ia oe Oh ees) Pe at Yn Tors tp ve —s Eat 1 eS ee OF iene et Spee fre oe les ah eee | Po eee rt wide gla Peet are was py at ura, alert * Oem ete wt: ; ee sa cake Conese bane an Xe Fa | Sarre res + Sane a et wees) ee ee ee ee eee a oro THE AUSTRALIAN MUSEUM, SYDNEY MEMOIR 16 Papers from the Echinoderm Conference THE AUSTRALIAN MUSEUM SYDNEY, 1978 Edited by FRANCIS W. E. ROWE The Australian Museum, Sydney Published by order of the Trustees of The Australian Museum Sydney, New South Wales, Australia 1982 Manuscripts accepted for publication 27 March 1980 ORGANISER FRANCIS W. E. ROWE The Australian Museum, Sydney, New South Wales, Australia CHAIRMEN OF SESSIONS AILSA M. CLARK British Museum (Natural History), London, England. MICHEL JANGOUX Université Libre de Bruxelles, Bruxelles, Belgium. PORTER KIER Smithsonian Institution, Washington, D.C., 20560, U.S.A. JOHN LUCAS James Cook University, Townsville, Queensland, Australia. LOISETTE M. MARSH Western Australian Museum, Perth, Western Australia. DAVID NICHOLS Exeter University, Exeter, Devon, England. DAVID L. PAWSON Smithsonian Institution, Washington, D.C. 20560, U.S.A. FRANCIS W. E. ROWE The Australian Museum, Sydney, New South Wales, Australia. CONTRIBUTIONS BIRKELAND, Charles, University of Guam, U.S.A. 96910. (p. 175). BRUCE, A. J., Heron Island Research Station, Queensland, Australia. (p. 191). CAMARGO, Tania Maria de, Institute of Oceanography, University of Sao Paulo, Brazil, (p. 165). CLARK, Ailsa, M., British Museum (Natural History), London, England, (p. 121). DAYTON, Paul, K., Scripps Institute of Oceanography, La Jolla, California, U.S.A. 93093. (p. 175). ENGSTROM, Norman, A., Department of Biological Sciences, Northern Illinois University, DeKalb, Illinois, U.S.A. 60115. (p. 175). GUILLE, Alain, Muséum National d’Histoire Naturelle, Paris, France. (p. 67). HARRIOTT, Vicki, Zoology Department, University of Queensland, St. Lucia 4067, Queensland, Australia (p. 53). JANGOUX, Michel, Zoology Department, Université Libre de Bruxelles, Bruxelles, Belgium. Gay Oe MARSH, Loisette, M., Western Australian Museum, Perth, Western Australia. (p. 89). MITROVIC-PETROVIC, Jovanka, Faculty of Mining and Geology, University of Beograd, Beograd, Kamenitka 6, Yugoslavia. (p. 9). NICHOLS, David, Department of Biological Sciences, Exeter University, Devon, England. (p. 147), PAWSON, David L., Smithsonian Institution, Washington, D.C., U.S.A. 20560. (p. 129). ROWE, Francis W. E., The Australian Museum, Sydney, New South Wales, Australia. (p. 89). SIMPSON, R. D., The University of New England, Armidale, New South Wales, Australia. (p. 39). Pais ie ae tal i ie meee FOREWORD Since the first major Symposium on Echinoderm Biology was held in London in 1966, sponsored by the Royal Zoological Society, at least six subsequent meetings have been organised by echinodermologists. These have been held in Washington D.C., U.S.A. (2), Rovinj, Yugoslavia (1), Sydney, Australia (1), London (1); the last two meetings (Sydney and London), within the same year (1978), and Brussels, Belgium. Also, at least four meetings are known to have been held in U.S.S.R. Such has been the surge of interest in the study of echinoderms over the past decade, that there is now a demand for the organisation of regular, and more frequent, meetings. The international representation at these meetings indicates the enormous involvement and co-operation which now exists between colleagues working in this exciting field, the world over. It is more than evident that the satisfaction and pleasure expressed by Professor Norman Millott, in his foreword to the first Symposium volume (1967), at the resurgence of interest in Echinoderm Biology has been clearly justified and can continue so to be. This volume presents twelve of the forty-one contributions offered at the Echinoderm Conference, Sydney, 1978. The papers are representative of the wide coverage of topics dealt with during the Conference, including echinoderm palaeontology, physiology, reproduction, ecology, behaviour and taxonomy. To the speakers and chairmen, and to all those who attended the Sydney Conference, I convey my thanks. I must also thank my Technical Officer, Ms Jan Marshall, and Dr Susan Oldfield (Queen’s Fellow at The Australian Museum, February, 1977-1979) for their unstinting assistance in the organisation of the Conference. Thanks are also due to the Department of State Fisheries (N.S.W.), Taronga Park Zoo, McWilliams Wines Pty, Leo Buring Wines Pty, Qantas Airways Ltd, and Trans-Australia Airlines (T.A.A.). To The Australian Museum Society -(TAMS) I extend a special thanks for assistance. This Conference could not have been held without the tremendous support and encouragement afforded to the organiser by Dr D. J. G. Griffin, Director, The Australian Museum, and the very generous financial support of the Trustees of the Museum, to both of whom I offer my very sincere thanks. DECEMBER 1979 FRANCIS W. E. ROWE 10. 1]. V2. CONTENTS . Etudes taphonomiques du gisement contenant la faune des échinides (l’éocéne d'Istrie). Oval VTP OVAGS Ne TROV Gi saa tke e ate ae er ticion aed ected wourdanteicraw atric aa ee Peete eet Ome 9 . Etude structurelle et fonctionnelle du tube digestif d’Asterias rubens Linnaeus (Echinodermata: Asteroidea). Teale cS UA BM ONEQOTVOe “Woon nm Ree fw Ae gare Riel fee Deen Unt ie e teet meee y Wre eer ash Ree 17 . The reproduction of some echinoderms from Macquarie Island. Aer OARS VETS OTT ee cate cee eee erasers ek Hack SIRE Ce rate ne erent clea och a's ANG aR F AS MBEE TAROT aT Ua 39 . Sexual and asexual reproduction of Holothuria atra Jaeger at Heron Island Reef, Great Barrier Reef. NST het ls LACE IO Le eden Pecdtcte trek Pitend bree Aer seat slic mies ascents pig saten nA eeisasee oth Sedna Ma 53 . A new genus and species of ophiacanthid brittlestar (Echinodermata: Ophiuroidea) from the Kerguelen Islands, with new taxonomic, biogeographic and quantitative data on the echinoderm fauna. PALA Dtnes (EMUMLL cae oe ez-a thats HE Rey AEA ryt NT ufersserris ale latactslsreW, eglelviata ete nalabu Reinet egelye dA oe pat race ose 67 . A revision of the asterinid genus Nepanthia Gray, 1840 (Echinodermata: Asteroidea), with the description of three new species. BRANCISHWW BE. NOM esAl mle OISehle Vie VALS cents s.e1maeesey cher eemdaaatacwecee sponser esis 89 . Inter-relationships of recent stalked, non-isocrinid Crinoidea. PAST Sr1g4\ eae Cl] tila ae CMRI seo ie co 8 ete Abarth Lonehayoel EN AES, Mies eile AS Sema ear stele ntieealelotee 121 . Deep-sea echinoderms in the Tongue of the Ocean, Bahama Islands: a survey, using the research submersible Alvin. DERG la) Lgl MERA SY Gl 3 cle wtedeioee ocean piste Ao ee aber tee ROSA 70g SATE PRR RGOnaT ay Hoa cen cre amteee aHaaod 129 . A biometrical study of populations of the European sea-urchin Echinus esculentus (Echinodermata: Echinoidea) from four areas of the British Isles. IDA TICMINT CHG Stews Bisbee ns Lie hore eds ect ch emneie eran lanl de satan mon Peenen cugatd glen oid 147 Changes in the echinoderm fauna in a polluted area on the coast of Brazil. “ata Nia hate OS CHt oy exo ea lease eksooeneerce otiry. omen de aaciealnns sete tac osteo eee onne orem ee abe 165 A stable system of predation ona holothurian by four asteroids and their top predator. Charles Birkeland, Paul K. Dayton and Norman A. Engstrom ................:esseeees 175 The shrimps associated with Indo-west Pacific echinoderms, with the description of a new species in the genus Periclimenes Costa, 1844 (Crustacea: Pontoniinae). sNoa dye | EVEN ELEY, 0 3 3 AROS ARN TIAR ARES Sez pine (AGM C OY AA aCe Rea Sere ASC HAmE RARER toon rath Kiln anne Aas 191 1. ETUDES TAPHONOMIQUES DU GISEMENT CONTENANT LA FAUNE DES ECHINIDES (L’EOCENE D’ISTRIE). JOVANKA MITROVIC-PETROVIC University of Beograd, Beograd, Yugoslavia SUMMARY Bacva Spring, near Pican Village in Istria (in the extreme north-west of Yugoslavia) is noted both for the large number of fossil echinoids and number of species to be found there. They can be collected directly from the stratum in which they occur. Many loose specimens may also be found mixed with scree material. In this paper an account is given of the lithological composition of the fossil-bearing stratum; position in which the echinoid fauna occurs in the rock, i.e. the position in relation to the bedding plane, interrelation of fossils and their random orientation; range of size of fossils found; degree and kind of damage observed and palaeoecological characteristics of the echinoids and accompanying fauna. ____ Judging by their ecological characters the genera Conoclypeus and Echinolampas were found im situ. The massive test, the enlarged solid base and the reduced jaw system in Conoclypeus suggest a habitat of turbulent water and gravelly substrate. _The representatives of other genera e.g. Cylaster, Linthia and Macropneustes among others, having thinner-walled tests, well-developed labrum and ambulacra occurring in funnel-shaped depressions suggest a habitat of deeper water with a silty-clay substrate. That these two ecologically different groups of echinoids occur together is accounted for by a secondary concentration of the fauna due to redeposition by sediment flow. INTRODUCTION Lune des localités les plus riches en échinides fossiles en Yougoslavie c’est la source Bacva, non loin du village Pican en Istrie. Cette localité est caracterisée non seulement par un grand nombre d’espéces, mais aussi par un trés grand nombre d’individus. C’est aussi une des localités rares chez nous ou les échinides sont ramassés directement du gisement (dans les plupart des cas on les trouvait naturellement preparés dans les éboulis). La concentration riche de la faune des échinides dans le gisement a permis |’étude minutieuse a) de la position de la faune en relation a la couche; b) du nombre d’exemplaires sur Punité de la surface; c) de la relation réciproque de la faune es échinides et de la faune accompagnantes; d) des types d’endommagements des squelettes produits au cours de la fossilisation. Tout cela nous a permis de former certaines conclusions relatives a la position du gisement fosilifére par rapport aux gisements voisins, ainsi qu’a la maniére de sa formation. REPRESENTATION DU GISEMENT FOSSILIFERE Dans la region de Pican (Istrie centrale) les sediments de l‘Eocéne sont développés sur un grand espace dans le facies calcaire et facies du flysche. Le gisement fossilifére qui est le sujet de cet exposé, se trouve dans le cadre de la série de flysche tout prés de la source Bacva a 3 km NW du village de Pican. Examinant due céte de la source vers la route principale Pican-GraciSte, on peut remarquer le profil suivant: directement au-dessus de la source on a découvert la couche de Australian Museum Memoir No. 16, 1982, 9-16. 10 JOVANKA MITROVIC-PETROVIC conglomerat avec le ciment calcaire, comblée des échinides et des nummulites. Les nummulites sont lavés, naturellement préparés et on les trouve en grand nombre sur la surface, tandis que les échinides sont disposés chaotiquement a!’intérieur du gisement. Leur position est différente par rapport au gisement, mais dans la plupart des cas ils sont retournés par la face aborale en haut, ou ils sont, dans certaine mesure obliques. On les trouve trés rarement la face orale retournée en haut. Les relations réciproques et densité sur lunité de la surface varient sensiblement. Quelquefois ils sont tellement resserrés qu’ils se touchent par la moindre ou de la plus grande partie de squelette et forment la masse fondamentale du gisement. Les cailloux roulés sont ceux qui predominent ailleurs, tandis que les squelettes des échinides se trouvent dans une position subordonneée. Les exemplaires de dimensions différentes se trouvent c6te a céte non assortis d’apres leur grandeur (fig. 1 a, b). Quant a la faune accompagnante concentrée dans le gisement méme ce ne sont que des rares representants des lamellibranches et des gastropodes. Parmi les échinides ce sont les représentants du genre Conoclypeus qui prédominent. D’aprés le nombre d’individus ils sont représentés avec 50% approximativement. Les” representants des genres: Echinolampas, Linthia, Prenaster, Pericosmus, Cyclaster et Macropneustes sont présent avec plus ou moins grand nombre d’espéces, et chaque espéce avec un nombre considerable d’individus. La liste compléte des échinides déterminés appartenant a ce gisement se compose de 16 espéces: Conoclypeus conoideus (Leske), C. pyrenaicus Cotteau, Echinolampas eurysomus Agassiz, E. calvimontanus (Klein) Loriol, Linthia vilanovae Cotteau, L. ducrocqui Cotteau, L. ybergensis Loriol, L. subglobosa (Lamarck) Desor, L. orbignyi Cotteau, L. mflata (Desor) Cotteau, Prenaster alpinus Desor, Pericosmus nicasei Pomel, P. spatangoides (Desor) Loriol, P. hispanicus Cotteau, Cyclaster ovalis Cotteau, et Macropneustes brissoides (Leske) Desor. Neuf parmi elles sont caracteristiques pour |’Eocéne moyen et sept autres indiquent seulement qu'il s’agit de l’Eocéne, mais sans possibilité d’analyse précise. Comme la plupart des especes indique l’Eocéne moyen et comme on a trouvé un grand nombre de représentants de l’espéce C. conoideus — l’espéce trés importante au point de vue biostratigraphique pour I’Eocéne moyen on peut considérer l’age de gisement comme celui de Eocene moyen. Les conglomeérats forment la base de la serie de flysche et leur épaisseur s’éléve a 15 m_ environ, Au dessus d’eux on trouve alternativement les marnes argileux et les argiles, plus rarement les gisements calcaires, mais sans la faune (fig. Ic). . D’apres sa structure lithologique, ce flysche correspond absolument au flysche carbonates. / L’absence des grauwacke est caractéristique. CARACTERISTIQUES PALEOECOLOGIQUES DE LA FAUNE ETUDIEE La faune des échinides des conglomerats de la série du flysche du Pican, au point de vue écologique est nettement hétérogéne. Les représentants des genres Conoclypeus et Echinolampas avec des squelettes grands et massifs peuplaient, sans doute, le fond caillouteux. A en juger dapres le squellete tres grand et massif des Conocylpeus et d’aprés la présence des machoires (quoique tres réduites), ce genre était certainement située dans la zone littorale. La carapace grande et massive, resistait efficacement aux coups des vagues. La machoire quoique assez atrophicée permaittait une alimentation composée tout d’abord des plantes qu’on trouve surtout dans les régions littorales (algues de mer, herbes etc). L’Echinolampas presente encore un genre typique pour l'eau peu profonde, D’aprés J. Cottreau (1913) on rencontre les Echinolampas souvent dans la zone de Litothamnium qu’on peut comparer aux champs de Zostera dans la Méditerranée (a la profoundeur d’environ 30 m). T. Mortensen (1948) estime que les ECHINIDES L’EOCENE D’ISTRIE Fig. 1. a. Le gisement du conglomérat, b. Le gisement du conglomérat (detail), c. Le profil complet au-dessus de las source Bacva. al 12 JOVANKA MITROVIC-PETROVIC Echinolampas actuels sont trouyvés dans les régions tropiques-subtropiques d’ Atlantique et celles du Paciphique Indo-occidental, a la profoundeur de 10-500 m. Les autres échinides, representants des genres: Cylaster, Pericosmus, Linthta, Macropneustes, Prenaster, avec les carapaces relativement tendres et fines habitaient un milieu un peu plus profound au fond vaseux. Leurs carapaces relativement fines et tendres, l’absence totale de la machoire, le labrum bien développé chez Linthia et Pericosmus témoignant une alimentation contenant la vase, tout cela sont des caractéristiques des échinides qui habitent l'eau un peu plus profonde. D’aprés T. Mortensen (1951) les espéces actuelles du genre Pericosmus vivent au fond vaseux et gresseux a la profondeur de 18-486 m. Un tres grand nombre de nummulites-foraminiféres benthoniques de l’eau peu profonde temoigne de la petite profondeur de l’eau et de la proximité de la cote. Les gastropodes et les lamellibranches rares d’aprés leurs caractéristiques morphologiques — les spires bases et les squelettes tres gros des gastropodes, ainsi que les squelettes massifs avec une ornamentation trés marque chez les lamellibranches, témoignent aussi de l’eau peu profonde. TYPES DDPENDOMMAGEMENTS LES PLUS FREQUENTS DE LA FAUNE DES ECHINIDES ETUDIEE Les endommagements des squelettes résultent des facteurs chimiques, mécaniques et biologiques qui par leur action dans la mesure plus ou moins grande les produisent. En ce moment les endommagements mécaniques provoques le plus souvent sous l’influence des vagues et des courants nous intéressent tout spécialement. Les études de tels endommagements peuvent tres bien étre appliquées en taphonomie, si l'on prend ce mot au sens plus large — comme un ensemble de facteurs qui ont influé sur la formation du gisements fossiliféres. Chaque élément de l’orictocénose peut donc étre une source d’information sur les facteurs qui ont agi lors du passage de la biocénose a l’orictocénose. La faune des échinides examinée dans l’ensemble est bien conservée, c’est a dire les squelettes sont en general tout entiers; proportionnellement il y a peu de squelettes cassés ou conservés fragmentairement. Ona remarque pourtant les sortes différentes d’°endommagements qui peuvent étre classées en quelques types principaux: 1. Consomption-erosion du test. On y distingue plusieurs cas: (a) Le consomption approximativement uniforme de toute la surface due test avec les tubercules et les granules conserves encore, mais érodés considerablement. (b) Plus haut degré d’érosion du test dont le résultat est le manque presque complet des tubercules et des granules sur la plus grande partie du test. Ils ne sont conservés que sur les surfaces proportionnellement trés petites et surtout sur la face orale. (c) Le plus haut degre d’érosion d’ou résulte le manque total des tubercules et des granules. 2, Le manque, plus ou moins grand, du test d’ou proviennent les dépressions plus ou moins profondes sur la surface du squelette. Dans certains cas il ne manque qu’un morceau du test, dans les autres il s’agit des depressions assez profondes. 3. Les fissures. 4. Les fracteurs qui ont emporté une partie du squelette la fragmentation. ECHINIDES L’EOCENE D’ISTRIE Fig. 2. a-b. Macropneustes brissoides: a. face orale, b. face aborale; c-e. Conoclypeus conoideus: c. face orale, d. face aborale, e. le profile; f. Echinolampas calvimontanus face aborale, on voit le manque d’une seule partie de test; g. Macropneustes brissoides, face aborale, on voit le manque dune grande partie du test; h. Cyclaster ovalis face aborale, dépression remarquable sur la partie droite. 13 14 JOVANKA MITROVIC-PETROVIC d’€rosion plus haut ou les tubercules et les granules sont conservés sur les surfaces relativement petites et presque toujours uniquement sur la face orale, plus rarement au bord du test ou sur la face aborale dans la dépression d’ambulacre impair (M. brissotdes fig. 2a,b). A la deuxiéme place on peut citer le cas ot I’érosion de la surface tout entiere du test est plus ou moins uniforme, et les tubercules et les granules sont conservés encore, mais considérablement érodés. Autrement dit, l’ornamentation n’est pas en relief mais en méme niveau avec la surface du test. Les tubercules et les granules sont nivelées ou bien on ne voit que leurs coupures. On trouve cela trés souvent, surtout parmi les nombreux exemplaires d’espéce C. conoideus (fig. 2c,e). Le manque d’une partie du test est aussi un phénoméne fréquent. Jugeant d’aprés le nombre d’exemplaires ainsi endommagés, cette sorte d’endommagement occupe le troisiéme place. On peut distinguer deux cas fondamentaux: le manque d’une seule partie du test (plus rare) et les dépressions plus ou moins profondes (plus fréquent). Le premier cas est remarqué, par exemple chez les espéces: Echinolampas calvimontanus (fig. 2f), et Conoclypeus conoideus (fig. 2c,e). Le second cas est présent chez les nombreux exemplaires de beaucoup d’espéces: Macropneustes brissoides (fig. 2g), et Pericosmus hispanicus, P. nicasei, Cyclaster ovals, C. Parmi les types cités d’endommagement du squelette on rencontre le plus souvent le : conoideus et d'autres (fig. 2h). ) Le plus haut dégré d’érosion d’ou resulte le manque total des tubercules et des granules est aussi present chez un nombre considérable de représentants: Pericosmus nicaset, Linthia vilanovae (fig. 3a,b), L. inflata, Cyclaster declivis et d’autres (fig. 3c,d). 7 ; 1 1 Ona rarement remarqué les fissures sur le matériel examiné. Chez quelques exemplaires de lespece Conoclypus conoideus la fissure oblique coupe la face orale et aborale dans la partie anterieur du test et elle est remplie de calcite, ce qui indique qu’elle provienne immédiatement aprés la mort de animal et qu’elle est remplie au cours du procés de la fossilisation (fig. 3e,f). Enfin, le plus rarement on peut rencontre la fragmentation ou le manque total d’une partie du fossile. Ce phenoméne est certainement la conséquence de la fracture arrivée immédiatement aprés la mort de l’animal ou pendant le transport (Fig. 3g,h). DISCUSSION Sur la base de tout ce qu’on a dit; des caractéristiques lithologiques du gisement fossilifére; de la position ou la faune des échinides se trouve dans le rocher (relation vers le sédiment, relation réciproque, l’absence d’orientation absolue); de la grandeur des fossiles (les formes grandes et petites mélangées); du dégré et de la sorte d’endommagement du squelette (la faune proportionnellement bien conservée avec les endommagements uniquement sur les parties les plus exposées du squelette); les caractéristiques paléoécologiques de la faune des échinides et de la faune accompagnante; on peut conclure qu'il s’agit de l’orictocénose allochtone, mais que le transport n’a pas été long. Quoique dans l’ensemble examinée l'association est allochtone elle continent une composante autochtone aussi. D’apres leurs caractéristiques écologiques les genres Conoclypeus et Echinolampas seraient d'origine “in situ’. Leurs tests massifs, ensuite la base élargie et solide et la présence de la machoire réduite chez les Conoclypeus permettaient la vie dans l’eau agitée dans le fond caillouteux. Les représentants de tous les autres genres avec un squelette mince, un labrum bien développe, les ambulacres retirés dans les depressions indiquent la vie sur le fond argileux et dans eau un peu plus profonde. ECHINIDES L’EOCENE D’ISTRIE Fig. 3. a-b. Linthia vilanovae: a. face orale, b. face aborale; c-d. Cyclaster dechvis: ¢. face orale, d. face aborale; e-f. Conoclypeus conoideus: e. face aborale, f. le profil; g-h. Conoclypeus pyrenaicus: g. face orale, h. face aborale. 15 16 JOVANKA MITROVIC-PETROVIC La présence commune de ces deux groupes d’espéces des échinides liés aux fonds et aux) conditions biotiques différentes résultent de la concentration secondaire de la faune sous influence “sediment flow” qui a apporté et redéposé les sédiments et la faune. REFERENCES Cottreau, J., 1913. Les échinides néogeénes du bassin méditerranéen. Ann S Inst. oceanog. Monaco 6(3): 1-192, 4 figs, 15 pls. Dimitrijevicé, M. N., 1967. Paleogeni flisevi spoljagnjih Dinarida. Karp. Balk. geol. asoc. VIII Kongres. Vodit ekskurzije. Geolo’ki problemi Dinarida. Pp. 30-33. Mitrovic-Petrovié, J., 1969. Biostratigrafski i paleoekoliski zna€aj eocenskih ehinida u Dinaridima. Inst. za geol.istr.III. Simpozij Dinarske asocijacije. Zagreb. Pp. 117-134. 1970. Eocenski ehinidi Jugoslavije. Geolo’ki Anali balk. Poluost, 35: 151-190. Mortensen T., 1948. Monograph of the Echinoidea. IV (1 and 2). Holectypoida and Cassiduloida (1); Clypeastroida (2). Copenhagen. Pp. 371, 326 figs, 14 pls (1): Pp. 471, 256 figs, 72 pls (2). 1951. Monograph of the Echinoidea. V(2). Spatangoida II. Copenhagen: Pp. 593, 286 figs, 64 pls. Salopek, M., 1954. Prilozi poznavanju geoloske gradje Labinskog i Picanskog basene Istre. Prirodosl. Istraz. 26: Pp. © 1-58. 2. ETUDE STRUCTURELLE ET FONCTIONNELLE DU TUBE DIGESTIF D’ASTERIAS RUBENS L. (ECHINODERMATA: ASTEROIDEA). MICHEL JANGOUX Université Libre de Bruxelles, Bruxelles, Belgium SUMMARY The digestive tract of A. rubens is composed of three morphologically and physiologically main regions: the floor of the cardiac stomach, the diverticula of the pyloric caeca and the rectal caeca. These regions are linked by transit zones (ciliary channels of the pouches and the upper part of the cardiac stomach, of the pyloric ducts, of the pyloric stomach and of the intestine). During the meal the everted stomach — cardiac floor — is in intimate contact with the soft parts of the prey. The cardiac zymogen cells secrete their enzymes and extra-oral digestion occurs (extracellular digestion). Some particles of food are embedded in mucus and passed to the pyloric ducts by the ciliary channel of the cardiac stomach. At the same time the rectal current carries some small food particles directly into the rectal caeca where they are absorbed (intracellular digestion), The pyloric enzymes digest the food that has passed into the pyloric diverticula. The digestive products are then absorbed (extracullar and intracellular digestion). The digestion by Asterias is virtually complete, little faecal matter being passed through the anus. Defecation is the result of the contraction of the rectal caeca wall, associated with the relaxing of the anal sphincter. The average duration of a meal is between five and six hours. INTRODUCTION L’anatomie digestive des astérides, particuliérement celles des Asteriidae (Asterias, Marthasterias, Pisaster . . .), est bien connue (Hamann, 1885, Cuenot, 1887 et 1948, Chadwick, 1923, Hyman, 1955. . .). A l’opposé notre connaissance des structures microscopiques et/ou des fonctions de certains organes digestifs est encore assez fragmentaire. Des quatre principaux organes digestifs, ce sont les caecums pyloriques qui ont été le plus étudiés. On sait qu’ils sont un site d’élaboration et de sécrétion d’enzymes digestives (voir entre autres Sawano, 1936, Anderson, 1966, Peng et Williams, 1973). Leur structure histologique ainsi que leurs réles dans Vabsorption et la mise en réserve de nutriments ont été clairement établis par Anderson (1953). Les caecums rectaux sont également des organes absorbants (Jangoux, 1972 et 1976). La structure de l’estomac cardiaque des Asteriidae est surtout connue par le travail d’Anderson (1954) et ’estomac pylorique a été décrit de fagon succincte par Jangoux et al. (1972). Le présent travail récapitule et compléte les notions acquises sur l’organisation digestive d’A. rubens. Le but poursuivi est d’établir un schéma fonctionnel de |’appareil digestif de cette espéce. MATERIEL ET METHODES Les A. rubens ont été récoltés sur un brise-lames de la céte belge 4 Knokke. Elles ont été gardées en captivité dans un aquarium marin en circuit fermé. Pour l’observation histologique et les tests histochimiques les différents organes digestifs ont été fixés au Bouin acétique ou au formol 10% tamponné a la neutralité. Les fragments d’organes sont ensuite enrobés soit a la paraffine 57-60° C et coupés a 7 ,, soit au polyéthyléne glycol-nitrocellulose (PEG) et coupés a 10-12 1 (technique de Reid et Taylor 1964 pour la préservation des graisses). Australian Museum Memoir No. 16, 1982, 17-38. 18 MICHEL JANGOUX Le détail des colorations histologiques et des tests histochimiques effectués se trouve exposé dans les ouvrages de Gabe (1968) et Gantes et Jolles (1969): Hématoxyline couplée a la phloxine et au vert lumiére (topographie), Trichrome de Masson (topographie), Hématoxyline phosphotungstique de Mallory (topographie), Acide périodique — Schiff (APS, mucosubstances), Bleu alcian pH 2.6 et 0.5 (BA, mucosubstances), Couplage APS-BA pH 2.6, Te de Danielli (protéines), Noir Soudan B controlé par extraction a la pyridine ipides). Pour l'étude ultrastructurelle, de fins fragments d’organes sont fixés 30 min. a4°C dans une solution de glutaraldéhyde (2 ml de glutaraldéhyde 6% 1 ml tampon cacodylate 0.4 M, 1 ml NaCl 7%). Aprés lavage dans le tampon (30 min.), ils sont postfixés 1 h a 4°C au tétroxyde d’Osmium (2 ml 0s04 2%, 1 ml tampon cacodylate 0.4 M, 1 ml NaCl 10%) puis 4 nouveau lavés dans le tampon (15 min.) pour étre enfin déshydratés dans de bains d’ethanol de degré croissant. Linclusion se fait dans |’épon (Luft, 1961) ou l’araldite (Glauert et Glauert, 1958) et les coupes ‘ sont réaliseées a laide d’un ultramicrotome Reichert Om U2. Les coupes semi-fines (0.5 pm) sont colorées au bleu de toluidine ou au bleu de méthyléne et observées en microscopie photonique. Les ultracoupes (+400 A) sont recueillies sur grilles de cuivre (300 trous), contrastées a l’acétate d’uranyle et au plomb (Reynolds 1963) et observées au microscope électronique Philips EM 300. Des observations in vivo et des vivisections ont permis de suivre les mouvements du tube digestif lors d’un repas. La détermination du sens des courants ciliaires digestifs a nécessité également des vivisections: des organes ou fragments d’organes vivants sont placés dans un ana paystohoeiae contenant une suspension de carmin ou d’encre de Chine et observes au inoculaire. OBSERVATIONS ET RESULTATS 1. ANATOMIE DE L’APPAREIL DIGESTIF (fig. 1) La bouche s’ouvre directement dans un vaste estomac subdivisé en deux étages: le cardia, ventral, et le pylore, dorsal. L’estomac cardiaque est une large cavité occupant les trois-cinquiemes du volume du disque de l’astérie. I] présente une symétrie pentaradiée par le développement de cing poches radiales aux parois extreémement plissé¢es. Les poches cardiaques sont chacune rattachées aux vertébres ambulacraires d’un bras par une paire de ligaments triangulaires. L’ensemble de ces ligaments forme le systeme rétracteur stomacal, systeme dont la nature (collagéne, elastine et muscles) et le trajet furent étudiés en detail par Anderson (1954, Asterias forbesi). Les paires de ligaments prennent appui de part et d’autre des premiéres vertebres ambulacraires et de la rejoignent le centre de chaque poche cardiaque au niveau dune nodule. De chaque nodule partent ventralement des faisceaux accolés a la paroi stoma- cale et qui, aprés s’étre plusieurs fois dichotomisés, s’enfoncent dans l’épaisseur du cardia. Ils fusionnent alors avec la couche conjonctivo-musculaire de lorgane, Les faisceaux accolés a la paroi stomacale forment le systéme rétracteur intrinséque, les ligaments triangulaires constituant le systeme extrinséque. La configuration du systéme rétracteur permet de distinguer trois régions cardiaques. II s’agit du plancher (portion comprise entre la bouche et la terminaison du systéme intrinséque; oesophage sensu lato), des poches (portion recouverte du systéme intrinséque) et du plafond cardiaque (portion comprise entre la région des nodules et |’étranglement du pylore). Lors d’un repas toute la partie du cardia comprise entre la bouche et la partie supérieure du systéme intrinseque peut s’évaginer, mais généralement |’évagination ne concerne que le plancher stomacal. ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 19 Gs \ <— 6» ~ DOAN) | Fig. 1. Anatomie digestive d’Asterias rubens. CS = estomac cardiaque, I = intestin, PC = caecum pylorique, PD = canal /pylorique, PDi = diverticule pylorique, PS = estomac pylorique, RC = caecum rectal. Le pylore est séparé du cardia par un étranglement. La zone pylorique ventrale est d’un point de vue anatomique semblable au plafond du cardia dont elle est le prolongement direct. L’estomac pylorique se particularise d’avantage par une face dorsale pentagonale qui présente au contact de la lumiére digestive une paroi lobée et trés circonvoluée. Chaque angle du pentagone pylorique donne naissance a un conduit de section ovalaire (canal pylorique) se dirigeant vers un bras. A la base des bras les cing canaux pyloriques se dichotomisent et les dix canaux résultants donnent naissance a un caecum pylorique (deux -caecums par bras). Les caecums pyloriques sont de longs appendices en cul-de-sac suspendus dans la cavité brachiale par chaque fois deux mésentéres longitudinaux reliés a la paroi aborale | du bras. Chaque caecum est formé d’un long canal médian qui n’est que la prolongation du canal 20 MICHEL JANGOUX pylorique correspondant. Le canal caecal apparait comme une cavité comprimée latéralement ¢ reguli¢rement perforée. A chaque perforation correspond un diverticule pylorique, petite poch plurilobée et aveugle branchée sur le canal central. Le tres court intestin issu du pylore se jette dans les caecums rectaux en leur point de réunion. Ces caecums, au nombre de deux, sont rattachés a la paroi aborale du disque par u mésenteére. Ce sont de petits sacs d’aspect variable. Extérieurement ils présentent de nombreu lobes, intérieurement ils envoient dans la lumiére digestive de larges villosités. Un rectum a peine visible relie les caecums rectaux a |’anus. 2. HISTOLOGIE GENERALE La structure de la paroi digestive est relativement constante. On distingue de dedans en dehors |’épithélium digestif, une zone nerveuse intraépithéliale (nerf interne, dépendance d nerf radiaire superficiel), du tissu conjonctif, une musculature parfois fort développée et double orientation (circulaire et a longitudinale), des filets nerveux répartis dans la couche musculaire (nerf externe, dépendance du nerf radiaire profond) et un épithélium coelomique. L’épithélium digestif est toujours la couche tissulaire la plus développée. C’est un épithélium monostratifié fait de cellules hautes et étroites (palissade). Toutes les cellules de revétement sont ciliées et munies d’une bordure en brosse. Selon les organes elles acquiérent Pune ou l’autre spécialisation: production de courants d’eau, accumulation de produits de réserve . . . On distingue également différents types de cellules sécrétrices. La zone nerveuse interne est assez discréte. On la remarque surtout sous les cellules épithéliales spécialisées dans la production de courants d’eau (plancher du cardia, canaux pyloriques, intestin . . .). Ailleurs elle est formée de fins prolongements axoniques s’immiscant entre les bases des cellules. Le développement du tissu conjonctif est variable: particuli¢rement épais dans le plancher cardiaque et les caecums rectaux, il est quasi virtuel au niveau des diverticules pyloriques. La couche conjonctive est continue tout le long du tube digestif. Elle cloisonne de ce fait longitudinalement la paroi digestive et isole parfaitement le plexus nerveux superficiel du plexus nerveux profond. L’importance de la couche musculaire varie également selon la région considérée. Lorsque la musculature est bien développée (plancher cardiaque et caecums rectaux surtout), elle se compose de plusieurs assises cellulaires et forme alors un tissu distinct de |’épithélium coelomique. Lorsqu’elle est faiblement développée (diverticules pyloriques par exemple), elle ne constitue plus une couche tissulaire distincte mais participe directement a la formation de ’épithélium coelomique (cellules myoépithéliales disposées entre les cellules épithéliales coelomiques). Le développement du plexus nerveux profond est directement li€é a celui de la musculature. L’épithélium coelomique se compose d’une rangée de petites cellules cubiques ciliées pourvues d’une collerette de microvillosités (‘“‘choanocyte-like cell’? de Ngrrevang et Wingstrand 1970). Ces cellules coiffent extérieurement la musculature digestive. 3. ESTOMAC CARDIAQUE (a) Plancher cardiaque. C’est la portion du cardia comprise entre la bouche et les terminaisons du systéme rétracteur intrinséque, elle correspond grosso-modo a la partie stomacale évaginable. L’épithélium digestif du plancher s’accompagne d’une zone nerveuse trés importante. Le tissu conjonctif est ici particuli¢rement riche en fibres de collagéne. La couche musculaire est bien développée et orientée longitudinalement; le nerf externe est développé en conséquence. Nerve plexus Nerve plexus Fig. 2. Schéma de |’épithélium digestif du plancher cardiaque. ce = centrioles, de = desmosome, er = ergatoplasme, fl = cil, go = appareil de Golgi, MC = mucocyte, mi = mitochondrie, mu = plage muqueuse apicale, mv = microvillosités, my = myofibrilles, n = noyaux, ra = racine, sp = sphérule muqueux, VC = cellule vibratile, ZC = cellule zymogéne, zy = granules zymogénes. 