclosed bvby the overlying egg envelope, trochophores of all species of sipunculans arcare kcithotrophic.lecithotrophic. Rudiments of most internal adult organs are present at this stage. Three regions of the gut (esopha­( esopha- gus, stomach, and intestine) are differentiated and mesodermal bands, two cell layers in thickness, arcare present on either side of the gut. The pattern and functional significance of ciliation varicsvaries among differ­differ- ent species. Directly developing species, such as PhascolionPhascolioJl cryptus and 4. SIPUNCULA 113

Themiste JJyroides,pyroides, lack cilia entirely; Golfingia mintltaminuta shows a weak and temporary ciliation on pretrochal,pretrochaI, posttrochal, and rosette cells, but not on prototrochal cclls.cells. All othcrother species possessposscss ciliated prototroch cells and an apical tuft of cilia. Prototrochal cilia are well developed and serve as the means of locomotion in species of Phascolosoma and in Golfingia pugettensis,JJugettensis, G. peUucida,pelltlcicla, and Paraspidosiphon ßscheri.fischeri. In trochophorcstrochophores of G. elongata, G. vulgaris,mdgaris, and PhascoJopsisPlzascolopsis goiddi,gOldeli, pre-pre­ trochal and metatrochal bands of cilia are also present; metatrochal cilia are by far the longest in the two former spcciesspecies and most significant for larval locomotion, whereas in the latter the pretrochal cilia are the longest and the metatrochal vestigial. MetatrochalVIetatrochal cilia are present in the trochophore of SiPlltlC1l111SSipunculus ntllltlsntidiis but, enclosed by the serosa, are not functional (Fig.( Fig. 19d).19d ).

4.4.3.2 METAMORPHOSIS OF TIlETHE TROCHOPHORE

Metamorphosis of the trochophore may result in one of several devel­devel- opmental stages. In those species with only one pelagic stage, it may end the pelagic phase of development, the trochophore metamorphosing into a vermiform stage whiehwhich gradually transforms into the juvenile form. It may result in a second pelagic larval stage, thcthe lccithotrophicIccithotrophie pelagosphera, which swims for a short.short time in the plankton or near the bottom before transforming into the vermiform stage. Finally, it may give rise to a planktotvophicplanktotrophic peJagospherapelagosphera which swims in the plankton for a prolonged period before a second metamorphosis into a juvenile form.form, Main events of metamorphosis into a planktotrophic peJagospherapelagosphera are iJJllstratedillustrated in the diagram of developmmtdevelopment of Phascolo-Plw8colo­ 801MsonuL perhtcensperl/lcens (Fig. 16). Metamorphosis is characterized by reduction or loss of the prototroch, formation or expansion of the coelom, transformation or shedding of the egg envelope, and, in those species with pelagosphera larvae, by thcthe elaboration of the metatroch as the principal locomotory organ. Muscular activity is initiated either at metamorphosis or shortly there­there- after with the introversion and cxtrusionextrusion of the introvert. In species which undergo dircctdirect development without a pelagic trocho­trocho- phorephore stage, changes OCCllroccur at the time of transformation into the vermi­vermi- form stage which are comparahlecomparalile to those of the metamorphosis of thcthe trochophore, e.g., elongation, formation of the coelom, dissolution of prototroch cells, and the beginning of muscular activity. The coelom is formed by schizocoely or a splitting of the meso­meso- dermal bands into two layers, one of which forms the inner laverlayer of the hodybody wall and the othcrother an outer covering of the gut. In ;peciesspecies 114 MARY E. RICE

with relatively micromicrolecithallecithal eggs and planktotrophícplanktotrophic pelagosphera the coclomcoelom is formed before metamorphosismetaniorphosis of the trochophore, but under­under- goes a great expansion at this time. The coelom of species with yolky eggs forms simultaneously with other events of metamorphosis. The egg envelope of most species appears to be transformed at the time of metamorphosis into the larval cuticle. Transformation is ac­ac- companied by a loss of porosity and lamellation, first immediately posterior to the prototroch in the region of initial elongation, and later throughout. At the same time, as the larva stretches and the muscles become functional, the previousl)'previously rather rigid egg envelope becomes elastic and flexible. T"voTwo species, PhascolopsisPhaseolopsis gouldi and Golfingia vulgaris, have been reported to shed the egg envelope at metamor­metamor- phosis and another species, PhuscoUonPhaseo/ion eryptus,cryptus, loses the prototrochal and pretrochal portions (Figs. I8b-d18b-d and 2Ia,b).21a,b). Sipvneu/usSiininculus nt/dllSnudus shcdssheds not only the egg envelope, but also the underlying ciliated cells (Figs. I9d,19d, and 22a). Prototroch cells, usually heavily laden with yolk, release their yolk granules dnringduring metamorphosis into the coelomic cavity, thus contribut­contribut- {nging a substantial source of nutrition for the developing larva. The signifi­signifi- cance of this contribution is greatest in species.species with macrolecithal eggs and lecithotrophic development. Entire cdlscells of the prototroch of Gol­Gol- fingia vulgaris and PhascolopsisPhase%!,sis gOldcli,gouldi, including nuclei and cytoplasm, are cast into the coelom at metamorphosis and the region of the proto-proto­ troch is overgrown by ectoderm cells. In Golfingia elongata and Phas­Phas- calioncolion strombi the prototroch cells degenerate, releasing lipid and yolk granules into the coelom, after which they are replaced with ectoderm. Prototroch cells of G. [Jugettensispugettensis are marked by charactcristiccharacteristic large lipid globules which at the time of metamorphosis are released into the coelom where they are readily recognizable; the cells, although re­re- tained as a prototrochal band on the head of the pelagosphera, arc much reduced in size. The large yolk-laden prototroch cells of G. minuta and TllemisteThemiste pi/roidesprJroides begin to degenerate at the time of coelom forma­forma- tion releasing their granules into the coelom. In species of Phascolosoma,Phase%soma, release of yolk from the prototrochpl'Ototroch cells begins at an early stage of deVelopmentdevelopment when granules are passed from the cells into cavities formed along the inner sides of the prototroch.pl'Ototroch. As the prototrochal cells decrease in size, the size of the "prototrochal cavities" becomes larger. The gran­gran- ules appear to break down in the cavities where presumably they provideproVide nutritionnutritioll for the developing embryo. At metamorphosis the prototrochalprototl'Ochal cells, reduced in size, persist in the pelagosphera as a weakly developed dorsal band of cilia 011on the head. Before metamorphosis in SipuneulusSipunctdus nudus,Iludt/s, the size of the cells of the serosa is diminished as they gradually 4. SIPUNCULA 115 liberate nutrients into the anterior and posterior cavities (Figs. 19 and 20). By the time of metamorphosis of P. agassizi and S. 1IuduoSnudus all of the prototrochal yolk has beenl)een discharged into embryonic cavities. At metamorphosis of the troehophoretrochophore the metatroch assumes the role of the principalprinc.:ipal organ of locomotion. In GolfmgiaGolfingia vulgarisvtdgaris and G. elon­elon- gata the metatroch is present and well developed in the trochophore stage along with the prototroch, but at metamorphosis the prototrochpvototroch diminishes and the metatroch remains as the sole locomotory organ. In G. TJugettellsispugettensis the metatroch is established late in the trochophore stage, but is elaborated at the timetimc of metamorphosis when the proto-proto­ troch regresses. In species of FhascolosomaPhascolosoma metatrochal cilia are devel­devel- oped at the time of metamorphosis. In SipuncultlsSipunculus nudtlsnudus metatrochal cilia appear before metamorphosis but become functional only when the serosa is shed.

4.4.3.3 PELAGOSPHEHAPELACOSPHERA LAHVALARVA

Following the stage of tJ'Ochophore,trochophore, the pelagosphcra larva, whether lecithotrophic or planktotrophic, is characterized by a prominent band of metatrochal cilia and a regionalizationrcgionalization of the body into head, meta­meta- trochal region or "thorax,""thora.x," and elongated trunk (Figs. 23-25).2.'3-25). The head and thorax are retractable into the trunk and later become the introvert of the adult. Dorsally the head bears at least one pair of eyespots. Posterior to the eyespots and along eithercither side of the head, a prototrochal ridge forms a U-shaped band of weakly developed cilia. ThcThe ventral head is ciliated and divided into two lobes by a median groove which continues into the mouth. Posterior to the head the metatrochal region or thorax is greatly distended when the larva is swimming and at its maximum extension is by far the widest portion of the larva (Fig. 25a). The posterior boundary of the thorax is marked by the metatrochal sphincter muscle, which is usuallyusuallv contracted when the larva is swimming.swimming and when the anterior end is completelycompletelv retracted (Figs. 23-25).23-2.5). The elongated trunk may vary somewhat in shape depending on the degree of extension or contraction of the flexible larval body. At the posterior tip of most larvae there is a terminal attachment organ. Terminal organs arc absent in a few planktotrophicplanktotrophie larvae and in the 1ccithotrophiclecithotrophic larvae of GolfillgiaGolfingia vulgaris and ThemisteTheviiste alutaceaahitaceti (Figs. 21d and 24a), al­al- though the posterior end of the latter species does possess adhesive properties. Terminal organs of planktotrophic pelagospheras, in contrast to those of lecithotrophic larvae, are retractile, being withdrawn into the body by the contraction of a single pair of retractor muscles, originat­originat- ing from the dorsal bodvbody wall near the anus (Fig. 24b,c). In lecitho- 116116 MARYMARY E.E. HlCERICE

FIG. 23. PelagospheiaPelagosphera laivalarva of PhrincolosuniaPh(/scolo~vllla pcrluceiis.peril/cells. DiagramLJiagral11 showing internal structures at 1 week of age. Left, lateral view; right,light, ventral view; a, anns;anus; bg, buccal groove; bo, buccal organ; c, cuticle; dnn,drm, dorsal retractor muscle; e, eye; ep, epidermis; es, esophagus; in, intestine;intestine; 1,I, lip; 19,Ig, liplip gland; Ip,lp, lip pore; m,Ill, metatroch; n, nephridiuni;nephridiul11; p, prototrocli;prototroch; pr,pI', posterior retractors;retractors; s, stomach; to,to, terminalterminal organ; vnc, ventral nerve cord; vrm, ventral retractor mu.scle.muscle. (From Rice, 1975.) trophictrophic pelagosphera.spelagospheras thethe gutgut cavitycavity isis incompleteincomplete and, althoughalthough thethe mouthmouth isis usuallyusually open,open, thethe anusanus hashas notnot brokenbroken throughthrough thethe larvallarval cuticle.cuticle. TheThe gutgut ofof planktotrophicplanktotrophic larvaelarvae isis complete,complete, thethe anusanus openingopening dorsallydorsally onon thethe middlemiddle oror thethe anterioranterior regionregion ofof thethe trunktrunk andand thethe larvaelarvae arearc in-in­ dependentdependent feeders.feeders. AssociatedAssoci,ltcd withwith thethe feedingfeeding processprocess arcarc structuresstructures notnot presentpresent inin lecithotrophiclecithotrophic larvae:larvae: thethe lowerlower lip,lip, aa projectionprojection formingforming thethe posteriorposterior boundaryboundary ofof thethe mouth,mouth, andand twotwo organsorgans associatedassociated withwith thethe lip,lip, thethe buccalbuccal organorgan andand liplip glands.glands. TheThe buccalbuccal organorgan isis extrusibleextrusible throughthrough aa transversetransverse slitslit oror buccalbuccal groovegroove atat thethe basebase ofof thethe liplip andand • I thethe liplip glandsglands openopen throughthrough aa porepore onon thethe lip.lip. InIn additionaddition toto thethe 77 speciesspecies ofof planktotrophicplanktotrophic larvaelarvae knownknown fromfrom developmentaldevelopmental studiesstudies (Table(Tahle II),II), numerousnumerous larvaelarvae ofof uncertainuncertain speciesspecies 4. SIPUNCULA 117