22 MICHEL JANGOUX L'importance de la zone nerveuse interne confére a I’épithélium digestif du plancher un aspect remarquable. Les portions infranucléaires des cellules épithéliales sont en effet de largeur extrémement réduite, elles forment de fins faisceaux d’ancrage attachant les cellules a la lame basale et entre lesquels s’étend le nerf interne (fig. 7c). Ces faisceaux renferment essentiellement des myofibrilles et quelques petites mitochondries, ils sont toujours en étroit contact avec des prolongements nerveux (fig. 7c). Les myofibrilles ne sont pas limitées a la portion infranucléaire des cellules, on les rencontre entourant et surmontant les noyaux, parfois méme dans la région apicale (fig. 7a). ; L’épithélium digestif du plancher renferme trois types cellulaires distincts: des cellules palissadiques, des cellules glandulaires spumeuses et des cellules glandulaires granuleuses (fig. 2 et 8a). Toutes présentent un amincissement infranucléaire et contiennent des myofibrilles. Elles sont en outre pourvues d’une bordure en brosse apicale et d’au moins un cil. Les cellules palissadiques sont bien stir de loin les plus nombreuses. Ce sont des cellules | cylindriques, relativement étroites et munies d’une importante ciliature (2 a5 cils par cellules). Il est certain qu'une de leurs fonctions est la production de courants d’eau, raison pour laquelle je propose de les nommer cellules vibratiles. Les cellules vibratiles sont attachées entre elles et aux deux autres types cellulaires par des desmosomes de deux sortes: des desmosomes classiques (macula adhaerens) visibles a l’apex et des desmosomes septés ne s’apercevant que dans le tiers supérieur des cellules. Le cytoplasme apical des cellules vibratiles renferme trois éléments — caracteéristiques: des organites ciliaires basaux (centrioles, racine et microtubules), des mitochondries tres allongées et une 4 deux plages muqueuses. Les organites cilaires sont tout a fait classiques. La présence a ce niveau de grandes mitochondries permet une activité ciliaire certainement trés intense. Enfin les plages muqueuses, le plus souvent au nombre de deux, ont un contenu qui présente les réactions caractéristiques du mucus. L’épithélium du plancher est en réalité un épithélium glandulaire. Entre l’apex et le noyau le cytoplasme ne présente guére de particularités. Quant aux noyaux, leur forme allongée (“en cigare’’) et la forte densité de leur chromatine font qu’on les reconnait aisément (fig. 7c). On rencontre des cellules vibratiles a différents endroits du tube digestif. Elles se caracteérisent toutes par l’existence d’au moins deux cils et la présence de longues mitochondries apicales. Leur noyau est toujours allongée, elles sont toujours amincies basalement et accompagnées d’une zone nerveuse bien développée. Les cellules vibratiles du plancher se différencient des autres cellules de méme type par la présence de myofibrilles basales et de plages muqueuses apicales. Les cellules glandulaires spumeuses, relativement nombreuses dans la région du plancher, ne sont rien d’autre que des mucocytes tout a fait typiques: leur sécrétat est constitué de mucopolysaccharides acides carboxylés et sulfatés. Mises a part les caractéristiques déja citées (bordure en brosse, cil et myofibrilles basales) elles n’offrent pas de particularités. On les rencontre également en différents endroits du tube digestif, le plus souvent en association avec des cellules vibratiles. Les cellules glandulaires granuleuses sont particuliérement intéréssantes. Moins nombreuses que les mucocytes on les trouve cependant sur toute l’étendue du plancher. Elles renferment de gros grains phloxinophiles qui s’avérent étre essentiellement de nature protéique. Observées au microscope électronique elles rappellent fortement les cellules zymogénes du pancréas de vertébrés (Herridge et Loschman, 1972). Elles présentent en effet un reticulum endoplasmique trés développé, surtout dans la région infranucléaire, entre les saccules duquel se remarquent quelques petites mitochondries. Leur noyau se distingue nettement de celui des cellules vibratiles; il est de forme arrondie et a le nucléoplasme plus clair. Dans son voisinage s’apercoit un volumineux appareil de Golgi formé de nombreuses cisternes et présentant des ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 23 vacuoles de condensation. Le contenu vacuolaire est de densité électronique variable et tout indique que ces vacuoles sont a ’origine des grains de sécrétion (fig. 7b). Ces derniers, présents en grand nombre dans la région apicale (fig. 7a), sont, rappelons-le, du nature proteique. Il est trés vraisemblable que les cellules granuleuses du plancher cardiaque ont pour fonction de synthétiser et sécréter des enzymes digestives. Leurs caractéristiques ultrastructurelles permettent en tout cas de la supposer. Il est certain que le plancher cardiaque occupe dans les phénomenes digestifs une place prépondérante. La composition cellulaire de son épithélium interne (fig. 2) ainsi que le fait qu’il entre trés intimement en contact avec les zones digestibles des proies (évagination) l’indiquent a suffisance. (b) Poches cardiaques. Les poches cardiaques correspondent a la région stomacale recouverte par les faisceaux conjonctivo-musculaires du systéme rétracteur intrinséque. Elles sont au nombre de cing et font légérement saillie dans les cavités brachiales. L’histologie des poches cardiaques est trés différente de celle du plancher, au moins en ce qui concerne les tissus périphériques: la zone conjonctive et la couche musculaire sont fortement réduites sauf aux endroits ou s’accolent les faisceaux intrinséques (fig. 8b). Les parois des poches ‘sont extrémement plissées, replis qui bien entendu s’atténuent voire disparaissent lors de ’évagination du plancher. L’épithélium digestif des poches cardiaques renferme également trois types cellulaires: des cellules vibratiles, des cellules de revétement banales et des mucocytes. Les cellules vibratiles des poches s’organisent en rubans orientés ventro-dorsalement: ils naissent de l’épithélium du /plancher et se dirigent vers le haut. Au fur et a mesure que |’on s’éloigne du plancher le nombre de rubans vibratiles diminue progressivement par suite des fusions successives des rubans entre eux. Les cellules vibratiles forment ainsi de nombreux chemins ciliés au sein desquels se remarquent des mucocytes. Les cellules vibratiles des poches cardiaques sont accompagnées dune zone nerveuse interne bien apparente (fig. 7d). Elles sont semblables a leurs consoeurs du plancher a ceci prés qu’elles ne contiennent ni myofibrilles basales, ni plages muqueuses apicales. Entre les chemins ciliés, |’épithélium interne des poches est forme de cellules de /revétement banales cili¢es (un cil), pourvues d’une bordure en brosse et dont la fonction | essentielle semble étre de faire palissade (fig. 7c, d). Les cellules banales sont plus larges que les _cellules vibratiles et constituent de ce fait un épithélium d’aspect plus lache. (c) Plafond cardiaque. C’est la région comprise entre les nodules du systéme rétracteur et létranglement séparant le cardia du pylore. Mise a part la disparition des faisceaux intrinséques, Vhistologie du plafond cardiaque est semblable 4 celle des poches. Le nombre de chemins ciliés de l’épithélium interne diminue de plus en plus pour n’étre plus que cing au niveau de Vétranglement pylorique. Ces cing chemins vont chacun contacter la gouttiére ciliée ventrale dun canal pylorique. 4. ESTOMAC PYLORIQUE (fig. 3) L’estomac pylorique se particularise essentiellement par sa face dorsale trés circonvoluée, la face ventrale n’étant rien d’autre qu’une zone de transition entre la plafond cardiaque et les canaux pyloriques. L’histologie de l’estomac pylorique est relativement simple. L’épithélium digestif se compose de nombreuses bandes de cellules vibratiles et de mucocytes qui confluent toutes autour de orifice intestinal. Le trajet de ces bandes cilio-muqueuses s’observe trés bien lorsqu’on regarde le pylore par son cété coelomique. Entre les chemins ciliés |’épithélium est formé de cellules de revétement banales. L’épithélium digestif du pylore est trés semblable a 24 MICHEL JANGOUX ORAL OR DIGESTIVE SIDE ABORAL OR COELOMIC SIDE Fig. 3. Vues orale et aborale de l’estomac pylorique. celui des poches cardiaques a ceci prés que les bandes cilio-muqueuses sont ici plus nombreuses mais que leur ciliature est moins importante (rarement plus de deux cils par cellule). Le nerf interne est bien stir trés apparent et le conjonctif, d’épaisseur variable selon !’endroit, est riche de fibres collagéne (fig. 9). La musculature est fortement réduite. 5. CANAL ET CAECUMS PYLORIQUES Chaque canal pylorique donne naissance 4 deux caecums pyloriques. Chaque caecum se compose d’un canal central (portion caecale du canal pylorique) dans lequel se jette un grand nombre de diverticules pyloriques (figs. 1 et 4). (a) Canal pylorique. La portion libre et la portion caecale du canal pylorique ont la méme structure histologique. L’épithélium digestif est constitué par deux gouttiéres cilio-muqueuses (pole oral et péle aboral) séparées par des cellules de revétement banales (paroi laterale). Le nerf ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 25 PDi Fig. 4. Coupe transversale du canal (a) et du caecum (b) pyloriques. ET = tissus périphériques (conjonctif, muscles et epithelium coelomique), Me = mésentére, N = plexus nerveux, PDi = diverticule pylorique. interne s’apercoit au niveau des gouttiéres. Le conjonctif et la musculature sont peu développés. Le pole aboral se distingue de son vis-a-vis par la présence de deux petites lacunes hémales situées dans le conjonctif au point de rencontre canal pylorique — mésentéres d’attache. (b) Diverticules pyloriques. L’épithélium digestif des diverticules pyloriques occupe plus des neuf-dixiémes de |’épaisseur de la paroi, les tissus périphériques sont extrémement réduits (fig. 9c). Les caecums pyloriques ont été étudiés en détail (Anderson 1953, Nimitz 1971, Jangoux et Perpeet 1972) et leur composition cellulaire est bien connue. On y remarque des cellules de revétement spécialisées dans |’accumulation de réserves énergétiques, des cellules zymogénes a large vacuole et, assez rarement, des mucocytes. La fonction de réservoir énergétique des cellules de revétement a été établie par Anderson (1953). Tout comme lui j’ai pu observer des goutelettes lipidiques et des granules de glycogéne dans leur cytoplasme, principalement dans la région infranucléaire. Assez étonnamment il arrive que certains caecums soient dépourvus d’inclusions graisseuses et dans tous les cas il s’agit de caecums d’astéries récoltées en été. Cela s’explique par le fait que tant la quantité que la composition des réserves pyloriques varient au cours de l’année (Jangoux et Van Impe 1977). Le noyau des cellules de réserve est arrondi ou légérement ovale et renferme une trame chromatique lache, rien ne le distingue du noyau des cellules de revétement banales. Les cellules de réserve sont également absorbantes. Elles possédent une bordure en brosse trés serrée et le cytoplasme apical contient de nombreuses petites vesicules claires, résultat de l’activité pinocytaire (fig. 9b). On y remarque aussi quelques lysosomes et de petites mitochondries. Toutes les cellules de réserve sont ciliées (un cil). 26 MICHEL JANGOUX Décrites pour la premiére fois par Anderson (1953), les cellules zymogénes des caecum pyloriques se caractérisent avant tout par la présence d’une large vacuole claire (fig. 9c). Cett vacuole semble vide et seules certaines préparations ultrastructurelles ont permis d’y observe un fin feutrage de nature inconnue. Le noyau est trés reconnaissable par son gros nucléol central, il se place toujours sous la vacuole. Les grains protéiniques (grains zymogénes s’observent sur toute le hauteur des cellules. Les organites responsables de leur fabrication son localisés de fagon plus précise, généralement au voisinage du noyau. II n’est pas rare de rencontrer a cet endroit des complexes d’organites composés d’un appareil de Golgi, de saccul ergastoplasmiques et de mitochondries. Ces derniéres entourent trés souvent les grain zymogenes en formation. Les cellules zymogénes pyloriques sont ciliées et coiffées d’un bordure en brosse. 6. PARTIE POSTERIEURE DE TUBE DIGESTIF (a) Intestin. Sa structure est particuliérement simple. C’est un trés court conduit tapiss entierement de cellules vibratiles parsemées de quelques mucocytes. II reproduit l’organisatio des bandes cili¢es des poches et du plafond cardiaque, du pylore et des canaux pyloriques. (b) Caecums rectaux. En coupe histologique les caecums rectaux se reconnaissement aisément: leur lumiére est envahie de nombreuses villosités formées par I’épithélium interne (fig. 8e). Ces villosités sont sous-tendues par des lames conjonctives riches en fibres collagéne. La musculature rectale est fortement développée et composée de deux couches de muscles d’orientations circulaire et longitudinale. ; L’€pithélium interne des caecums rectaux comprend deux types cellulaires: des mucocytes typiques et des cellules de revétement d’aspect banal en microscopie photonique. L’étude ultrastructurelle des cellules de revétement montre cependant qu’elles présentent de trés intéréssantes particularités (Jangoux 1972, 1976). L’existence d’une bordure en brosse serrée, de vésicules de pinocytoses et de lysosomes dans le cytoplasme apical indique qu’il s’agit de cellules absorbantes, fonction qu’elles partagent avec les cellules de réserve des diverticules pyloriques, Tout comme ces derniéres elles sont ciliées. Toutefois, et a la différence de leurs homologues pyloriques, elles présentent de trés longues mitochondries apicales. Le cytoplasme median et le noyau, semblable a celui des cellules de revétement cardiaques ou pyloriques, ne se caractérisent en rien. Par contre la zone cellulaire basale renferme d’importants replis membranaires ( B-cytomembranes) limitant de fines bandes cytoplasmiques riches en mitochondries. (c) Rectum. C’est un conduit trés fin et trés court. L’épithélium digestif n’est constitué que de cellules de revétement banales. La seule particularité du rectum est l’existence d’une importante couche musculaire formant sphincter. 7. MOUVEMENTS DIGESTIFS LORS DU REPAS (a) Observations externes. Le comportement d’une Asterias s apprétant a se nourrir est bien connu. L’astérie enserre la proie de ses bras et se bombe le dos du disque (réflexe du “gros-dos” ou “humping reflex’’). Ce réflexe est trés caractéristique et précéde toujours |’évagination stomacale. I] peut également apparaitre lorsqu’on place une astérie dans un milieu riche en aliments en suspension (Heeb 1973, Jangoux 1976). L’évagination stomacale n’intéresse en général que le plancher cardiaque. Toutefois lorsque les proies sont de grande taille les extrémités ventrales des poches cardiaques peuvent faire saillie a l’extérieur. L’estomac évaginé ressemble a une vessie dilatée, gonflée de liquide coelomique sous pression, Cette surpression coelomique fait que lorsque le plancher cardiaque a pu simmiscer entre les valves d’un bivalve par exemple, il se moule trés exactement sur les ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS PF. an ( Fig. 5. Estomac cardiaque en place (a) et dévaginé (b). CF = plancher cardiaque, CP = poche cardiaque, PD = canal pylorique. Les replis stomacaux longitudinaux ne sont pas représentés sur le dessin b. 28 MICHEL JANGOUX parties molles de sa proie et les contacte intimement. La durée de l’évagination est trés variable et depend bien stir des tailles relatives de la proie et de l’étoile de mer. On peut cependant estimer le temps moyen d’€vagination a 5-6 heures, l’estomac n’étant remis en place que lorsque la coquille est parfaitement vidée. Pendant toute la durée de |’évagination l’astérie reste étonnamment immobile et semble se concentrer entiérement sur son repas. Durant tout le repas le dos du disque est bombé. Seule une observation patiente permet de remarquer de temps a autre l’éjection d’eau par I’anus. De tels rejets s’observant trés rarement chez une astérie au repos; ils peuvent se voir avec une certaine fréquence chez les individus s’alimentant (2 a 6 gections/heure). L’eau ainsi rejetée n’est pas limpide mais renferme des féces brunatres d’aspect muqueux (Jangoux 1976). (b) Observations Internes. Lors d’un repas deux régions digestives présentent des mouvements modifiant plus ou moins leur configuration: ce sont l’estomac cardiaque et les caecums rectaux. Estomac Cardiaque. Le mécanisme de l’éversion stomacale a déja été explicité par Anderson (1954). Cela se déroule comme suit: ouverture de la bouche, relachement du systéme rétracteur stomacal et des muscles de la paroi de l’estomac, contraction des muscles de la paroi du corps entrainant une augmentation de la pression intracoelomique et I’évagination du plancher cardiaque. Anderson ne précise toutefois pas ce qui est a l’origine de l’étonnante extension de la paroi du plancher. Selon moi cette élasticité est principalement le résultat de la contraction des myofibrilles intraépithéliales (voir 3.a), ce qui permet I’étirement transversal de la paroi stomacale. Il est evident que l’evagination stomacale modifie complétement la topographie du cardia (figs. 5, a et b), La modification la plus spectaculaire est bien stir le considérable étirement du plancher cardiaque, mais la partie cachée de l’estomac va également changer de configuration. L’etranglement pylorique et les replis des poches du cardia s’estompent: poches cardiaques, plafond cardiaque et plancher pylorique s’alignent sur un méme plan incliné allant de la bouche aux orifices des canaux pyloriques. La surface de ce plan incliné stomacal est couverte de chemins cilio-muqueux dirigés ventro-dorsalement et séparés par des plages de cellules banales. Les mouvements stomacaux n’ont lieu qu’en début et en fin de repas. Pendant toute la durée de l’alimentation l’estomac reste “‘déroulé’’, tout au plus peut-on voir se modifier la longueur de la portion évaginée (contractions — relaxations des myofibrilles épithéliales). Caecums Rectaux. Les seuls organes digestifs a présenter des mouvements répétés au cours d’un repas sont les caecums rectaux. Ce sont de petits organes contractiles aux parois trés musculeuses. Leur contraction entraine un rejet anal d’eau (relachement du sphincter rectal) et autorise donc la défécation. Des dissections d’étoiles de mer en train de s’alimenter ont permis de comprendre le fonctionnement des diverticules rectaux. Les deux caecums se comportent indépendamment l’un de I’autre, chacun est capable non seulement de se contracter mais aussi de se dilater tres fortement. Lors d’un repas les caecums agissent en alternance de la facon suivante: alors que l'un d’eux semble au repos, l'autre se dilate trés progressivement. La phase de dilation peut durer plusieurs dizaines de minutes, le caecum se gorgeant petit a petit d’un liquide brunatre. Au bout d’un certain temps l’organe dilaté se contracte violemment et rejette son contenu par l’anus. A la suite de cela le deuxieme caecum entame sa phase de dilatation, le premier se mettant au repos. Alors que la période de dilatation peut durer plusieurs dizaines de minutes, la contraction est trés rapide et ne dépasse jamais 10 secondes. Le mucus brunatre qui s’‘accumule dans la lumiére rectale provient pour une bonne part de la cavité stomacale. Cela se remarque fort bien en dissection ou on observe trés souvent la présence d’un film muqueux continu allant des parois cardiaques a l’intestin. Ce phénoméne n'est explicable que si on admet qu’ils existe un courant d’eau, sorte de pompage rectal, allant ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 29 { An [re ca RECTAL 4! = PYLORIC PYLORIC STOMACH —— a9 DIVERTICULUM —¥ = = fink \ —— / ne | / la \| | ) ’ 1% ~~ s | \ PYLORIC CARDIAC | CARDIAC CRE CUES WALL POUCH (interradius) 1 \ r (radius ) 2g \ / oh: eae CARDIAC SSS FLOOR Mo Fig. 6. Schéma fonctionnel de l’appareil digestif (en traits gras, les trois régions digestives principales). An = anus, Mo = bouche. dans le sens bouche-anus. Le premier réflexe alimentaire significatif d’Asterias est celui du “‘gros-dos”. Des dissections d’astéries au stade “gros-dos”’, c’est-a-dire avant |’évagination stomacale, ont permis de remarquer qu’un disque bombé renfermait toujours un caecum rectal dilate. [I] apparait done que le pompage rectal précéde dans le temps |’évagination stomacale. La répetition de certaines des manipulations de Heeb (1973, adjonction dans |’eau de l’aquarium de diverses dilutions de broyat de moule) a montré que le réflexe du “‘gros-dos”’ n’est pas nécessairement suivi de l’évagination cardiaque. Celle-ci ne s’observera que si la dilution ajoutée au milieu est suffisamment concentrée en stimuli alimentaires. Le pompage rectal peut donc étre déclenché lorsque I’astérie se trouve dans un milieu suffisamment riche en micro-aliments. 8. SENS DES COURANTS CILIAIRES C’est au niveau des zones tapissées de cellules vibratiles que les courants sont les plus intenses. Leurs directions sont les suivantes (fig. 6): — plancher cardiaque: courant ventro-dorsal — bandes cilio-muqueuses des poches cardiaques, du plafond cardiaque et du plancher pylorique: courant ventro-dorsal — plages de cellules banales des poches ou du plafond cardiaques: courant faible latéral ou dorso-ventral — gouttiere orale des canaux pyloriques: courant centrifuge — gouttiére aborale des canaux pyloriques: courant centripéte — diverticules pyloriques: courant tourbillonnaire — bandes cilio-muqueuses du plafond pylorique: courant centripéte — intestin: courant ventro-dorsal — caecums rectaux: courant tourbillonnaire Ces résultats confirment et complétent ceux obtenus par Budington (1942) et Anderson (1954) sur Asterias forbesi. DISCUSSION Les cellules épithéliales digestives présentent deux caractéristiques fondamentales: elles sont en principe toutes cili€ées (au moins un cil) et absorbantes (bordure en brosse). Cela ne signifie évidemment pas qu’il n’existe pas de sites absorbants privilégiés, sites qui sont les 30 MICHEL JANGOUX Fig. 7. Micrographies de la paroi du cardia (a ac plancher cardiaque, d poche cardiaque). (a-b) = Cellules a grains protéiques (cellules zymogénes), c = cisternes de l’appareil de Golgi, er = ergastoplasme, g = granules protéiques (zymogénes), my = myofibrilles, p = plage muqueuse apicale d’une cellule vibratile; (c-d) = Base de l’épithélium digestif, co = tissu conjonctif, cv = cellule vibratile, g = granules protéiques (zymogénes), my = myofibrilles, n = plexus nerveux. ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 31 Fig. 8. Coupes transversales au travers du plancher (a) et des poches cardiaques (b-d), et des caecums rectaux (€). c = tissu conjonctif, eb = épithélium banal, ev = épithélium vibratile, ed = épithélium digestif, m = muscles, mu = mucocytes, n = plexus nerveux, sr = systéme rétracteur intrinséque, z = cellule a grains protéiques (zymogénes). oy MICHEL JANGOUX Fig. 9. Micrographies de la paroi du pylore (a) et de l’épithélium digestif des diverticules pyloriques (b-c). c = coelome, ce = épithélium coelomique, co = tissu conjonctif, f = fibres musculaires, g = granules zymogénes, i = inclusions lipidiques, mu = mucocytes, mv = microvillosités, n = tissu nerveux, p = vésicules de pinocytose, v = vacuole de cellule zymogéne pylorique, vi = cellule vibratile. ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 33 caecums pyloriques et rectaux. Bien que composé de cellules diversement spécialisees, Pépithélium digestif a un aspect semblable dans tous les organes: cellules hautes, étroites et assez serrees, formant une veritable palissade et rendant assez difficile ! observation microscopique. La microscopie électronique a permis de constater l’omniprésence du plexus nerveux interne dont les filets s'intercalent toujours entre la membrane basale et les membranes des cellules digestives. Le nerf interne est particuliérement bien apparent sous les bandes de cellules vibratiles; son trajet correspond a celui observé par Smith (1937) chez Marthasterias glacialis. Le tissu conjonctif, riche en fibres de collagéne, se développe différemment selon les régions. C’est au niveau des zones extensibles ou contractiles (plancher cardiaque et caecums rectaux) que son épaisseur est maximale. Le développement de la musculature digestive et du nerf externe qui l’innerve est paralléle a celui du tissu de soutien: le plancher cardiaque et les caecums rectaux sont les régions les plus musclées. Je rappellerai enfin que l’epithelium externe (coelomique) se compose de petites cellules cubiques ciliées et garnies d’une couronne de microvillosités (““chaoanocyte-like cells’’). (a) Estomac cardiaque. Anderson (1954) s’est surtout attaché a étudier la structure du systéme réetracteur et des poches cardiaques, beaucoup moins celle du plancher du cardia (oesophage sensu lato). Les cellules a gros grains réfringents décrites par lui dans l’oesophage d’A. forbesi sont trés vraisemblablement identiques aux cellules 4 grains proteiques du plancher cardiaque aA. rubens. Par contre les cellules muriformes (‘‘mulberry cells”) observées par Anderson (1959) dans la paroi stomacale de Patiria miniata sont certainement d’un autre type; il s’agit probablement de coelomocytes. Il est incontestable que le plancher cardiaque est la région stomacale la plus importante du point de vue digestif. C’est la partie évaginable et toutes les cellules épithéliales non glandulaires qui la constituent sont vibratiles (2 a 5 cils par cellule). Le fort courant ciliaire qu’elles produisent est unidirectionnel, et entraine l’eau et son contenu vers les organes digestifs internes. L’estomac évaginé contacte trés intimement les parties molles des proies: Anderson (1966) nous dit que l’espace séparant l’estomac de la proie est microscopique, pour ma part j’en estimerais la largeur 4 quelques dizaines de microns. Dans ces conditions le courant d’eau produit par les cellules épitheliales est trés étroitement canalis€é, ce qui augmente évidemment son efficacité. Outre une ciliature intense, les cellules vibratiles sécrétent du mucus (plages muqueuses apicales): elles vont donc pouvoir enrober les fragments de proie en méme temps qu’elles les véhiculent vers le reste du tube digestif. Etant donné qu’il y a digestion extra-orale, il faut qu’au niveau du plancher parviennent des enzymes digestives. La ciliature de la partie stomacale évaginée créant un fort courant ventro-dorsal, il est physiquement impossible que les enzymes nécessaires proviennent des caecums pyloriques. I] est donc logique, en dehors de toute observation microscopique, de supposer |’existence de cellules zymogénes dans le plancher cardiaque. Cette supposition est vérifiée par |’étude histologique et ultrastructurale qui y a révélé la présence de cellules a grains protéiques présentent les caractéristiques classiques des cellules zymogeénes. Ces cellules sont d ailleurs trés semblables aux cellules gastriques exocrines observées dans la premiere spire digestive de l’oursin Strongylocentrotus par Holland et Lauritis (1968). En conclusion on peut penser que chez A. rubens le phénomene de la digestion extra-orale se réalise exclusivement a Pintervention d’enzymes digestives originaires du plancher de l’estomac cardiaque. Il convient ici d’insister sur une particularité propre a toutes les cellules de l’épithélium digestif du plancher: la présence de myofibrilles dans leur cytoplasme. Les contractions de ces myofibrilles permettent d’expliquer la remarquable élasticité du plancher stomacal, élasticité que la seule sur-pression du liquide coelomique ne pouvait expliquer (la surpression étant essentiellement responsable de la turgescence de |’estomac évaginé). 34 MICHEL JANGOUX Les poches et le plafond cardiaque ne sont importants que par les chemins cilio-muqueux qui les tapissent. Ce sont, avec la face ventrale de l’estomac pylorique, des zones de transition dont le role est d’acheminer vers le pdle oral des canaux pyloriques les petits fragments de proie enrobés de mucus. A l’état de repos les poches cardiaques sont extrémement plissées. En coupe histologique elles apparaissent comme une succession de creux et de crétes, ces derniéres étant toujours tapissces de cellules épithéliales vibratiles (Anderson 1954, Jangoux et al. 1972). A la suite de levagination les poches s’étirent et les replis s’estompent. Les chemins vibratiles sont a ce moment tous disposés sur un méme plan et séparés les uns des autres par des zomes couvertes de cellules de revétement banales. (b) Estomac pylorique. On peut comparer la face dorsale du pylore a une roue a rayons, chaque rayon €tant constitué par une bande de cellules ciliées. Les courants d’eau créés par les bandes ciliées se dirigent tous vers l’orifice intestinal, centre de la roue. L’estomac pylorique est donc ¢galement un organe de transit. Dans le cas des Asteriidae, contrairement a ce que pensait Ferguson (1969), ce n’est pas le pylore qui distribue aux caecums pyloriques les produits de la digestion cardiaque. Ces produits leur sont en effet directement acheminés par les chemins ciliés du cardia. L’hypothése de Ferguson est toutefois partiellement vraie dans la mesure ou les microparticules entrainées vers le haut par le courant rectal (courant buco-anal des caecums rectaux) peuvent étre réceptionnées, enrobées de mucus et vehiculées par le pylore. Ces particules ne sont cependant pas destinées aux caecums pyloriques mais bien aux caecums rectaux, via l’intestin, En outre le pylore peut prendre en charge les matériaux rejetés par les caecums pyloriques (gouttiére ciliée aborale des canaux pyloriques) et les diriger également vers les caecums rectaux. (c) Caecums pyloriques. Les caecums pyloriques sont les organes digestifs les plus étudiés (Anderson 1953, Karnovsky et al. 1955, Bargmann et Behrens 1968, Chia 1969, Chan et Fontaine 1971, Nimitz 1971, Jangoux et Perpeet 1972). Ils sont formés d’un canal médian de section ovalaire (canal pylorique) parcouru par deux gouttiéres ciliées, une orale a courant centrifuge, l'autre aborale a courant centripéde. Les parois latérales du canal pylorique présentent de part de d’autre et réguli¢rement des petits sacs plurilobés: les diverticules pyloriques. La gouttiére orale alimente les diverticules en matériaux nutritifs originaires du cardia, la gouttiére aborale ¢limine vers le pylore les matériaux perdus ou refusés par les diverticules. Trois fonctions physiologiques importantes sont remplies par l’épithélium digestif des caecums pyloriques: l’absorption alimentaire, la mise en réserve de substances énergétiques, l’élaboration et la sécrétion d’enzymes digestives (voir entre autres Anderson 1966, Hori et al. 1977, Jangoux et Van impe 1977). Comme permet de le penser la structure fine des cellules pyloriques et comme l’ont montré les experiences réalisées in vitro a l'aide de fragments de caecums pyloriques (incubation dans une solution marine de ferritine, Jangoux 1976), les cellules de réserve endocytent trés activement. L’endocytose pylorique semble étre discriminante et on peut supposer qu’une ~ partie des matériaux rejetés par les caecums le sont a la suite de leur refus par les cellules — épitheliales (Jangoux 1976), r Mise a part l’existence d’une large vacuole, les cellules zymogénes pyloriques offrent les mémes caractéristiques générales que leurs homologues du plancher cardiaque. Cette vacuole — n’est pas vide mais remplie d’un fin feutrage de nature encore inconnue. Le réle de la vacuole reste mysterieux alors que la nature zymogéne des grains de sécrétion a été clairement démontrée _ par Horier al. (1977). Les enzymes digestives produits par ces cellules agissent sur les matériaux _ acheminés dans les diverticules pyloriques par la gouttiére orale; ces enzymes poursuivent donc la digestion stomacale. On I’a vu, il est physiquement impossible que les enzymes pyloriques ETUDE DU TUBE DIGESTIF D’'ASTERIAS RUBENS 35 iennent au plancher cardiaque, les sens des courants ciliaires sy opposent. d) Caecums rectaux. Les fonctions des caecums rectaux ont été récemment discutées (Jangoux 976). Ce sont des organes contractiles, leurs contractions ayant pour principal effet de permettre la défécation (contraction caecale couplée au relachement du sphincter anal). Au urs du repas les phases de contraction alternent avec des phases de dilatation lente. C'est ndant les dilatations que les caecums rectaux se gorgent de substances diverses (particules imilables, mucus d'origine cardiaque et/ou pylorique . . .). Contrairement a ce qui avait ete uppose (Jangoux 1976), il est peu probable que les dilatations des caecums rectaux soient msables de I’existence d’un courant d’eau buco-anal. Au contraire elles en seraient plus olontiers la conséquence. Le “pompage rectal” est selon toute vraisemblance le résultat de deux €canismes distincts: (1) aprés contraction, le retour des caecums rectaux a l'état de repos relachement de la musculature des parois) entraine un appel d'eau; (2) la trés importante iliature intestinale draine I’eau de la cavité stomacale vers les caecums rectaux, entrainant ainsi eur dilatation. En d'autres termes le pompage rectal serait principalement la conséquence de "activité ciliaire de l'intestin (mécanisme de pompage ciliaire). Le pouvour absorbant des caecums rectaux est trés élevé et il s'y déroule une intense igestion intracellulaire. Contrairement a ce qui se passe dans les caecums pyloriques, *endocytose rectale ne parait pas discriminante (Jangoux 1976). e) La mécanique digestive d’A. rubens (fig. 6). Aprés avoir localisé sa proie (une moule par emple) l’astérie la contacte, l"enserre de ses bras et applique trés intiment un grand nombre de entouses ambulacraires sur ses valves. A ce stade l’astérie se bombe le dos du disque (dilatation ides caecums rectaux, déclenchement du pompage rectal) et peu aprés débute l"évagination du plancher cardiaque. On peut penser que ces différents réflexes alimentaires (pompage rectal et éVagination stomacale) sont le résultat de chemoréception. Dés le déclenchement du pompage rectal, les caecums rectaux peuvent recevoir un certain nomibre de particules alimentaires mais c'est bien sur l'évagination stomacale qui permettra ‘apport nutritif le plus important. L’évagination cardiaque est un processus complexe mécéssitant succéssivement l’ouverture de la bouche, la relachement du systeme rétracteur macal ainsi que des muscles de la paroi du plancher cardiaque et la contraction des muscles de paroi du corps (raidissement de lastérie). Cette contraction va entrainer une surpression du liquide coelomique qui va pousser le plancher stomacal, poussée qui entrainera la saillie omacale dans le milieu extérieur. L’estomac évaginé revét l’aspect d’un lobe blanchatre gescent, il va s appliquer tres étroitement sur les valves de la proie. La taille de 'estomac évaginé peut varier: c est la conséquence des contractions-relaxations des myofibrilles intra€épithéliales du plancher cardiaque. La facon dont l’estomac pénétre entre les valves de la proie n’est pas trés bien comprise. Sans doute profite-t-il a la fois des échancrures naturellement présentes entre les deux valves et tractions exercées par les pieds ambulacraires sur les valves pour les écarter (Nichols 1964, (Péquignat 1970). Une fois l’échancrure repérée, l’estomac est poussé dans le bivalve par la ression du liquide coelomique et va se mouler trés exactement sur les parties molles de la proie. ce moment les cellules zymogenes cardiaques déversent leurs enzymes et la digestion tomacale extra-orale débute (digestion extracellulaire). Les fragments de proie libérés par les ferments digestifs sont enrobés de mucus par les cellules du plancher et dirigés vers les chemins ciliés des poches cardiaques qui les transporteront jusqu’au niveau des canaux pyloriques. Pendant ce temps la pompe rectale fonctionne et entraine dans le courant qu'elle crée de trés fins fragments de proie qui alimenteront directement les caecums rectaux (digestion intracellulaire). Il n'est pas interdit de penser que le pompage rectal favorise également l’acheminement des fragments de proie le long des chemins ciliés du cardia. Arrivés dans les diverticules pyloriques, J les fragments de nourriture vont subir l’attaque des enzymes pyloriques et les produits de cette» digestion serbnt absorbés (digestions extracellulaire et intracellulaire). 36 MICHEL JANGOUX Comme I’a signalé MacBride (1909), la digestion de l’astérie est tres complete et peu de materiaux sont rejetés par l’anus. Les substances défécées ont une double origine: mucus brunatre provenant de l’estomac et matériels refusés ou perdus par les caecums pyloriques et rectaux. Ces substances s’accumulent dans la lumiére des caecums rectaux et sont éliminés par | l’anus a l'occasion d’une contraction rectale. La durée moyenne d’un repas est de 5 26 heures, la proie n’étant abandonnée que lorsque sa coquille est parfaitement nettoyée. On assiste alors ala remise en place de l’estomac: relachement de la musculature des parois de corps et diminution de la pression intracoelomique, contraction du systéme rétracteur et des muscles du plancher — cardiaque, fermeture de la bouche. Le schéma de fonctionnement du tube digestif proposé ci-dessus a été établi a partir de> l'étude d’Asierias rubens. Etant donné la grande homogénéité anatomique des Asteriidae, il est tres probable qu’il puisse, a quelques détails prés, s’appliquer a toutes les espéces de cette importante famille. A l’opposé je ne crois pas qu’on puisse |’étendre aux autres familles de la classe. Les structures digestives sont en effet trés variées dans le groupe des étoiles de mer (voir entre autres Anderson 1960, 1966, 1978) et 4 chaque type morphologique correspond sans nul doute une mécanique digestive particuliére. En conclusion on peut dire que le tube digestif des Asteriidae est formé de trois régions importantes par leurs caractéristiques morphologiques et physiologiques, régions ou se déroule l’essentiel des phénoménes digestifs (plancher cardiaque, diverticules pyloriques et caecums rectaux). Ces régions sont reliées entre elles par des zones de _ transit (chemins cilio-muqueux des poches et du plafond cardiaque, des canaux pyloriques, du _ pylore et de l’intestin). REFERENCES Anderson, J. M., 1953. Structure and function in the pyloric caeca of Asterias forbesi. Biol. Bull. mar. biol. Lab. Woods | Hole 105: 47-61, 1954. Studies on the cardiac stomach of the starfish Asterias forbesi. Biol. Bull. mar. biol. Lab. Woods Hole 107: | 157-173. 