a b

FIG. 24. Photographs of living pelagosphera larvae, reared from spawnings in the laboratory.laboratory, (a) Lecithotrophic pelagospherapclagosphera of ThemisteThemiate ailltacea;ahitacea; 3 days. Lateral view. Tenninal organ is ahsent.ah.sent. (From Rict',Riee, 1975.) (b) Lecithotrophic pelagosphera of GolfillgiaGolfingia ptigettcnsis;]JI/gettellsis; 13 days. Dorsal vic\\'.view. :\"oteXote presence of terminal organ. (From Rice, Hl67.)1967.) (c)(e) PlanktotrophicPlanktotrophie pelagospherapelagnsphcra of rhascolo.wmapfwseolosol/lll perlucem;perll/cens; about 7 days. Lateral view. (From( From Rice, 1975.)1975. ) Scale, 25 JLm.ßm. a,a. Anus; m, mctatroch; p, prototroch; to, terminal organ. have been reportedlepoited from theth<> oceanic plankton. Such larvaehirvae were firstfir.st recognized as belonging to the Sipuncula bybv HiickcrlUickcr (1898)(LS9S) from plank­plank- ton collectionscollection.s in the North and South Atlantic Ocean. Comparing them with Si]JlIllculusSipuncuhis itt/rIllSnudus known from Hatschek's studies, hche describeddescrilied three larval types which he named "Baccaria oliva," "Baccari"Baccariaa eitrindla,"citrinella," and "Baecaria"Baccaria pirum," distingllishingdistingin'shing them by bodybodv shape and form of cuticll­cuticu- lar papillae. These larva('larvat- are less spectacular in size and form'form than the large, transparent planktonicplanktoniC larvae described by later workers. Unaware of earlier work, \lingazinniMingazinni (190,5)(1905) describcddescribed a larva from one preserved and contractedcont¡acted speeimenspecimen collected in the plankton from a depth of 500 m between New Caledonia and New Zealand. Ill'He errone­errone- ouslyousK' assunwdassumed the larva to be an adult, creating ;1a new genus and species, PelagosJlhacraPelagosphuera alo!Jsii.aloiisii. Senna (1906),( 1906), working with material from the sanwsame expedition but from waters off India and Ceylon, suggested that struc­struc- ture'stures identified bybv \;IingazinniMingazinni as gonads were instead tlwthe glandular appendages of the esophagus which are now referred to as lip glands (Jagersten, 1963). Although !\'Iingazinni'sMingazinni's error was soon realized, the name pelagospherapelagosphcrn remains entrenched in thcthe litcraturcliterature and is toc1a~'todav used to designate the larval stage of sipuncnlanssipunculans which succeeds the stage of trochophorctrochophore (Rice, 1967). 118 MARY E. RICE

FIG. 25. Photographs of living plaiiktotrophicplanktotrophie pelagospherapchigosphcra larvaelai'vae from oceanic plankton,plankton. (a) Larva of unknown adult. From plankton of Florida Current. Antero­yVntero- dorsal vicwview of swimming.swimming larva with extended mdatroch.motatroch. (h) Same.Same speeimcn.spcciiricn asa.s (a) in quiescent position with terminal organ t'xtendt'd.extended. 1\'letatroehMetatroch partially extended.extended, (c) Larva of SiP"IlClllsSipunciihi.s pohjmyotufi.polymyotlls. From plankton of Florida Current. Larva swimming with metatroch extended.extended, (d) Same.Same specimen.specimen as (c). Larva 'juieseent;quiescent; metatroehmetatroch withdrawn. Scale, 250 ",m.p.m. 4.4, SIPUNCULA 119