1959. Studies on the cardiac stomach of a starfish Patiria miniata (Brandt). Biol. Bull. mar. biol. Lab. Woods Hole 117: 185-201, | 1960. Histological studies on the digestive system of a starfish Henricia, with notes on Tiedemann’s pouches in — starfishes. Biol. Bull. Mar. biol. Lab. Woods Hole 119: 371-398. 1966. Aspect of nutritional physiology. In R. A. Boolootian (ed.), Physiology of Echinodermata. Intersciences | Publ., New York: 329-357. q 1978. Studies on functional morphology in the digestive system of Oreaster reticulatus (L.) (Asteroidea). Biol. Bull. mar. biol. Lab. Woods Hole 154: 1-14. Bargmann, W, and B. Behrens, 1968. Ueber die Pylorusanhange des Seesterns (Asterias rubens L.) insbesondere ihr Innervation. Z. Zellforsch. 84: 563-584. Berridges, M. J. and J. L. Loschman, 1972. Transporting epithelia. Academic Press, New York, 91 pp. Budington, R. A., 1942. The ciliary transport-system of Asterias forbesi. Biol. Bull. mar. biol. Lab. Woods Hole 83: 438-450, Chadwick, H. C., 1923. Asterias, L.M.B.C. Mem. typ. brit. mar. Plants Animals 25: 63 pp., 9 pls. Chan, V. G. and A. R. Fontaine, 1971. Is there a 6-cell homolog in starfish? Gen. comp. Endocrinol. 16: 183-191. Chia, F. S., 1969. Histology of the pyloric caeca and its changes during brooding and starvation ina starfish, Leptasterias hexactis. Biol. Bull. mar. biol. Lab. Woods Hole 136: 185-192. ETUDE DU TUBE DIGESTIF D’ASTERIAS RUBENS 37 \Cuénot, L., 1887. Contribution a I’étude anatomique des astérides. Arch. Zool. exp. gen. 5: 1-144, 9 pls. 1948. Anatomie, éthologie et systématique des Echinodermes. In P. P. Grasse (ed.), Traité de Zoologie, vol. 11. Masson, Paris: 3-372. Ferguson, J. C., 1969. Feeding, digestion and nutrition in Echinodermata. In M. Florkin and B. T. Scheer (eds.), Chemical Zoology, vol. 4. Academic Press, New York: 71-100. Gabe, M., 1968. Techniques histologiques. Masson, Paris, 1113 pp. Gantes, P. and G. Jolles, 1969. Histochimie normale et pathologique. 2 vols., Gauthier-Villars, Paris. Glauert, A. M. and R. H. Glauert, 1958. Araldite as embedding medium for electron microscopy. 7. biophys. biochem. Cytol. 4: 191-194. Hamann, D,, 1885. Die Asteriden anatomisch und histologisch untersucht. Fisher, Jena, 126 pp., 7 pls. ‘Heeb, M. A., 1973. Large molecules and chemical control of feeding behaviour in the starfish Astenas forbest. Helgolander wiss. Meeresunters. 24: 425-435. 'Holland, N. D. and J. A. Lauritis, 1968. The fine structure of the gastric exocrine cells of the purple sea urchin, Strongylocentrotus purpuratus. Trans. amer. Soc. micr. Sci. 87: 201-209. Hori, S. H., K. Tanahashi, and N. Matsuoka, 1977. Morphological and cytochemical studies on the secretory granules of the pyloric caeca of the starfish Asterias amurensis. Biol. Bull. mar. biol. Lab. Woods Hole 152: 64-74. ‘Hyman, L. H. 1955. Echinodermata. In The Invertebrates, vol. 4. MacGraw-Hill, New York: vii + 763 pp. Jangoux, M., 1972. La structure fine des caecums rectaux de deux Asteriidae: Marthasterias glacialis (L.) et Coscinasterias tenuispina (Lam.). Z. Zellforsch. 126: 366-384. 1976. Fonctions des caecums rectaux chez I’étoile de mer Asterias rubens L. Thalassia Yugosl. 12: 181-186. Jangoux, M. et C. Perpeet, 1972. Etude comparative de la structure fine des caecums pyloriques de trois espéces d’Asteriidae. Cah. Biol. mar. 13: 401-420. \Jangoux, M.,C. Perpeet et D. Cornet, 1972. Contribution al’étude des poches stomacales d’Asterias rubens L. Mar. Biol. 15: 329-335, Jangoux, M. and E. Van Impe, 1977. The annual pyloric cycle of Asterias rubens L..F. exp. mar. Biol. Ecol. 30: 165-184. Karnovsky, M. L., S. S. Jeffrey, M. S. Thompson and H. W. Deane, 1955. A chemical and histochemical study of the lipids of the pyloric caecum of the starfish Asterias forbesi. 7. biophys. biochem. Cytol. 1: 173-182. | Luft, J. H., 1961. Improvements in expoxy resin embedding methods. 7. biophys. biochem. Cytol. 9: 409. MacBride, E. W., 1909. Echinodermata. In S. F. Hormer and A. E. Shipley (eds.), Cambridge Natural History, vol. 1. Macmillan, London: 425-623. Nichols, D., 1964. Echinoderms: experimental and ecological, Oceanogr. mar. Biol. ann. Rev. 2: 393-423. Nimitz, A. M., 1971. Histochemical study of gut nutrient reserves in relation to reproduction and nutrition in the sea-stars, Pisaster ochraceus and Patina miniata. Biol. Bull. mar. biol. Lab. Woods Hole 140: 461-481. Ngrrevang, A. and K. G. Wingstrand, 1970. On the occurrence and structure of choanocyte-like cells in some Echinoderms. Acta Zool. 51: 249-270. Peng, R. K, and D. C. Williams, 1973. Partial purification and some enzymatic properties of proteolytic enzyme fractions isolated from Pisaster ochraceus pyloric caeca. Comp. Biochem. Physiol. 44B: 1207-1217. _ Péquignat, C. E., 1970. Biologie des Echinocardium cordatum (Pen.) de la Baie de Seine. Nouvelles recherches sur la digestion et l’absorption chez les Echinides et les Stellérides. Forma Functio 2: 121-168. Reid, J. D. and D. Taylor, 1964. An improved method for embedding tissues, using polyethylene glycols, with incorporated low-viscosity nitrocellulose. Amer. 7. Clin. Pathol. 41: 513-516. MICHEL JANGOUX Reynolds, E., 1958. The use of lead citrate at high pH as an electron opaque stain in electron microscopy. 7. Cell Bio f 17: 208-213. Sawano, E., 1936. Proteolytic enzymes in the starfish Distolasterias nipon (Déderlein). Sci. Rep. Tokyo Bun. Daig. B2: 179-199. Smith, J. E., 1937. On the nervous system of the starfish Marthasterias glacialis (L.). Phil. Trans roy. Soc. Lond. B 227: 111-173. 3. THE REPRODUCTION OF SOME ECHINODERMS FROM MACQUARIE ISLAND R. D. SIMPSON Department of Zoology, University of New England, Armidale, NSW, Australia SUMMARY Three species of starfish, Anasterias directa (Koehler), Anasterias mawsoni (Koehler), and Cycethra macquariensis Koehler, and one species of holothuroid, Pseudopsolus macquariensis (Dendy) were collected over a period of one year from rocky sub-littoral shores at Macquarie Island. Study of preserved collections showed that both Anasterias species and the holothuroid have a brooding mode of reproduction and distinct reproductive cycles. For Cycethra macaquariensis no cyclic reproductive pattern was evident. Ovarian condition in Cycethra macquariensis suggested a non- pelagic development; however the mode of this development was not observed. In addition, examination of some specimens of Pseudocnus laevigatus (Verrill), Trachythyone macphersonae Pawson (holothuroids) and Pseudechinus novaezealandiae (Mortensen) (echinoid) showed a brooding habit in Pseudocnus and ovarian conditions that suggested non-pelagic larval development for Trachythyone and pelagic larval development for Pseudechinus. INTRODUCTION Apart from prerequisite taxonomic studies, collections of marine invertebrates in the sub-Antarctic regions have been examined mainly for zoogeographical interpretation and reports on general ecology. Collections have usually been made in summer months and records of reproduction have been restricted to descriptions of non-pelagic development, especially via brooding or ovoviviparity. Simpson (1977) lists sources for the above studies. At Macquarie Island, systematic collections were made each month for one year of some echinoderms from littoral and sub-littoral zones on rocky shores, in order to determine reproductive cycles and to categorise the mode of larval development as either pelagic or non-pelagic. Apart from some nearby rocky outcrops, Macquarie Island (54° 38’ S; 158° 53’ E; Fig. 1) is isolated in the Southern Ocean. For studies of marine invertebrates, Macquarie Island is important in that (a) it marks the limit of southerly 1 ice-free littoral zones and (b) its oceanic isolation bridges a geographic gap in any comparisons of littoral invertebrates over all southern latitudes in the Australian region. MATERIALS AND METHODS Specimens of the asteroids Anasterias mawsoni (Koehler) and Anasterias directa (Koehler), and the holothuroid Pseudopsolus macquariensis (Dendy) were collected at approximately monthly intervals between March, 1968 and March, 1969. Another asteroid, Cycethra macquariensis Koehler, was collected monthly over the same period but collections were not obtained in the months of May, June and November. Within the size category designated for each species (see later), the first five specimens of each sex were examined from each monthly collection. Four specimens of the holothuroid Pseudocnus laevigatus (Verrill), ten specimens of Australian Museum Memoir No, 16, 1982, 39-52. 40 R. D. SIMPSON 60 2 9 INDIAN OCEAN KERGUELEN Dt ire ire CROZETS.,. Yi 2 23 0° 120° x zg S a < Be (6) A ES PRINCE EDWARD eu “MARION nt ; \ MACQUARIE 1 CAMPBELL, *AUERLAND arm BALLENYS ie s : 3 ATLANTIC i OCEAN i \ WEDDELL SEA f \~ sicny ie SU uCeOnGTa 4 PENINSULA Stet. 2295 may Ke FALKLANDS& “ty “S PACIFIC a ry OCEAN 8 8 SOUTH AMERICA 60° 90° 120° Fig. 1. The sub-Antarctic and Antarctic regions. EXPOSED SHELTERED LITTORAL ZONE E.H.W.S. EL.W.S. + Fig. 2. Zonation of Macquarie Island rocky shores correlated with a universal scheme (from Simpson, 1976(a)). E.H.W.S.: extreme high-water spring tidal mark; E.L.W.S.: extreme low-water spring tidal mark. REPRODUCTION OF SUBANTARCTIC ECHINODERMS 4] Trachythyone macphersonae Pawson and four specimens of the echinoid Pseudechinus novaezealandiae (Mortensen) were examined for mode of reproduction only. All specimens were preserved immediately after collection and the material was examined approximately one year after collecting ceased. The methods of preservation were those described in Simpson (1977). That paper also outlines the rationale, method, and terms used in describing the reproductive condition of a large collection of preserved specimens. The same procedures were applied, as appropriate, to the echinoderms. Briefly, for those species collected over a period of one year, reproductive conditions of specimens are described by egg sizes, the state of gonads and of broods, and stages in spermatogenesis determined by microscopic examination of smears of testes. An individual was labelled as being in a particular reproductive stage which, in most cases, was clear. When there was overlap (e.g. for spermatogenesis) the most predominant state was assigned as the stage for that individual. For the brooding species, progression in a brood was described by classification into eggs, embryos, and juveniles: eggs — when no embryonic differentiation was evident; embryos — when eggs showed embryonic differentiation; juveniles — when embryos appeared as fully formed juveniles, The position of collecting sites was maintained by reference to the shore zonation scheme, as defined by Simpson (1976a), (see fig. 2). Anasterias mawsont, Anasterias directa, Cycethra macquariensis, and Pseudopsolus macquariensis had a vertical range from the kelp zone of Durvillea antarctica (Chamisso) Hariot holdfasts in the lower eulittoral zone, down to a depth of 10 metres — the maximum depth investigated. Extension of the range of the above echinoderms up into the kelp zone depends on the cover provided by the living kelp (Simpson 1976a). While Durvillea antarctica itself could be classed as a lower eulittoral species, the zone it creates justifies a biological classification into a “‘sublittoral fringe’. Both Anasterias had average densities on rocky surfaces of one per m? in the upper sub-littoral zone and two per m? in deeper water (Simpson, 1976b). Pseudopsolus macquariensis was often found in patches of high density on rocky surfaces. These patches were more common in the sub-littoral zone, and immediately below it, than in deeper water. Cycethra macquariensis was not abundant in any of the areas investigated. RESULTS Anasterias mawsoni (Koehler) Anasterias mawsoni is endemic to Macquarie Island. It is a six-armed starfish and has a number of colour forms, any one specimen having a single colour. Distinctive features of this species are outlined in Clark (1962). A. mawsoni was found on solid, rocky substrata. Specimens were collected from channels, gutters and pools at the top of the sub-littoral zone. All animals used for reproductive investigations had a central disc diameter of at least 20 mm. The sexes are separate; out of 157 specimens sexed, 61 were males and 96 were females. A pair of gonads was situated in each of the six interbrachial regions. In the males, the testes had a botryoidal appearance while in the females each ovary consisted of two compact, round sacs. When ripe, the testes greatly enlarged and extended down into the arms; in the females, the sacs expanded to accommodate the enlarging eggs. Figure 3 shows the annual reproductive cycle of A. mazvsoni. The female brooded the young from egg to juvenile stage, the brood forming a compact cluster overlying the oral region (see fig. 4). Females assumed a distinctively arched posture when carrying a brood. The central disc was raised, the proximal parts of the arms being at a steep angle to the substratum and the distal parts horizontal and still attached to the substratum. This created a protected cavity at the oral R. D. SIMPSON 42 ‘93k1S J]TUZAN( 1B POOIg YIM tuosmou spUaIsDUY “p “317 raw oz (Cuonelaid13iut yensta Iq 1OJ sINBIy IY] UO poles se JOQWIAS YORI JO Slas 3YL, “31818 aANonpoidas JueUTWMOpaid sit Aq palfissejo Us9q sey yorym Jo peuosd oy ‘uswidads auo siuasaidas yoquiAs yorq) “SuIgIOSAI = W ‘SUNSII = ‘stsouasolvmsads Ajivo = gy ‘spneutiads yo suepunge = Q¢) ‘adi = @ ‘sutumeds iuada1 = X ‘yquour yowa ¢ = NX ‘soyeur -q SAreAo ul sddo = gm “poolq ul s839 = Q ‘poo1g ul soAIquIa = 7 ‘poorq ul sayiuaan! = Ww ‘gsvajai aftuaan! JuUada1 JO JDUapIAa = KX SyUOW YoRa ¢ = N “sa[eUlay “Be “MOSMDU SHLIAISDUY JO SaJIAO aaToNpoAday *¢ “SIT 6961 SHLNOW 8961 Ww J it Q N 0) 3 v f ic Ww Vv vv voy v vv v v oo o o fel fel {i o06UG o o s a 6 s s s ] oe) Oo) 6. 3 oO 0 ° {oe 5) ° ° ee ie e e e ee @ e ee e e x Kai X= EX x q 6961 SHLNOW 9961 ZZ OZ ot mH te 90 70 t (ww) $993 40 YSLSWVIC 3OVYSAV ATHLNOW Ss B es s os a es Ss eB se s 6s Ss see ss a es s @ as a s ea 6 se ss @ so sa @ s ses 8 so ese e a °o.|.U0°0 ° ° fo] ° fe} Vv v Md vey oy : x eee x x x dD REPRODUCTION OF SUBANTARCTIC ECHINODERMS 43 region. Larger specimens could form a more spacious, protected cavity, and hence could presumably accommodate a larger brood. The number of eggs per brood varied from 168 to 296. In July, the eggs in the ovaries averaged 0.4 mm in diameter. The eggs progressively increased in size, reaching 2.0 mm in December. They were transferred to the brood clusters in January-February at which time the diameter of eggs in a brood averaged 2.2 mm. Development proceeded in the brood toa juvenile stage in May and June when the juveniles were released. A female was classed in a “juvenile release” stage either when a very reduced brood was at the oral region or when there was no brood but there were membranous remnants of the brood sac around the oral region. In males, the testes were generally reduced from February to May. There was some spermatogenic activity during this period, but occurrences were few. The testes then progressively increased in size and large, ripe testes were common from September to November. In December and January, they were reduced and mainly consisted of mature spermatozoa, indicating that spawning had recently occurred, The monthly reproductive conditions of the testes suggested that the breeding season occurred in early summer. During this period, the eggs were being transferred to the brood clusters. Anasterias directa (Koehler) Anasterias directa is endemic to Macquarie Island. It is a five-armed starfish and has a predominant dark olive colour form. Distinguishing features of the species are outlined in Clark (1962). A. directa was found on solid, rocky substrata. Specimens were collected from channels, gutters, and pools at the top of the sub-littoral zone. All starfish used for reproductive investigations had a central disc diameter of at least 19 mm. The sexes are separate. Out of 128 specimens sexed, 62 were females and 66 were males. A pair of gonads was situated in each of the five interbrachial regions. In the males, the testes had a botryoidal appearance, while in the females each ovary consisted of two compact round sacs. When ripe, the testes greatly enlarged and extended down into the arms. In the females, the sacs expanded to accommodate the enlarging eggs. Figure 5 shows the annual reproductive cycle for A. directa. The female brooded the young from the egg to the juvenile stage, the brood forming a compact cluster overlying the oral region (see fig. 6). Small eggs were present in the ovaries in April and May, the average diameter in April being 0.5 mm. The Fernerst of the eggs increased to 1.1 mm in early June and progressively increased until they were transferred to a brood at the oral region in July, at which time diameter of the eggs ranged from 1.8 to 2.0 mm. Development of the embryos proceeded in the brood to a juvenile stage (October-November). The juveniles were released in the November-December period. Again, a female was classed in a “juvenile release’’ stage using the same criteria as for A. mawsoni. In January, the ovaries were small and the average egg size was 0.3 mm. The eggs progressively enlarged to an average diameter of 0.6 mm in mid-March. In the specimens examined, the number of eggs per brood varied from 174 to 220, depending on the size of the starfish. Females exhibited a distinctively arched posture when carrying a brood (like that of A. mawsoni). Again, larger starfish were able to create a larger cavity and hence could accommodate a larger brood. In males, during the corresponding period of brooding in the females, the testes were generally reduced in size and showed little signs of spermatogenic activity. Growth and spermatogenic activity were evident from November to February. Ripe testes were predominant in late February and March. They were present in April and May of the preceding year, but at “DIIEAIp SDUAISDUY JO ‘98eIS a[TUSANI 1k “poolg “9 “3Iy 6961 SHLNOW 8961 WwW Jd f a N (e) Ss Vv if if W Vv T a al ae | ‘ce! Vv Vv v Vv v v v v fa) o fa] G oH o fa] o oo fa] oO oa o a oO s 8s s8 s a Z 2 9 (re) ° ee) mo ° = ° ° et ee e e e 2 n ee : ee Q e . x x x x ~ x x x x x x 6961 SHLNOW 9961 WwW Jd f ie} N 0 Ss v iP cr WwW Vv Fee Sle ee ee ele Ae eo ele 90 7-0 £0 6+ 64 M4 40 $0 2 (Ww) S993 30 H3L3WVIC 39VY3SAV ATHLNOW se a s se s au s 8 8 ee s s s 8 s ee a es s 8 se a s 8 8 se eS s ° o oO ° ° ‘(uoneiaidsoqut yensta i A 1919q JOF IINSTy sy] UO patyNess ore [oquIAs YoRa JO slas sy], ° “1e1s aANONpoidas 1uvUTWOpaid sit Aq parfisse[o usaq sey YY. ‘ Vv YoIyM Jo peuosd ayi ‘uawoads auo siuasaidai oquiAs yoeq) v ek “SUIqIOSAaI = W ‘SuNSsaI = [] ‘stsouasoleurtods Ajieo = v an jo aourpunge = () ‘adi = @ ‘Sutumeds juasa1 = x. Mi M syuoul ae S = N ‘soyeur -q Are. AO UT S339 = gy ‘pooiq v = $3839 = GQ ‘pooiq ul sOAIquIa = V7 ‘poolq UT SsotIuaAnl = me ES Vv ‘asevayar aytuaani{ 1uada1 Jo aduapIAd = X SylUOU Yea ¢ = a N ‘sayeulay *B “V1Ie41Ip spiuaispUY JO 39D DANINpoIday *¢ “B1y dD REPRODUCTION OF SUBANTARCTIC ECHINODERMS 45 this time testes were mainly reduced and consisted of mature spermatozoa with little signs of spermatogenic activity, indicating that spawning had recently occurred. The monthly reproductive conditions of the testes suggested a breeding season in the autumn. Yet, in this period, the eggs were still mainly held in the ovaries. From this pattern it appears that male spawning was too early if the eggs were to be fertilized during the seemingly opportune time of transfer to the brood region. Fertilization may occur in some other way. However, it is more likely that the above is a discrepancy in the pattern owing to either (a) a bias resulting from the small number of specimens used or (b) an ecologically-induced difference, that is, males being taken from habitats different to those of females in the first part of the collecting period (April-June). Cycethra macquariensis Koehler The nomenclature for this species is not clear. The collected specimens were firstly identified as Asterina hamiltoni Koehler. However, on examination of type specimens held by The Australian Museum, Sydney, the specimens from Macquarie Island were then identified as Cycethra macquariensis in the family Ganeriidae. Yet, on the basis of the arrangement of plates and spines, the specimens could be identified as a genus in the family Asterinidae, in accordance with the key in Clark (1962). Clark (1962) remarked that the separation between Cycethra and Asterina is indistinct. Cycethra macquariensis is endemic to Macquarie Island. It is a small, five-armed starfish. Specimens were collected from pools and channels in the sub-littoral zone. All animals used for reproductive studies were at least 22 mm in total body diameter. The sexes are separate. Out of 141 specimens sexed, 96 were male and 45 were female. A pair of gonads was situated in each of the five interbrachial regions. The testes had a botryoidal appearance and when ripe, were large and extended into the arms. Each ovary consisted of a number of small sacs (typically seven) and each sac contained eggs of various sizes. The number of eggs per sac varied from 19 to 24, the average being 20. There were three distinct size categories (diameter): (a) less than 0.3 mm, (b) 0.3 to 0.49 mm, and (c) 0.5 to 0.8 mm. The smallest eggs were more plentiful than the largest. In grouping the figures from all specimens, ee of numbers in the three size categories was (a) 11 to 16 (x = 13), (b) 4 to7 (X = 5), (c) 1 fone x= "2). The above condition of the ovaries was maintained throughout the collections. About 40% of the males had ripe testes in any one month. Thus C. macquariensis appeared to breed continuously throughout the year. It is not clear whether females released ova to the sea for external fertilization. The average number of eggs per female is 20 x7 x 10 = 1,400 (number of eggs per sac x number of sacs per gonad x number of gonads). However, from the size range of the eggs, it was apparent that only about 140 of these were mature ova. The release of such a number would hardly ensure successful fertilization or survival rate if eggs were released to the open sea. No brood was found either enclosed in the body cavity or on the surface of the starfish. From the above evidence, it appears reasonable to assume that Cycethra macquariensis lays egg cases. Pseudopsolus macquariensis (Dendy) Pseudopsolus macquariensis is endemic to Macquarie Island. A description of the species is given by Pawson (1968). Specimens were collected from channels and gutters encrusted with coralline algae at the top of the sub-littoral zone. Adult holothuroids were commonly 25 mm long from the base of the tentacles to the anus. For plotting the reproductive pattern, specimens of at least 15 mm were used. 46 R. D. SIMPSON The gonad consists of a cluster of unbranched caeca uniting at acommon base in mid-dorsal mesentery, approximately one quarter body length from the anterior end of the animal. From the base, a single genital duct passes To an opening in the oral disc. Ludwig (1898) grouped the caeca into left and right tufts, presumably using the mesentery wall as the divider. The species has previously been described as hermaphroditic (Ludwig, 1898; Mortensen, 1925) but this requires further examination. For a maturing gonad, there are a number of large caeca that contain the developing gametogenic material which, for each animal, is either male or female — not both. (Any further designation here of an individual as male or female has used the sexual status of the large caeca as the criterion). The contentious point is the role of the smaller genital caeca at the base of the cluster. Ludwig (1898) described one specimen, which was obviously a male with a ripe gonad, and noted much smaller caeca containing “‘small eggs” 0.23 to 0.28 mm in diameter. In a description of two specimens, Mortensen (1925) reported a condition similar to that found by Ludwig in one and, in the other (a small female), he observed ovarian caeca each with two eggs (1.5 mm diameter) and some smaller caeca which he classed into both male and female types. In the present study, 18] specimens (measuring 15 mm or more from anus to the base of the tentacles) were sexed as 102 males and 79 females. This bias in sex ratio was greater in a collection of 29 smaller specimens that were 12 to 15 mm in length (20 males, 8 females, 1 immature). Further observations on the 181 larger specimens showed that the small genital caeca were divisible into two groups: (1) very small buds and (2) caeca 1 to 3 mmin length which nearly aways contained egg-like forms. Four male specimens had small non-differentiated caeca containing spermatogenic material. Three specimens had developing ovarian caeca and thin, male caeca that were in a regressed stage. Two females with broods had genital caeca containing egg-like forms in some and, in others, developing testes arranged ina bead-like pattern down the caeca. These last five specimens confirmed the occurrence of successive hermaphroditism as reported by Ludwig (1898) and Mortensen (1925). However, the deductions on this point by Ludwig and Mortensen were ill-founded in using the presence of smaller genital caeca with egg-like forms to indicate that the next sexual role would be female. Such caeca appear to contain precursor material for either male or female formation. Histological work on specimens in key reproductive stages is required for further elucidation of such a process. It is unlikely that each individual would change sex each year with a resultant switch in the sexual bias. This was confirmed by a similar state of sexual bias showing at both the start and end of the period of collection, which covered two successive reproductive cycles (see later). Thus, some individuals must retain the same sex in successive cycles. If sex change is an option for an animal, there is the speculative hypothesis that the number of females increases in response to some form of feedback that favours a population increase, and vice versa. In a description of 8 specimens (11 to 19 mm in length) of a new sub-species, Pseudopsolus macquariensis gruai, from the Kerguelen Islands, Cherbonnier (1974) found gonads with 4 to 5 thick, white, male caeca and “about ten” very fine, long tubules with a “few eggs” less than a micron in diameter. Such a gonadal condition was not found in specimens of Pseudopsolus macquariensis in the present study. The number of large caeca varied from 12 to 24 in both males and females. The caeca were divided into two groups by mid-dorsal mesentery usually with half the number in each group. Smaller caeca and buds at the base of the gonadal cluster were more numerous. When sexually ripe, large male caeca were looped and the total length of some exceeded the body length of the animal. In females, the number of eggs in caeca varied from 73 to 154 and numbers of eggs/embryos in a brood varied from 61 to 130. These figures apply to specimens over 15 mm from anus to the base of the tentacles, and the number of eggs was related to the number of ' REPRODUCTION OF SUBANTARCTIC ECHINODERMS 47 a x a a a a a a Ae: ° ° ° O° oO ° fo} fe) es a os Ss se se ) Ss 8 s a a eo 8 a r r tl] 1] | a @ | es ss a a a a es a a as o a a a a 68 MONTHLY AVERAGE DIAMETER OF EGGS (mm) : 1300 4548 08 10 4 ans eae Hes |e [SP ie A M J J A S 0 N D J F M 1968 MONTHS. 1969 b x x x xX x x x e e : ° e e ® e e e ej; . ° ° fo} ° fe} ° ie} ° ° ° fo} ° ° a os a a a a a a Ss oO Oo Gr Oo Oo ‘aly 12) o OP i oO Oo A a a 4 a 4 | | | | | | [em 23) ee ete | A M 4 J A iS) (0) N D J te M 1968 MONTHS. 1969 Fig. 7. Reproductive cycle of Pseudopsolus macquariensis. a. females, N = 5 each month (except October); X = no brood found in forty specimens — all juveniles released, A = juveniles in brood, A = embryos in brood, O = eggs in brood, = eggs in ovary; b. males, N = 5 each month; X = recent spawning, @ = ripe, O = abundance of spermatids, ll = early spermatogenesis, [1 = regressed, A = resorbing. (Each symbol represents one specimen (except for females in October), the gonad of which has been classified by its predominant reproductive state. The sets of each symbol are stratified on the figure for better visual interpretation.) tubules and animal size; that is, the more tubules and/or the larger the animal, the more eggs. Eight females of 12 to 15 mm in length (in the above dimension) had egg numbers of 25 to 52 in the large caeca. Figure 7 shows the annual reproductive cycle for Pseudopsolus macquariensis. Eggs were developing in the ovaries from November to June. In December, the eggs (averaging 0.8 mm diameter) were distinctly separate down the length of the caeca, resembling a string of beads. The size of the eggs progressively increased in the following months and in May, their diameter was 1.5 mm. In May and June, eggs appeared in internal brood sacs and at this stage were 1.8 mm in diameter. There was negligible variation in the size of eggs in an individual at any particular phase of the cycle. In the first week of July, all specimens assuming the female role had eggs in the internal brood sacs. There was no evidence of internal connection to these sacs. External transference was not observed nor did any specimen collected show an intermediate 48 R. D. SIMPSON : : phase, with some eggs in caeca and some in incubatory sacs. However, two females were found! with a small stone in a brood sac that also housed eggs. This suggested insertion of eggs into they brood sacs from the outside. Further development proceeded in the brood sacs, the September samples showingy advanced embryos and juveniles. Release of juveniles occurred in late September-October, There was a high degree of synchronization during juvenile release. On 22nd October, juveniles were found underneath adults and large numbers of adults were collected on this date. Forty, specimens from this collection were dissected and not one contained any juveniles. The resting period of the testes was from July to September. Early spermatogenesis wag predominant in October. Spermatogenesis and growth continued progressively with large, ripey testes predominating in March-April. In May and June, testes were in an obvious post-spawning condition, i.e. reduced, not firm, and full of mature spermatozoa with little spermatogeni¢ activity. The cycle in the testes indicated that the breeding season occurred in May-June during the transfer period of the eggs from gonadal caeca to incubatory sacs. If the eggs are transferred) externally, as findings here suggest, fertilization may depend on the coincidental timing off transfer and male spawning. Two large, often convoluted, protuberances appeared on the ventral surface of many, individuals, about half-way down the body (fig. 8). At first it was thought that these were: “brood pockets” from which the young were released as the bulges coincided with the openings to the incubatory sacs. However, the protuberances were present in specimens assuming both) male and female roles, The number of specimens with these growths increased during they juvenile-release phase of the cycle but with no bias to those acting as females. Histological sections of these areas showed that the extra growth was largely a result of increased connective. tissue. The significance of these protuberances remains unexplained. The incubatory sacs were deeply internal and not surface pockets. There were two sacs, , divided into compartments, situated in the left and right ventral interambulacral areas. Each sac) had a ventral opening via a single duct. Figure 9 shows a brood sac with eggs and a transverse section through the ventral duct. The opening was a simple hole, half-way down the ventral body wall and coinciding with the previously described protuberances (if they were present), Juveniles were released through these ventral pores and moved out from underneath the parent on to the surrounding rock surfaces. The walls of the sacs were transparent and of light texture. ‘ Pseudocnus laevigatus (Verrill) Only four specimens were collected from rock pools in the lower eulittoral zone. The largest individual (35 mm in length, from the base of the tentacles to the anus) contained 93 young i internal brood pouches. The brooding habit in this species has been previously reported b Pawson (1968). Trachythyone macphersonae Pawson Ten specimens of this species were collected from rock pools in the lower eulittoral zone. The sexes are separate. There were no small gonadal caeca opposite in sex to large caeca as in Pseudopsolus macquariensis. The females had a small number (110 to 130) of eggs in long unbranched ovarian caeca; these eggs were large, diameters ranging from 0.2 to 0.8 mm witha _ fairly uniform size in any one specimen. The size of the egg appears to depend on the stage in the reproductive cycle. However, the small number and the 0.8 mm size eggs indicated that thi species does not have a planktonic larval stage, most likely developing via a brooding habit, REPRODUCTION OF SUBANTARCTIC ECHINODERMS Fig. 8. Protuberances on the ventral surface of Pseudopsolus macquantensts, (p. = protuberances). mm ie Fig. 9. Side view of brood sac and ventral duct of Pseudopsolus macquariensis. b. = ventral body wall, c. = compartments in the brood sac, d. = ventral duct, e. = eggs. 49 30 R. D. SIMPSON Pseudechinus noyaezealandiae (Mortensen) Four specimens of P. novaezealandiae were examined. Two were from collections during the present study (one from the sub-littoral zone, the other from a diving station at a depth of seven metres) and two were obtained from the National Museum of Victoria. As in other regular echinoids, there were five gonads suspended by mesenteries along the inner surface of the interambulacra. Three of the specimens were males, the other female. In the female, the ovaries were large and projected well down ventrally, filling a large part of the available space. The ovaries contained numerous eggs (diameter = 0.1 mm) suggesting that this echinoid has a planktonic larval stage. DISCUSSION : For the seven species investigated, Anasterias mawsoni, Anasterias directa, Pseudopsolus macquariensis and Pseudocnus laevigatus were found to have a brooding habit. Ovarian conditions - suggested non-pelagic development for Trachythyone macphersonae (most likely via brooding) and for Cycethra macquariensis (most likely via the laying of egg masses), and a pelagic larval development for Pseudechinus novaezealandiae. These predictions of larval development for the latter three species are based on the number and size of the ova. Invertebrate species with a non-pelagic larval development characteristically have a small number and large size of eggs (Thorson, 1950 and Mileikovsky, 1971.), The suitability of this type of prediction is discussed elsewhere (Simpson, 1977) as well as the less positive nature of the converse — that a large number of small eggs indicates a pelagic development. | For the four species collected at regular intervals, both Anasterias and Pseudopsolus | macquariensis were found to have an annual reproductive cycle. Cycethra macquariensis appeared to breed continuously throughout the year. From the respective numbers examined, the sex ratios varied among the four species, that is female: male was approximately 1.5:1 (Anasterias — mawsoni), 1:1 (Anasterias directa), 0.5:1 (Cycetkra macquariensis), and 0.8:1 (Pseudopsolus , macquariensis). However, no attempt was made to randomize collecting localities and hence the - rauios may be affected by any sexual bias in distribution. The reproductive condition found in Cycethra macquariensis is atypical for starfish. They generally have a well defined short breeding season, usually as part of an annual reproductive cycle (Boolootian, 1966). In a list of breeding seasons of asteroids compiled by Boolootian (1966), the two exceptions to this rule were in the family Asterinidae (Patiria miniata and | Asterina exigua) which were reported as breeding continuously throughout the year. Lawson-Kerr and Anderson (1978) confirmed that Patiriella exigua (named Asterina exigua | by Mortensen (1921) in Boolootian’s list) was potentially capable of breeding at any time of the year. A further species reported as being capable of breeding throughout the year is also in the family Asterinidae — Patiriella vivipara (Dartnall, 1969). ! Although some species of asterinid starfish have limited breeding seasons, (Boolootian, 1966; Lawson-Kerr and Anderson, 1978; Komatsu, pers. comm.), it is curious that all starfish _ reported as capable of continuous breeding are in the Asterinidae. To elucidate the reasons for _ the reproductive strategies of marine invertebrates, many factors require examination, One of | these factors is phylogenetic affinity for a particular reproductive type. It would seem that asterinid starfish present appropriate material for Investigating such a factor. Anasterias mawsoni and Anasterias directa had the same distributional range, occupied similar habitats and had very similar prey (Simpson, 1976b); yet their reproductive cycles were — different in that there was a time difference of four months between the peak release of young REPRODUCTION OF SUBANTARCTIC ECHINODERMS 51 nd hence recruitment of the two species. This may be important in alleviating any competition vetween the recruitment of the two species into such similar ecological niches. For the species collected, there is a predominance of a protective mode of larval levelopment with a comparatively small number of offspring. This conforms with ‘“Thorson’s ule” that there is an increase in frequency of non-pelagic development with increasing latitude. (he adaptive significance of this phenomenon has been the subject of much speculation and any hypotheses, which are linked to changing physical and biotic ecological conditions with |ncreasing latitude (Thorson, 1950; Mileikovsky, 1971; Menge, 1975). The sub-Antarctic provides an important link in records of reproduction of marine nyertebrates from tropical and temperate regions to the Antarctic. Gathering of data on the node of reproduction of invertebrates from selected groups across the range of southern atitudes should result in better interpretation of zoogeographical origins and lines of istribution. This will provide a more complete historical background to possible investigations yf adaptational advantages to be gained by animals with a protective mode of development in igher latitudes. Details of the timing of the events, when reproductive patterns of such animals san be obtained, will allow insight into key areas for such investigations. ACKNOWLEDGEMENTS I should like to thank Ms. I. Bennett for her advice on the known life histories of »chinoderm species at Macquarie Island prior to the expedition. Mr S. Harris greatly assisted in collecting specimens. The collections were made possible by the support of the Antarctic ivision, Department of Science, Australian Government. I am indebted to Professor J. S. earse for helpful comments on some of the data. REFERENCES 30olootian, R. A., 1966. Reproductive Physiology. In R. A. Boolootian (ed.), Physiology of the Echinodermata: 561-613, 4 tables, 37 figs. John Wiley and Sons, New York. sherbonnier, G., 1974. Holothurides et Echinides. Jn Invertebres de Vinfralittoral rocheux dans |’Archipel de Kerguelen. Vol. III: 27-31, 1 fig. Com. natn. fr. Rech. Antarct. No. 35, Jlark, A. M., 1962. Asteroidea. Rept. B.A.N.Z. antarct. Res. Exped., Ser B, 9: 104 pp, 14 tables, 5 plates, 18 figs. Jartnall, A. J., 1969. A viviparous species of Patiriella (Asteroidea, Asterinidae). Proc. Linn. Soc. N.S.W. 96: 294-296, I table, 1 plate, 1 fig. -awson-Kerr, C. and D. T. Anderson, 1978, Reproduction, spawning and development of the starfish Patiniella exigua (Lamarck) (Asteroidea: Asterinidae) and some comparisons with P, calcar (Lamarck), Aust. 7. mar. freshw. Res. 29: 45-53, 1 table, 13 figs. -udwig, H., 1898. Holothurien. Wiss. Ergenbnisse Hamberger Magalhaensische Sammelreiss 3: 98 pp, 3 plates. Menge, B. A., 1975. Brood or broadcast? The adaptive significance of different reproductive strategies in the two intertidal sea stars Leptasterias hexactis and Pisaster ochraceus. Mar, Biol. 31: 87-100, 7 tables, 7 figs. |Mileikovsky, S. A., 1971. Types of larval development in marine bottom invertebrates, their distribtution and ecological significance: a re-evaluation. Mar. Biol. 10: 193-213, 4 tables. Mortensen, Th., 1921. Studies on the development and larval forms of echinoderms. G.E.C. Gad Copenhagen: 261 pp. 1925. Echinoderms of New Zealand and the Auckland-Campbell Islands. IJ-V: Asteroidea, Holothuroidea and Crinoidea. Vidensk Meddr dansk. naturh. foren. 