Other reports of oceanic planktotrophic larvae of sipunculans have been made by Heath (1910), Dawydoff (1930), Stephen (1941), Fisher (1947), Akcsson (1961b), Damas (1962), JagerstenJägersten (1963), Murina (1965), ScheltemaSchcltema and Hall (196.5),(1965), and Hall and SchdtemaScheltcma (1975). Most of the earlier obscrvationsobservations wercwere made on preserved and contracted specimens, and only in more recent studies on living material (Jagersten,(Jägersten, 1963; Murina, 196;3;1965; Hall and Sche1tcma,Scheltcma, 197;3)1975) has the morphology of oceanic larvae been correctly interpreted. JiigerstenJägersten gavcgave the first accu­accu- rate description of the head region, pointing out the relationships be-be­ tween lip, lip glands, and buccal organ. He examined several larval types which he divided into two groups, those with a "smooth" body surfacesinface and those with a "rough" or papillated body cuticle. Characters generally used to distinguish pclagosphera larvae externally are texture of cuticle, presence and form of cuticular papillac,papillae, pigmenta-pigmenta­ tion, shape of head and lip, number and color of cyespots,eyespots, general body form, and presence or shape of the terminal organ (Fig. 25). Some distinctive internal characters arcare number of lobes of lip gland, number of retractor muscles, shape and pigmentation of nephridia, and presence and number of longitudinal muscle fibers. Pigmentation in different larvallarva! forms may vary from brilliant orange to various shades of pink, yellow, green, and brown. The body may be white or transparent \vithwith pigmented nephridia and buccal groove. Eyespots are either red or black and usually two in number, butbllt in some spedesspecies there are accessory spots lateral or anterior to the larger ones. The retractable terminal organ may be a thick elongated cylinder or thin narrow rod, or in some instances no more than a rounded knob. The terminal organ in some species may be extended frequently, rarely in others, and in a few no terminal organ has hcenbeen obscrVL'dobserved (Hall and Scheltema,Scheltcma, 197.5).1975). SpedficSpecific affinities have been tentativdytentatively assigned to three of the oceanic planktotrophic larvae. Fisher (1947), basing his identification on the number of muscle bands, designated a larva collt'ctedcollected off Cape Hatteras, North Carolina as SiPl.l11CllltlsSipunculus polijmyatus.polyl7lyoltlS. Murina (1965) identified larvae from the Gulf of Aden as S. aClfllahilisaeqiiabilis and from the Northwest Pacific as S. 1l0rvCgiCIIS.norvegicus. Hall and Sdw1temaScheltcma (1975) described external and internal morphology of 10 larval types from the North Atlantic. Of these, 7 were previously undescribed and the following 3 were rede­rede- scribed: Sipl.lIlclIlllsSipunculus pohjmijoltis,P01Y11lYOll/s, "Baccaria oliva," redcscribedredescribed as Type C, and "13accaria"Baccaria citrinclla,"citrinella," redcseribedredescribed as Type A. The latter was as­as- signed to the genus Asjyidosiphon,ASJlidosiphon, but no other adult affinitiesaflBnities were detenndetermined.ined. The pclagospherapdagosphera larva of SiPllllCull/sSiptmculus polymyotus]Jolymyotlls is large and trans­trans- parent (Fig. 25e,d),25c,d), often 4-5 mm in contracted length, commonly found 12012ü MARY E. RICE

in thetlie Gulf Stream (Fisher, 1947; Hall and ScheltelJla,Scheltema, un')).1975). The head is simple with a single pair of blackblaek eyespots and accessory pigment spots. The lip is hifurcatctlbifurcated hvb~' a groove marked hyby a yellowish-greenvellowish-green pigment. Similarly pigmentée!pigmented arcare the duct of the lip gland, the metatro­inetatro- eh<1Jchal band at the base of the cilia, the entire recurvcdrecurved gut, and the small rounded nephridia. The euticlccuticle is smooth and iridescent.iridescent IntcrnallyInternally the body wall mlISculntmemusculature is divided into approximately 50 longitudinal muscle bundles. There arcare three pairs of retractor muselesmuscles of th(~the introvert, two dorsal and one ventral. The most commonl~;connnonlv occurring larva in the North Atlantic is "Baccari"Raccariaa citrinella" or Type A (lIall(Hall

4.4.3.4 LARVAL ORGANOGENESISORGANOGéNESIS

Derivation of tissues and organs has been studied in 7 species of sipunculans, representing all developmental patterns: GolfingiaColfingia mimita,mint/ta, PhascolionPhaseDlion stromhi (Akesson, 1958), PhascolopsisPhascDlopsis goiildi,gnuldi, Golfingia vulgarismilgaris (Gerould, 19(7),1907), G. elongataelollgata (Akesson, 1961a), PhascolosomaPhascolosonw a{!,assi::..iagdssizi (Hice,(Rice, 1973), and SiJllIllClIllIsSijuinciihis 1lIIllIISnuchi.s (llatschek,(ITatschck, 188.3).1SS3). As in all protostomous coelomates, the endomesoderm is derived from the 4d cell and the coelom originates bybv a splitting of the mesodermal bands of the troehophOl'etrochophore into splanchnic and somatic layers.lavers. The coelom forms before troehophornltrochophoral metamorphosis in species with planktotrophic development and relatively later in species with more yolk),yolky eggs and lecithotrophic development (Section 4.4.3.2). In all species yolkvolk granules are released from the regressing prototroch cells into the dev~lopingdevçloping coelom where thevthe:-' furnish an important sourccsource of nutrition for the larva. The stomodcumstomodeum develops as an ectodermal invagination at the ante­ante- rior site of the closed blastopore and pushes farther anteriorlyanteriorlv into the region of the prototroeh.prototroch. At the time of metamorphosis of the trocho­trocho- phore,phon" the stomodeum ruptures through the egg envelope to form the mouth and the ventral ciliated surface of the head. The esophagus is derived from the stomodeum and, in planktotrophic species, the huccalbuccal organ (in part) and the lip glnnd,gland, both appendages of the mouth, are also derived from the stomodeum. Entomeres,ICntomeres, frequentlyfrequenth- marked hybv distinctive pigmentation, give rise to the stomach and intestine. A stomach is present only in planktotrophic larvae and disappears at meta­meta- morphosis of the pelagosphera. The anus and rectum are derived from 122 Y1ARYMARY E. RICE a proctodeal invagination which is located 111in a dorsal position in the middle or posterior trunk. Larval retractor muscles.muscles, retained as the muscles of the adult, are reported to arise from ectomesoderm in all species except SipllllculusSiptinculus ntulus.niidtis. IIatschekHatschek (1883) proposed that in the latter species the retractors originate from somatic mesoderm, although he was unahleunable to observe their early developmcnt.development. Reports of the origin of bodvbody wall musculature are conflicting. In S. nudusIll/dus both circular and longitudinal muscles are supposed to form from endomcsodermendomesoderm (llatschek,(Hatschek, 1883). Circular mus­mus- cles of Golfingia vulgaris and Phascolopsis gouldi are presumed to be derived from ectomesodenn,ectomesoderm, hutbut no information is available on the source of the longitudinal muscles in these species (Gerould,(Cerould, 1907). Circular musculature of G. minminutauta and PlwscoliollPhascolion strom])istrombi is believed to originate from ectomcsodermectomesoderm and longitudinal musculature from endomesoderm (Akesson, 1958). Descriptions of the derivation of the nephridia vary for different spe­spe- cies. Observations on development in PhascolionPhascoliol1 strondrí,strombi, Golfingia min­min- uta, G. vulgaris, and Phascolopsis gouldi (Gerould,(Cerould, 1907; Akesson, 1958) indicate that the nephridia arcare composite stl'llctures,structures, conforming to the concept of mixoncphridiamixonephridia by Goodrich (1945). In these species the nephridium proper has been reported to arise from an ectodermal in­in- vagination and the internal funnel from peritoneum. In Si/JlmctdllsSipuncuhis nl/dusnudus and Phascolosoma agassizi, on the other hand, a purely mesodermal origin of the nephridia has been suggested (Hatschek, 1883; Rice, 1973). The larval cuticle in most species is formed bvby a transformation of the egg envelope at metamorphosis of thcthe trochophorc.trochophore. Howevt'r,However, in threcthree species, PlwscolopsisPhascolopsis gouldi, Golfingia vulgaris, and Sipunculus mulus,nudus, the egg envelope is shed at thcthe time of metamorphosis and the cuticle of thcthe larva is a newlynewlv formcdformed secretion of the epidermis (Section( Section 4.4.3.2).