79: 261-420, figs. 1-70, pls. 12-14. }Pawson, D. L., 1968. Some holothurians from Macquarie Island. Trans. R. Soc. N.Z. (Zool.) 10: 141-150, 13 figs. 52 R. D. SIMPSON Simpson, R. D., 1976a. The shore environment of Macquarie Island. A.N.A.R.E. Rep. (Ser. B.I) No. 125: 1-41, 3: tables, 10 figs. 1976b. Physical and biotic factors limiting the distribution and abundance of littoral molluscs on Macquarie Island (sub-Antarctic). 7. exp. mar. Biol. Ecol. 21: 11-49, 16 tables, 7 figs. 1977. The reproduction of some littoral molluscs from Macquarie Island. Mar. Biol. 44: 125-142, 1 table, 14 figs. Thorson, G., 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol. Rev. 25: 1-45, 6 figs. 4. SEXUAL AND ASEXUAL REPRODUCTION OF HOLOTHURIA ATRA JAEGER AT HERON ISLAND REEF, GREAT BARRIER REEF VICKI HARRIOTT* Zoology Department University of Queensland St. Lucia, Queensland Australia SUMMARY H. atra is the most common epifaunal holothurian on the reef flat at Heron Island. The gonad maturation cycle of H. atra was followed using gonad index estimations and histological examination. Gonad samples were taken at intervals of 6 to 8 weeks for 18 months, Mature gonads occurred in most samples, and gonad maturity peaked twice; in early winter and in summer. Sex ratio of female to male animals was not significantly different from a 1:2 ratio. Sex ratio ranged from 1:8.5 in animals weighing less than 100 g, to 1:0.7 in those over 1,000 g. H. atra commonly reproduces asexually by transverse binary fission. In 21 samples, each of approximately 50 animals, 6% to 70% of individuals were detectable products of asexual reproduction. Occurrence of frequent asexual reproduction compounds difficulties in estimation of growth parameters from data such as size-frequency distributions and growth increments. INTRODUCTION Holothurians are amongst the most common coral reef macro-invertebrates, but little information is available on their reproductive biology (Bakus, 1973). An understanding of reef systems requires data on the population dynamics and patterns of recruitment of these, and many other reef species. Temperate holothurians, in common with other temperate marine invertebrates, generally spawn for a limited period during spring or summer (Boolootian, 1966). Tropical species, however, exhibit a variety of spawning patterns. Holothurians are also known to reproduce asexually, by transverse binary fission (Hyman, 1955; Bonham and Held, 1963). Holothuria atra Jaeger is widely distributed in the Indo- West Pacific region, and is the most common epifaunal holothurian on the reef flat at Heron Island. In this habitat, it is generally found on sandy substrata. Pearse (1968) studied sexual reproduction of H. atra at several low latitude sites in the tropical Indo-Pacific. Because individuals with mature gonads were found throughout the year, he concluded that spawning was asynchronous. He predicted that populations distant from the equator would have more restricted spawning periods. Bonham and Held (1963) reported asexual reproduction by fission in H. atra at Rongelap Atoll, Marshall Islands, and suggested that fission occurred commonly. Ebert (1978) interpreted the apparently high rate of asexual reproduction in H, atra at Enewetak Atoll, as an adaptation enabling the species to span periods of unsuccessful recruitment from the sexual phase. The relative frequency of recruitment from sexual and asexual modes of reproduction is a potentially important life history parameter. This paper reports on sexual and asexual reproduction in H. atra at different sites on Heron Island reef, in the Capricorn Group, at the southern extremity of the Great Barrier Reef (Lat. 23° 27’ S). *Present address: Dept. Marine Biology, James Cook University, Townsville Queensland. Australian Museum Memoir No. 16, 1982, 53-66. 54 VICKI HARRIOTT Fig. 1. Heron Island reef showing sample sites. A = shallow lagoon; B = S. W. reef flat, gutter; C= S. Wi reef flat, mid-reef; D = S. W. reef flat, crest; E = N. E. reef flat, inshore; F = N. E. reef flat, rubble crest. REPRODUCTION IN H. ATRA a | | METHODS Sites sampled on Heron Reef during the course of the study are shown in Figure 1. 1. SEXUAL REPRODUCTION The reproduction cycle of H. atra was studied over 18 months, from December, 1976 to June, 1978. Gonads were sampled at intervals of 6 to 8 weeks during this period. Sexual reproducton was studied in 2 ways; (a) by use of gonad index estimation, and (b) by histological examination of excised gonads. (a) Gonad index estimation. For each sample, individuals of H. atra were collected from the lagoon of Heron Island reef (fig. 1; site A). Sample size ranged from 12 to 24, with a mean sample size of 16 individuals. The animals were taken to the laboratory and maintained overnight in aerated aquaria to allow emptying of the guts. Wet weight of the animals was measured to the nearest 5 g, and gonads were exised and weighed to the nearest 0.01 g. The gonad index was calculated as the ratio of wet gonad weight to wet body weight, expressed as a percentage, for each individual. The mean gonad index (+ S.E.) was calculated for each of the 13 samples obtained. (b) Histology of gonads. Excised gonads were preserved and fixed in alcoholic Bouin’s fixative and stored in 70% ethyl alcohol. They were dehydrated, cleared, embedded in paraffin wax, and sectioned at a nominal thickness of 10 pm_ Sections were stained with haematoxylin and eosin. Gonads were assigned to one of three classes of maturation, derived from those recognised by Tanaka (1958) — G) Resting stage and indeterminate gonads. (ii) Recovery stage and growing stage gonads. (iii) Mature stage and shedding stage gonads. The sex ratio was recorded for individuals of a wide range of weights. These animals were collected from several habitats. 2, ASEXUAL REPRODUCTION Samples of 33 to 100 individuals, each weighing less than 90 g, were collected from 6 habitats, A to F (fig. 1). Habitats A, B and C were sampled five times (between August, 1977 and June, 1978), habitat D was sampled four times, and habitats E and F were sampled once only. Each individual was dissected and examined for the following signs of recent binary fission, as recognised by Crozier (1917). An abnormally large or small calcareous ring. A change in colour or texture of the internal body wall at the region of regeneration. The absence or smallness of either anterior or posterior body organs. A narrowing of the five longitudinal muscle bands at the line of division. aoe For each fissioned individual, the degree of regeneration following fission was determined, and scored on a scale of 0-3 by the following criteria — 0. No visible regeneration; very recent division. 1. First signs of regeneration of gut, mouth or anus; very small calcareous ring present (less than 5 mm diameter); body wall regrowth less than | cm in length. . Body organs regenerated but small; body wall regrowth 1-3 cm. . Complete regeneration of body organs; discernable as fission product by narrowing of muscle bands and difference in colour of body wall; body wall regrowth over 3 cm. Wh 56 VICKI HARRIOTT The percentages of fissioned individuals in early stages of regeneration (stages 0 and 1) were plotted against time to indicate changes in the frequency of recent division. 3. POPULATION SIZE STRUCTURE Size-frequency distributions were used to ascertain population size structure in the habitats from which specimens had been examined for asexual reproduction. Samples of 80-300 individuals were weighed in the field to +5 g, and weights plotted to produce size-frequency histograms. RESULTS 1. SEXUAL REPRODUCTION From Figure 2, the plot of gonad index over time, seasonal variations in gonad index are apparent. Gonad index peaked at 1.8%, 2.7% and 1.2% in May, 1977, December, 1977 and May, 1978 respectively. Each peak was followed by a decrease in gonad index to 0.3% and 0.8% and 0.4% respectively, indicating possible spawnings during the months of May/June and December/January. Seasonal changes in histological state of the gonads are recorded in Figure 3. Changes in percentages of mature gonads closely parallel changes in gonad index, indicating that, for this species, gonad index is a good measure of gonad maturity. The percentage of mature gonads peaked at 75% and 100% in May and December, 1977 respectively. Histological data confirm that spawning follows these peaks in gonad maturity. A decrease in the percentage of mature gonads is accompanied by an increase in the percentage of _ histologically deterntined spawned gonads. Spawning occurred later in summer 1977-78 than it did the previous year, when individuals had spawned by mid-December. _ _ Differences in reproductive cycle between male and female individuals were indeterminable because of the small numbers of females in some samples (Table 1). Hermaphroditic gonads, that is, those containing both male and female gonad elements, — were detected in 2 of the 155 individuals examined. ; Sex-ratio for individuals of H. atra from different samples is shown in Table 1. The sex ratios were tested for homogeneity using a X? test, and were found to be sufficiently — homogeneous to permit pooling (p > 0.05). Pooled data were tested, and the ratio was found to differ significantly from 1:1 (X? = 24.25, p < 0.01), but was not sifnificantly different froma 1:2 ratio of females to males (X* = 0.04, p > 0.5). Figure 4 shows the relationship between sex and size (weight), for all individuals examined. Because of small sample size, results have been pooled for the large size classes. The relationship between sex and weight was tested using X? contingency table analysis, and sex was found to be dependent on size class (p < 0.02). In individuals weighing less than 100 g for which gonads are recorded, the ratio of females to males was 1:8.5 (n = 21); while in those individuals over 1,000 g, the ratio was 1:0.7 (n = 11). 2. ASEXUAL REPRODUCTION Table 2 shows the percentages of detectable products of asexual reproduction in each sample, from the habitats indicated, Small individuals, i.e. those less than 100 g were sampled because fission products will eventually regenerate and become unrecognisable. Fission Re “JOIIO Prepurls Jo IUeI MOYs SOUT] [kONIAA “(8/6] PUN[-9/6] Jaquiacaq) v4 “Fy JO xapuT peuos ueaut UI sastiey [euoseas °7 “31 eZ6l Z£L61 BS Q SS © = = cs wo) Z a S TZ e = F < S a eC S e) &. (a4 Ay jea) [a4 VICKI HARRIOTT 58 ‘aseis SUIPpoys pue aimey (11) ‘a8eis Surmos3 = Seen I 1 I pue Ar1sA0d0y (IT) ‘aleuTUTIA]OpUT pue a8eIs BUTISay (I) saseIs 9 I Gory 2[Gures Yea JO S[ENPIAIpuT fo saseyusoiad ay} SuIMOYs “(9/6 T [dy - 9/6] Jequiasaq) nav ‘yy Jo uoneIMeW Ase ul sadueyd eee € 4 ee (1!) (| 35D4s aBnis 36D\s Ot 02 001 REPRODUCTION IN H. ATRA 59 TABLE 1. Abundance of male and female H. atra, sampled from lagoon, unless specified. & 3 December 1976 6 3 January 1977 4 10 March, 1977 2 9 May, 1977 6 10 July, 1977 3 6 August, 1977 4 i October, 1977 0 8 October, 1977 — Crest 6 4 November, 1977 5 16 November, 1977 — Flat 4 6 November, 1977 — Gutter 5 14 December, 1977 5 8 January, 1978 4 12 January, 1978 — Crest 4 + February, 1978 5 10 April, 1978 4 9 202) 66 136 products were rarely detected in the individuals greater than 100 g that were sampled for gonads. Differences in frequency of fission among these samples were tested using X* contingency table analysis, at the 95% confidence level. Several patterns are apparent from the data, e.g.: a. The frequency of asexual reproduction is significantly dependent on habitat for all months, with p(X?) < 0.001, except in February 1978 when p(X’) = 0.006. b. In the lagoon and reef crest, the percentage of fission products is dependent on time p(X2) > 0.005). Figure 5 shows changes in the percentage of recent divisions (stage 0 and 1), over the period sampled, in the gutter habitat (fig. 1, site B). Division is more frequent in the period preceding August 1977 than in the period preceding June 1978. This result is supported by the changes in fission frequency with time in some habitats, and suggests that fission frequency is not constant throughout the year, but is maximal during limited periods. The ratio of former posterior to former anterior ends was approximately 1:1, indicating a similar mortality rate for each section. Fission rate appears to be approximately equal in males and females, although only a small percentage of fissioned individuals examined contained detectable gonads. Of the 38 individuals with visible gonads, 12 were female, 15 were male, and 11 were indeterminate. Although no detailed histological examination was made, gonads examined appeared to include all stages of gonad maturity. 3. POPULATION SIZE STRUCTURE The size-frequency distributions obtained were variable over time and habitat, but frequently conformed with one of two general patterns, unimodal or bimodal (fig. 6). No consistent age-class groups were detectable. Samples trom the inshore gutter area always had a VICKI HARRIOTT 60 ‘(sadins [euozt0Y) s;euUIay pure ‘(sadins [eonsaa) afew sem IBY} SSeS 1Y48I9M Yea Jo adeJUaoIad ay] s1uasaidad aeUIPIO ay], “IYSIaM SUISaIOUT YIIM OFeI Xas UT sasuBY Jo Ydey “p -817 SS 1G ees STAN Oool< O00!-00S O0S-O0E O0€-002 O0¢-08! astoag9! 091-0 7 | O7l-O2! G2t-001 00l > TOE COME COE DOO TE DOO TTT EE LOTT TTT TTT ITE Pa PSTPATSTE ciate loses Col Cie) (Sei ( = (82 ) (OZ ) (te) (IS), (27) (€7) (ke) ee REPRODUCTION IN H. ATRA 61 TABLE 2. Percentages of H. atra samples (sample numbers in brackets) that were discernable products of asexual reproduction in 6 habitats of Heron Island reef from August 1977 to June 1978. HABITAT August October Nov./Dec. February April June 1977 1977 1977 1978 1978 1978 S. W. Reef Flat Gutter 61% (51) 58% (56) 58% (52) 56% (50) 51% (50) Crest 11% (44) 22% (58) 37% (48) 22% (50) 6% (50) Mid-Reef 40% (33) 62% (51) 60% (45) 70% (46) Lagoon 20% (65) 14% (50) 34% (50) 52% (50) 52% (52) N.W. Reef Flat Rubble crest 47% (108) Inshore 50% (50) n=(30) ( 31) ( 30) ( 28) ( 23) ORO Ts 90 80 4 70 6 0 es % 5). (0) 40 T 2310) 710) 10 ) : , : BI A S O N D J ie M A M J J 1978 Fig. 5. Graph of percentage of fissioned individuals that were in early stages of regeneration i.e. stages 0 and 1, sampled from the S. W. reef flat gutter, from August 1977 to June 1978. 62 VICKI HARRIOTT bimodal distribution, those from the reef-crest had a unimodal distribution, and samples trom the lagoon and mid-reef had distributions that varied with time. DISCUSSION H., atra has a semi-annual reproductive cycle at Heron Island reef, and a large proportion of the population breeds in both winter and summer. Until 1966, there was no record of semi-annual reproductive cycles in holothurians. (Boolootian, 1966). Since then 2 species have been reported with this breeding pattern. A, scabra Jaeger in India spawns predominantly semi-annually, in summer and autumn (Krishnaswamy and Krishnan, 1967). The authors suggest that salinity changes, caused by fresh” water influxes following monsoons, may induce spawning in H. scabra and other tropical marine animals. There are no comparable semi-annual salinity changes at Heron Island reef. More recently, a temperate hermaphroditic species, Leptosynapta tenuis (Ayres), has been | reported to have a semi-annual reproductive cycle (Green, 1978). The population spawned in the spring and in the fall, with a mid-summer cessation of reproductive activity. Green suggested _ that external factors, possibly temperature, might regulate the reproductive cycle. A semi-annual reproductive cycle is well-known amongst other tropical marine invertebrates. The Great Barrier Reef Expedition reported a semi-annual spawning cycle (spring and autumn) in the echinoid Tripneustes gratilla (Linnaeus) ( Stephenson, 1934). Both T. gratilla and T’. ventricosus (Lamarck) vary in reproductive cycle with geographic location, but exhibit a tendency for semi-annual reproduction, and in some locations spawn in winter and surmmer (Pearse, 1974). In his study of reproduction in tropical H. atra, Pearse (1968) found that spawning was asynchronous, mature gonads being present throughout the year. He suggested that breeding ; season would become more restricted with distance from the equator. This proposition is supported by the present study, since mature gonads were found in almost every sample, but the majority of individuals spawned in two limited periods. The semi-annual reproductive cycle of H. atra is not typical of other species of holothurians studied at Heron Island reef. H. impatiens (Forskaal) (unpub, data) and H. leucospilota (Brandt) (S. Franklin, pers. comm.) are both synchronized annual breeders, spawning in spring or summer. Spawning in H. edulis (Lesson) is apparently non-synchronized and non-cyclic (unpub. data). Several hypotheses may be proposed to account for the change in sex ratio of H. atra with increased size. Three hypotheses are presented here. a. Mortality rates of male and female animals may change throughout their life-span. Mortality rate may be high in female individuals before sexual maturity, and in male | individuals after sexual maturity. It is difficult to postulate biotic or abiotic factors that might contribute to this differential mortality. b. A relatively higher rate of fission in male individuals than in females would lead to a_ higher incidence of males than females among small animals and a higher incidence of females than males among large animals. However, available data on sex of fission products (15 male, 12 female, 11 indeterminate) suggest that the sex ratio of fission products is approximately 1:1. Because of the small number of fission products for which sex data are available, this theory cannot be entirely discounted. c. Some proportion of the population of H. atra may exhibit protandrous ————e REPRODUCTION IN H. ATRA 63 45 Lagoon 40 Feb. 78 a 2s 20 40 GORE 60s O00: e140 60" 21180 = 200 220 = 240 2:60 280 S00 Individuals Gutter of Feb, 78 Number ~~ (2. 0 20 40 60 SO 00. 120) e000 160) 200° 2205 240) 260! 2:80) 300 Weight (g) Fig. 6. Typical size-frequency distribution patterns of H. atra; unimodal (lagoon) and bimodal (gutter). 64 VICKI HARRIOTT hermaphroditism. Hermaphroditism is common in holothurians and ot] echinoderms (Hyman, 1955), and protandrous hermaphroditism has been recorded several echinoderms (Bacci, 1965), At some stage during their growth, a cert proportion of individuals may change from males to females. This hypothesis is diffi to test as gonads regress after each spawning, and may disappear before the developmy of new gonads. It is impossible to determine the sex of previous gonads. hermaphroditic gonads have been’ detected amongst those sectioned to determine reproductive cycle. Itis possible that individuals may change sex more than once duri their lifetime. Although transverse fission in holothurians is a commonly recorded phenomenon (Hymz 1955; Bonham and Held, 1963), few studies include any quantitative data on the importance asexual reproduction in the life-histories of the species. Crozier (1917) found evidence of bin fission in 11% of individuals of H. surinamensis Ludwig examined, and concluded that fissi\ represented a regular means of multiplication in the species. Deichmann (1922) found 56 regenerating specimens of H, difficilis Semper, and 65% of H. parvula (Selenka), in samp) studied. The mode of division in these species is apparently identical with that of H. ap Fissioned and fissioning specimens of Stichopus horrens Selenka, S. chloronotus Brandt andi edulis have been found by the author at Heron Reef, the latter commonly. | The frequency of asexual reproduction is related to habitat. In general at any one tiny fission frequency is greater in the gutter and mid-reef habitats than in the S.W. reef-crest al lagoon habitats. From the present study, it is impossible to distinguish causative factors in tl’ relationships between habitat and frequency of asexual reproduction. Pearse (1968) suggeste that fission may be more frequent in H. atra in surf-swept intertidal areas. This is supported t the low fission frequency of the lagoon, but contradicted by the unusually low rate for the S.W reef-crest. It is probable that many factors contribute to the variation in fission rate with habitat, am these may include temperature, exposure, current flow, food availability, or more comple factors such as “patchiness” of resource distribution within a habitat, or possible genet? differences between populations. The latter would depend on a limited flow of genetic materi between populations, and may not be compatible with the existence of free-livin) planktotrophic larvae of H. atra. Ebert (1978) suggested that asexual reproduction accounts for the absence of large A. ain on the reef flat at Enewetak Atoll, and that fission is promoted by environmental factors. H interpreted the high rate of asexual reproduction in H. atra as an adaptation to the lo recruitment rate from the sexual phase. H. atra is common at both Enewetak Atoll and Hew Island, and the high frequency of asexual reproduction may contribute to its abundance. Data on changes in frequency of asexual reproduction over time in some habitats, and da on changes in regeneration states over time, indicate that fission is more frequent during limiter periods of the year. Whether this change in frequency follows a yearly cycle cannot be determined from the data available. ) Bonham and Held (1963) suggested that sexual reproduction may be seasonal, whill asexual reproduction by binary fission could occur throughout the year. This study shows thas both sexual and asexual reproduction may occur at low frequencies throughout the year, bi. with highest frequency of each during one or more limited periods. Size-frequency distributions can be correlated with frequency of asexual reproduction. Ip areas of high fission frequency, samples were generally bimodally distributed, and in areas ob low fission frequency, a unimodal distribution was most common. The lower peak of the REPRODUCTION IN H. ATRA 65 bimodal distributions were composed of individuals weighing less than 90 g, and a large proportion of these were products of asexual reproduction, Absence of age classes in the size-frequency distributions could be related to the relative importance of recruitment of fission products to the population. In some organisms, growth has been estimated by obtaining growth increment data from tagged individuals, and applying growth equations such as the Brody-Bertalanffy equation. This method has been successfully applied to echinoids (Ebert, 1977) using tetracycline tagged skeletal elements, and has been tested on holothurians (Ebert, 1978). These analyses of growth are complicated when individuals undergo asexual reproduction. Calcification rates, used as an estimator of growth, may not be uniform in former oral and former anal ends. The size of the calcareous plates will have no relationship with the overall size of the animals, so most growth equations become inapplicable. If individuals divide more than once in their lifetime, mortality may be impossible to define or measure. ; Overall, the unusual growth and reproductive characteristics of this species indicate that traditional growth analyses have little value in obtaining meaningful data on growth and mortality. ACKNOWLEDGEMENTS I thank Dr Ann Cameron and Mr Russell Reichelt for their valuable advice and assistance during the course of this study, and for critically reading the manuscript. I acknowledge Mr Richard Martin and Dr Peter Dwyer for their helpful suggestions, and thank my friends for assistance in the field. The use of facilities at the Heron Island Research Station is gratefully acknowledged. REFERENCES Bacci, G., 1965. Sex Determination. Pergamon Press, Oxford. \Bakus, G. J., 1973. The biology and ecology of tropical holothurians. Jn Jones and Endean (eds.), Biology and Geology ) of Coral Reefs, Vol. 2: 325-367. Academic Press, New York. Bonham, K. and E. E. Held, 1963. Ecological observations on the sea cucumbers Holothuria atra and H. leucospilota at Rongelap Atoll, Marshall Islands. Pacif. Sci., 17: 305-314. ‘Boolootian, R. A. 1966. Reproductive physiology. Jn R. A. Boolootian (ed.), Physiology of Echinodermata: 561-614. Interscience Publishers, New York. Crozier, W. J., 1917. Multiplication by fission in holothurians. Am. Nat., 51: 560-566. Deichmann, E., 1922. On some cases of multiplication by fission and of coalescence in holothurians. Vidensk. Meddr. dansk naturh. Foren., 73: 199-204. Ebert, T. A., 1977. An experimental analysis of sea urchin dynamics and community interactions on a rock jetty. 7. exp. mar. Biol, Ecol., 27: 1-22. 1978. Growth and size of the tropical sea-cucumber Holothuria (Halodeima) atra Jaeger at Enewetak Atoll, Marshall Islands. Pacif. Sct., 32(2): 183-191. Green, J. D., 1978. The annual reproductive cycle of an apodous holothurian, Leptosynapta tenuis: a bimodal breeding | season. Biol. Bull., 154: 68-78. Hyman, L. H., 1955. The Invertebrates: Echinodermata. The Coelomate Bilateria. Vol. 4: 1-763. McGraw-Hill, New York. Krishnaswamy, S. and S. Krishnan, 1967. A report on the reproductive cycle of the holothurian Holothuria scabra Jaeger. Curr. Sci., 36: 155-156. 66 VICKI HARRIOTT Pearse, J. S., 1968. Patterns of reproductive periodicities in four species of Indo-Pacific echinoderms. Proc. India Acad. Sci., Sec. B., 67: 247-279. 1974. Reproductive patterns of tropical reef animals: three species of sea urchins. Proc. 2nd Int. Coral Ree Symp. 1: 235-240. Stephenson, A., 1934. The breeding of reef animals. Part I]. Invertebrates other than corals. Scient. Rep. Gt. Barrie Reef Exped., (1928-29), 3(9): 117-128. Tanaka, Y., 1958. Seasonal changes occurring in the gonad of Stuchopus japonicus. Bull. Fac. Fish Hokkaido Univ., 9 29-36. 5. ANEW GENUS AND SPECIES OF OPHIACANTHID BRITTLESTAR (ECHINODERMATA: OPHIUROIDEA) FROM THE KERGUELEN ISLANDS, WITH NEW TAXONOMIC, BIOGEOGRAPHIC AND QUANTITATIVE DATA ON THE ECHINODERM FAUNA ALAIN GUILLE Muséum National d’Histoire Naturelle, Paris, France SUMMARY Forty-two species have been sampled on the continental shelf of Kerguelen islands during MD04/Benthos cruise of M.S. ‘‘Marion-Dufresne” (March 1975). Among the species, the taxonomic position of a few animals is still uncertain. Four ophiuroids however, allow interesting taxonomic and biogeographical comments: a new genus and new species in the family Ophiacanthidae, Ophioparva blochi, is described; Ophiomisidium speciosum Koehler was known previously only from the tropical deep Atlantic; Ophiocten hastatum Lyman and Ophiocten amitinum Lyman are placed in the genus Ophiura. The fauna has a generally wide antarcuc and sub-antarctic distribution, but a few species are endemic to the Kerguelen province. Ophiura hastata and the newly recorded Ophiomisidium speciosum are also known from the deeper parts of the sub-tropical Atlantic. On the continental shelf, the average density of echinoderm individuals is 52.8 m? at depths between 10 and 180 metres. In fjords, the density of individuals is high but the number of species is low. On the contrary, outside of the fjords in open sea, the density of individuals is low but the specific diversity is high. Around the Kerguelen islands, the specific and quantitative composition of the echinoderm fauna is correlated with hydrological conditions, in particular with the west wind drift. INTRODUCTION The Kerguelen islands are located approximately 50°S latitude, 70°E longitude, at the limit of the antarctic convergence, in the southern part of the Indian Ocean, half-way between South Africa and Australia. The bionomic and physiographic characteristics of the continental shelf of the archipelago, as well as the fjords and interior gulfs and bays, have been the subject of several publications (Desbruyeres and Guille, 1973 and 1977; Guille and Soyer, 1976; Guille, 1977a and b; Murail, David and Panouse, 1977). Since 1972, an intensive programme of bionomic and biological research has been in operation on the benthic fauna of the continental shelf of the Kerguelen islands, in particular on the echinoderms. After the study of the qualitative and quantitative composition of the echinoderm fauna of the Morbihan gulf, virtually an enclosed sea with distinct hydrological and substrate conditions (Guille, 1977a), a similar study has been carried out on the open sea, on the continental shelf surrounding the archipelago, during the MD04/Benthos cruise of “Marion-Dufresne”’ (Guille, 1977b). Before the present programme, the echinoderm fauna of Kerguelen was essentially known through the major expeditions at the turn of the century and shore collections by Rallier du Baty published by Koehler (1917). More recent studies have encompassed a wider geographic area and added to our taxonomical understanding of related faunas, Hertz (1927), Mortensen (1936), Madsen (1955), and A. M. Clark (1962). The only ecological data, pertaining to the Kerguelen echinoderms, has been given by Arnaud (1974) and Cherbonnier and Guille (1974). The collection studied here concerns only the samples of the MD04/Benthos cruise taken by an Okean grab with a 0.5 m?* opening. In fact, of the many kinds of benthic samplers used in Australian Museum Memoir No. 16, 1982, 67-87, 68 ALAIN: GUILLE 68°F 70°E als ] 7 ee I7e 95) 094 0°66 092 065 48°S|_ Al 49°S DO2LS)S Fig. 1. Map of grab samples of MD04/Benthos cruise. KERGUELEN ECHINODERMS 69 this expedition (grabs, trawls, dredges, large diameter corer), only the grabs provided a precise quantitative estimate of the fauna present. Because of the great extent of hard substrates (rock platforms and basalt pebbles), especially in the south-east region of the archipelago, the grabs were used at only 63 of the 120 stations. Of these 63 stations (fig. 1) only 45 yielded echinoderms; 34 were at depths of 10-180 metres, 8 at 180-390 metres, and 3 on the continental slope, between 843 and 1390 metres deep (cf Guille, 1977b, for geographic co-ordinates, depth, nature of substrate of each station). SYSTEMATIC ACCOUNT Among the 42 species collected (Tables 1 and 2; fig. 1), four species require taxonomic discussion: the discovery of the new species and genus of ophiacanthid, Ophioparva blochi; the species Ophiomisidium speciosum Koehler, previously known only from the deep tropical Atlantic; and the change of generic position of the species Ophiura hastata and Ophiura amitina, previously placed in the genus Ophiocten. Family OPHIACANTHIDAE Ophioparva n. gen. DESCRIPTION: Disc covered entirely by small well-calcified plates, imbricated, finely granular but naked, among which are visible, only dorsally, small, widely separated radial shelds, at the edge of the disc. Oral papillae contiguous, the distal one enlarged, completely closing the buccal orifice; one unpaired infradental papilla; a row of dental papillae. Genital slits elongated and narrow. Arm length up to six times the disc diameter. Arm spines erect, cylindrical. A small tentacle scale. TYPE SPECIES: O. blochi n. sp. Ophioparva blochi n. sp. Figs 2 and 3a, b MATERIAL EXAMINED: 13 specimens (d.d. 2.5-5.5 mm), st.5, February 22nd, 1975, 49°30.0'S-70° 56.0’E, 147 m basalt pebbles and shelly sand (holotype d.d. 5.5 mm) coll. MNHN n° ECOS 20371 and 12 paratypes (d.d. 2.5-5.0 mm) coll. MNHN n° ECOS 20371, 4 specimens (d.d. 3-4 mm), Kerguelen, st. 54, March 3rd, 1975, 48°19.0'S-67°56.5'E, 192 m, basalt gravel and muddy sand. All type and non-type material deposited in the Muséum National d’Histoire Naturelle, Paris. DESCRIPTION: The disc diameter of the holotype measures 5.5 mm; the arms, broken, measure at least six times the disc diameter. The disc is completely calcified, pentagonal, the interradial edges straight or more or less excavated. The disc plates and the arms all have a finely granular appearance. The dorsal side of the disc is swollen, covered with small imbricated plates, rounded or oval, among which neither the centrodorsal nor primary plates are apparent. The radial shields are at the edge of the disc, encasing the arm base, and are widely separated by several rows of plates which extend onto the first arm segments. The radial shields are small, three times longer man oe more or less triangular, approximately equal to the one third of the disc radius (figs. a and 3a). The ventral interradial areas are covered with plates similar to, but larger than, those of the dorsal face of the disc, The genital slits, mostly very narrow, are usually bordered by two or three elongated plates with either fine granules or traces of them. The oral shields are cordate, as long as broad, the distal edge broadly convex, the proximal angle subacute. The adoral shields are large, trapezoidal, more than twice as long as broad, the proximal edge shorter than the distal AAAAAALN NIL SAKE “Uy “(ZDETT (71006 (16 {E09 “(D6T {ODOT (26 (Zg1)8 SCOL)s sapnig vousupIuD “yy “(2006 *(1)68 *(1)62 *(1)6Z *(17Z)8 “Hs Zia} v10a104d sunjnduv “vy “(1s UBWIAT SLuDjnduD suDjnsuD Damiyduy “(hrs ELS -ds-u “3° 14901q vawwdoydG “CDETT “(206 *(7ZES SAI1S uewAT] osput ‘CQ “Des ‘(01 “(Ds uewsunl] vindiaia vyjunsn1ydQ :eapromniydg “(Des ‘ds pupunony “(106 “ds snjosg “(D8 ‘ds plopuryy “(106 Qieduie]) vuniss0ad suungnz04312 FT “(Ds (92U.L) wnidasui wnipyos gy “(Des (ueWIYyW) DIDjnNIUNUAp *T “(D6L “DL “(DSS “(zr (3IMpN’T) Dasnd auokyiXyvs J “(DETT Shs (Aue A) 1ay]1an01 snunono04nD1S “(Dss “(IDES *h)6Z “CDOT $16 “(P78 +s ([[1J9A) Snvs1aan] snuzopnas_J “(2068 *(E)Z8 “(798 “(DIB (E)6L “CDPL SCD) “(ZZ *(€)0Z “(LOT *(h)8 (JapNig) 0321012 vippdjowny :eaplomnyojoH, “(Es (uapeS) Diafudjpos spuaispjiu > (D6Z UaPe[S smIDjnuuD saIsDIpiqvT “(MEL (O)8 (1211I3g) sypuoipuam spuaispjdiq “(1)6Z (yiIuIg) uated spuaispuy “(©)8 (1S JOYstf 49qgDjs DINI4DIUD “J “(162 WWW voUI4DIUD vINIADIUD DIUDIOg “(2s Jalleg tunigay swiffp saispsai gg ‘Ws (WIIG) stDUoIpuaU s21sDIUOPC) "(£06 §(Z)b8 {(1)T9 *(Z)09 YIU sisuauajanduay stsuauajansiay saispyrsidaT “(D68 *(1)99 UaPR]S snsaqo saduoy saisorqcying Baplolalsy “(DLOT S(D)SOT “(16 “(D6 “(O76 $68 *(DLL WED TH Siuauajansiay wispsug “(Q)IZT SDETT SOTOTT *(1)06 S(9)98 S(Z6L (SPL (SDL “(1109 S@eér “(TE (OOE “Der “(HBT “(ZZ “COT “(88 (129A) SnIpp4s02 snipgy “(DPIT S(DETT £(106 *()P8 “(ZZ S)SZ #129 *(€)8 “(Ds (Jepnig) Duapoip snuiyse121¢ “(CDFIT SCDETT “(9s (WOsuIOY TL “K\) XMiInu suppioualy) :eaplourysy “(Drs (DES (DST {CD8t 216 (D8 Cr)s Joquadie) “Hg Sstsuajandiay snuuyr0yrvuos gy -EopIOurt suaurtdads jo (stsayjuered ul) Jaquinu puke Jaquinu uoNels satsads *() gaquinu uaurtoads puke Joquinu uonels yitM satoads Jo ist] “[ 3]qeL 71 KERGUELEN ECHINODERMS uw 9T “ALT 80002 : $8 Tho6b Icl Ul QS *A,S°95.69 : S,€°80.6b $8 w ¢9 'A,7'0S089 : S,7 8ho8b 6r Wi 7S~ *A,9°6700Z : $,0°6S.6¢ SII Ww O¢Z *H,0°872.69 : $,9° L684 18 w CPI £A8°9b89 : $,€° 60064 (Gy W g9I ‘Ab yZ00L * S/S S064 FIT Ww Ep ‘A, £17.69 ? $,0°F0.6¢ 6L W 08 *H,0'7S089 : $,8° 80.64 6£ W Cr SA8° 61002 : $,0°1S.6h €Il W C8i SA€'01069 : SETS 08h Wass uw 89 SAS 9.89 : S,€°L7.6b I€ W gy SA,0°PE.0Z : $,0° 016+ OIT wild ,0°80.69 : S$, 1° 8So8h cL Ww ce SA, T0S089 : S,L°L2.6¢ 0€ WO6El ‘A,/S'8IoIL : S.