4.44.4.3.5..3.5 METAl\.fORPHOSISVIETAMORPHOSIS OF TIlETHE Pl,LAGOSI'HEnAPiîLACosi'HEnA

The end of the pelagospherapclagosphera stage is marked by the loss of metatrochalmetatroehal cilia in both lecithotrophic and planktotrophic pelagospheras. Lecitho­Lecitho- trophic larvae, which live in the plankton for only a short time, change to the adult form gradually, passing through a vermiform stage and sometimes requiring scveralseveral weeks to attain thcthe form of the juvenile. Dl1l'ingDuring this timctime the coelomiccoclomic yolkvolk is entirely absorbed, the gut is com­com- pleted, and that portion of the body anterior to the postmetatrochal sphincter is elongated to form the introvert with terminal ciliated tentacll-tentaeu-

_I 4. SIPUNCULA 123

lar lobes. In GolfingiaGolßngia pugettensispllgettensis 7 weeks may elapse before these changes are completed; in TlJemisteTJiemiste alutacea the change may take 3-4 weeks. Planktotrophic pelagospheras, after a prolonged period in the plankton, undergo a relatively rapid metamorphosis into the juvenile form. After a month in the plankton, the larva of SipuncullisSipunculus nlldusnudus elongates and sinks to the bottom where metamorphosis takes place over a period of 1 or 2 days (Fig. 22c).22e). The metatroch and associated organs of the mouth are lost, the mouth moves anteriorly, a tentacular lobe forms on either side of the mouth, and the head becomes proportionately smaller (Hatschek, 1883). Unidentified planktotrophicplanktotrophie pelagospheras from the oceanic plankton have been reported to undergo a similar meta­meta- morphosis with regression of the lip and formation of tentacles on the rim of the mouth (Jiigersten,(Jägersten, 1963). Metamorphosis of "Baccari"Baccariaa citri­citri- nella" or Type A larvae is marked by an elongation of the introvert to a length three times that of the trunk with 35 rings of hooks at its anterior end. Over a period of several weeks, the juvenile develops anterior and posterior shields which characterize it as belonging to the genus Aspidosiphon (Hall and Scheltema,Scheltenia, 1975). Metamorphosis of oceanic pelagospheras, of unknown adult affinities, has been observed in the laboratorylaboratorv from 11 to 129 days after collection. Maximum duration of larval life has been calculated from approximations of age at the time of collection to be 254 days (Scheltema( Scheltema and Hall, 1975). Metamporphosis of planktotrophic pelagospheraspclagospheras reared from spawnings of known adults in the laboratorylal:)oratorv has not heenbeen observed, although larvae of PliascolosomaPlwscolosoma agassizi have been maintained in culturecultiu^e to an age of 7 months.

4.4.4 Summary of Developmental Patterns

The most common developmental patterns in the phylum Sipuncula are summarized in the schematic diagram (Fig. 26). As .shownshown for Cate­Cate- gory I, the embryo may develop directly within the egg coats with no ciliated stage, hatching out as a small, crawling worl11worm which graduallygraduallv transforms into a juvenile, or it ma~'mav develop into a lecithotrophic trocho­trocho- phore with ciliated prototroch and apical tuft which then develops further along one of three pathways, represented bybv Categories II, III, and IV. In Category II the troehophoretrochophore transforms directly into the vermiform stage, ending the pelagic phase, or the trochophore maymav meta- • Inorphose,morphose, as in Categories III and 1\',IV, into a second larval stage, the pelagospherapclagosphera with metatrochal band of cilia. The pelagosphera of Cate-Cate­ gory III is lecithotrophic, remaining in the plankton only a short time