€° TEP LOT wi¢ A, 110069 ? S,€° 70064 bl ul ¢7 ‘A L'11.69 : $,0'1€.6¢ 67 Ul EPs “AS 9001L : S,6' EFSF Sol WOE *A,0'70069 * S,€°90064 eZ UW 68 SAS 71069 : $,0'0€06¢ 87 Ww 076 ‘ALT €E002 ¢ $,2°7S.9¢ L6 ul 707 °H,0°00069 : SS Tholb 99 w /TI °A,7'072069 : $,6°LE.6h 07 Ww O/T SA 8 EZ0L * SS 97LP $6 w 19 *A,€°11069 * $9 OFo8F 79 Ww QZ *A,L'07.69 : S,19€.6¢ 61 w /9T ‘ALL SIOOL * S,8' PHL 6 wos °1,9°€0069 : S/S‘ Ob Sb 19 Ww 8T 'H,0°07.69 : S,7°S€.6¢ 81 UW 87 *H,0°60002 : $,6'078¢ 06 w /T AZ 70069 ? S,0' 1 ho8h 09 Wi O¢ SALT €069 : §,8°6£.6P al w col *A,0°90.0Z : $,0'8€.8F 68 ul 6¢ SALT 872049 : SF S084 9S Ww g¢ £A,8°0b069 : S,7' £06 Or Ww 90T £H,0°0000Z : S,7°SSo8F L8 Ww ¢ol7 'H6 1poL9 ? S,7 T1o8h gs Ww Il SA 6£069 : SZ 7Eo6h 6 Ww 6 *H,0°8S.69 : $,6°70.6F 98 Ww 76l *A,S°9S0L9 : $,0' 61084 vs wi 77 SALE 8069 : ST TE.6b 8 wos¢ "AZ ETOOL : $,7'90.6F $8 w csi *H,0°91.89 : $,0°0€.8F €S Ww /pT 'H,0°9S.0Z : $,0°0£.6¢ S yidoq So] BUTP1O-07) “ONS yidaq SO1BUTPIO-07) “ON IS widaq SO]BUTPIO-07) “ONUIS syidap pur saleurpi0-o7) uoTelsg *Z 3]qe I, “(1)SOT (uRWAT]) DIDISDY °C “(CODSIT £(9)26 *(1)76 (1)S8 “(LSS “(bps (uewAT) DuNNUD “CE “PIT “DET “(h)06 *(E)S “(106 (0168 £(9)P8 *(E)6Z §(1)09 *(1)0Z $16 *(1F)8 “(1S “(106 *(€)s “(SLOT *(S)s8 “(61009 *(O)ZP “CODE “(ZT *(E)OT +18 “CDI £(1)99 *(Z)9S *(1)rS *(6)s “(Ds “CE)PS “(ZS “LOT “(ESOT “(h)9S “(Ess (URUIAT) DNsIquD ‘CE (yituig) Durdsiaasq vamiydC (FaPMg) vIDULLDI sidajosmiydC Iapmsg vipsiean] Duoso1yd— (Matus) sx9pxay snjouo1yd— Ja]Yys0y wnsoweds wnipisrmoiydd (uBWAT) D2194DIUD DIAOY10dG0 | ueUWIAT Dsojuawol “vy JaTYPOY magnol -y = Nh ALAIN GUILLE ods Cenk LG. eee +-oean imm Fig. 2. Ophioparva blochi n. gen., n.sp.: a. dorsal view; b. ventral view; c. dorsal view of an arm (13th segment); d. ventral view of an arm (13th segment); e. arm spines (7th segment); f. dorsal view of the end of an arm. which is broadly contiguous with the first lateral arm plate without arm spines. Proximally the adoral shields and oral plates leave an obvious diastema, the latter only joined in their proximal third. These oral plates are subtriangular, more than twice as long as wide; they bear three lateral oral papillae, the distal papilla the largest rectangular, the middle one also rectangular but shorter, the proximal papilla conical. This proximal papilla is separated from the middle one by a short diastema when the buccal orifice is broadly open in larger sized specimens. There is an unpaired infradental papilla, more developed than the proximal lateral papillae, conical or more or less rounded at the tip; this is immediately beneath a row of 4-5 oval, well-developed dental papillae. The oral and dental papillae are bordered by very fine denticles. The contiguous oral papillae are capable of completely closing the buccal orifice (figs. 2b and 3b). The first two or three dorsal arm-plates are separated by one or two small rectangular plates, the following dorsal arm plates by lateral plates which become more and more broadly joined on the median line (fig. 2c). The dorsal arm-plates are triangular, the distal edge broadly convex, the proximal angle acute, as wide as long. Towards the distal end of the arm, they become smaller and smaller, although the segments become longer. (fig. 2f). The first ventral arm-plate is pentagonal; the next, larger, is triangular, as wide as long, the distal edge almost straight, the lateral edges convex to the level of the scarcely visible tentacle pore which is without tentacle KERGUELEN ECHINODERMS 73 scales, the proximal angle sub-acute. The other ventral arm-plates are oval, more than twice as broad as long, the distal edge broadly convex, the proximal edge composed of two sides united by a short acute angle, slightly excavated at the level of the tentacle pore, marked by a tny pointed tentacle scale (fig. 2d). Towards the end of the arm, the ventral plates become similar in appearance to the dorsal plates, triangular, as long as broad, the distal edge convex, the straight proximal sides joined by a sub-acute angle; the tentacle scale, always single, becomes comparatively more developed, lanceolate and finely denticulate at its extremity. At the end of the arm (fig. 2f) the last two segments become abruptly shorter; the penultimate dorsal and ventral arm-plates become broader while the last ones are much reduced. A long cylindrical, non-segmented section ends the arm. The arm-spines are raised, 3 in number, cylindrical, slightly narrower and finely denticulate at their extremity. In the proximal part of the arm the dorsal spine is twice as long as the ventral one, which is slightly shorter than the length of the segment. Towards the distal end of the arm the spines become sub-equal, more pointed and more denticulate (fig. 2e). REMARKS: Specimens of this species were found only in grab samples of two stations and not in the dredge and trawl samples from the same stations or any other stations of the MD04/Benthos cruise. Although the external skeletal plates are well calcified, their fine granules suggest that the 17 specimens collected were juvenile. But those, whose maximum disc diameter is only 5.5 mm, do not appear to correspond with any species from the sub-antarctic region, or even from the antarctic or Indian Ocean areas. These specimens have not been idenufied with any known genus, so it is with much hesitation that I place them in a specific family. The characters of Ophioparva blochi suggest affinities with several families: the genus Ophiochyira of the Amphilepidae by the buccal structure and several other characters, the juvenile Ophiocomidae (for example Ophiocoma erinaceus) by the dorsal side of the disc, the Ophiolepidinae by the ventral side of the disc. The link with the Ophiacanthidae is however, most suggested by the arm and buccal structures; sub-rectangular oral papillae occur in certain species of this family. Family OPHIURIDAE Ophiomisidium speciosum Koehler Figs. 3c, d and 4 as speciosum Koehler, 1914: 34-36, pl. 3, figs 3-4. — Schoener, 1969: 131-133, figs MATERIAL EXAMINED: 9 specimens, st. 5, February 22nd, 1975, 49°30.0’S- 70°56.0'E, 147 m, basalt pebbles and shelly sand; 1 specimen, st. 54, March 3rd, 1975, 48°19.0'S-67°56.5'E, 192 m basalt gravel and muddy sand; 2 specimens, st. 56, March 3rd, 1975, 48°05.4'S-67°28.1'E, 390 m, gravel and organic mud; | specimen, st. 66, March 12th, 1975, 47°41.5'S-69°00.0'E, 202 m; 1 specimen st. 114, March 15th, 1975, 49°54.5'S- 70°24.4'E, 168 m basalt pebbles and sand with bryozoans. DESCRIPTION: The disc diameters of the 14 specimens measure from 1.5 to 4 mm; the arms, mostly broken, are 5 mm on the largest specimen (incomplete) which is described here. In dorsal view, the disc appears circular, covered by a small number of large, regular, symmetrical plates (figs. 3c and 4a). The centrodorsal is pentagonal, surrounded by 5 primary plates, much wider than long, trapezoidal, the proximal and distal edges semi-circular. The radial shields are longer than wide, joined proximally, diverging and rounded distally, separated by two small plates: the first of which is triangular, the second semi-circular. A single row of two ALAIN GUILLE 74 TPSIOP *9 :19TYIOy wnsorads wnrpismuo1ydd “MATA [EMUDA *q {MOIA [eSIOp ‘ek : “MOTA [BIJUDA *p SMITA ‘ds'u “uaz -u 149019 vaswdorydg *¢ “BLY KERGUELEN ECHINODERMS 75 plates covers the interradial spaces. the first rectangular, much longer than wide, the second tapezoidal and enlarged. In ventral view, the disc appears pentagonal due to the great development of the first three lateral arm-plates. The interradial spaces are much reduced. covered by a large elongated, fan-shaped, plate, the convex distal edge revealing, subjacent. the edge of the marginal dorsal interradial plate; the ventral interradial plate is bordered by two very narrow plates, sometimes almost entirely hidden by the first lateral arm-plates. The oral shields are small, pentagonal, with an acute proximal angle, the subequal sides joined by an obtuse angle. the distal edge straight. Their sides are distally bordered by an elongated genital plate marking a very narrow genital slit. The adoral shields and oral plates are well-developed; the former are trapezoidal. contiguous, the lateral sides slightly indented by the first arm tentacle pore; the latter sub-triangular, bordered by a very long and narrow distal oral papilla. rectangular, preceded by amuch shorter proximal papilla, usually small and difficult to see. The pointed single terminal papilla is likewise small (figs. 3d and 4b). The first dorsal arm-plate is larger than the following ones, rectangular, sometimes _ pentagonal, wider than long. The following plates are triangular, with an acute proximal angle, the distal edge slighily convex; further along the arm they become smaller and smaller and have almost disappeared after the 10th segment. The first five ventral arm-plates are large, becoming smaller towards the distal end of the atm, hour-glass shaped, hexagonal with an obtuse proximal angle, the distal edge broadly convex, the sides excavated at the level of the tentacle pores. The tentacle pores are wide, circular, bordered by a large, sometimes double. oval tentacle scale. Beyond the fifth segment the tentacle pores disappear and the ventral arm-plates become abruptly triangular, of a similar shape to the dorsal arm-plates. They become smaller along the arm, and disappear beyond the 10th segment. In ventral view, the first four lateral arm-plates are much larger than the rest. of decreasing | Size, trapezoidal. The distal end of the second pair, sometimes the spines of the first. are visible in dorsal view, beyond the edge of the disc. These first four pairs of lateral arm-plates are barely | joined ventrally; dorsally only the fourth pair 1s much in contact, like the following plates which become much narrower and longer, giving the arm a fusiform appearance. After the fifth Segment, like the dorsal and ventral arm-plates. the appearance of the lateral arm-plates is similar on both dorsal and ventral surfaces. There are two spines on the first arm-segment, then three on all others. These spines are subcylindrical, squat, and are armed with spinelets distally (fig. 4c). They become shorter and shorter along the arm. Seen under a binocular microscope. the disc and arm-plates appear granular. REMARKS: Sim species of the genus Ophiomisidium are presently known, but none has been commonly collected. The species appear to have restricted distributions (fig. 5): O. pulchellum (W . Thomson) from the south Atlantic, close to the South Africa coast (275 to more than 3000 m), O. flabellum (Lyman) from the Sydney coastal region (60 m), O. leurum Ziesenhenne from similar coastal regions of the Galapagos Islands and Chacahua Bay (Mexico) (80-140 m), O. mene Fell from the Chatham Islands and Pegasus canyon (New Zealand) (230-1006 m), O. speciosa Koehler from the tropical Arlantic (587-1562 m) and O. mzrabile Smirnov from the Ob’Bank (Antarctic part of the Indian Ocean) (240 m). Three of these species. O. flabellum, O. trene, and O. murabile are characterised by a great development of the first pair of lateral arm-plates which. ventrally. are jomed at the back of the ALAIN GUILLE arm spines (2nd segment). Koehler: Fig. 4. Ophiomisidium speciosum 77 KERGUELEN ECHINODERMS “Ig]YIoY wnsorsads wnrpisimoryd¢ *s ‘(uOsuIOY |, *\) unjaysnd wnipisnuoityd, *d <(ueWwAT) Dunnun vantyd¢C “wi SauUaqUasel7Z wnina] wnipisnuciydy *| Tea uaa wnipisimoryd¢ “1 <(UkWIAT) DIDISDY pantydg¢ *y {(ueWIAT) wnpjagoyf mnipisnuo1yd, “J SAOUITUIS a7rqnsn wmipisnuoiydg *q S(Uasua1I0W) supjnuus swiffo vantydQ *e :paniydC snuas jo satsads 20141 Jo pue umipisimo1ydd snuas jo satdads Jo uonnqimsip s1yde1s0en) *¢ “31.7 78 ALAIN GUILLE much reduced oral shield. The three other species, as is characteristic of the genus Ophiomisidium, also have the first lateral arm-plates well-developed, but these are no longer joined and reveal, distal to the oral shield, one to three interradial plates. The genital slits are present only in O. pulchellum and O. speciosum but Fell (1960) and Baker ( 1977) record, in new material of O. irene, the appearance of tiny genital slits on the largest specimen. The specimens examined here, agree closely with the original diagnosis of O. speciosum, However, Koehler’s holotype and five paratypes, housed at the USNM and kindly lent to me by Maureen Downey for examination, are missing a well-developed and rectangular first dorsal arm-plate; this is also missing from my larger specimens. Also, the surface of the disc plates, like the arms of the types, are much more granular than those of the Kerguelen specimens. These two differences however, seem merely to represent intraspecific variation, The evolution of the skeleton structure of O. speciosum as a function of size as shown by Schoener (1969), and the appearance of genital slits in the larger specimens of O. irene suggests that the criteria used to distinguish the species of the genus Ophiomisidium may need reconsidering. The discovery of O. speciosum on the continental shelf of the Kerguelen islands and the variability described above, suggests that some of the other “species” of Ophiomisidium may not, in fact, be separable from one another. Thus, the only other antarctic species of the genus, O. mirabile Smirnov (1977), is very close to O. irene, and according to Baker (1977) O. flabellum and O. irene can be distinguished only by the number and the size of the upper disc-plates and in the shape and relief of the plates on the underside. The other two genera at present placed in the family Ophiuridae (Ophiurinae), Astrophiura Sladen and Ophiophycis Koehler are also characterised by the large first lateral arm-plates. With the genus Ophiomisidium they could be considered to form a distinct family, the Astrophiuridae Sladen (Cherbonnier and Guille, 1976). Ophiura amitina (Lyman) Figs 6a-c and 7c, d Ophiocten amitinum Lyman, 1878: 100, pl. 5, figs 129-130; 1882, 79, pl. 9, figs 7-9. — Studer, 1885: 16, pl. 2, figs 8a-f. — Koehler, 1907: 288; 1923: 122. HOE. Clark. 1915; 325s Mortensen, 1936: 335, fig. 48a. — Madsen, 1967: 138. — Fell, Holtzinger and Sherraden, 1969: pl. 26, map 2. — A. M. Clark and Courtman-Stock, 1976: 192. MATERIAL EXAMINED: 1 specimen, st. 85, March 9th, 1975, 49°06.2'S-70°13.2'E, 50 m, basalt gravel and pebbles, organic muddy sand; 1 specimen, st. 92, March 10th, 1975, 47°44.8'S-70°15.7'E, 164 m, basalt gravel and pebbles, green muddy sand; 4 specimens, st. 54, March 3rd, 1975, 48°19.0'S-67°56.5'E, 92 m, basalt gravel and muddy sand; 16 specimens, st. 115, March 15th, 1975, 49°59.0'S-70°29.6'E, 252 m, fine sand with bryozoans; | specimen, st. March 3rd, 1975, 48°11.2'S-67°41.9'E, 275 m, fine sand, 5 specimens, st. 97, March 11th, 1975, 46°52.7'S-70°33.1'E, 920m, mud. (d.d. range 3-7 mm). REMARKS: Two species of the genus Ophiocten are known from the southern part of the Indian Ocean: O. amitinum Lyman from the Kerguelen islands and O. hastatum Lyman. Examination of the ‘‘Challenger” types, deposited in the British Museum and loaned to me by A. M. Clark, reveal several errors in Lyman’s diagnoses and figures, which have doubtless led to confusion of the species. Lyman, in effect, distinguishes the two species mainly by the presence of sub-equal arm-spines, a smaller, more triangular first ventral arm-plate in O. amitinum, and oral papillae of different shapes. In fact, in all the specimens of O. amutinum from the “Challenger’’, the first arm-segments bear an upper spine thicker and twice as long as the next two; the oral papillae and KERGUELEN ECHINODERMS 79 OW i) (2 Kf aN oe, WePU0 KE ry _—— acde b 1 mm Fig. 6. Ophiura amitina (Lyman): a. ventral view; b. dorsal view; c. arm spines (Sth segment). Ophiura hastata (Lyman): d. one jaw; e. arm spines (Sth segment). the first ventral arm-plate are similar to those of O. hastatum. As a further point of confusion, two sub-species of O. amitinum, O. amitinum microplax and O. amitinum simulans were distinguished by Mortensen (1933, 1936) and their distribution seems limited to South Africa. However, A. M. Clark and J. Courtman-Stock (1976) showed that these two sub-species should really be reunited and linked with the north Atlantic Ophiura affinis. These two authors stress that the distinction between the genera Ophiura and Ophiocten is “clearly very slight with Ophiura affinis bridging the gap between them”’. While the forms hastatum, amitinum and affinis simulans are very distinct (Table 3), they appear too close to be generically separated. Until there has been a worldwide revision of the generic limits of the family Ophiuridae, these three forms must be placed in the same genus Ophiura, as A. M. Clark and J. Courtman-Stock have already concluded for one of them. _ Ophiura amitina is considered by Madsen (1967) as a common circumpolar species, principally sub-antarctic, but in the South African region it is replaced by O. affinis simulans (Fig. 5) which is very abundant in some biotopes (Day, Field and Penrith, 1970). 80 ALAIN GUILLE Ophiura hastata (Lyman) Figs 6d, e and 7a, b | | Ophiocten hastatum Lyman, 1878: 103, pl. 5, figs 133-134; 1882: 82, pl.9, figs 10-11. — Koehler, 1898: 42, pl.7, figs 32-33. — Fell, 1958: 29. Ophiocten longispinum Koehler, 1896a: 204; 1896b: 243. MATERIAL EXAMINED: 1 specimen, st. 105, March 13th, 1975, 48°43 .0’S-71°06.5’E, 843 m, mud (d.d. = 13 mm). REMARKS: Ophiura hastata does not appear to have been found since the expeditions of \ the “Challenger” and “Hirondelle”, and is known from widely separated localities, and from - only a few specimens: the central Atlantic (Azores), the southern Indian Ocean and New - Zealand, always at depths of more than 1800 m. The discovery of O. hastata at only 843 m isan important extension of bathymetric range. As T have indicated in the discussion about the taxonomic problems relating to O. amitina, the similar characteristics of these two species and the errors in Lyman’s descriptions have caused confusion between them, and this explains to some extent the absence of data on O, hastata. In presently available collections O. hastata can be easily distinguished from O. amitina by the ornamentation of the dorsal face of the disc, the wider than long oral shields, and the very long upper arm-spine. But one must also note that all the known specimens of O. hastata are ofa | larger size than those of O. amitina. BIOGEOGRAPHY OF SPECIES COLLECTED Except for the discovery of Ophioparva blochi and Ophiomisidium speciosum, the species collected have a wide antarctic and sub-antarctic distribution, or are endemic to the Kerguelen islands or the Kerguelen province, as defined by Koehler (1912), which includes Heard, Crozet, Marion and Prince Edward Islands. However these endemic species are very closely related to circum-antarctic species. Thus, the single species of crinoid collected, Promachochrinus kerguelensis, is the most widespread and abundant crinoid in the antarctic and sub-antarctic region between 10 and 1080 metres deep. Three of the four echinoids are endemic to the Kerguelen islands; one, Ctenocidaris nutrix, has also been found around the Crozet islands. The number of asteroid species is low compared with previous studies, probably due to the sampling method used. Their collection was in fact, almost limited to two stations: one at the south-east of the archipelago, at the base of the fjord of Table Bay (st. 29, 23 m), the other at the entrance of Royal Pass leading to the Morbihan Gulf (st. 5, 147 m). Of the asteroids, only Anasterias perrieri is endemic. Two of the ophiuroids are similarly endemic, and are also the most common and most abundant ophiuroids — in the littoral sedimentary substrates: Amphiura antarctica (synonymous with A. eugeniae Koehler 1917) and Ophiura brevispina. Amphiura joubini is only found at Kerguelen at depths of more than 275 metres, although this species, littoral in the Antarctic, is considered by Fell et al. (1969) as the only eurythermal form of a stenothermal genus (Hemilepis). Of the holothurians identified, only Eumolpadia violacea is endemic and similarly very common in the muddy substrates to a depth of 250 metres. ASSESSMENT OF QUANTITATIVE DATA OF SPECIES COLLECTED Thirty-four of the forty-two species collected were only present at one to five of the sixty-three stations sampled, and their densities were always low; eight other species were more 81 KERGUELEN ECHINODERMS ‘(OsIp Jo JolaWIRIP = “p'p SosIp Jo AvI = J Sspyarys [eIpes Jo YIpIM pue yIsus] = M pur J) wut ¢*/ 01 “p"p aieurres wae Jo sejins Joddq Jenbo-qns soutds-wiy apIM URY] JOSUC] SPaTYys [PIO OPIM Se BUC] SB SOUT] 9dIY] URY) aiow wio0sTuIds squios-uIe Jo sautds snoraqo ‘pedoyaaap-[]a" sajeyd Areuttid pue [esroponus) Ig/[+ = 1 S1/Z = “/] uasIeAIp Anystys ‘poieredas spyarys yerpey UWI g 01 ¢ “p’p XOAUOD We Jo adejns JoddqQ OM) 1XOU IY] SB SUC] se 991M] UR) 21OU PoUdySTY] s]usuIses [euTxo1d \sIIJ JO sutds-we soddy apIM UR] 1oBU0] SpPpTYs [RIO apIM se SUC] SB 991M) UR] 2IOUI JOU SquIOd-WuR Jo sauTds snotago ‘padoyaaap-]jam saieyd Areutid pure [esioponus) ate = [I {¢/¢ = M/] SWuesr9ATp AT[eWITXO1d ‘paieredas ‘]]eurs Spjatys [eIpey LORD ILC Sh. 1212) XOAUOD UIR JO JdeJANs Jaddq OM] 1xXoU ay] URY) JasUO] SAWN p UR dIOW PIUdYSTY] sJusuIsIs [eutxo1d isu Jo sutds-ue 1oddq apIM se SUOT sv 10 SUOT UBY] IPIM IIOUW SPyPTYs [eI 2PM Se SUC] SB ddIM] UR) dIOU JOU squIOs-wIe Jo sauIdg SUSI Ajoreq pur ‘padoyaaap-apiqy sajeyd Areutid pur [es1oponuay Ig/[+ = T ‘1/€ = M/] SWUesIIAIp Apysys pareredas pyarys yerpey G-2Z “3y) supjnuas suiffo piniydC (P-9Z S9-e9 “s3IJ) punuun paniydC (G-B/ £9-Pg “S3IJ) pivispy viniydcQ supjnuas suiffo °C pue vurup °C “vIDIspy vanrtydC UdIM aq sdrysuoNnRpPY “€ 219",1, 82 ALAIN GUILLE Fig. 7. Ophiura hastata (Lyman) (type Challenger 82, 12, 23, 352, British Museum): a. dorsal view; b. ventral view. Ophiura amitina (Lyman) (type Challenger Kerguelen 82, 12, 23, 346, British Museum): c. dorsal view; d. ventral view. Ophiura affinis simulans ( Mortensen) (type Discovery South Africa 1936, 12, 30 144, British Museum): e. dorsal view: f. ventral view. common. The crinoid Promachochrinus kerguelensis, present at eight stations between 18 and 200 metres depth, had a maximum density of 8 specimens/m? (st. 5, 147 m). Abatus cordatus is the most common echinoderm at Kerguelen, present at 17 stations, at 8-147 metres depth and reaching a density of 32 specimens/m? at the mouth of Lac Marville, at the east of the archipelago, at 18 metres depth. The density of this euryhaline urchin is also relatively high in the sands of the lower intertidal zone where many juveniles occur (Guille and Lasserre, 1979). The ophiuroid Ophionotus hexactis, common throughout the antarctic and sub-antarctic, was present in only six stations. It had a maximum density of 38 specimens/m? in mud, at 18 metres depth, in the small bay KERGUELEN ECHINODERMS 83 of Port Christmas at the far north of Kerguelen (st. 60). One station (st. 8) in the south of Kerguelen at the base of the fjord of Swains Bay, ata depth of 22 metres and in mud, yielded the highest densities of the five other most common species: Sterechinus diadema (6 specimens/m*, present at 8 other stations), Amphiura antarctica (314 specimens/m?, present at 8 other stations), Ophiura brevispina (82 specimens/m-, present at 7 other stations), Eumolpadia violacea (8 specimens/m*, present at 10 other stations) and Pseudocnus laevigatus (248 specimens/m2). This last holothurian, present at 6 other stations, was more abundant at one of these others (st. 53), situated at the north west of Kerguelen, 162 metres depth, in pebbles and basalt blocks (282 specimens/m?). Station 8 also had the highest total density of echinoderms with 958 specimens/m? for 14 species present. The station yielding the highest number of species, with 25 species/0.5 m? for 218 specimens/m2, was in the open sea, to the east of the archipelago, opposite the entrance of Royal Pass which gives access to Morbihan Gulf, a pass usually swept by strong oceanic currents (st. 5, 147 m). Because of the sudden shoaling here, upwellings are frequently produced. The average densities of species and specimens for all the 49 stations where echinoderms were present, are respectively 3.7 species/0.5 m? and 52.8 specimens/m?. These results, or more precisely those limited to the same bathymetrical range as the data obtained in Morbihan Gulf, indicate, by comparison with the latter (Table 4), a greater density of species and a lower numerical density in the stations outside the gulf, really a separate sea, with a surface of 700 km2. The significance of these differences is even more evident if the data obtained in the MD04/Benthos stations situated in fjords is separated from those from stations on the continental shelf (Table 4; fig. 8). The diversity of the echinoderms is thus greater, and their numerical density lower on the exterior continental shelf than in the protected fjords and bays. This qualitative and quantitative distribution is related to the topography and its effect on hydrological circulation. In fact, the fauna of fjords and interior gulfs is relatively isolated from the strong oceanic currents where as the exterior continental shelf benefits from the supply of nutritive salts and planktonic larvae from the oceanic environment. The south coast of Kerguelen, comprised partly of fjords (e.g. the very rich station 8) is enriched by the general south-west to north-east direction of hydrological circulation (Murail et al., 1977; fig. 8). The kind of distribution shown by the echinoderms at Kerguelen has been noted for other groups of benthic invertebrates, for example the ascidians (Monniot, 1979). However, it is sull only a preliminary observation, obtained from few samples. Some stations, moreover, are exceptions such as station 53 (162 m) at the north east of Kerguelen where the numerical density is markedly raised due to the abundance only of the holothurian Pseudocnus laevigatus, whose mode of reproduction produces a patchy distribution. Thus, the collection from the MDO4/Benthos cruise confirms again the qualitative and quantitative richness of the echinoderm fauna of the Kerguelen islands (Guille, 1977a). ACKNOWLEDGEMENTS The programme of investigation of the benthos of the continental shelf of the Kerguelen islands is made possible under the scientific Director, Jean-Paul Bloch, of the administration of “Terres Australes et Antarctiques Francaises’’. I thank G. Cherbonnier for identification of holothurids species, Miss A. M. Clark (British Museum) and Miss Maureen Downey (Smithsonian Institution) for loans of type species and advice, Miss Janet Marshall (Australian Museum) and Professor Lowell Thomas (University of Miami) for the translation of my French paper. 84 OO OU 1 Oke ut ¢'/ds Zp €€ GE @ (Q6€I-81 vas uado ut suoneig ORY ON AS VL cur ¢0/ds 9°¢ 81 LI inet (OREO spsol} ul suoneisg suonels JIN ueyIqiop 01 -Wu/"pul 79 ug g/ds €"p OF 8E WI Q8T-O1 afqeseduros syidep ie suoneig (YIOM sTy1) -W ¢*~Q ‘quid URZAC ae OUS ST GS ctug'o/ds /*¢ a4 6¥ @ 06eT-O1 ‘J ugjensI9y (BLL6I “ATTIN) “WO qeis aiAupoy yMUsS aD OM (Lee cur [0/ds 6°7 Ns OE! 12! OBIS (J wapensiay) JINH ueyrqiow s]eNprArpul jo saisads Jo satoads sajdures a3url AISUap IBPIOAY Alisuap o8e0AYy jo JoquinN jo Joquiny saw Aue g YOM sTy] puke (&//6]) [INH wos sontsusp usurtdads pue satoads Jo Arewung “p IqGeL, 85 S[ENPIATPU! Wapouryog Jo Jaqumu *q ‘-W ¢°¢ 10J UoTNeIs YoRa ur salads wapourype Jo Jaquinu ‘e :asmm19 soylusg/fOdW jos “WI ¢°() JO} uONeIS Yoea ut a[dures qviz jo sdey °g -31q So0S So6V So8P al Fo89 FoOL 2 F089 86 ALAIN GUILLE REFERENCES ’ Arnaud, P. M., 1974. Contributions a la bionomie marine benthique des régions antarctiques et subantarctiques. Tethys,, 6 (3): 467-653. Baker, A. N., 1977. Some deep-sea Ophiuroidea from New Zealand. Nan. Mus. N.Z. Rec., 1(10): 149-160, 3 figs, Cherbonnier, G. and A. Guille, 1976. Sur la présence, a ile Heard, del’ Ophiure Astrophiura permira Sladen. Bull. Mus, natn. Hist. nat., Paris, 3e Sér., 351 (Zool. 244): 17-21, | fig. Clark, A. M., 1962. Asteroidea. Brit. Australian N.Z. Antactic Res. Exped. 1929-1931 Rep. Adelaide (ser. B) Zool. Bot., 9: 1-104. Clark, A. M. and J. Courtman-Stock, 1976. The Echinoderms of South Africa. London, pp. 277, 276 figs. Clark, H. L., 1915. Catalogue of recent Ophiurans. Mem. Mus. comp. Zool. Harv., 25: 165-376, 20 pl. Day, J. H. J. G. Field, and M. J. Penrith, 1970. The benthic fauna and fishes of False Bay, South Africa. Trans. R. Soc, S. Afr., 39(1): 1-108, 1 fig. Desbruyéres, D. and A, Guille, 1973. La faune benthique de l’Archipel de Kerguelen. Premiéres données quantitatives. | C. r. hebd. Séanc. Acad. Sc. Paris, 276(D): 633-636. : 1977. Bionomie benthique du plateau continental des iles Kerguelen. Macrofaune. 8. Variations spatiales et | temporelles dans le peuplement des vases a spicules. In Biology of Benthic Organisms. Pergamon Oxford and New | York: 185-196, 5 figs. Fell, H. B., 1958. Deep-sea Echinoderms of New Zealand. Zool. Pubns. Vict. Univ. Coll., 24: 1-40. 1960. Biological results of the Chatham islands 1954 Expedition. 2. Archibenthal and littoral Echinoderms. Bull. : N.Z. Dep. scient. ind. Res., 139 (2): 55-74, 10 pl. } Fell, H. B., T. Holtzinger, and M. Sherraden, 1969. Ophiuroidea. Antarctic Map Folio Series. American Geogr. Soc. folio 11: 42-43, pl. 26-27. : Guille, A., 1977a. Benthic bionomy of the continental shelf of the Kerguelen islands: quantitative data on the Echinoderms of the Morbihan Gulf. Jn Adaptation within Antarctic Ecosystems: Proceedings of the Third SCAR Symposium on Antarctic Biology. G. A. Llano (ed.) Gulf Publ. Comp., Houston: 253-262, | fig., 4 Tab. 1977b. La campagne MD04/Benthos du “Marion-Dufresne’”’: Bionomie du plateau continental des iles Kerguelen. Stations prospectées et résultats préliminaires. Com. natn. fr. Rech. Antarct., 42: 3-40, 3 figs. Guille, A. and P. Lasserre, 1979. Consommation d’oxygne de l’oursin A batus cordatus (Verrill) et activité oxydative de son biotope aux iles Kerguelen. Mem. Mus. natn. Hist. nat. Paris, — sér C, 63: 211-219, 3 figs, 2 tableaux. Guille, A., and J. Soyer, 1976. Prospections bionomiques du plateau continental des iles Kerguelen. Golfe du Morbihan et Golfe des Baleiniers. Com. natn. fr. Rech, Antarct., 39: 49-82, 1 carte. Hertz, M., 1927. Die Ophiuriden. Deut. Sudpolar Exped. 1901-1903, 19, Zool., 11: 1-56. Koehler, R., 1896a. Note préliminaire sur les Ophiures recueillis pendant les campagnes de I'Hirondelle. Mem. Soc. gool. Fr., 9: 223-233. 1896b. Note préliminaire sur les Ophiures des premiéres campagnes de la Princesse Alice. Mem. Soc. zool. Fr., 9: 98-103. 1898. Echinides et Ophiurides provenant des campagnes du yacht |’Hirondelle (Golfe de Gascogne, Terre-Neuve, Acores). Res. Camp. Sci. Monaco, 12: 12-59, 9 pls., 50 figs. 1907. Révision de la collection des Ophiures du Muséum d’Histoire Naturelle de Paris. Bull. scient. Fr. Belg., 41: 279-351, pl. 10-14. 1914. A contribution to the study of Ophiurans of the United States National Museum. Bull. U.S. nat. Mus., 84: 1-173, 18 pls. KERGUELEN ECHINODERMS 87 1917. Echinodermes recueillis par M. Rallier du Baty aux iles Kerguelen en 1913-1914. Ann. Inst. Oceanogr. Paris, 7(8): 1-87. 1923. Astéries et Ophiures. Further Zool. Res. Swed. Antarct. Exped., \(1): 1-145, 15 pls. Lyman, T., 1978. Ophiuridae and Astrophytidae of the Challenger Expedition. Bull. Mus. comp. Zool. Harv., 5(7): 65-168, 10 pls. 1882. Ophiuroidea. Rep. Scient. Results Voy. ‘Challenger’. Zool., 5: 1-386, 46 pls. Madsen, F. J., 1955. Echinoderms other than Holothurians collected in sub-antarctic and antarctic seas, mainly by the “Norvegia” Expeditions, 1928-1930. Sct. Res. Norw. Antarctic Exped., 1927-1928, 3(37): 1-17. 