I j 124 MARY E. RICE

•

J n III 11II111II111I11I1II1

FIG. 26. Diagrammatic .....sumniaiyummary otof cliicfchid de\'l']opnll'nla)dcvi-lopmcntal patti'in.spaUl'rns inill tlH'the .Sipiinciila.Sipuncula. CategoryCatf'gory I. DirectDircct dcxclopnicnlde\"(~Jopmcnl with no pelagicl'elagie stages.stages, CategoryCatl'l.:ory II.)I. Pelagic,l'(']agie. leeitho-Ieeitho­ trophic trochophoretrnchophofl' which transformstrall~fonns into vermifonnwrmiform stage.stagl·. Catt'gOlYCategor\' III.Ill. Pelagic,l'ela.~ie. !ecithotmphiclecithotrophic trochopllOretrochophore Itidalliorph"Mielaiiiorpho.ses....es intllinto se(~lIldsecond larval stag",stage, ]ecithotrophiclecithotrophic pdagnsJlhera.pelagosphera, which thcnthen trallsformstransforms into \'('rmifonn\erniiforni stage. CategoryCall'l~ory IV. P"lagic,Pelagic, It't"ithotrophiclecithotrophic troehophoretrochophore nwtanwrphosesmetaniorpho.ses iotoinfo second lal'\'allarval stage.stage, ¡ilanklotrophicphlllk(otrophic pelagospliera.peJagosphera. Aftcr.-Vfter a prolonged periodperiocJ in th"the plankton the larva.larva, haviltghaving incfl'asN]increased in sizc,size, undergocsundergoes a second metamorphosismetamorpho.sis into the juvenilejllvenile form. (From( From Rice.Rice, 1975. ) before transformingtian.sfonning into the vermiform stage, then grac1l1all:'gradiiallv as~mmingassuming the features of the juvenile. TheTlic planktotrophic pt'!ngosp]H'rapelagosphera of CategoryCategorv IV lives for a prolonged period in the plankton, increasing considerablyconsiderablv in size, before undergoing a st'condsecond metamorphosis directly into the juvenile form.

Acknowledgments I am gmtefulgrateful to Harhor Branch Foundation Laboratory, Fort PiercC'.Pierce, Florida, where much of thisthi.s chapter was written.written, farfor lise\ise of their facilitics,facilities, and the cooperation of tlwirtheir stalfstaff in preparation of the figures. ~ral'YMary Galdizen.Coldizen, Harbol'Harbor Branch Foundation Lahoratory,Laborator)', drew Figs. 17, 19.19, 20.20, 21, antIand 22. Douglas PutnalllPutnam ancland Williall)William Daven- port assisted in the photo~raphicphotographic dTorts.efforts. • Carolyn H.B. Cast,Ga.st, scientific illustmtorillustrator illin the Department of Inwrll'hratl'Invertebrate Zoology, National MuseumMu.seum of I'\aturalNatural IIi.~tory,History, SlIlithsouian.Smithsonian Institution.Institution, is gratefullygratcfulK acknowl- edgcdedged forfor her iIIustratioll'illustrations which appear as Figs. 1,3,1, .3, 16,2.3.16, 23, and 26. 4. SIPUNCULA 12.5125

4.5 References

'. Akesson, B. (]\')58).(1938). A stndystudy of thcthe nervous system of the SipnnculoideaeSipunculoideae with some !'t'lllarksremarks on the development of two species PimscolionPhaseoliOIl stromlJislrom}>i i\lontagllMontagu and GolfillgiaColfmi^ia millHtaminuta Kefcrstein.Keferstein. UndcrsiikllingarVndcrsokningar iieerover OresllIlllOresund 38,3S, 1-249. Akesson,Âkes,s<)n, B. (1961a). The development of Golfingia elungatadongaia Keferstein (Sipuncll­(Sipuncu- lidca)lidea) withwitli some remarks on the developmentdevekipment of neurosecretorynenroseeretory cells in sipllnclI­sipuncu- lids.Ms. Ark. Zool.ZooT. (Ser. 2),13(2),511-531.2), 13(2), 511-53!. Alesson.Akesson, B. (1961b). Some observationsoljservations on pelagosphaera larvae. Galathea Rept. 5, 7-17. Andrews, E. A. (1889). ThcThe reproductive organs of PhascoloxomaPhaseoTosoma goiddii.gouldii. ZuoT.Zool. Allz.Anz. 12, 140-142. Andrews,Andrew.s, E. A. (1890). Notes on the anatomy of SipUlIeldlisSipunciihis goiddiigouTdii PO\ll'tales.Pourtalès. Johns HopkillsHopkins Ulde.Univ. Biol.RioT. Lab. StluT.Slud. 4, 390-4:30.390-430. Awati.Awati, P. R. and Pradhan, L. 13.B. (HJ36).(1936). The anatomy of Dendro.'itomaDellc!rustoma signifcrsigllifer Selenka and de Man. J./. Univ. BomhaijBumbay 4(5),114-1:31'.4(5), 114-131. Carlis!l',Carlisle, D. l:l.B. (HJ,s9).(1959). On the ncurosecretorynenroseeretory system of the l)rainbrain and associateda.ssociated strncturesstructures in SiplillclITusSipuncidus IlIUltl,l'nttdiis with a note on the cuticle. GIIIlIllGGunma ].}. Med. Sci.Sei. 8, 183-HJ.4.183-194. Clapan\le,Clapart^le, E.É. (1863). "Anatomie undmid Entwicklungsges(·hiclt·Entwicklungsgesehicte wirbelloser ThiereTliiere an derdel' KiisteKüste von Normandie angestcllt:'angestellt." vii + 120 pp. Leipzig. Clark, R. B. (HJ69).(1969). SystematicsSvstemalies and Phylogeny: Annelida, Echiura,Eehiura, SipnncuJa.Sipuncula. In "Chemical Zoology" (M.( M. Florkin and B. T. Scheer, eds.)eds. ) Vol. 4, Chapter I,1, pp. 1-68. AcademicAcadémie Press.Press, New York and Londoll.Lcmdon. Cole.Cole, J.[. B. (1952). The morphology of GoTfingiaGolfingia piigcttensi.'i:pugettcnsis: A sipunculid worm. f-,l.M. S. Thesis. 78 pp. l:niversityl'uiversitv of \Vashington,Washington, St'attle.Seattle. Damas, H. (1962). La collection de PelagosphaeraPclagosphaem du "Dana". Dana Repi.Rept. 1\'0.No. 5~J>59, 22 pp. Da\\'ydofr.Dawydufl, C.C'. l'\.N. (1930). Quelques ohservationsobservations sursm- Pehgospliacra,Pelagosp}lI/era, larvl'larve de sipllneulitlt'sipuneulide des C

Conse,Gonse, P. (1957a). L'ovogeneseL'ovogenèse chez Phascolosoma vulgare.vulgäre. III. Respiration cxogimeexogène et endogeneendogène de I'ovocyte.l'ovocyte. ElfetEffet de I'eaul'eau de mer. Siochim.Biochim. Biophys. Acta 24, 267-278. Gonse, P. (1957b).( 1957b ). L'ovogeneseL'ovogenèse chczchez Phascolosoma vulgare.vulgäre. IV. Etude chromatique des sucres du plasma, action de differentsdifférents substrats et du malonate sur la respira- • tion de I'ovocyte.l'ovocyte. Biochim. Biophys. Acta 24, 520-.531.520-5.31. Goodrich, E. S. (1945). The study of the nephridia and genital ducts since 1895. Quart.Qiuirt. J. Microsc. Sci.Set. 86 (342-344), 113-392. Green, W. (1975). The annual reproductive cyclecyc]e ofoí Phascolosoma lurco (Sipuncllla).(Sipuncula). Proc. Intern. Symp. Biol. Sipuncula and Echiura, in press. Naucno Delo Press, Belgrade. Hacker, V. (1898). Die pelagischen Polychaeten- und Achaetenlarven der Plankton­Plankton- Expedition. Ergeh. Plankton-Exped. HlImholdt-Sti/tungHumholdt-Stiftung 2, 1-50. Hall, J. R. and Schcltema, R. S. (1975). Comparative morphology of open-ocean pclagosphaera.pelagosphaera, Proc. Intem.Intern. Symp. Biol. Sipu1lculaSipuncula (I1ldand Echiura,Echiul"O. in près.?.pres.... Naucno Delo Press, Belgrade. '" Halschek,Hatschek, B. (1883). Ueber Entwicklung von SiptmculusSipunculiis nudus. Arh.Arb. Zool. lnst.In.H. Univ. WicnM^ien Zool. Stat. Triest 5, 61-140. Heath, II.H. (1910). Pelagosphaera,Pelagosplwcm, a larval Gephyrean. Biol. Bull. 18, 61-140. Hernbel,Hérubel, M. A. (1908). RecherchesRécherches sur les sipunculides. Mêm.Mem. Soc. Zool. Fr. 20, 107-419. Hyman, L. H. (1959). "The Invertebrates: Smaller Coelomate Groups," Vol. 5. !v[cGraw-HiII,McGraw-Hill, NewNcw York. Jagersten, G. (1963). On the morphology and behaviour of Pelagosphaera larvae (Sipuncllloidea).(Sipunculoidea). Zool. Bidr.Biclr. Uppsala 36, 27-35. Keferstein, W. (1863). BeitrageBeiträge zur Kenntnis der Gattung Pha.tcolosoma.Plwscolosoma. Z. Wiss.VViis. Zool. 12, 35-51.3.5-51. Mingazinni, P. (1905). Un GeRreoGefireo pelagico:pelágico: Pelagosphaera AloysiiAloy.sü n. gen., n. sp. Rend. Acad. Naz. LincdLincei 14,713-720.14, 713-720, Murina, V. V. (1965). Some data on the structure of pelagospherexpelagosphere.~-sipunculid - sipuncuJid larvac.larvae. Zool. Zh. 44(11),1610-161944( U ), 1610-1619 (in Russian).Ru.ssian). Pasteels,Pasteeis, J. J.]. (1935). Recherches sur.sur Iele determinismedéterminisme de I'entreel'entrée en matmationmaturation de I'oeufl'oeuf chez divers Invert/\bn\sInvertébrés marins. Arch. Biol. 46, 229--262.229-262. Rajulu,Rajulll, G. S. (1975). AsexualAsexnal reproduction by Imddinghudding in the Sipllncula.Sipuncula. Proc. Intern. Sy1llp.Symp. Biol. SipunclIlaSipuncula and Echiura,Echillra, in press. NaucnoNallcno Delo Press, Belgrade. Rajulu,Rajuhi, G. S. and Krishnan, N. (1969). Occurrence of asexual reprodnctionreproduction by budding in Sipunclllida.Sipunculida. Nature (London) 223, 186-187. Ricl',Rice, M. E. (1966). ReproductiveRci)roductive biology and development in Sipnncllla.Sipuncula. Doctoral Dissertation, University of Washington. Seattle. 322 pp. Rice, M. E. (1967). A comparative study.study of thcthe development of PhascolosomaPlwscolosollw agassizii,ogassizii, GolfingiaGolfijigia pugettensis,pllgettcllsis, and TlremisteThemiste pyroidesPYl'oides with a discussion of dcvelopmentaldevelopmental patterns in the Sipuncula. Ophelia 4, 143-171.14.3-171. Rice, !vf.M. E. (1970). AscxualAsexual reproduction in a sipunculan worm. Science 167, 1618--1620.1618-1620. Rice, M. E. (1973). Morphology, hehavior,behavior, and histogenesis of the pelagospherapelagcsphera larva of PhascolosomaPllascolosoma agassizii (Sipnncllla).(Sipuncula). Smithson. Contrib. Zool. No. 132, 1-51. Rice, !vf.M. E. (1974). GametogenesiSGanietogene.sis in three species of Sipuncula: Phascolosoma agassizii,agas.tizii, Golfingia pugetten.sis,pugettensis, and ThelllisteThemi.ite pyroidcs.pyroides. La Cellule, 70(2), 295-313. r