1967. Ophiuroidea. B.A.N.Z. Antarct. Res. Exped., ser. B, 9(3): 123-144, 8 figs., 1 pl. | Monniot, C., 1979. Répartition des Ascidies autour des iles Kerguelen. Mem. Mus. nain. Hist. nat. Paris, sér C, 635 249-253. | Mortensen, T., 1933. Echinoderms of South Africa (Asteroidea and Ophiuroidea). Vidensk. Meddr. dansk. naturh. | Foren., 93: 215-400, 91 figs., 12 pls. 1936. Echinoidea and Ophiuroidea. Discovery Rep., 12: 199-348, 53 figs., 9 pls. | Murail, J. F., P. David, and M. Panouse, 1977. Résultats scientifiques de la campagne MD04/Benthos: Hydrologie du | plateau continental des iles Kerguelen. Com. natn. fr. Rech. Antarct., 42: 41-64, 8 figs., 11 pls. | Schoener, A., 1969. Atlantic ophiuroids: some post-larval forms. Deep-Sea Res., 16: 127-140. Smirnov, I. S., 1977. A new species Ophiomisidium (Ophiuroidea) from the sub-antarctic waters of the Indian Ocean. Acad. Sc. U.S.S.R., Zool Inst. Explorations of the fauna of the seas. XXI (X XIX). New species and genera of marine Invertebrates: 105-108, 1 fig. Studer, T., 1885. Ubersicht tiber die Ophiuriden, welche wahrend der Reise S.M.S. “Gazelle” um die Erde 1874-76 | gesammelten wurden. Abh. preuss. Akad. Wiss., 1883: 1-37, 3 pls. 6. A REVISION OF THE ASTERINID GENUS NEPANTHIA GRAY, 1840 (ECHINODERMATA: ASTEROIDEA), WITH THE DESCRIPTION OF THREE NEW SPECIES FRANCIS W. E. ROWE. The Australian Museum, Sydney, N.S.W., Australia and LOISETTE M. MARSH The Western Australian Museum, Perth, W.A., Australia SUMMARY The genus Nepanthia Gray, 1840 is revised and eight species recognised, including three new species, one from New South Wales, one from north of Sabah, Borneo and one form the Sulu Sea, Philippines. Fisher’s opinion that Parasterina crassa should be referred to Nepanthia is supported and the species is here included. Variation in the species N. belchert (Perrier) is discussed and the species N. brevis (Perrier), N. suffarcinata Sladen, N. joubini Koehler, N. variabilis H. L. Clark and N. magnispina H. L. Clark are considered conspecific with it. A key is given for the eight species recognised. INTRODUCTION Gray (1840; 1866), described the genus Nepanthia, in the family Pentacerotidae, for two species N. tessellata (from an unknown locality) and N.maculata (From Migupou, Philippine Islands). Miiller and Troschel (1842) referred these species to Chaetaster which they described a few months prior to Gray’s Nepanthia in 1840, considering Gray’s species to be congeneric with C. subulata (Lamarck) (type-species of Chaetasier by monotopy). Perrier (1875), placed Chaetaster in the family Astropectinidae, synonymising Nepanthia tessellata with C. longipes (Retzius). He considered Nepanthia to be a subgenus of Asterina (family Asterinidae; in which family it has since remained) recognising the validity of A. (N.) maculata, the type-specimen of which he examined. He described two new species A. (N.) belcheri and A. (N.) brevis. This action leaves N. maculata as type species of Nepanthia, failing Gray’s nomination of a type-species from either of the two he included in his genus. Viguier (1878), placed Chaetaster in the family Linkiadae of his subclass of asteroids (Stellérides) ‘‘Astéries adambulacraires’’ (characterised by the predominence of the adambulacral plates in the mouth ring), listing Nepanthia as a synonym of Chaetaster. He did not discuss the genus or its species. Perrier (1884), in his remarkable classification of asteroids, based on the form of the pedicellariae, without comment listed Nepanthia in the family Asterinidae, Order “Stelleridae Spinulosae” (p. 164, referred to as Echinulatae in subsequent pages). Sladen (1889), considered the skeletal features of the species of Nepanthia to be sufficiently distinctive to “warrant the retention of Nepanthia as an independent genus’, in the family Asterinidae (subfamily Asterininae) of the new order Phanerozonia. Gray, 1847 described Patria? crassa from Western Australia in a third group of species (P. ocellifera, P. obtusa and P. ? crassa) of his genus Patiria. These were characterised by having 5 Australian Museum Memoir No. 16, 1982, 89-120. FRANCIS W. E. ROWE AND LOISETTE M. MARSH 90 86 “RIO “Tl “HSV 6161 “OUST SY 8E6l “RIO “TH SV ulaIoy “1aysjag “Nj JO wAuouAs y 9S61 RID “WV JeYe “prylupda Ny Jo satsads pre, uraioy “uay2jaq ‘N Jo wAuouds Vy ula10Y “uayojaqg “N Jo wAuoUAS Y 6881 “Uapels Jaye ‘piyjupdan Jo satseds prea Tr6l “PYsty Pye ‘piylundaN 01 UIa19Y peisajoy OPS ‘AvryH rs1jeB piyjunda Ny JO satsads pre A 8e6l “ARID “TH “1uay2jaq “NY JO wAuOoUAS Y 6161 “194sty “piylungauniv J 01 prs12J2yY 861 RI “TH “M249/9q "N Jo wAuoUds y 961 RID “WV ‘oupundan or perepey 6881 “UApERIS ‘rpunday O1 pasesYy 6881 “UAPETS ‘PUpundaN O1 pr1}PY 9461 “MARIO “I (H ‘puuaisping 1 ‘Ip6l “IYsIy ‘vIylundan 01 £2061 ‘IAUSIA ‘Duuasping O1 §S/Q] ‘latueg ‘(piylundan) DuLuaisy 01 patlayjay 6881 ‘UapRIS ‘vyupdany O1 {C/Q] ‘Jatag ‘(p1yiuvydany) Duuaisy 01 SZ7PRT “JaYOSOLL 2 JOT[INW “vasp1avyD 01 prlayey SL8T Jolag *99eR sadisuo] “> Jo wiAuouAs & S7PQ] “JaYOSOLL %® JI[NW “vaIspiapyH 01 prapy eljensny ‘purjsusen?) ‘uoidueyyxs0y /9Z61 “U[Iapod xnpdxjod viyrundan UkdIDQ) URIPU] ‘spurys] UeWepuYy ROTEL “191490 VIMY IDIG DIYIUDdaN Rag URUISe] ‘pUuR[S] IMOF{ PlO'T /6061 “AFBID “TH Suspsalay vistuua zy (WIBUIIT A) PUTYD) UTYDOD ‘sanboef “1g dey /8061 “19T420y tmrgnol pipunde N Bag BUTT) YlNOS ‘yURg PJelysopsoeyy /PO8I “Ted Snasvug DunDg Ruling ‘osejadrypiy msi0y 16881 ‘UapelsS vivurzsvf{fns vrylundan elyetisny ‘purfsusen) “N ‘Wellg Sato, “WeINS saeA\ JO IU G/¢/ 8] ‘Jetta g s1aaig (piyjuDda Ny) DUuLuaISy UMOUYUN/S/8] ‘IITA uusya]ag (piysunda Ny) Duuaisy eIelsny UIAISIA\//P8 SABID) Vssp49 ¢ DUDE saurddrryg ‘nodnsty/0pPgl ‘ABID vipjnopwu piyluddan UMOUYUN/OPR] SARIN) vID]JaS8a1 DIyIUDdaN piyjundany UI sM1e1g 1UasaIg 1Uasatg 0] AIOISTE, Ayyeso’T ada p/satsadg AeIny piyjupdan 01 paiiajel satsads jo smeis “| FIGVL }I| ASTERINID GENUS NEPANTHIA GRAY uloray satoeds MaN ulorsy satoads MON UldI9Yy sotdads MON 8961 ‘proydays sy 9961 “ARID “WV SV ure19Y “piIpjnipu *N JO wIAUOUAS Y uTdIOY “14ay2]aq *N JO wiAuOUAS Y ulo194 “1uaysjaqg *N Jo wiAuouds Y 9961 “ARID “WV SY 8961 ‘psoydoys ‘imolysnou “Ny JO wWAUOUAS Y 9961 SAARID “OW 'W “2uolysno.u * Ny JO wAUOUAS YW 9961 RID W “VW “DiyundaN o1 paraypY sourddryryg /ysiep pure smoy syipis viylundan BIssuOpU] SYyRqes /Ysieyw pure smMoy waysif niyiundan elyeLsny “saTB AN YINOS MON ‘spurys] Areiyos /Ysiep pur smMoy vauunigoisiu viyiundan BleLsny eo, ‘Aeg diymyg og 9961 “ARID “WV 2Ymunovippy viyluDde N BYPLISNY “K\'S/8E6I ‘YIRIQ “TH Sypiuapi990 puasaispin gq BIPUSNY Uls1saA\ ‘sWO0Ig /S€61 “APIO “TH Sinuan vrysundan BIeLISNY UaIsaA\ SOWOOIg /S€61 “ARID TH Suyiqvuve vupundenN elyensny “AN Spurysy (Ausedureys 10) snisnsny /SE61 “ARID “TH vuidsusnwu vipunden BIyeAISNY UldIsaA\ SAURGTY/PE61 ‘QUOISSUTAY] 1MO]YSNOM DULIAISDLD oe FRANCIS W. E. ROWE AND LOISETTE M, MARSH arms and “‘dorsal ossicles, especially those at the ends of the arms, broad rounded, the back covered with 2 or 3-beaked pedicellariae nearly hiding the tubercles”. Perrier (1875) restricted the genus Patiria to the species ocellifera and crassa, relegating coccinea (the type-species of | Gray’s Patiria), granifera and obtusa to Asterina. Fisher (1908), noted that Perrier had excluded the type-species, P. coccinea, from his genus _ Patiria, which is therefore not the Patiria of Gray, and renamed Perrier’s genus Parasterina, with | type-species Patiria crassa. Verrill (1913), revised the subfamily Asterininae of the Asterinidae. He described several new genera and constructed a table (key) of genera and subgenera. He listed N. maculata as — type-species of Nepanthia and included brevis in the genus. He included Fisher’s Parasterina with type-species crassa and P. obesa H. L. Clark from Peru. The latter species has subsequently (Bernasconi 1973) been referred to Patiria. In his key Verrill allied Parasterina to his new genus spacer tae with type-species Patiria ocellifera Gray, erroneously attributed to Australia by errill Fisher (1940) doubted the validity of Parasterina after comparing a specimen of P. crassa (the type-species) with Nepanthia variabilis and N. belcheri and in 1941 used the combination Nepanthia crassa. H. L. Clark (1946), however, considered that crassa, troughtoni and occidentalis formed a homogeneous group easily distinguished from Nepanthia by having non-crescentic, crowded, often swollen abactinal plates and inconspicuous papulae; he referred the three species back to Parasterina. Spencer and Wright (1966) include Parasterina in the synonymy of Nepanthia. A. M. Clark (1971, in Clark & Rowe) indicated that some synonymisation of the tropical species of Nepanthia may be necessary when sufficient material has been examined. After a study of most of the type-specimens of species referred to Nepanthia and collections housed in several Australian and international institutions, we have concluded that only 5 of the previously described species of Nepanthia warrant recognition, of these four occur around the coasts of Australia. Three new species are described, one each from north of Borneo, the Sulu ees a New South Wales, Australia. Table 1 summaries the species referred to the genus epanthia. ABBREVIATIONS AM The Australian Museum, Sydney, N.S.W. Australia. BM British Museum (Natural History), London, England. MCZ Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts, USA. NMV National Museum of Victoria, Melbourne, Victoria Australia. USNM United States National Museum (Smithsonian Institution), Washington D.C., USA WAM Western Australian Museum, Perth, Western Australia. ASTERINID GENUS NEPANTHIA GRAY 93 SYSTEMATIC ACCOUNT Family ASTERINIDAE Genus Nepanthia Gray, 1840: 287 | DIAGNOSIS: An asterinid genus with five or more distinct, elongate subcylindrical to tapering arms and a relatively small disc; abactinal surface strongly convex, actinal surface usually flat; adult size between 30 and 80 mm R. The abactinal plates appear crescentic to rhomboidal and are usually in distinct abactinal and lateral ‘fields’ on the arms; the primary _ plates imbricate throughout or only in certain areas, small secondary plates are often present. Marginal plates small, not prominent though they may delimit the lateral/actinal line. Actinal plates in 4-9 rows at base of rays, decreasing distally. Adambulacral plates bear a fan or comb of furrow spines backed by a fan or group of variously sized subambulacral spines which are sometimes larger than the furrow spines. All plates evenly covered or bearing tufts of spinelets which vary, between species, from short and thick set with many points to slender and hyaline with 1-3 points. Simple fasciculate pedicellariae are sometimes present. Several species fissiparous. Eight species recognised. TYPE SPECIES: N. maculata Gray, 1840, restricted by Perrier, 1875, designated by Verrill, 1913. OTHER SPECIES INCLUDED: N. crassa (Gray, 1847), N. belcheri (Perrier, 1875); N. briareus (Bell, 1894), N. troughtoni (Livingstone, 1934); N. nigrobrunnea n. sp., N. fisher’ n. sp. and N. gracilis n. sp. KEY TO SPECIES OF NEPANTHIA 1. Abactinal spinelets short, thick with many (more than eight) equal points; no prominent convex abactinal plates; distribution — southern Australia; two species 2 — Abactinal spinelets slender with few (not usually more than five) unequal points; one species (crassa) with spaced convex abactinal plates; distribution — Australia (except the south coast) and Indo-Malay region; six SPeCl€S ..........cccssevecersseeseseeseseesuses 3 2. Exposed portion of plates of dorsal field rhomboidal, scarcely notched for papulae; 4-6 plates across the field at 1/2 R; papulae single; colour whitish-pink LOLG CEP eLOSe mpApiide mec eSOMLMerIeVUST halide yg meen dhe tats etenc ake iki ke nace N. troughtoni — Plates of dorsal field small, irregulary shaped, 8-10 plates across the field at 1/2 R; usually 2-3 papulae between plates; colour dark brown with black papulae; MIG PRIOEL LA tse Mesa PETRA AREAL Ouen AP 25 -edtilot its owl wie aids tua ntitete ees nak « N. nigrobrunnea 3. Arms arched abactinally, actinal surface flat; craratein) plates form a distinct ACU Gr AP LAISE CD Owe, Sok ses sete dusters Fats §.a0 dn toad oh ach een purdraad. shares shisha gatas 4 — Arms cylindrical or terete, marginal plates not forming a distinct actino-lateral edge . 5 4. Usually prominent convex primary abactinal plates with smaller secondary and granule-like tertiary plates between them; not fissiparous; colour usually mottled red/brown or brown, sometimes (at Abrolhos Is.) bright blue; west CORSLEO LO ESTONIA ATISTIG La eeta Mineman, eos vein at, eve tat she ve ne arm ey ARR CE, Deka ae N. crassa — Primary abactinal plates narrow, crescentic with small granule-like secondary plates around papulae; often fissiparous; mottled, variously coloured, often shades of grey or dull green or orange, red or brown; Indo-Malay region to TOLL M EPEAT G ial cares wer tekes st esta etree tse nts te har ava teats substdans adveunetethe mite degree N. belchent PMS PING CLG a The ecaOl IG Smimtd rec ntirdie arreaubhaker Sante at aE Ree sleet tain e.c4 hetrad Gls 6 ES SHIMCOTS AVAt IT THO TAM APOlUSs pecs deten cet wesc petlecn ot Saaetal poy ames rad chapel ety esg 7 94 FRANCIS W. E. ROWE AND LOISETTE M. MARSH 6. Arms terete, tapering to an acute tip; spinelets in 2-3 small tufts on the primary plates; colour (holotype) grey-blue; Philippines to Timor Sea .......ec.eece00c2000. N. fisheri: — Arms long, cylindrical or slightly tapering to a blunt tip; spinelets evenly covering primary plates; colour cream to buff, sometimes with dark spots; Phalsppunessto opine -AWs tala So. +r fs oestecrt-cutsby More eeecet eat ti. eee N. maculata | 7. Five rays, not known to be fissiparous; cleaned abactinal plates flat; spinelets q slender with 7-8 points; Philippine area and N.S.W. ..ccccccccccececceecececeeeceee N. gracilis” — Multirayed, fissiparous; cleaned abactinal plates moderately convex; spinelets | | short with 5 or 6 points; South China sea to Moluccas ...ccc..ceeeceeesececceccce. N. briareus Nepanthia troughtoni (Livingstone) Figs 1; 2d, e; 5k; 61. Parasterina troughtoni Livingstone, 1934: 179, pl. 18 figs 1-6 — H. L. Clark, 1938: 180; 1946: 143.—Rowe and Pawson, 1977: 346. Parasterina occidentalis H. L. Clark, 1938: 180, pl. 21 fig. 5; 1946: 143.—Rowe and Pawson 1977: 346. Parasterina sp. c.f. troughtoni.—A.M. Clark, 1956: 378, text fig. 3, pl. 11. Nepanthia hadracantha. A. M. Clark, 1966: 320, text fig. 3, pl. 3, figs 4-6. Nepanthia troughtoni. A. M. Clark, 1966: 322.—Shepherd, 1968: 748.—Rowe and Pawson 1977: 348. MATERIAL EXAMINED: 1 specimen (R/r = 16/4.5 mm = 3.5), holotype, N. troughtoni, AM No. J3978; 1 spec. (R/r = 34/7 mm = 4.8), paratype, Parasterina occidentalis AM No. J6178; 1 spec. (R/r = 67/12 mm = 5.6), paratype, P. occidentalis WAM 46-32; 1 spec. (R/r = 37/9 mm = 4.1), paratype, P. occidentalis, WAM No. 606-31; 1 spec. (R/r = 55/12 = 4.6), holotype, N. hadracantha, NMV No. H14. In addition 77 specimens from the W.A. Museum, 7 specimens from the Australian Museum and 10 specimens from the $.A. Museum were examined. A summary of data for the 99 specimens examined is given in Table 2. Table 2. Variation in size and R/r ratio of specimens of N. troughtont from five areas of the coast of southern Australia. Distribution Number of Max. Min. Range Mean Mode Specimens R/r, Rir, R/r R, mm. R, mm. mm. mm. W.A. 30°05'S to 26 80/17 19/5 3.3-5.8 48.6 41-S0 32°20'S W.A. 33°10'S to 27 64/14 25/6 3.6-5.0 40.7 31-40 34°0'S South Coast W.A. 28 65/11 16/4.5 3.5-5.9 47.7 41-50 South Australia 16 62/15 38/9 3.7-5.9 46.5 31-40 Victoria 2 57/11 55/12 4.6-5.2 — — 99 80/17 16/4.5 3.3-5.9 45.9 41-50 ASTERINID GENUS NEPANTHIA GRAY 95 Table 3. Size distribution (R) of the specimens of N. troughtom in 10 mm class intervals. 1-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 0 2 8 25 34 22 7 l 0 DIAGNOSIS: A species of Nepanthia with 5 cylindrical to slightly tapering blunt ended arms, non fissiparous. Maximum known R’/r of 80/17 mm, mean R of 46 mm, range of R:r from 3.3 to 5.9; colour pinkish-white to rose with red papulae; abactinal plates rhomboidal, slightly convex; crystal bodies, secondary plates and pedicellariae absent; abactinal spinelets all over plates, short and stout with ca 10-20 points surrounding a prominent hemispherical to slightly pointed boss; furrow armature in overlapping combs of 4-5 subequal furrow spines with 6-10 -subambulacrals in 2-3 rows; intertidal to 73 metres on rock, southern Australia. COLOUR: N. troughtoni varies in colour from pinkish white-through salmon-pink to rose with red papulae and red skin visible between the plates in life. HABITAT AND DISTRIBUTION: The habitat is on open coasts exposed to considerable wave action where it is found under reef ledges, grazing on encrusting organisms such as compound ascidians, under boulders or on vertical rock faces in the sublittoral. N. troughtomi is recorded from Wilsons Promontory, southern Victoria, around southern Australia to Green Head (30° 05’ S) on the west coast of Western Australia. The known depth range is from the intertidal to 73 metres. REMARKS: A. M. Clark (1966) referred Parasterina troughtoni to the genus Nepanthia, at the same time synonymising P. occidentalis with it and describing a new species, N. hadracantha. A. M. Clark (1966) also showed that there is quite a significant difference between the form of the spines of the cool temperate species N. hadracantha and its tropical congeners. Shepherd (1968), after examining 47 specimens from Victoria and Western Australia, together with those collected during extensive underwater surveys of the coastal waters of South Australia, discussed the variation in arm taper, size and crowding of the abactinal plates and variation in the spinelets concluding that there was no evidence for maintaining N. hadracantha and synonymised it with N. troughtont. In comparing the holotypes of N. troughtoni, N. occidentalis and N. hadracantha the only detectable difference is in the slightly stouter, squatter shape of the spines of N. occidentalis, a difference not considered here of specific importance in view of Shepherd’s (1968) observations. Examination of a large series of specimens (99) shows considerable variation in the size of the abactinal plates and in the regularity of their arrangement. The specimens range in size from R/r = 16/4.5 mm (holotype of N. troughtoni) to 80/17 mm, while the R/r ratio varies from 3.3 to 5.9, the mean R measurement is 45.9 mm and the mode lies in the class interval 41-50 mm. The figures (Table 3) show a normal size distribution with a complete lack of very small specimens as with many other asteroids. No geographical differences can be detected in the size of specimens or in R/r ratios. There is considerable variation, however, in the length and breadth of arms within a population. Nepanthia nigrobrunnea n. sp. Figs lz 2a,%b; ce S13 oy, k. MATERIAL EXAMINED: 1 specimen (R/r = 65/15 mm = 4.3) holotype, AM AS J10147, Groper I, Coffs Harbour, N.S.W., on reef, 20 m, N. Colennany, September, 1976; 6 FRANCIS W. E. ROWE AND LOISETTE M. MARSH specimens (R/r = 69-75/14.5-16 mm = 4.6) paratypes, AM No. J9885, Julian Rocks, Byro Bay, N.S.W., 10-30 m, S. Parish, May, 1976; 3 specimens (R/r = 56-66/13-14 mm = 4.3-4.7, paratypes, AM No. J9135 locality as J9885, on reef, 18 m, N. Coleman, 30.3.75; 2 specimens (R/r = 53-64/14-15 mm = 3.5-4.4) paratypes, AM No. J10920, Groper I, Coffs Harbour, N.S.W., 16 m, J. Ogg and C. Short, 19.8.77. DIAGNOSIS: A species of Nepanthia with 5 subcylindrical to slightly tapering blunt ended) arms, non fissiparous. Maximum known R/r of 75/16 mm, mean R of 62 mm, range of R:r fro 4.3 to 5; colour dark brown with black papulae; abactinal plates somewhat irregular in shape, often double notched with 2-3 papulae (sometimes | or 4) to an area; crystal bodies, secondary, plates and pedicellariae absent; abactinal spinelets all over plates, short and stout with 12-160 points, the central one sometimes enlarged; furrow spines in a comb of 3-4 subequal spines with’ 6-7 subambulacral spines in 2 rows; sublittoral, 10-30 metres, on rock, known only from northern N.S.W., Australia. DESCRIPTION: The holotype has 5 arms, subcylindrical tapering evenly to a narrow but}, blunt up; R/r = 65/15 mm = 4.3; br = 15.5 mm at base and 9 mm at two-thirds R. The» madreporite is inconspicuous, lying in one interradius about 4 mm from the centre of the disc, The abactinal plates are all similar in size (about 1.1 mm maximum diameter). There are twot ‘fields’ of plates, the dorsal field, where the order is irregular, and a lateral field each side, with / the plates forming 10-11 longitudinal rows (at least proximally). The shape of the plates from the - dorsal field is variable, from transversely elongate to triangular or rounded (fig. 5i). The proximal edge of the majority of plates is concave so that the plates can be said to be generally crescentic, Proximally there are about 3 plates between the lateral fields, 8-10 plates at half R and 6-7 plates at the arm up. The plates of the lateral fields are more regularly triangularly crescentic. The first row extends to the arm tip, the succeeding rows extending to shorter distances so that the 11th row comprises only 3-4 plates at the arm base. The actino-lateral edge is rounded. The inferomarginal plates form a regular row of longitudinal plates (about 1.7 mm x 0.7 mm proximally) along the arm, becoming shorter, rounded and convex distally. The superomarginals are smaller than the inferomarginals and from the 12th-14th inferomarginal there are 2 small superomarginal plates aligned per inferomarginal. The superomarginals are irregular distally, and difficult to distinguish. The adambulacral plates bear 3-4 subequal, flat-tipped furrow spines backed by six or seven subambulacral spines arranged in 2 rows. The first actino-lateral row of plates extends almost to the arm tip. The second row comprises 7-8 plates, the third row 3-4 plates and 3-4 plates are present in the distal triangle of the actinal surface. The oral plates bear 5 furrow spines and 7-8 suboral spines. The proximalmost suboral spine is largest. The spinelets on the abactinal and actinal plates are spaced and coarse. Abactinally the largest plates bear up to 45 spines, the actinal plates bear about half of this number. The abactinal spinelets are 2.5-3 times as long as their maximum width (base) and some have a median, large, blunt process between the terminal points (fig. 6}, k). The spinelets are remarkably even in size (360-375 pm long x 120-150 ym wide), The actinal spines are larger, 2.7-2.8 times as long as wide (570-600 pm long x 200-240 pm wide) (fig, 2c), Between the plates of the dorsal field are 1-4, usually 2-3, papular pores but between the plates of the lateral fields only 1 pore per plate occurs. There are no pores between the plates of the 10-11th rows of lateral plates or actinally. There are no fasciculate pedicellariae. Besides the holotype, there are eight paratypes which are similar in all respects to the holotype. COLOUR: In life the animal is very dark brown, with black papulae. HABITAT AND DISTRIBUTION: Known from Byron Bay to the Solitary Islands, New South Wales, Australia, in 10-30 m depth. ASTERINID GENUS NEPANTHIA GRAY 97 ETYMOLOGY: nigrobrunnea (Lat.) refers to the colour of the animal. REMARKS: N. nigrobrunnea is most closely allied to the Flindersian species N. troughtoni ‘rom which it is most easily distinguished by the arrangement of marginal plates, the shape of slates in the dorsal field, groups of 2-4 papulae and colour. These features alone would distinguish nigrobrunnea from its tropical congeners but additionally the shape of the spines Sovering the skeletal plates distinguish this species and troughtoni from the tropical species. The closer relationship of nigrobrunnea with the Flindersian species would indicate that though this species occurs in the northern parts of New South Wales (where there is a known overlap of tropical and warm temperate species) it can be considered a Peronian species. Its presently known restricted distribution is difficult to assess though a somewhat similar situation yecurs on the western coast of Australia where Nepanthia crassa occupies the whole west coast. It might be expected that nigrobrunnea will be found further south along the New South Wales coast. Whether nigrobrunnea is derived from Flindersian stock is difficult to determine. It might be speculated that troughtoni and nigrobrunnea have developed independently from the progression of a northern species southward along either side of the continent. However, the form of nigrobrunnea compares much more closely with that of the Flindersian troughtont than its tropical congeners belcheri and maculata. That nigrobrunnea has developed after a possible isolation from troughtoni due to the separation of populations by the Bassian isthmus during the Pleistocene epoch is possibly more likely than a development from the small, high!y fissiparous belcheri from the north, Nepanthia crassa (Gray) Figs 15. 3a, b;-ci 5g, ]3 6d, e. Patria? crassa Gray, 1847: 833; 1866: 17. Patina crassa. — Perrier, 1875: 326-327. Parasterina crassa. — Fisher, 1908: 90; 1940: 270-271. — H. L. Clark, 1923: 243; 1938: 179-180; 1946: 143. Nepanthia crassa. — Fisher, 1941: 451-455, figs 20, 21, pl. 70, fig. 2. non Patiria crassa. — Bell, 1884: 131. — Whitelegge, 1889: 201. MATERIAL EXAMINED: See Table 4 for a summary of data for the 127 specimens examined from the Australian and Western Australian Museums. DIAGNOSIS: A species of Nepanthia with 5 subcylindrical to slightly tapering, stout, blunt ended arms. Maximum known R of 72/17 mm, mean R of 40 mm, range of R:r from 2.8 to 5.1; colour variable, unicolorous blue, blue-green or orange, more often mottled browns or pink to red and brown; ‘papulae single; primary abactinal plates subcrescentic, imbricating when young, tumid to hemispherical, surrounded by numerous secondary plates when large; few crystal bodies on plate margins; fasciculate pedicellariae usually on plates of the lateral field; thorny abactinal spinelets with 5-8 points cover primary plates and occur in tufts on secondary plates; furrow spines in a graduated fan of 7-10 (usually 9) spines with a fan of 7-10 subambulacral spines and 5-7 spinelets; intertidal to 38 metres, on rock, sand and muddy sand, on the west coast of Western Australia, from Point Cloates to Cape Naturaliste. COLOUR: In colour N. crassa is variable, usually brownish, sometimes mottled with darker shades; a specimen from Shark Bay was pinkish-buff, mottled with red-brown and dark brown while at the Abrolhos Islands most specimens are blue or blue-green. 98 FRANCIS W. E. ROWE AND LOISETTE M. MARSH Table 4. Variation in size and R/r ratio of specimens of N. crassa from four areas in Western Australia. Distribution Number Max. Min, Range, Mean, Mode, (Western of R/r, R’r, R/r Rs R, Australia) specimens mm. mm. mm. mm. 22°40'S to 29°55'S 23 60/14 23/7 3.1-4.6 38.0 31-40 Abrolhos, 28-29°S 45 72/17 18/6 2.8-4.4 38.1 31-40 Fremantle area, 32°S 52 65/15 15/5 3.0-5.1 42.2 41-50 Geographe Bay, 3373575 7 56/14 34/9 3.7-4.7 43.6 41-50 127 72/17 15/5 2.8-5.1 40.5 31-40 Table 5. Size distribution (R) of the specimens of N. crassa in 10 mm class intervals. 1-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 0 4 24 44 38 16 3 ] 0 HABITAT AND DISTRIBUTION: N. crassa is found in sheltered bays, commonly under jetties and on piles but is also found on sand amongst seagrass, mud or amongst algal covered rocks. In the Abrolhos N. crassa occurs near islands on the western platforms e.g. around Gun Island (Pelsaert group), Rat I. (Easter Group) and Pigeon I. (Wallabi group) in areas protected by seaward reefs from much wave action. The substrate is rock with a thin covering of algae and silty sand.-During the day the animals are concealed in crevices or under dead coral slabs and are active at night on the reef platform. They are rarely found elsewhere in the Abrolhos. N. crassa is confined to the west coast of Western Australia between Cape Naturaliste (33°35'S) and the Point Cloates area (22°40'S) where it is separated by North West Cape from populations of N. belchert and N.. maculata in Exmouth Gulf. The recorded depth range is from the intertidal to 38 metres. REMARKS: The history of generic changes undergone by this endemic Western Australian species are given in Table 1 Fisher (1940, 1941) re-described N. crassa, showing (1941, figs 20, 21) the imbrication of the abactinal plates, thus removing the main character by which his genus Parasterina was distinguished from Nepanthia. H. L. Clark (1946) stressed the shape of the abactinal plates as characterising Parasterina crassa, troughtoni and occidentalis. Since the latter two species have already been included in Nepanthia (A. M. Clark, 1966) and the shape of the plates does not differ significantly in small specimens of crassa, H. L. Clark’s view cannot be upheld. In small specimens of N. crassa, e.g. WAM 54-79, which has R/r of 30/9 mm, (fig. 5j), the abactinal plates imbricate regularly and are notched for papulae; they are slightly tumid with secondary plates only around the papular pores and few tertiary plates in the dorsal field. The plates of the lateral field also imbricate regularly but most lack secondary plates. A large specimen, WAM. 625-75 with R/r of 57/15 mm, (fig. 5g), has the primary abactinal plates ASTERINID GENUS NEPANTHIA GRAY 99 widely separated, linked by secondary plates with numerous small tertiary plates in the dorsal field and 2-5 small plates around the papulae of the lateral field. In this specimen the primary plates are not greatly enlarged nor very tumid but a few examples e.g. WAM. 756-75, (fig. 3a), the primary plates are almost hemispherical, on short stout arms. The abactinal and actinal plates are covered in a dense coat of thorny spinelets which radiate from the small secondary and tertiary plates but give an even coating to the primary plates. Detail of the abactinal spinelets is shown in figure 6d, e; they most closely resemble those of N. belcheri. The furrow armature, actinal surface and pedicellariae all closely resemble N. belcheri, pointing to a close relationship of the two species. The fact that both Bell (1884) and Whitelegge (1889) identified specimens of N. belcheri, from the east coast of Australia, as N. crassa, highlights the resemblance of the two species. The principal difference is in the shape of the abactinal plates, crescentic in N. belcher, very convex and rounded or irregular in shape in adult N. crassa, and in the presence of tertiary _as well as secondary plates in N. crassa. Among the specimens examined the arms vary in shape from slightly tapering to quite cylindrical, apart from a slight ventro-lateral angle, but are always blunt ended. The majority of specimens have moderately convex primary plates and slightly tapering arms. The variation in size of specimens from four areas on the Western Australian coast (Table 4) shows a geographical trend with higher mean and modal sizes (R) in specimens from the southern part of the range (Fremantle to Cape Naturaliste). Measurements of the 127 specimens as a whole, however, show a similar trend to that shown in N. troughtoni, but witha slightly lower mean R measurement, 40.5 mm, the mode falling in the class interval 31-40 mm (Table 5), there being a similar lack of very small specimens. Nepanthia belcheri (Perrier) Figs. 1; 3d, e; 5a, b; 6a, b. Asterina (Nepanthia) belcheri Perrier, 1875: 320. Asterina belcheri. — Bell, 1884: 131. erin (Nepanthia) brevis Perrier, 1875: 321. — Bell, 1884: 131, pl. 8, figs A, a. — Studer, Nepanthia suffarcinata Sladen, 1888: 328, pl. 28, figs. 9-12. — Koehler, 1910a: 133; 1910b: 288. — Clark and Rowe, 1971: 38 (distribution), 66 (key).. Nepanthia belcheri. — Sladen, 1889: 387. — H. L. Clark, 1938: 169; 1946: 141. — Kenny, 1969: 51, figs. 1-4. Endean, 1953: 54; 1956: 125; 1957: 240; 1961: 291. — Clark and Rowe, 1971: 38 (distribution), 66 (key). — Otteson, 1976. Nephanthia brevis. — Sladen, 1889: 387, pl. 63, figs 3-5 — Déderlein 1896: 40. — H. L. Clark, 1921: 95, pl. 6, figs 3-4 (col.); 1938: 172; 1946: 141, — Endean, 1953: 54; 1956: 125; 1957: 240; 1965: 230. — Clark and Rowe, 1971: 38 (distribution), 66 (key). Patria crassa. — Bell, 1884: 131. — Whitelegge, 1889: 201 (non P. crassa Gray). Nepanthia joubini Koehler, 1908: 232, figs 1-4. — Fisher, 1919: 423, pl. 113, figs 1-2. — H. L. Clark, 1938: 172 (footnote). — Clark and Rowe, 1971: 38 (distribution), 66 (key). — Domantay, 1972: 55. Henricia heteractis H. L. Clark, 1909: 530, pl. 49, figs 1-2; 1926:2. Nepanthia polyplax Déderlein, 1926: 20, pl. 4 figs 2-2a. 100 FRANCIS W. E. ROWE AND LOISETTE M. MARSH Nepanthia ? brevis. — Livingstone, 1932: 262, pl. 5, figs 8-9. Nepanthia magnispina H. L. Clark, 1938: 174, pl. 20, figs 1-2; 1946: 142. Nepanthia variabilis H. L., Clark, 1938: 176, pl. 10, figs 4-5 (col.), pl. 20, figs 4-5; 1946: 141. — Fisher, 1941; 454, fig. 22. — Clark and Rowe, 1971: 38 (distribution), 66 (key). MATERIAL EXAMINED: | specimen (R/r = 15/3.8 mm = 3.9) holotype of N. belchen, BM No. 1847.3.10.4; 1 specimen (R/r = 24/5.7 mm = 4.2) holotype of N. brevis, BM 1854.11,15.290; 1 specimen (R/r = 25/8 mm = 3.1) holotype of Henricia heteractis, AM No, G11430; 1 specimen (R/r = 31.9/8.5 mm = 3.8) holotype of N. magnispina, MCZ No. 3230; 2 specimens (R/r = 30-50/6.7-10 mm = 4.5-5.0) paratypes of N. vaniabilis, MCZ No. 322532 specimens (R/r = 40-47/10-11 mm = 3.6-4.7) paratypes of N. variabilis, WAM No. 119-39, atc 4 specimens (R/r = 32-62.5/7.5-12.5 mm = 4.3-5.3) paratypes of N. variabilis, AM No, 6187. In addition 138 specimens from the Western Australian and Australan Museums were examined. A summary of data for the 150 specimens is given in Table 6. DIAGNOSIS: A fissiparous species of Nepanthia with 4-7 (usually 5-6) subcylindrical to slightly tapering, moderately stout arms. Maximum known R’r of 65/14 mm, mean R of 31 mm, range of R:r from 2.0 to 5.3; colour highly variable, often grey-green or fawn ground colour mottled with red, green, brown or black; papulae single, lying in notch of abactinal plates which have a raised crescentic ridge carrying spinelets, with crystal bodies on the lower parts of the plates; fasciculate pedicellariae usually present on plates of the lateral field; abactinal spinelets, with 3-5 acute points, on ridges of primary plates and on 2-4 secondary plates, adjacent to the papulae; furrow spines in a graduated fan, usually 8-9, with a subambulacral fan of 7-12 spines and 3-7 additional spinelets; intertidal to 46 metres on a muddy sand and rock substrate, Burma to northern Australia. COLOUR: Highly variable, often grey-green or fawn ground colour mottled with red, green, brown or black. HABITAT AND DISTRIBUTION: N. belcheri is found from the intertidal to a known depth of 46 metres, Near low tide mark it may be found clinging to the underside of boulders on a muddy sand or rock substrate. The species appears to be always associated with somewhat muddy conditions. Geographically N. belcheri ranges from the tropical coast of Australia, northwards to the Philippines, Cochin China (Vietnam) and west to Burma. REMARKS: Although he was convinced that his species Henricia heteractis from Lord Howe Island and Déderlein’s Nepanthia polyplax from Rockhampton, Queensland, were conspecific with Nepanthia belcheri, H. L. Clark (1938) described two new species N. magnispina and N. variabilis from north-western Australia. The similarity of specimens of N. variabilis and N. brevis led the present authors separately to question the basis for separating them and a comparison with specimens of N. belchen indicated that this species could not be distinguished morphologically from the other two species. The results of an examination of 150 specimens of the three nominal species, ranging from Lord Howe Island and Port Jackson, N.S.W. through Queensland and Torres Strait to Exmouth Gulf, Western Australia, including the holotypes of N. belcheri, N. brevis, H. heteractis and N. magnispina and eight paratypes of N. variabilis, are presented in Table 6. It is evident that two of the key characters used by Clark (1938) to separate N. belcheri, N. brevis and N. variabilis i.e. the number of arms and number of furrow spines, do not provide a basis for distinguishing them. Samples from the three areas, east coast of Australia, Torres Strait to Darwin and north western Australia (Table 6), corresponding to the distribution of the three nominal species ASTERINID GENUS NEPANTHIA GRAY 101 Table 6. Variation in Nepanthia belchert. Number of specimens East coast and Torres Strait Kimberley to Lord Howe I, to Darwin Exmouth Gulf i Size Rin mm 1-10 8 1 0 11-20 22 3 3 21-30 17 11 12 31-40 11 25 10 41-50 2 4 15 51-60 0 0 3 61-70 0 l 2 N = 60 45 45 Mean 22.2 33.4 38.1 Mode 11-20 31-40 41-50 Number of Furrow spines 5-6 6 0 0 6-7 16 0 2 7-8 13 1 1] 8-9 19 22 17 9-10 5 6 5 10-11 0 6 6 IN} = 5O) 45 4] Mean 7.5 8.8 8.5 Mode 8-9 8-9 8-9 ' Number per plate, Abactinal spines <30 8 i 0 31-40 35 5 5 41-50 13 ly 3 51-60 1 12 13 61-70 1 5 7 71-80 0 4 8 >81 il 0 7 Ne—59) 44 43 Mean 38.15 SILLS 61.56 Mode 31-40 41-50 51-60 Number of arms 4 0 1 2 5 21 38 39 6 28 3 4 7 11 3 0 102 FRANCIS W. E. ROWE AND LOISETTE M. MARSH N = 60 45 45 Mean 5.8 Sar Sell) Mode 6 5 5 R/r ratio Max. Rir 46.5/15.0 63.0/13.5 65.0/14.0 Min. R/r 8.0/3.0 10.0/3.0 12.0/4.0 Range R/r 2.0-5.0 2.8-4.4 3.0-5.3 Table 7. Percentage of specimens of N. belcheri with various arm numbers. N <5 5 6 7 >7 Moreton Bay (Kenny, 1969) 837 10 8 55 25 ] Townsville (Otteson, pers. comm.) 613 22 =2 25 aa <6 East Coast Queensland OLE: 0 35 47 18 0 Torres Strait to Darwin 45 2 84 ft 7 0 Kimberley to Exmouth Gulf 45 4 87 a 0 0 respectively, show close similarity in the number of furrow spines, all having a modal number of 8-9. The slightly lower mean number in the east coast population is related to the smaller mean and modal R measurements of these specimens. The mean and modal size of specimens increases progressively northwards and westwards, with a corresponding decrease in fissiparity. The sample of the east coast population is 65% multibrachiate and 35% five rayed, the north coast sample is 13% multibrachiate and 84% five rayed and the northwest sample is 9% multibrachiate and 87% five rayed. A few specimens from northern Australia and the northwest are four rayed. A comparison of the arm numbers of specimens from Moreton Bay, Townsville and the present, smaller, sample from the whole east coast of Queensland (Table 7) shows a preponderance of 6 rayed specimens in Moreton Bay and on the east coast overall while 7 rayed specimens predominate at Townsville. A much higher proportion of 5 rayed specimens occur in the present sample from the Queensland coast than in the other two studies. The reason for the preponderance of small fissiparous specimens in the east coast population is unknown. Kenny (1969) showed that members of the Moreton Bay population of N. belcheri reproduce asexually by fission in their second or third year. Otteson (1976), studying the reproductive pattern of N. belchen, found the Townsville population to be serially protandric hermaphrodites in which sexual reproduction appeared to be modified by fissiparity. A parallel to the east coast population of small sized fissiparous N. belcheri is found in Coscinasterias acutispina Stimpson, where a dwarf race, reproducing fissiparously was reported from the island of Maui, Hawaii while normal specimens were found in Kaneohe Bay, Oahu, Hawaii (Edmondson, 1935). Edmondson also referred to a small fissiparous Nepanthia? sp. However, one of us (F.W.E.R.) has re-examined this specimen and found it to represent Asterina burtoni Gray. Clark’s third distinguishing character, the shape of the abactinal plates, varies within each population and is not a reliable distinguishing character. The number of spinelets per abactinal ASTERINID GENUS NEPANTHIA GRAY 103 plate also varies within each population, but since it is related to the size of the specimen, the number is higher in the populations of larger individuals from north and northwestern Australia (Table 6). The shape of the abactinal spinelets is identical in the three populations (Figs 6a, b). The type locality of N. belcheri is uncertain though Bell (1884) recorded specimens from Port Jackson, N.S.W. and there are three specimens in The Australian Museum and one in the Stockholm Museum from Lane Cove River, Port Jackson, N.S.W. These were reported by Whitelegge (1889) as Patiria crassa, redetermined by H. L. Clark as Henricia heteractis (1926) and later as Nepanthia belcher (1938). No other records of this species exist from New South Wales. A. M. Clark (pers. comm.) has thoroughly researched the literature and found that the type locality for the species could not be New South Wales since Belcher did not visit Australia on his voyages. His specimens were more likely to be from Indonesia. Its occurrence in the Moluccas is confirmed by Koehler (1910a), as N. suffarcinata (2 five rayed specimens taken from 15 m in Aru) and by specimens recorded here from 25-45 m in the Aru Islands. Nepanthia joubini was described from Cap St. Jacques, Cochin China (Vung Tau, Vietnam). Both Fisher (1919) and Clark (1938) doubted the validity of N. joubini, Fisher considering it to be probably conspecific with N. brevis and Clark with N. belcheri. A. M. Clark (in Clark and Rowe, 1971) also noted that N. joubini would probably prove to be conspecific with N. belchert. Livingstone (1932) identified a specimen of N. belcheri, with 6 rays and 5 madreporites, from the vicinity of Low Islands, North Queensland as N. brevis, remarking on the resemblance to Fisher’s example of N. joubint. A. M. Clark (1971) also indicated that N. suffarcinata might be conspecific with brevis or variabilis. In comparing Sladen’s (1888) description of N. suffarcinata, collected from Burma, with our data, we consider the species to be conspecific with N. variabilis or N. brevis, and therefore falling within the range of N. belcheri. Examination of the holotype of N. magnispina by one of us (F.W.E.R.) shows it to be virtually identical with specimens of N. belcheri from Queensland and New South Wales. There is evidence of a third regenerating arm (H. L. Clark notes Only 2), a small madreporite is present (H. L. Clark missed it) and the number of spines on the abactinal plates and adambulacral plates is similar to the specimens of N. belcheri collected from Lane Cove River, N.S.W. The “enlarged oral spine” recorded by H. L. Clark, is not particularly prominent and falls certainly within the range of variation of sizes of spines seen within the 150 specimens examined. This study has led to the conclusion that N. brevis, N. suffarcinata, N. joubini, N. magnispina and N. variabilis should be regarded as junior synonyms of N. beleheri, a somewhat variable species occurring on the east, north and northwest coasts of Australia from Lord Howe Island and southern Queensland to Exmouth Gulf and in Indonesia, the Philippines, Vietnam and Burma. Nepanthia fisheri n. sp. Figs 1; 4a, b, c; 5d, h; 6g. «Nepanthia maculata. — Fisher, 1919: 423 (part) (non N. maculata Gray). MATERIAL EXAMINED: 2 specimens, the holotype, WAM 102-78, and paratype WAM 101-78, from west Banguey channel, northeast of Sabah, dredged on mud, 25 fms (46 m), B.R. Wilson on ‘Pele’, 11.II1.1964; 1 specimen WAM 42-69 from 8 km west of Agal Bay, N.W. Sabah, dredged on mud, 16 fms (29 m), B. R. Wilson on ‘Pele’, 12.III.1964; 1 specimen WAM 100-78, from 14 km west of Cape Melville, Balabac I., Philippines, dredged on sand, 27 fms (49 m), B. R. Wilson on ‘Pele’, 9.IIT. 1964; 2 specimens, USNM40290 & 40371, from Linapacan Strait, N. of Palawan I., Philippines, 11°37’ 15°’ N: 119° 48° 45” E, Albatross st. 5335, 46 fms (84 m) sand, mud, 18.XJI.1908 (included in N. maculata by Fisher, 1919); 1 specimen AM 104 FRANCIS W. E. ROWE AND LOISETTE M. MARSH J12649, from st. 1154, 11° 48’ S: 129° 19° E, 68 m (Timor Sea), R. Martin (CSIRO). DIAGNOSIS: A species of Nepanthia with 5 regular, strongly tapering arms, not known to be fissiparous. Maximum known R’/r of 40/9.5 mm, mean R of 19 mm, range of R:r from 3.4 to 4.1; colour of holotype blue-grey; abactinal plates broadly crescentic slightly notched for single papulae with 2-3 raised areas bearing spinelets separated by lower areas with crystal bodies; 1-2 secondary plates usually occur in the papulae areas; abactinal spinelets in radiating groups of short sharp single pointed spinelets on the raised areas of each plate and on the secondary plates; furrow spines in a graduated fan of 8 with a fan of 8 subambulacral spines and 10-15 additional spinelets; sublittoral on mud or sand, 29-84 metres, Philippines to the Timor Sea. DESCRIPTION: The holotype has 5 subequal tapering arms, R/r = 38-40/9.5 mm = 4.1, br = 9-10 mm at base of ray decreasing to 5 mm at half R and 2.5 mm just proximal to the terminal plate; R/br at base = 4.1. Rays elongate, broad at the base, tapering to a narrow up. Abactinal surface of the disc and rays convex, depressed interradially, interbrachial arcs acute; actinal surface plane but margin of disc and rays not distinct. There are two madreporites both situated nearer the centre of the disc then the margin, one is radial in position, 2 mm in diameter, the other interradial and 1.5 mm in diameter. The anus is central, surrounded by about 8 small granules. The abactinal spinelets are in radiating groups arising from 2 or 3 raised areas on each plate, giving the impression of a large number of small convex plates. The imbricating abactinal plates are arranged in dorsal and lateral fields; when denuded the exposed part of the plates of the dorsal field are seen to be broadly crescentic, 1 to 1.2 mm in diameter on the proximal part of the arm, and slightly notched for single papulae, beside which are two small rounded supplementary plates. Secondary plates are absent distally. Plates of the dorsal field are fairly regularly arranged with alternating transverse rows of plates across the ray. There are 6 plates in a diagonal series across the dorsal field at half R. Plates of the dorsal field have 2-3 raised areas separated by lower areas with embedded crystal bodies. Each raised area bears a radiating group of 20-25 short, sharp spinelets, tapering to a single point while each secondary plate has a group of 5-10 similar spinelets, fig. 6g. The lateral field of plates is in six longitudinal rows at the base of the ray decreasing to three at half R. Two rows extend to the arm end. The terminal plate is rounded, 1 mm in diameter. The spinelets of the lateral arm plates are arranged in a horseshoe shape proximally giving the impression that they surround a pedicellaria pit but no modified or enlarged spinelets are present, nor are pits present on the denuded plates. Superomarginal plates, small and rounded, alternate with elongate, angled inferomarginals; occasionally small supplementary plates lie between the superomarginals. The proximal end of each inferomarginal slightly overlaps the distal end of the preceding one; they project very slightly, scarcely forming an angled margin to the rays. Interradially they are less conspicuous and tend to lie on the actinal surface. Spinelets on the marginals are similar to those of the abactinal plates, about 20 on each superomarginal and 40 to each inferomarginal, proximally, Actinal plates in 4 series at base of rays with an extra plate or two interradially in some arm angles. The innermost row of squarish plates, each opposite a similar adambulacral plate extends nearly to the arm end; distally they become compressed and similar in shape to the inferomarginals; the second row extends to about half R, with odd plates extending further; the third row extends to between one third and half R or to about the 15th inferomarginal; the fourth row extends to the 5th inferomarginal and three interradial plates represent a fifth series of actinal plates. The convex actinal plates are closely covered by about 50 radiating, short, ASTERINID GENUS NEPANTHIA GRAY 105 pointed, glassy spinelets. Adambulacral plates bear a webbed fan of 8 graduated furrow spines, of which the central one is up to 1 mm in length, followed by a webbed fan of 8 slightly smaller, blunt subambulacral spines while 10-15 tapering, pointed spinelets cover the remainder of the plate; these spinelets are considerably thicker than those of the adjacent actinal plates (fig. 4c). Oral plates have a similar armature to the adambulacrals each with a marginal series of 8 spines, of which the innermost pair are longer and stouter than the remainder; there are 6-7 suboral spines, of which the last two are very small and 10-14 smaller thorny spinelets on the actinal surface of the plates. The paratype (WAM 101-78) has five equal rays and one madreporite, R/r = 20/5.5 mm = 3.6, br = 5.5 mm at base of ray, 4 mm at half R (Fig. 4b). The specimen is similar to the holotype although little more than half the size. The arms are less attenuated, there are 4 instead of 6 series of plates in the lateral field at the base of the rays and fewer papulae; secondary abactinal plates are few and scattered. The innermost row of actinal plates extends nearly to the arm end, the second row varies between the 7th to 12th inferomarginal, the third row extends to the 2nd or 3rd inferomarginal and the fourth is represented by 1-3 plates in the arm angle. The furrow and oral spines are as described for the holotype. OTHER SPECIMENS: Among the other specimens, WAM 42-69 has five equal rays anda single madreporite, R/r = 18/5 mm = 3.6, br = 5.5 mm at base of ray, 4 mm at half R. This specimen differs from WAM 101-78 only in having the plates of the dorsal field on the rays less | distinctly subdivided. WAM 100-78 has six rays, two of 11 mm Rand four of 8 mm, r = 3 mm, R/r = 3.7, br at base = 3 mm, 2.5 mmathalf R. There are two madreporites, The characteristic features of the species are less developed in this small specimen but it resembles the larger | specimens more closely than it does N. belchen. The characteristic appearance of the abactinal plates, their surface subdivided and bearing tufts of spinelets is progressively less clearly seen in the smaller specimens and is scarcely distinguishable in the smallest. One specimen (AM J12649), from the Timor Sea, has five unequal rays, two of 27.5 mm, | two of 26.5 mm and one of 12 mm, r = 5-7 mm, R/r = 2-3.5, br at base = 7 mm, 4 mm at half R. There are 2 madreporites. The appearance of the abactinal plates is characteristic. However, there are often 3 secondary plates proximal to the crescentic abactinal plates. Two small specimens (USNM 40290 and 40371) from the Philippines were doubtfully referred by Fisher (1919, p. 423) to N. maculata. These are both five rayed with a single madreporite and anus and have R/r of 12/3.5 mm and 15/4 mm. Fisher noted that ““A peculiarity of these two specimens is the grouping of spinelets of the abactinal crescentic plates in 3 or | sometimes 2 distinct tufts to each plate. This gives the appearance of numerous small plates. These small specimens are distinct from N. brevis and N. suffarcinata, and of course may represent a third species. Their affinities are close to N. maculata’’. These two specimens have been examined and are here referred to N. fisher. Had Fisher seen a larger specimen he would have been in no doubt that they represented an undescribed species. Fisher’s specimens agree closely with the holotype and other specimens examined. COLOUR: Uniform blue-grey with madreporites cream in life. HABITAT AND DISTRIBUTION: The holotype and paratypes were all taken in the area north of Sabah, Borneo, dredged on mud or sand at 29-49 metres. Fisher’s specimens were taken 106 FRANCIS W. E. ROWE AND LOISETTE M. MARSH north of Palawan Island, Philippines, on sand and mud at 84 metres. ETYMOLOGY: The species is named in honour of W. K. Fisher. REMARKS: Nepanthia fishert has affinities with both N. belcheri and N. maculata but hase very distinct facies of its own. It differs from N. maculata in having tapered arms and in the character of the abactinal plates which bear tufts of spinelets on two or three raised areas on each plate in contrast to the uniform covering of spinelets on each plate in N. maculata. Like N. maculata the pointed spinelets are in radiating groups but are shorter and less numerous in N, fishert. The sumilarity in general form of the small specimens to those of N. belchert points to the close relationship between the two species. N. fisheri has crystal bodies embedded in the skeletal plates, as in N, belchen, The adambulacral armature of N. fisher’ differs little from either species. The fact that one specimen is six rayed and that the holotype has two madreporites indicates that the species is potentially fissiparous although the other specimens are five rayed with a single madreporite. The most distinctive feature of N. fisheri, apart from the shape, is the grouping of the short, sharp glassy spinelets on several raised areas on each abactinal plate. A comparison with N, gracilis n. sp. is made under the latter species. Nepanthia maculata Gray Figs 1; 2f; 5e; 6h,i. Nepanthia maculata Gray, 1840: 287; 1866: 15. — Studer, 1884: 42. — Sladen, 1889: 388, pl. 64, figs 1-4, — Fisher, 1919: 422, pl. 113, figs 3,4. — A. M. Clark, 1956: 377, text fig. 2. — Clark and Rowe, 1971: 38 (distribution), 66 (key). — Domantay, 1972:55, Chaetaster cylindratus Mobius, 1859: 3, pl. I, figs 3, 4. Asterina (Nepanthia) maculata.— Perrier, 1875: 322, Nepanthia tenuis H. L. Clark, 1938: 175, pl. 20, fig. 3; 1946: 142.— Clark and Rowe, 1971: 38 (distribution), 66 (key).— Rowe and Pawson, 1977: 347. MATERIAL EXAMINED: 1 specimen (R/r = 38.6/6.8 mm = 5.7) holotype of N. maculata, BM No, 1953. 4.27.40, Migupou, Philippines 7-12 fms (2-5.5 m), fine sand, coral; 1 specimen (R/r = 38,5/5.5 mm = 7) paratype of N. tenuis, AM No. J6176, Broome, Western Australia; 2 specimens (R/r = 94/15 and 83/10 mm), Darwin Museum, off Weipa, Gulf of Carpentaria; 1 specimen (R/r = 75/11 mm) Darwin Museum, off Christmas Creek, Gulf of Carpentaria, 14°30'S: 141°30’E; 1 specimen (R/r = 72/13 mm), Darwin Museum off Tasman Pt., Gulf of Carpentaria; 1 specimen (R/r = 47/8 mm), Exmouth Gulf, W.A., WAM 1825-75, coll. J. Penn on ‘Flinders’ 1. VIII.1975, trawled 9-20 m; 1 specimen (R/r = $0-70/13 mm), 6 km west of Dampier, W.A., WAM 587-75 coll. L. Marsh, 28.X.1972, on muddy sand flat, exposed at low spring tide; 2 specimens (R/r = 50/7 mm and 35/6.5 mm), N.E. of Malus I., Dampier Archipelago, W.A., Mariel King Exped., 31.V.1960, WAM 586-75, dredged on sandy rubble, 18 m; 2 specimens (R/r = 38/6 mm and 23/4 mm) 11-16 km W.N.W., of Cape Melville light, Balabec I., Philippines, WAM 585-75, coll. B. R. Wilson on ‘Pele’, 9.III.1964, dredged on coarse sand, 37-49 m; 1 specimen (R/r = 44/6 mm) off Elat Bay, west coast of Nuhu Tyut, Kai Is., Indonesia, st. KN IT, 5°40’S: 132°59'E, WAM 57-79, M. King Mem. Exped., 13 VI.1970, dredged on sand and rubble, 49-84 m; 1 specimen (R/r = 31/4 mm) north of Du Rowa I., Kai Is., Indonesia, st. KR VI/1, 5°32'S: 132°41'E, M. King Mem. Exped., 10.VI. 1970, dredged on sand, 33-37 m; 1 specimen (7 rays, R/r = 18-28/r mm) off Tg Tutuhuhur, Piru Bay, Ceram, st. CPI/1-6, 3°15'S: 128°8'E, M. King Mem. Exped., 1.VI.1970, dredged on coarse sand, 42-64 m, ASTERINID GENUS NEPANTHIA GRAY 107 DIAGNOSIS: A species of Nepanthia usually with 5 regular, subcylindrical arms, non fissiparous. Maximum known R’r of 94/15 mm, mean R of 51 mm, range of R:r from 5.4 to 8.3; colour cream to buff unicolorous or with the central disc area dark blue, brown or green with a few flecks of the same colour on the arms; abactinal plates broadly crescentic to rhomboidal, notched for single papulae and often pitted for pedicellariae, a few crystal bodies on plate margins, often absent; fasciculate pedicellariae, when present, in the dorsal field; secondary plates few, usually absent; abactinal spinelets slender, tapering with 1-3 (rarely 4) acute points, all over plates; furrow spines in a graduated fan of 7-8 with a fan of 9-12 subambulacrals and 9-12 additional spinelets; intertidal to 84 metres, on muddy to coarse sand with rubble or coral, Philippines to northern Australia. COLOUR: Specimens from the Philippines and Indonesia vary in colour from cream witha large dark blue to violet spot on the disc centre and small spots on the arms to mottled light orange and cream with a few dark brown spots on the arms or light brown witha dark brown spot in the centre of the disc and small brown spots on the arms. Northern Australian specimens are either uniformly cream or buff with or without dark brown spots. The holotype of N. tenuis was light grey with scattered flecks of deep green. HABITAT AND DISTRIBUTION: N, maculata is a rather uncommon species found on mud, sand or sand and rubble bottoms from the intertidal to at least 84 metres. Studer’s record of a specimen of R = 10 mm from 400 fms (731 m), McCluer Gulf, New Guinea is questionable since the maximum chart depth in the Gulf is 56 fms (102 m) although deep water (to 1000 fms) is found between the Gulf and Ceram. N. maculaia is known from the Philippines, Moluccas and northern Australia, from the Gulf of Carpentaria to Exmouth Gulf, Western Australia. REMARKS: A. M. Clark (in Clark & Rowe, 1971) commented that N. tenuis might prove to be a synonym of N. maculata. Direct comparison of the holotype of N. maculata and a paratype of N. tenuis shows them to be conspecific, neither possessing secondary abactinal plates, and both being similar in size. The presence of secondary plates may well be related to size of the animal since large specimens from northern Australia (Gulf of Carpentaria and northwestern Australia) possess scattered secondary plates. The other characters (R/r ratio and number of furrow spines) used in H, L. Clark’s 1938 key do not provide distinguishing features, since the R/r ratio of one of the paratypes of N. tenuis is identical to that of the holotype of N. maculata. Likewise, although the holotype of N. tenuis has 5-6 furrow spines, a paratype has 7-8. The R/r ratio of specimens of N. maculata examined varies from 5.8 to 7.8 in those from the Philippines and Indonesia and from 5.5 to 8.3 in those from northern Australia. There are 8 furrow spines in the specimens from the Philippines and Indonesia, 7 to 8 in those from northern Australia. We can see no valid reason, therefore, for not considering N. tenuis and N. maculata to be conspecific. Pedicellariae, not previously described in N. maculata, occur on some of the northern Australian specimens, usually on the dorsal field of arm plates with a few on the disc and lateral field of arm plates. The fasciculate pedicellariae consist of 4-7 stout, thorny, tapering spinelets surrounding a furrow on the plate just distal to the papular pore. Nepanthia gracilis n. sp. Figs 1; 4d, e, f; 5f; 6f. MATERIAL EXAMINED: Two specimens, the holotype, WAM 103-78 from 14 km and 242° from Zal I., $.W. of Pearl Bank, Sulu Sea, Philippines, dredged from 122-124 m, heavy sponge, B. R. Wilson on ‘Pele’, 22.X1.1964 and the paratype WAM 104-78, from 15 km and 242° from ZalI.,S.W. of Pearl Bank, Sulu Sea, dredged from 100 to 110 m, heavy sponge, B. R. Wilson on ‘Pele’, 22.X1.1964; two specimens (R/r of 65/14 and 57/12 mm) trawled off Crowdy Head, N.S.W., 31°59’S: 152°57’E to 31°56'S: 152°58’E, ‘Kapala’ st. 78.05.08, 110 m, AM No. 108 FRANCIS W. E. ROWE AND LOISETTE M, MARSH J11880. DIAGNOSIS: A species of Nepanthia with 5 regular strongly tapering arms, non fissiparous. Maximum known R/r of 65/14 mm, mean R of 49 mm, range of R:r from 4.6 to 5.3; colour unknown, cream when dry; abactinal plates triangular to rhomboidal, flat except for bosses for spinelets, sometimes a few crystal bodies on plate margins; single papulae; no secondary plates; abactinal spinelets subcylindrical, with thorny tips (7-8 acute points), all over plate; furrow spines in a comb of 4-5 subequal spines and 6-7 shorter subambulacral spines sometimes arranged in two rows; sublittoral, 100-124 metres, with sponges; Philippines and eastern Australia. DESCRIPTION: The holotype has 5 equal tapering arms, R/r = 37/7 mm = 5.3, br = 7 mm at base of ray decreasing to 4 mm at half R and 1.5 mm just proximal to the terminal plate (Fig. 4d, e, f). Rays elongate, pointed, tapering from the base to a very narrow tip. Abactinal surface of the dise and rays convex, depressed interradially; interbrachial arcs acute; actinal surface plane but margin of disc and rays not distinctly angled. The single madereporite is interradial in position, nearer the margin than the centre of the disc. The anus is central, concealed amongst the dise spinelets. The skeletal plates are covered in slender spinelets standing vertically, not radiating. Arm skeleton composed of dorsal and lateral fields of imbricating plates; the exposed portion of those of the dorsal field varies in shape from broadly crescentic to rounded or squarish, often irregular in outline; surface of plates flat except for minute bosses for spinelet attachment; few crystal bodies. The plates are scarcely notched for single papulae; secondary plates absent (fig. 5f). There are 3 plates across the dorsal field at base of ray, 6 in a diagonal series at about half R; near the arm end dorsal and lateral fields not clearly distinguished. At the base of the ray there are 10 rows of triagnular to squarish plates in the lateral field, decreasing to 6 at half R; distally the series are no longer differentiated from the dorsal field but on some rays one row extends to the arm end. Papulae occur singly between all plates of the dorsal and lateral fields except near the arm tip and in the arm angle. Plates of the dorsal field bear 25-35 slender, more or less cylindrical spinelets with thorny tps (fig. 6f) while those of the lateral field have 15-20 similar spinelets, near the base of the ray, decreasing in number distally. No pedicellariae. Superomarginal plates large and rounded, more prominent than inferomarginals, usually lying opposite them, occasionally alternating; inferomarginals elongate, angled, proximal end of one overlapping the distal end of the preceding one; inferomarginals lie entirely on the actinal surface of the rays so that the superomarginals form the ventrolateral margin of the arms and disc although not forming a conspicuous angle; spinelets on marginals similar to those on abactinals, about 20 per superomarginal and 16 per inferomarginal. Actinal plates in 5 rows, with | or 2 plates of a 6th row, at base of ray; innermost row extends to 0.8 R (S mm from arm tip), second row to nearly half R or the 18th inferomarginal, third row to 9th or 10th inferomarginal, fourth row to 4th inferomarginal, with 1 or 2 plates in the arm angle representing a 6th row. The actinal plates are moderately convex and carry 5-10 slender thorny tipped spinelets standing vertically. Adambulacrals usually bear 4 (3-5) elongate cylindrical furrow spines, up to 1 mm in length, the middle two slightly longer than the others; 6-7 similar but shorter subambulacral spines, sometimes arranged in two rows. Oral plates each have a marginal series of 6 spines decreasing in size from the innermost towards the furrow and 7 elongate suboral spinelets. ASTERINID GENUS NEPANTHIA GRAY 109 The paratype (WAM 104-78) has R/r of 37/7 mm = 5.3, br = 7 mm. It is identical to the holotype except for the presence of 2 small secondary plates in the dorsal field on one ray and the furrow spines are more frequently in combs of 5 rather than 4. OTHER SPECIMENS: Two specimens from N.S.W. are provisionally referred to this species. They are badly distorted and not well preserved but agree in most respects with the description of N. gracilis. The abactinal plates of the dorsal field are less regular in shape and arrangement, and are more numerous than in the holotype or paratype but near the ends of the arms the cleaned plates match those of the holotype very closely. At the base of the ray there are 8 rows of actinal plates in the N.S.W. specimens compared with 5-6 in the holotype of N. gracilis. These differences could well be attributable to the larger size of the N.S.W. specimens. The abactinal and actinal spinelets and the furrow spines agree closely with N. gracilis. The specimens are referred to N. gracilis with some hesitation but they are not sufficiently distinctive to describe as new; further specimens should clarify the position. They were taken at the same depth as the Philippines specimens and the occurrence of N. gracilis at 31°S is not impossible. COLOUR: Not recorded in life, but dry it is cream. HABITAT AND DISTRIBUTION: This species is known only from the two specimens described above, both taken with sponges from 100 to 124 m near Pearl bank in the Sulu Sea and two specimens provisionally referred to this species, from 110 m, off N.S.W., Australia. ETYMOLOGY: The species is named from the Latin gracilis in reference to the slender, tapering arms. REMARKS: N. gracilis resembles N. fishert in size and form but closer examination shows them to be very different. The disc of N. gracilis is smaller and the rays more slender than those of N. fisheri but the most distinctive difference is in the nature of the spinelets which are cylindrical and slender, standing more or less vertically in N. gracilis in contrast to the short, tapering acutely pointed spinelets radiating from 2-3 elevations on cach abactinal plate in N. fisheri. Crystal bodies are present in N. fisheri, absent in N. gracilis. Papulae extend further into the lateral field in N. gracilis than in N. fishert and there are more rows of lateral plates in N. gracilis. Vhe superomarginals are more prominent than the inferomarginals in N. gracilis while the reverse is true of N. fisher. The furrow spines are in combs of 4-5 spines of nearly equal length in N. gracilis in contrast to the fans of 8 graduated spines in N. fisher. N. gracilis is most nearly related to N. briareus from which it differs in being non fissiparous and in the form and covering of the abactinal plates. In N. briareus the plates tend to be crescentic particularly on the proximal part of the arms and spinelets are borne on*the convex ridge of each plate. The proximal plates of N. gracilis are not convex or crescentic and tend to be rhomboidal or irregular in shape. The dise plates of N, briareus are smaller than the dorsal arm plates and irregular in shape, in N. gracilis they tend to be larger than the dorsal arm plates. The dorsal field of plates is more distinetly set off from the lateral field in N. gracilis than in N. briareus. The arrangement of marginal and actinal plates is similar in both species. The abactinal spinelets of both species are shown in figs 6f and 6c. The sponelets of N. gracilis are longer than those of N. briareus, with up to eight acute points in contrast to up to six rather blunt points in N. briareus. Nepanthia briareus (Bell) Figs 1; 4@3 5c; 6c. Patria briareus Bell, 1894: 404, pl. 25, figs 1-3. Nepanthia briareus.— A. M. Clark, 1956: 374-377, text fig. 1, pl. 10.— Jangoux, 1978: 297-298. 110 FRANCIS W. E. ROWE AND LOISETTE M. MARSH MATERIAL EXAMINED: | specimen, lectotype, Macclesfield Bank, South China Sea, 55 to 83 m. BM No. 1892. 8.22.267,7 rays, (3 long, R = c. 35 mm, 4 short, R = 17-20 mm,r= 6.5 mm, R/r = 5.4); 1 specimen from Taluk Dodinga, Halmahera, Indonesia, st. Hd 1/4-5, O'49'N: 127°31'E, M. King Mem. Exped., 20:V.1970, dredged with sponge and alcyonarians, 82 m, 7 rays, (3 long, R = 29 mm, 4 small, R = 17-20 mm, r = 6mm, R/r = 3.0) .a madreporites; 2 specimens, off Elat Bay, West coast Nuhu Tjut, Kai Is, Indonesia, st. KN II, 5°40’S: 132°59'E, M. King Mem. Exped., 13.VI.1970, dredged on sand and rubble, 27-46 m,7 rays, (4 long , R = 36-44 mm, 3 small, R = 25-26 mm, r = 9 mm, R/r = 5.0), 2 madreporites; and 7 rays, (3 long, R = 30 mm and 4 smaller, R = 25-26 mm, r = 7 mm, R/r = 4.3), 2 madreporites, WAM 56-79. DIAGNOSIS: A fissiparous species of Nepanthia with 7-10 (usually 7) slender, terete arms. Maximum known R/r of 44/9 mm, mean R of 34.4 mm, range of R:r from 4.3 to 5.4; colour uncertain, one specimen faded orange after drying from formalin; papulae single; abactinal plates oval, diamond shaped or irregular tending to have an oval or crescentic ridge bearing short thorny spinelets with 5-6 blunt points; no secondary plates, no crystal bodies; furrow spines ina comb of 4-5 with an oblique comb of 5 subambulacral spines followed by 1-4 spinelets. Found on sand and rubble sometimes with sponges and Alcyonaria; sublittoral, known depth range 27-83 m, South China Sea and Indonesia. COLOUR: The colour of the three Indonesian specimens (after drying from formalin) is pale orange. HABITAT AND DISTRIBUTION: This apparently uncommon species is known only from the South China Sea, Philippines (Jangoux, 1978) and the Moluccas, at a depth of 27 to 83 metres. REMARKS: A. M. Clark (1956) redescribed the eight specimens collected from Macclesfield Bank (South China Sea) and selected a lectotype. Re-examination of the lectotype by one of us (F.W.E.R.) has revealed it to possess three small madreporites and two anal openings, indicating fissiparity, as suggested by A. M. Clark from observation of regenerating multibrachiate specimens in the British Museum (Natural History). The specimens examined here (figs 4g, 5c) agree closely with the lectotype. Detail of the spinelets is shown in figure 6c. N. briareus appears to be more closely related to N. gracilis n. sp. than to N. maculata, the differences are shown in the key. DISCUSSION The genus Nepanthia traverses the Indo-Malay Australian region extending from Mergui Archipelago (Burma) eastwards to the Philippine Islands and southwards through Indonesia to circumscribe Australia (fig. 1), Altogether 17 species have been described and referred at one tume or another to the genus Nepanthia, either as valid species or as synonyms of those species (Table 1). In this study the number of previously described species has been further reduced, by synonymy, to five and three new species are described. Nepanthia tenuis is considered conspecific with maculata while suffarcinata, brevis, joubini, variabilis and magnispina are considered to be conspecific with belcheri. The synonymy of belcheri was realised after a study of 150 specimens ranging from New South Wales northward through Queensland and Torres Strait to Exmouth Gulf in Western Australia. Specimens vary in size from R of 8 mm to 65 mm and show a geographical cline from southern Queensland (?N.S.W.) to Exmouth Gulf (W.A.) with size and, concomitantly, spinulation increasing and fissiparity decreasing in that direction. ASTERINID GENUS NEPANTHIA GRAY 111 Both fissiparous and five rayed specimens have been recorded from Aru Is in the Moluccas (Koehler, 1910a and present study), fissiparous specimens from Vietnam (Koehler, 1908) and the Philippines (Fisher, 1919) and five rayed specimens from Burma (Sladen, 1888). Fissiparity appears to become predominant as a method of reproduction in response to certain ecological conditions, as yet undetermined. The origin of the genus may well have been in the Indo-Malay region from where it extended to the southern coast of Australia. Ekman (1953) has pointed out that a high proportion of species and genera of the southern Australian fauna is of tropical origin. He also concluded that Australia had a basically northern (tropical/subtropical) fauna and southern (warm-temperate/temperate) fauna. H. L. Clark (1946) found that the distribution of echinoderm species corresponded fairly well with Hedley’s (1904, 1926) four zoogeographical provinces, the Dampierian (Torres Strait to Geraldton on the west coast), Solanderian (Torres Strait to 26°S, on the Queensland coast), Peronian (south east coast) and Flindersian (southern and south western coasts). Clark ignored the Banksian province proposed by Whitley (1932) for the Queensland coastal fauna as distinct from the ‘Solanderian’ reef fauna. Wilson and Gillett (1971) prefer to divide the Australian molluscan fauna simply into northern and southern regions with a long overlap zone on the east and west coasts. The distribution of species of Nepanthia around Australia is therefore most interesting as it appears to support both of these views. That there is a northern and southern fauna is demonstrated by the belcheri/maculata versus troughtoni/nigrobrunnea distribution, with a distinctive difference in the more hyaline spinulation of the northern species compared with the coarse spinulation of the southern species. These species, on present known distribution, can be further assigned to zoogeographical provinces with belcheri extending across the Dampierian/Banksian regions, maculata being Dampierian in Australian distribution, troughtoni Flindersian and nigrobrunnea Peronian. N. crassa occupies the western overlap zone. Wilsons Promontory, Cape York and North West Cape all provide sharp dividing lines (?barriers) between the distribution of Nepanthia species. North West Cape separates populations of N. maculata and N. belcheri from N. crassa, Cape York is the other end of the Australian range of N. maculata and Wilsons Promontory marks the eastern limit of N. troughtont. It is curious that, if both N. maculata and N.. belcheri originated in the Indo-Malay region, N. belcheri has spread through Torres Strait and far down the east coast of Australia, while N. maculata is only found west of Cape York despite the fact that their habitat requirements appear to be identical and they occur together in north Western Australia. Neither species extends beyond Exmouth Gulf, which is the last suitable embayment until Shark Bay 500 km south of North West Cape is reached. However, Nepanthia crassa, which appears to be closely related to N. belcheri, replaces the latter on the western side of North West Cape extending from Point Cloates southwards to Cape Naturaliste, wherever suitable habitats occur. Most of the Nepanthia species are found in sheltered, sometimes muddy, situations from shore to the inner continental shelf perhaps explaining why they have not spread far into the Indian and Pacific Oceans. They are what Endean (1957) has termed ‘‘mainland” rather than “reef” species. N. troughtoni and N. nigrobrunnea differ from the other species in habitat requirements since they are both open coast species favouring rocky substrates exposed to considerable wave action. N. belcheri has a floating yolky egg suggesting pelagic lecithotrophic development but the larval life span is unknown (Otteson, 1976). Nothing is know of the reproduction of the other species. Further study of the ecological/physiological factors responsible for the predominance of 112 FRANCIS W. E. ROWE AND LOISETTE M. MARSH fissiparity over sexual reproduction in some populations of N. belcher: should be rewardin ACKNOWLEDGEMENTS We would ee to thank Miss A. M. Clark, British Museum (Natural ee Ge Londo England; Dr D. Pawson, Smithsonian Institution, Washington, D.C., U.S.A.; Dr Fe Woollacott, used of Comparative Zoology, Harvard, Massachusetts, U.S.A. and Dr FB Smith, National Museum of Victoria, Australia for the loan of type- -material. One of us (L.M also examined material in the Darwin Museum, Northern Territory, Australia, for which thr Director is thanked. REFERENCES Bell, F. J., 1884. Echinodermata. Jn Coppinger, R. W. Report on the Zoological collections made in the Indo- Pacifit Ocean during the voyage of H.M.S, ‘Alert’ 1881-2. London: pp.117-177 and 509-512, pls 8-17 and 45. ——— 1894. On the Echinoderms collected during the voyage of H.M.S. ‘Penguin’ and by H.M.S. ‘Egeria’ whe: surveying the Macclesfield Bank, Proc. zool. Soc, Lond, 1894: 392-413, pls XXIII-XXVIL. Bernasconi, I., 1973. Asteroideos Argentinos. VI. Familia Asterinidae. Revta Mus. argent. Cienc. nat, Bernadini Rivaclavia (Hidrobiologia) 3(4):335-46, 2 pls. Clark, A. M., 1956. A note on some species of the family Asterinidae (Class Asteroidea). Ann. Mag. nat. Hist. (12) on 374-383, 1 pl. — 1966. Port Phillip Survey 1957-63, Echinodermata. Mem. Nat. Mus. Vict. No, 27: 289-384, 10 figs, 3 pls, 1 i) Clark, A. M. and F. W. E. Rowe, 1971. Monograph of the shallow-water Indo-west Pacific echinoderms. London: pp, 238, 100 text figs, 31 pls. ) Clark, H. L., 1909. Scientific results of the trawling expedition of H.M.C.S. ‘Thetis’. Echinodermata. Mem. Aust. Mus.” + (11): 519-564, pls XLVU-LVIII. ) 1921. The Echinoderm Fauna of Torres Strait. Pap. Dep. mar. biol. Carnegie Instn. Wash. 10: pp. viii + 223, 38% pls. ~ 1923. Some Echinoderms from West Australia. 