4.4. SIPU~CULASIPUNCULA 127127

Rice,Rice, M.M. E.E. (1975). (1975). ObservationsObservations onon thethe developmentdevelopment ofof sixsix speciesspecies ofof CaribbeanCaribbean SipunculaSipuncula withwith aa reviewreview ofof dcvelopmcntdc\'elopmcnt inin thethe phylum.phylum. PTOC.Proc. Intern.Intern. SYlllp.Sijinp. Bioi.Biol. SipunculaSipuncula andand Eclliura,Echitira, inin press.press. NaucnoNaucno DeloDclo Press,Press, Belgrade.Belgrade. Sawada,Sawada, N.,N., Noda,Noda, Y.,Y., andand Oehi,Ochi, O.O. (1968).(1968). AnAn electronelectron microscopemicroscope stndystudy onon thethe oogenesisoogénesis ofof GolfingiaGolfinnia ikedai.ikedai. Mem.Mem. EhimeEhimc Univ.Univ. Sci.Sei. SeT.Ser. B6(B6(l),1), 25-39.25-39. Seheltema,Scheltema, R.R. S.S. (HJ75).(1975). TheThe frequencyfreijucncy ofof long-distancelong-distance larvallan'al dispersaldispersal andand thethe raterate ofof gene-flowgene-flow betwcenbetween widelywidely separatedseparated populationspopulations ofof sipullculans.sipunciilans. Proc.Proc. Intern.Intern. Symp.Stjmp. Bioi.Biol. SipttnC1llaSiptinctila andand F.cZ,iura,Rchiura, inin press.press. NaucnoNaucno DeloDelo Press,Press, Belgrade.Belgrade. Scheltcma,Scheltema, R.R. S.,S., andand Hall,Hall, J.j. R.R. (1965).(1965). Trans-oceanicTrans-oceanic transporttransport ofof sipnneulidsipnnculid larvaelarvae belongingbelonging toto thethe genusgenus P!la.vcolosoTT1ll.Phascolosoma. Amer.Amer. Zool.Zool. 5(5(2),2), 216216 (Abstr.).(Abstr.). Scheltema,Scheltema, R. S.S. andand Hall,Hall, J.|. R. (1975).(1975). TheThe dispcrsaldispersal ofof pelagospherapelagosphcra larvaelarvae hyl)V oceanocean currentscurrents andand itsits relationshiprelationship toto geographicalgeographical distrihutiondistribution ofof sipuneulans.sipunculans. Proc. Intern.Intern. Sym1J.Stjtiip. Bioi.Biol. SipunculaSipuncula andand Echiura, inin press.press. NaucnoNaucno DelaDelo Press, Belgrade.Belgrade. Sehleip.Schleip, W. (1934).(19.34). Die Regeneration desdes RiisselsRüssels Plwscolio/lPhascoUon strombi Mont. (Sipuneulidae).(Sipunculidac). Z. Wiss.Wi.çs. Zool. 145,145, 462-496. Senna, A. (lOOG).(1906). Sulla struttura di alcunealeune larvelarve (Pclagosphaera)( Pelagosphaera) di Sipuneulidi.Sipiuiculidi. Puhbl.Puhhl. Inst.Inst. Stuc/.Stud. SuperioriSuperior! Pract. PerfezionameTltoPerfczionamento Firellae;Firenae; Sez.Sei. Sci.Sei. Fis. Nat., .. PI"pp. 50-7850-78.. Shipley, A. E. (1890). On PhYlllosomaPhymosoma varians.varians. Qllart.Quart. J. Microsc.Microxc. Sci.Sei. 31, (1), ]-27.1-27. Stephen, A. C. (1941). The Echiuridae.Echiuridae, SipnnculidaeSipunculidae and PriapnlidaePriapulidae collectcollectedcd by the ships of the "Discovery" Committee during the ycarsyears 1926 to 1937. Discovert!DiscovMY Rep. 21, 237-260. Stephen, A. C. and Edmonds, S. J. (1972). "The Phyla SipuneulaSipuncula and Echiura." The British Museum (Natural History), London, 527.527 pp. Thorson, G. (1046).(1946). Reproduction and larvallar\al development of Danish marine hottolllbottom invertebrates, with special reference to the planktonic larvae in the Sound (0re.sund).(0resund). Medd.Meeld. KommnKOl1l71l11 Danm.Dall1/l. Fi.sk.Fisk. Ilavundcr.s.I-Iavu/l(ler.v. Plankton 4, 1-523. Towie,Towle, A. and Ciese,Giese, A. C. (1967). TheTI1l' annual reproductive cyclecyele of the sipnnculidsipllnculid PhascolosomaP/laSco!oso71la agassizii. Physiol.l'hysiol. Zool. 40, 229-237.220-237. Wegener, F. (1938). Beitrag zur7.ur KenntnisKcnntnis derdel' RusselregenerationRusseIre~encration derdel' Sipunculiden. ' Z. WÍ.SS.Wiss. ZoolZool. 150,ISO, 527-565. Williams, J.J. (1972). Development of a rock'rock burrowinghurrowin~ sipunculid inhabitingillhahiting stony coral.coral. Amer.Amcr. Zool.Zoul. 12(4),12(4),723 723 (Ab.str.(Ahstr.) )..

•