7. Linn. Soc. (Zool.) 35: 229-251, pl 13. 1926, Some sea-stars from the Riksmuseum Stockholm. Ark. Zool. 18 a (8): 1-8, 2 figs. 1938. Echinoderms from Australia. Mem. Mus. comp. Zool. Harv. 55: pp. viii + 596, 63 figs, 28 pls. — 1946, The Echinoderm Fauna of Australia. Publs Carnegie Instn. No. 566: 1-567. Déderlein, L., 1896, Bericht tiber die von Herrn Prof. Semon bei Amboina und Thursday Island gesammelten} Asteroidea, Denk. Ges, Jena 8: 301-322, pls 18-22. 1926. Uber Asteriden dem Museum von Stockholm. K. svenska VetenskAkad. Handl. (3) 2 (6): 1-22, 4 pls. Domantay, J. S., 1972. Monographie studies and check list of Philippine littoral echinoderms. Acta Manilana A 9 (15): 36-161, Edmondson, C, H., 1935. Autotomy and regeneration in Hawaiian starfishes. Occ. Pap. Bernice P. Bishop Mus. 11 no. 8: 1-20, figs 1-3. Endean, R., 1953-65. Queensland Faunistic Records. Parts 3, 4, 7, 8. Echinodermata (excluding Crinoidea). Pap. Dep. Zool. Unto, Qd. 1: 51-60 (1953); 123-140 (1956); 289-298, 1 fig. (1961); 2: 229-235 (1965). 1957. The biogeography of Queensland’s shallow-water Echinoderm Fauna (excluding Crinoidea), with re-arrangement of the faunistic provinces of tropical Australia, Aust. J. mar. Freshwat. Res. 8: 233-273, 5 figs. Ekman, S., 1953, Zoogeography of the Sea. London: pp. 417, 121 figs (reprinted 1967). ASTERINID GENUS NEPANTHIA GRAY 113 Fisher, W. K., 1908. Necessary changes in the nomenclature of starfishes. Smithsonian Misc. Coll. 52: 87-93. 1919. Starfishes of the Philippine Seas and adjacent waters. Bull. U.S. natn. Mus. 100 (3): pp. xii + 711, 156 pls. 1940. Asteroidea. Discovery Rep. 20: 69-306. 1941. A new genus of sea stars (Plazaster) from Japan, with a note on the genus Parasterina. Proc. U.S. natn. Mus. 90: 447-456, figs 20-22, pls 66-70. Gray, J. E., 1840. A synopsis of the genera and species in the class Hypostoma (Asterias Linn.): Ann. Mag. nat. Hist. (1) 6: 175-184 and 275-290. 1847. Descriptions of some new genera and species of Asteriadae. Proc. zool. Soc. Lond. 15: 72-83. 1866. Synopsis of the species of starfish in the British Museum (with figures of some of the new species). London: pp. iv + 17, 16 pls. Hedley, C., 1904. The effect of the Bassian isthmus upon the existing marine fauna: a study in ancient geography. Proc. Linn. Soc. N.S.W. 28: 876. 1926. Zoogeography. Jn Australian Encyclopedia 2. Sydney: p. 743. Jangoux, M., 1978. Biological results of The Snellius Expedition XXIX. Echinodermata, Asteroidea. Zool. Meded. 52 (25): 287-300. Koehler, R., 1908. Description d’une Astérie nouvelle (Nepanthia joubini) provenant du Cap Saint-Jacques (Cochinchine). Bull. Mus. Hist. nat. Paris. 14: 232-235, 2 figs. 1910a. Astéries et ophiures des iles Aru et Kei. Abh. senckenb. naturforsch Ges. 33: 265-295, pls. 15-17. 1910b. Shallow-water Asteroidea. Echinoderma of the Indian Museum. Calcutta: pp.1-192, 20 pls. Kenny, R., 1969. Growth and asexual reproduction of the starfish Nepanthia belcheri (Perrier). Pacif. Sct. 23: 51-55, 4 figs. Livingstone, A. A., 1932. Asteroidea. Scient. Rep. Gt. Barrier Reef Exped. 4 (8): 241-265, 2 figs, 12 pls. 1934. Two new asteroids from Australia. Rec, Aust. Mus. 19 (3): 177-180, pl. XVIII. Mobius, K., 1859. Neue Seesterne des Hamburger und Kieler Museums. Abh. Verh. naturw. Ver Hamburg 4(2): 1-14, 4 pls. Miller J. and F. H. Troschel, 1840. Uber die Gattungen der Asterien. Arch. Naturgesch 6 (1): 318-326 (Alsoin: Ber. K. Preuss. Akad. Wiss. 1840: 100-106). 1842. System der Asteriden. Braunschweig: pp.xx + 134, 12 pls. Ottesen, P., 1976. The reproductive patterns and population dynamics of two small intertidal starfish, Patiriella obscura Dartnall and Nepanthia belcheri (Perrier). B. Sc. Hons. thesis, James Cook University of North Queensland. Perrier, E., 1875. Révision de la collection de Stellérides du Muséum d’ Histoire Naturelle de Paris. Paris: pp. 1-384 (also published in: Arch. Zool. exp. gen. 4 (1875): 263-449; 5 (1876): 1-104, 209-304. 1884. Mémoire sur les Etoiles de mer recueillis dans la mer des Antilles et de golfe du Méxique durant les expéditions de dragage faites sous la direction de M. Alexandre Agassiz. Nouv. Archs Mus. Hist. nat., Paris. (2) 6: 127-276, pls 1-9. Rowe, F. W. E., and D. L. Pawson, 1977. A catalogue of echinoderm type-specimens in the Australian Museum, Sydney. Rec. Aust. Mus. 30: 337-364. Shepherd, S., 1968. The shallow-water echinoderm fauna of South Australia. Part 1. Rec, S. Aust. Mus. 15 (4): 729-756, 1 fig. 2 tables. Sladen, W. P., 1888. On the Asteroidea of the Mergui Archipelago. 7. Linn. Soc. (Zool.) 21: 319-331, pl. 28. 114 FRANCIS W. E. ROWE AND LOISETTE M. MARSH 1889. Asteroidea. Rep. scient. Results Voy. “Challenger” (Zool.) Vol. 30: xlii + 893, 117 pls. Spencer, W. K. and C. W. Wright, 1966. Asterozoans, Jn Moore, R. C. Treatise on Invertebrate Palaeontology, Part U, Echinodermata 3, 1: pp xxx + 366, 271 figs. Studer, T., 1884. Verzeichniss der Wahrend der Reise S.M.S. ‘Gazelle’ um die Erde 1874-76 gesammelten Asteriden und Euryaliden. Abh. preuss. Akad. Wiss.: 1-64, 5 pls. | Verrill, A. E., 1913. Revision of the genera of starfishes of the subfamily Asterininae. Amer. 7. Sci. 45 (4): 477- 485. Viguier, C., 1878. Anatomie comparée du squelette des stellérides. Arch. Zool. exp. gén. 6: 33-250, pls. 5-16. Whitelegge, T., 1889. List of the marine and fresh-water invertebrate fauna of Port Jackson and the neighbourhood. Subkingdom Echinodermata. 7. Proc. Roy. Soc. N.S.W. 23: 197-206. Whitley, G., 1932. Marine zoogeographical regions of Australia. Aust. Nat. 8: 166. Wilson, B. R. and K. Gillett, 1971. Australian Shells. A. H. and A. W. Reed, Sydney. p.168. 115 ASTERINID GENUS NEPANTHIA GRAY 00 “piyuvdan Jo satoads Jo uonnqmsiq “| ‘3ty AsOPUOWOAd S,UOS|IAA uosyoeP 340g e OF “| @MOH p07 At. “Keg uojai0yy ] /U0}Y6N01} eauunsgosb6iu elejnoew $1/198I6 /4ayslp esseso ViIIvVeisnyv adeg snaselsq Hay Ijaq STOSWAS OL AA» @BOO@C ADD ae ae ee vy: BB sesonjow ¢ a=) 116 FRANCIS W. E. ROWE AND LOISETTE M. MARSH Fig. 2. a-c, Nepanthia nigrobrunnea sp. noy., holotype, a. abactinal, b. actinal, c. adambulacral armature, R/r = 65/15 mm; d-e, N. troughtoni, WAM 55-79 d. adambulacral armature, e. abactinal, R/r = 70/14 mm; f. N. maculata, WAM 587-75, abactinal, R/r = 70/13 mm. ASTERINID GENUS NEPANTHIA GRAY Fig. 3. a-c, Nepanthia crassa, a. WAM 756-75, abactinal, R/r = 43/13 mm; b. WAM 12-77, abactinal, R/r = 42/10 mm; c. WAM 607-75, abactinal, R/r = 37/10 mm; d-e, N. belcheri, d. WAM 590-75; Dampier, W.A., abactinal, R/r = 43/12 mm;e. WAM 615-77, Thursday Island, Qld., abactinal, R/r = 4 118 FRANCIS W. E. ROWE AND LOISETTE M. MARSH Fig. 4. a-c, Nepanthia fisheri sp. nov., holotype, a. abactinal, c. actinal, R/r = 40/9.5 mm; b. paratype, abactinal, R/r = 20/5.5 mm, d-f, N. gracilis sp. nov., holotype, d. abactinal, e. actinal, f. adambulacral armature, R/r = 37/7 mm; g. N. briareus, WAM 56-79, abactinal, R/r = 44/9 mm. ASTERINID GENUS NEPANTHIA GRAY Fig. 5. Denuded arms of Nepanthia species. a-b, N. belchen, a. WAM 615-77, Thursday I., Qld., R/r = 40/11 mm, b. WAM 296-75, Dampier, W.A., R/r = 41/12 mm; c. N. briareus, WAM 56-79, Moluccas, R/r = 44/9 mm; d,h. N. fisheri sp. nov., holotype, Sabah, R/r = 40/9.5 mm; e. N. maculata, WAM 1825-75, Dampier, W.A. R/r = 70/13 mm;f. N. gracilis sp. nov., holotype, Philippines, R/r = 37/7 mm; g,j- N. crassa, g- WAM 625-75 Cockburn Sound, W.A., R/r = 57/15 mm; j. WAM 54-79 Houtman Abrolhos, W.A. R/r = 30/9 mm; i. N. migrobrunnea sp. nov., holotype, N.S.W., R/r = 65/15 mm;k. N. troughton, WAM 1486-74, Sorretono, W.A., R/r = 52/13 mm. 119 FRANCIS W. E. ROWE AND LOISETTE M. MARSH Fig. 6. Scanning electron photomicrographs of abactinal spinelets from plates of dorso-lateral area at base of ray of Nepanthia species. a. Nepanthia belchen, WAM 613-77, Townsville, Qld.; b. N. belchen, WAM 935-76, Dampier, W.A.; c. N. briareus, WAM 56-79, Moluccas, Indonesia; d, e. N. crassa, WAM;f. N. gracilis sp. nov., holotype, WAM 103-78, Philippines; g. N. fisheri sp. nov. holotype, WAM 102-78, Sabah; h. N. maculata, WAM 587-75, Dampier, W.A.; i. N. maculata, WAM 57-79, Moluccas, Indonesia; j, k, N. nigrobrunnea sp. nov., holotype, AM-J 10147 N.S.W.; 1. N. troughton., WAM 570-75, Esperance, W.A. 7. INTER-RELATIONSHIPS OF RECENT STALKED, NON-ISOCRINID CRINOIDEA AILSA M. CLARK British Museum (Natural History), London, England SUMMARY Outlines are given of the body form in the extant families of the stalked crinoid orders Millericrinida and Bourgueticrinida with particular notes on the few taxa exhibiting secondary arm branching. The recent nominal species of the Bourgueticrinida are listed in a table, together with their distributions and an indication of the size range of the often limited known material. The wisdom of division of these taxa into more than the one family Bathycrinidae is questioned, in view of recent observations on ontogeny and variation, particularly with regard to the stalk attachment. A new record of a particularly relevant species, Porphyrocrinus thalassae Roux, is included, with a photograph showing the secondary arm branching. INTRODUCTION Apart from the aberrant Holopodidae (order Cyrtocrinida), the remaining recent non-Isocrinid taxa of stalked Crinoidea are referable to two orders — the family Hyocrinidae to the Millericrinida and the remainder to the Bourgueticrinida. Most of these species are remarkable among recent crinoids for the conspicuous part the basal plates play in making up the calyx of the adult. SYSTEMATIC ACCOUNT The Hyocrinidae have thin-walled cup-shaped calyces, surmounted by the arms, which are more or less widely-spaced, approximately cylindrical in cross-section and unbranched in most genera, including Hyocrinus. However, Calamocrinus diomedae from near the Galapagos Islands, representative of a monotypic genus, has irregularly-branching arms, evidently formed by elaboration of up to five of the original pinnules on each side of a primary arm into secondary arms, themselves bearing pinnules. This kind of augmentation of arm number contrasts with the multiplication by what is called ‘adolescent autotomy’ at proximal syzygies followed by regeneration, with the first new ossicle becoming an axillary, found throughout the Comatulida. The single exception in this order is Comatula rotalaria Lamarck, from northern Australia, in which the second brachial of each of the ten primary arms of the post-pentacrinoid gradually transforms itself into a symmetrical axillary by modification of its appendage into an arm instead of a pinnule. Similar arm multiplication also occurs in some Isocrinida. The species of the Bourgueticrinida differ from the Hyocrinidae in having the calyx more compact and thick-walled, bearing closely approximating arms lacking pinnules on usually the first six to ten brachials. The nominal species currently recognised are listed in Table 1. However, some of these names are very likely to prove to be synonymous since many are abyssal and many are only known from incomplete specimens often of a limited size range, so that inadequate allowance has often been made for very wide geographical distributions and for growth changes when naming supposedly new taxa. The first five genera in Table 1 have been referred to the family Bathycrinidae, three of them: Rhizocrinus, Conocrinus and Democrinus, having five simple arms while Bathycrinus and Australian Museum Memoir No. 16, 1982, 121-128 122 AILSA M. CLARK Monachocrinus* have ten arms, the second post-radial ossicle being a primary axillary. All hav: the arm bases aligned vertically when in the non-feeding position with a marked lateral flange or each side, beyond which basal part the muscular and non-muscular joints between the brachial normally alternate regularly. The calyx ranges in form from conical to narrow vase-shaped ana the stalk is xenomorphic with some of the topmost columnals discoidal, their joint faces more on less smooth and not yet modified into the oval synarthries found between the succeeding more elongated columnals, the alignment between the synarthrial joints being twisted from one to th next through nearly 90°. The distal part of the stalk, when known, bears irregular rhizoid-like’ jointed appendages for attachment, though ina few cases, notably Democrinus brevis A. H. mes (see A. M. Clark, 1977) the rhizoid system is more or less completely replaced by an irregula flattened expansion at the end of the stalk. In 1907, A. H. Clark proposed a new family Phrynocrinidae for Phrynocrinus nudus from SE: of Japan, a species with all the columnals fairly short and twisted so that successive joints appear” alternately wide and narrow when seen in one plane, distal attachment is solely by an expanded! terminal plate, the calyx is markedly flared above bearing almost cylindrical arms which leaves exposed the relatively large disc or tegmen and which often have several successive musculary joints easily outnumbering the non-muscular joints by about 4:1. Judging from what is: left of the three least broken post-radial series of the holotype, the five primary arms branch | irregularly at least once, these three having axillaries at brachials 13, 20 and 25 respectively, all | preceded by pinnules. It is likely that this development of secondary arms is by modification from pinnules, as in Calamocrinus. Subsequent discoveries of recent Bourgueticrinida have tended to blur the distinction | between Phrynocrinidae and Bathycrinidae. In 1973, I described an Atlantic species, Zeuctocrinus gisleni, referring this new genus to the Phrynocrinidae, on account of its low, flared calyx, rounded arm bases and the similar form of the stalk to that of Phrynocrinus nudus with synarthrial joints throughout (at least in larger specimens), though unfortunately the stalk attachment is unknown. However, smaller specimens of Zeuctocrinus than the holotype of Z. gis/eni, show that earlier in the ontogeny the upper columnals are much shorter and the distal ones relatively longer. Possibly the same will prove to be true of P. nudus, when a better range of specimens is available. Z. gislem: parallels Bathycrinus in having a primary axillary — normally the second post-radial ossicle — and ten arms and also shows the same relatively high frequency of non-muscular joints in the arms as Bathycrinus, only a few proximal brachials having muscular joints at both ends. In 1912, A. H. Clark described Naumachocrinus hawatiensis, a species with stalk attachment *In 1970] noted that A. H. Clark’s figure of the holotype of Monachocrinus sexradiatus (1923) appeared to show muscular joints at both ends of brachials 3, 6 and 9, as characteristic of Bathycrinus, where Monachocrinus is diagnosed by Gislén (1938) as having complete alternation of muscular and ligamentary joints. However, thanks to Dr, Madsen, I have been able to see the holotype and find that only one arm out of those remaining has muscular joints at both ends of brachial 3 and its condition beyond brachial 5 is unknown due to breakage. Nevertheless, the joint sequence hardly seems of generic weight unsupported. Other distinctions cited by Gislén are the fusion of the basal ring in Bathycrinus, while some specimens at least of Monachocrinus (e.g. the paratype of M. sexradiatus but not the holotype) show distinct interbasal sutures, and the profile of the calyx showing an angle between the basal and radial rings in Bathycrinus but a straight line or smooth curve in Monachocrinus. The importance of the latter was stressed by Macurda & Meyer (1976). It seems to me likely that fusion of the basal ring may be correlated with a higher incidence of autotomy between the two rings and so be more frequent in Bathycrinus. The subsequent regeneration of the radial ring and arms would result in at least temporary discontinuity of the profile. Conversely, some specimens of Barthycrinus do show smooth profiles, along the radii if not also the interradii, either by slight constriction of one or the other ring near the junction point or by an even flaring of both rings. The second is true of the specimen of B, australis shown in Déderlein’s pl. 5, fig. 1 and pl. 6, fig. 7 (1912), which incidentally also shows distinet interbasal sutures as photographed in toluol in pl. 6. I consider therefore that the generic distinction of Monachocrinus from Bathyerinus is ill-founded. CRINOID INTER-RELATIONSHIPS 123 similar to that of Phrynocrinus, prompting him to refer the genus to the Phrynocrinidae although he said that the upper part of the stalk resembled that of Rhizocrinus (now Democrinus) webert and the calyx is almost perfectly cylindrical with extraordinarily long radials but a basal ring no higher than the discoidal uppermost columnal just below it (see A. M. Clark, 1973, Fig. 6h). The arms are unknown. A fourth genus — Porphyrocrinus — has also been referred to the family Phrynocrinidae, | after considerable deliberation by Gislén (1925), because the type-species, P. verrucosus from Indonesia, has simple arms, proximally flanged and concealing the tegmen, the calyx is almost cylindrical and the proximal columnals are discoidal. The stalk attachment is unknown. Consequently there is very little superficial resemblance to. Phrynocrinus, though more to Naumachocrinus. However, in 1973 I described a similar species from the SW Indian Ocean, -Porphyrocrinus polyarthra, from a specimen retaining the distal part of the stalk, showing that attachment is by a lobed terminal expanded plate. Even so, the anomalies between - Porphyrocrinus and Naumachocrinus, on the one hand, and Phrynocrinus on the other, seemed to me so great that I proposed a third family, Porphyrocrinidae, intermediate between the two others, characterized by the stalk attachment of the Phrynocrinidae and the calyx form of the Bathycrinidae. Subsequently, two factors affecting the validity of this third family and perhaps even of the Phrynocrinidae, have become evident. Recently both Roux and I have independently found a new bourgueticrinid which he described as Porphyrocrinus thalassae in 1977. Smaller specimens of this species (upper stalk diameter c.3 mm) have simple arms but the larger ones (s.d. c.4 mm) have the first pinnule, on the right side of the eighth brachial (Brs) modified into a secondary arm and may also have the first pinnule of the left side (on Br1o) similarly modified, converting these two ossicles into rather lop-sided axillaries. Possibly at a larger size still the secondary arms achieve equality with the primary ones and adopt a plane tangential to the vertical axis instead of being inclined obliquely like the pinnules; the axillaries would then become more nearly symmetrical. Since the holotypes of the other species of Porphyrocrinus were smaller, it is not unlikely that they too may show a similar augmentation in arm number with growth. Secondly, re-examination of the type material of Democrinus brevis, brought home to me in 1977 the great variation in stalk attachment shown by different species of undoubted Bathycrinidae, D. brevis showing expanded terminal plates in contrast to both D. parfain, the type-species of Democrinus, and some West Indian specimens which I have attributed to D. conifer, which consistently have slender branching rhizoids terminating the stalk. Macurda (in Meyer, Messing and Macurda, 1978) believes that D, brevis and conifer intergrade. Also McKnight (1977) has described a bathycrinid stalk from the Kermadec Islands which terminates in both an expanded plate and rhizoids. Gislén’s 1927 diagnosis of the Bathycrinidae as having stalks attached by rhizoids needs modification. As indicated in Table 1, in about a third of the nominal species of Bathycrinidae the distal part of the stalk is unknown. Hopefully, the current increase in the amount of deep-water biological collecting and the number of specialists interested in these animals may soon result in some degree of clarification of the inter-relationships of these recent Bourgueticrinida. 124 AILSA M. CLARK NOTE ON AN UNRECORDED SPECIMEN OF PORPHYROCRINUS THALASSAE) ROUX Porphyrocrinus thalassae Roux Fig. 1 Porphyrocrinus thalassae Roux, 1977: 34-38, 50-54, fig. 1B, pl. 1, figs, 1-5. MATERIAL EXAMINED: ‘Discovery’ station 8511/2, 41°49'N, 11°06'W (NW of Spain), 2574-2584 metres; 1 specimen. REMARKS: Only the uppermost 30 columnals remain, measuring 15 mm. The first 20 are: discoidal, the height of the uppermost one being 0.3 mm, while the thirtieth is 1.5 mm high, Stalk diameter at the top is 4.1 mm and at the bottom 3.2 mm. The basal ring height is 1.0 mm radially, 2.0 mm interradially. The radial ring is 1.4 mm) high radially. The total calyx height is 2.4 mm radially and 2.8 mm interradially. The top of the: basal ring is slightly constricted after a slight expansion so that both top and bottom are 4.1 mm} in diameter. The top of the radial ring is 5.2 mm in diameter. The interbasal sutures are not} distinguishable, the undulating basiradial suture only after removal of the skin but the* interradial sutures are more easily visible. The first eight post-radial ossicles are joined in pairs by non-muscular joints, 1+2, 3+4, , 5+6, 7+8ax., Brs bearing a secondary arm on the right. Two rays, C and E, have another © secondary arm on the left of Brio. Four of the arms from Brs have the first brachial divided longitudinally; non-muscular joints mostly alternate with muscular ones from 2+3 onwards $0 | that the first pinnule is on the outer side of Br3. The primary arms have their first pinnule on the right of Bri2. This consists of 13 very elongated pinnulars with a short gonad from segments 2-5; the length is 11.5 mm. The longest arm remaining is a secondary one; it measures 60 mm and consists of 50 brachials; probably c.20 mm is lost. Some of the more distal brachials have a muscular joint at both ends. The dark brown tegmen is widely exposed between the bases of the primary arms; it extends to about Bre. The numerology of the ossicles is debatable but, as the additional arms are clearly secondary, the first eight post-radial ones are not a true division series and are better counted as brachials. The numerology used by A. H. Clark in 1907 in describing Phrynocrinus nudus was that of Carpenter, counting the ossicles on both sides of the ligamentary joints as forming a single brachial. Translating to Clark’s later method of counting, as now generally adopted, the joints of the holotype of P. nudus are as follows: 142, 3, 445, 6, 7+8, 9, 10, 11, 12+13, 14, 15+16, 17, 18, 19+20, 21, 22, 23, 24+25ax. P28 SG etS, 6. 789. 0, ee eS TG at fer 1+2, 3, 4, 5+6, 7, 8+9, 10, 11, 12+13, 14, 15+16, 17, 18, 19+20ax. 1+2, 3, 4+5, 6, 7+8, 9, 10+11, 12, 13, 14+15, 16, 17, 18+ 1+2,.3,4+5, 6,7, 8, 9+10, 11, 12, 13ax. REFERENCES Clark, A. H., 1907. Two new crinoids from the North Pacifie Ocean. Proc. U.S. natn. Mus. 32: 507-512, 2 figs. 1912. Naumachocrinus, a new genus belonging to the crinoid family Phrynocrinidae. Proc. U.S. natn. Mus. 42: 195-197. 1923. Crinoidea. Dan. Ingolf-Exped. 4(5): 1-58, 58 figs. Clark, A. M., 1970. Echinodermata: Crinoidea. Marine Invertebrates of Scandinavia. Oslo. No. 3: 1-55, 19 figs. 1973. Some new taxa of recent stalked Crinoidea. Bull. Br. Mus. nat.‘Hist. (Zool.) 15(7): 265-288, 6 figs., 2 pls. 1977. Notes on deep-water Atlantic Crinoidea. Bull. Br. Mus. nat. Hist. (Zool.) 31(4): 157-186, 5 figs. CRINOID INTER-RELATIONSHIPS p25) Fig. 1. Porphyrocrinus thalassae Roux, ‘Discovery’ st. 8511/2, showing the secondary arm branching. The calyx has been partially bleached to clarify the sutures. ———————— ie uedef § OLI-OFT + + Sz SoIpuy 1seq 067 i = scl [1ze1g OSLI-091 = i sil BOY § BA OO8T-OIF ae oe Sc 1-S7'0 SOTPUT 189 /K\ OS ar eh Ciel=Sall purleaZ MON St6-0S9 7 a OT Avosig Jo Avg OIS-O€E = = $0 Avosig Jo Avg OL0Z-SL6I - = 01 SoIpuy 1Sey OO0ET (ei) = S70 osnueny N SLve-OFl = a3 ONES-O puro] SLO7 “i ia S'0 oOdD010; ‘Sa10ZY 09Z7+¢-00EZ + ta CLaleS Cal BOLIFY “S “So10zy 009¢-0291 (AE) = S70 jesuag jo Avg OLET ar (+) CaS, eIuOpayey) MON 0001 ete =: al) SOTPUyT 1S9/X\ 0971 = = eS0ed purleoZ MON OST7-090T or ar scl jesusg jo Arg OLLZ = = ond uedef § 0s9l + (ep) 20'1T¢ onueny AN OL0S-OStr sid 2 SL0-S7'0 se -PIw O0CCY 7 é c OTE Rag BSULINg OF8Z an oP CO°C-SE TL onuepy “N 91 OT87-09EI aR tg SETVESEO pueleaZ MON 8£8-£69 oF + OT uev29Q) UOYINOS OIC8-OEZI F a 0°C-0'T onueny-Pplw OS8S-OCEE ay a SL°0-S'0 asurl asuey yidoq lag yes (uu) asue1 ‘xoidde eorydeis0an puodsaq [e1siq :To]QUTeIp SULLY yes [BWIXO1g ua]sty snotuodpl UD[SI) supjngo]s CO'H'Y) 4afmoo (pod) lunyo COTH'W) St204q “SOW Snuvaion snumusowaq XNOY MaluUuogsays XNOY 14I01QDI snuuI0U0) (pod) snuauim SIV sisuajofo] snursoz1y YY ‘\YH'W Snipippaxas (1lag) snipsadnyad CA 2 M) Matuad COH'W) snxoppsvd UDISID) tasuaqiou CO'H'W) Snaqqiivs “NPOW Doavalopd snuLsa0yrDUO WW ‘YH V wospupoon ‘YH snoyfiood “LAM Syopss ‘YH V Ssypuoinba ‘YH’ Snipunjduos CM °C) Maiuaduvo “SPW $29n.190] D41sND CO'H'Y) sypsisno “LA Snuniysupyp snuwonying 126 *POATOAUT YIOMSSANS JO VoIZ9p 9] AJIUSIS 22, .¢, YM Solu pue paid us0q SARA\TP 1OU SBY JUIUIOINSeOU STY] ‘9SURI 9ZIS dy] JO d1eUITISO J]eUITxOIdde UB SIATS UUINIOS ISI SY] “[esIeU papsOdeI sy] ur sized SuNesisuase1 10 aj9;]dwWoouUT pue UdIYOIg BUNKTPUT S]oyYOeIQ UT SoTU ‘TeTYDeIQ ISITJ 9Y] puOk|q sue ay] Jo Led Isa] 1e JO pue yyeis ay) Jo led yeistp sy] Jo 9dUdIINI90 dy] SUIMOYS ‘epluloNansinog luadaI Jo satoadg * |] 3]qe L, 127 SoIpuy 1seq Avostg Jo Aeg uevId0Q UeIPUy A\S onueny onueny AN “sy ueITEMeyy uedef qs SOIPUT 189 /\ SoIpUyT 1S8ey onueny AN Sorpuy 1sey SVE O8Sc-O11Z OOF 00FC-00EI SZSC-OLIC OOE1-0€6 0611-009 OSOC-OIT OS9-OL OLSI 097¢b-009 OLST-O9TI | ++ ++ 444 4+ 4 | Se SLESL Sl 09-0'% Sc rs 0 Sal Sc 0 STO 1 Sc0 UD]SID) snsoonsiaa XNOY avssvjvyl WY 2eyisvsjod UQ]SID) SnIMsspsqur snumI04AY GLO “WY lWajsis snutss019ne7 ‘YH Sisuaupapy snuusoyopiunv jN ‘OHV Snpnu snuroouday gy (pod) maqam Cumog) mosan4 (pod) wnjnsod ‘Log umfavd (POC) sadipou 128 AILSA M. CLARK Déderlein, L., 1912. Die gestielten Crinoiden der Deutschen Tiefsee-Expedition. Wiss. Ergebn. dt. Tiefsee Ex “Valdivia” 17(1): 1-34, 9 figs., 12 pls. Gislén, T., 1925. Two new stalked crinoids from the Kei Islands. Vidensk. Meddr dansk. naturh. Foren. 79: 85-9 figs. 1927. Japanese crinoids. Vidensk. Meddr dansk. naturh. Foren. 83: 1-69, 80 figs. 1938. A revision of the recent Bathycrinidae, with a study of their phylogeny and geographic distribution. Lay Univ. Arsskr. N.F. 34(10): 1-30, 18 figs. McKnight, D. G., 1977. Some crinoids from the’ Kermadec Islands. NZOI Rec. 3(13): 121-128, 7 figs. Meyer, D. L. and Macurda, D. B. Jr., 1976. The identification and interpretation of stalked crinoids (Echinodermit from deep-water photographs. Bull. mar. Sci. 26: 205-215, 4 figs. Meyer, D. L., Messing, C. G. and Maeurda, D.B, Jr., 1978. Zoogeography of tropical western Atlantic Crinoidea. By mar. Sct. 28(3): 412-441. Roux, M., 1977. Les Bourgueticrinina du Golfe de Gascogne. Bull. Mus. natn. Hist. nat. Paris (Zool.) No. 296: 25-83%) figs., 10 pls. 8. DEEP-SEA ECHINODERMS IN THE TONGUE OF THE OCEAN, BAHAMA ISLANDS: A SURVEY, USING THE RESEARCH SUBMERSIBLE ALVIN. DAVID L. PAWSON Smithsonian Institution, Washington D.C. 20560, U.S.A. SUMMARY Deep-sea echinoderms of the Tongue of the Ocean, Bahama Islands, have been studied, jusing trawled collections made by the University of Miami together with observations from the deep submersible Alvin. Transect runs in the submersible permitted studies of population densities and behaviour of approximately 38 species of larger invertebrates, of which 27 were ‘echinoderms. Several echinoderm species show a patchy distribution pattern which is apparently not related to available food resources. Some species are exclusively herbivores, ifeeding on fragments of turtle grass, Thalassia testudinata and sargassum weed, Sargassum spp. Feeding habits of some Tongue of the Ocean echinoderms are compared with those of the same species from further north, where supplies of plant material are not nearly so abundant. Trawled collections of echinoderms do not include some of the most common and ecologically important holothurians; conversely, some burrowing species very common in the area were not observed from the Alvin. Further observations were made on swimming behaviour of holothurians. All swimming forms studied apparently derive nourishment from the seafloor. Short tracks on the seafloor indicate that swimming behaviour is a common means of transportation from one area to another. The ophiuroid Bathypectinura heros is capable of active swimming movements. Uniformly conical mounds on the seafloor are often built up around a central core of holothurian faeces. INTRODUCTION During January, 1977, a series of eight dives were made in the submersible D.S.R.V. Alvin, to depths in excess of 3,660 metres, in the Tongue of the Ocean, Bahama Islands. The purpose of the dives was to make “‘. . . first-hand observations . . on the biology of deepwater benthic fishes and larger invertebrates and to take qualitative and quantitative data by visual and photographic methods” (D. M. Cohen, 1 March 1977, Cruise Report NOAA — MUST dives with D.S.R.V. Alvin in the Bahamas — unpublished). I was able to participate in four dives, and on one, Dive 703, an excellent opportunity was provided to make quantitative studies of echinoderms and other large invertebrates, and to observe activities of echinoderms. This paper represents for the most part the results obtained during Dive 703, although some aspects of “‘natural history” of echinoderms were obtained during one or more of the other dives in which I participated. Additional information on echinoderms from the Tongue of the Ocean was obtained from the extensive collections of the Rosenstiel School of Marine and Atmospheric Science, University of Miami, whose staff members have occupied numerous trawl stations in the Tongue of the Ocean over the past several years. METHODS 1. LOCATION OF ALVIN DIVE 703 AND DESCRIPTION OF ACTIVITIES Dive 703 was made on January 12, 1977 in the Tongue of the Ocean, Bahama Islands Australian Museum Memoir No. 16, 1982, 129-145 130 DAVID L. PAWSON 24°54.9N, 77°41. W, at a depth of 1938-2141 metres. Total time on bottom 5.6 hours. Pilot, Dudley Foster, observers Daniel M. Cohen and David L. Pawson. During Dive 703 a total of 11 measured transects were run. Transect 9 was aborted because of a malfunction in the metre wheel which was used to measure the distance traversed by the submersible. Transect runs were essentially consecutive, and were made in a south to south-easterly direction (see Table 1). Thus, no overlap of transect runs occurred. Table 1. Transect runs covered by Alvin, Dive 703. Transect 9 was aborted due to equipment malfunction. Transect compass heading distance run (m.) area studied (m7?) l 180 914 3,654 2 180 679 2,716 3 154 202 808 4 164 716 2,864 5 145 419 1,676 6 143 445 1,880 7 148 138 552 8 121 267 1,068 10 138 188 752 11 138 289 1,156 4,257 metres 17,126 sq. metres It has been estimated that an observer can see out for a distance of approximately 4 metres from the observer’s viewing port in the pool of light thrown by the submersible’s floodlamps. Knowing this distance, and knowing the distance traversed by the submersible during a transect run, it is possible to calculate the total area studied during a run, and then to determine the number of specimens of a particular species occurring per unit area. In the table shown here, population densities are expressed numbers of individuals per 1,000 m+ (Table 2). In order to estimate the accuracy of population counts made through the viewing port, a comparison was made between numbers of animals photographed in early transects by an E.G. & G. camera mounted on the submersible and numbers of animals counted, and results obtained were closely similar. A complete photographic survey for all transects was not possible because of the limited amount of film in the single camera. The bottom was composed of a firm to flocculent greyish to light brown sediment, which contains very numerous pteropod mollusk shells of several species. Fragments of the floating alga, Sargassum spp., and roots and blades of turtle grass, Thalassia testudinata are ubiquitous on the seafloor (figs la, 1b). Ina few areas they form mats, but for the most part the fragments seem to be more or less evenly scattered. 2, IDENTIFICATION OF ECHINODERMS VIEWED FROM ALVIN Through the courtesy of Dr Gilbert L. Voss, I was able to examine the invertebrates collected by the Rosenstiel School of Marine and Atmospheric Science in the Tongue of the Ocean. This enabled a positive identification to be made of most of the echinoderms seen from the Alvin. Some systematic problems yet remain to be resolved. Furthermore, the opportunity was taken to investigate stomach contents of selected species of ophiuroids and echinoids, so that feeding propensities could be determined. The University of Miami collections also included specimens of echinoderm species which are apparently common in the study area but were not observed during Alvin Dive 703. These taxa are ECHINODERMS FROM THE TONGUE OF THE OCEAN, BAHAMAS 131 Fig. 1. Seafloor features. la, scattered fragments of Thalassia and Sargassum spp. Photo: D. M. Cohen. 1b, short track left by swimming holothurian, Benthodytes. 1c, Benthodytes lingua near rock outcrop. 132 DAVID L. PAWSON discussed below, RESULTS AND DISCUSSION 1. SPECIES STUDIED During the ten successful transect runs, counts were made of 38 species of larger invertebrates, These comprise one sponge (Luplectella suberea Thomson, included in Table 2 for purposes of comparison of population densities), four coelenterates (“‘pennatulacean”, Cerianthus sp., Anthomastus sp. and Umbellula sp.), three crustaceans (Glyphocrangon sp., penaeid shrimp — Hepomadus sp.?, hemit crab), one pycnogonid (Colossendets colossea Wilson), and two mollusks (white gastropod and octopus). The remaining 27 species were echinoderms, as follows— ECHINOIDEA: Hygrosoma peterst (Agassiz), Phormosoma placenta ‘Thomson, Plestodiadema antillarum (Agassiz), Salenocidarts profundi (Duncan), Brissopsis elongata Mortensen, dead tests. HOLOTHUROIDEA: Ellipinion delagei (Herouard), Benthodytes typica Théel, Benthodytes sanguinolenta Théel, Benthodytes lingua Perrier, Psychropotes depressa (‘Vheel), Enypniastes exima Théel, Deima validum Théel, Pseudostichopus species A, Pseudostichopus species B, Mesothunia vernill (Théel), Paelopatides sp. ASTEROIDEA: Nymphaster arenatus (Perrier), Ceramaster grenadensis (Perrier), Zoroaster fulgens (Thomson), Freyella species A, Freyella species B, “Luidia”. OPHIUROIDEA: Bathypectinura heros (Lyman), Ophiomusium species A, Ophiomusium species B, “red ophiuroid”’. CRINOIDEA: “ten-armed sea lily”. A few other species of echinoderms were seen but not.counted. These included numerous specimens of small pink ophiuroids (Family Ophiacanthidae?) clinging to sponges and small rocks, and pinkish five-armed euryalid ophiuroids, also clinging to sponges. 2. COMMON TONGUE OF THE OCEAN ECHINODERMS NOT COLLECTED BY THE UNIVERSITY OF MIAMI The following species of holothurians are common in the area studied, but are not represented in the University of Miami collections: Elipinion delage: (Herouard); Enypniastes eximia Theel; Paelopatides sp. The first and last species were found to be numerically dominant in some transects (see Table 2). These species were not collected in trawls because they are exceedingly fragile and would be reduced to gelatinous masses in trawl samples. Furthermore, they are barely negatively buoyant, and are easily dislodged by the bow wave of the submersible; the bow wave of a trawl would have the same effect, materially reducing the possibilities of capturing specimens of these species. These species are highly important elements in the benthic biota of the area and their influence upon ‘the composition and structure of sediments is undoubtedly important. 3, COMMON TONGUE OF THE OCEAN ECHINODERMS COLLECTED BY THE UNIVERSITY OF MIAMI, BUT NOT OBSERVED FROM THE ALVIN. The following species of holothurians are common in the area studied, but were not observed during Alvin transects: Molpadia barbourt Deichmann; Molpadia musculus Risso; i=) SONnASCS A on tm Som | SCSOtnNoe SO NOWOODO i . ™ Ver nies ls} \o ool oooocw°€c eee) ae i) SS. 0-95 Ss ISS) i | ome) +40 ono nS ee ooto SFG 61 9°02 9°0 VC 0 0 0 OT cv fede Site 0 0 8°0 VLI 6 v1 IZ 0 0 £0 0 0 9°0 ial 9°0 0 0 81 0 Ote 81 0 GE Gal se'0 0 0 0 $0 Gal 0 0 Gl 6 €& 9 S v Sel faa) na Coal ray No) Fes ens pian [NN an) ONDM~NO NaN loom vay aaa) wooo “86 e rl tOtOSoS ee S 0 + 00T ye) a S°6 So + Wai NOE CO) Us) A°-NnNSoS 6 €OL eI wWepy — JoquInu lOasuRI TL, pasagns pjja12ajdnq *| (sasuods) erayiIiog A]I] B98 powie-udy, “| RapIouly) ‘ds sapundojavg ‘Tt q/V sndoyousopnasd “01/6 YJM1I2 DUNYIOSAPW °g unpYyvaA DUiag */ puuixa saispiuddug °9 DNSUY] saIKpoylUag °¢ pssaidap sa1odosyIMs] “$ ouinsuvs/pI1dK] saldpoyluag *¢/Z ladvjap uowidy) sy *| vaploimyojoy (peep) vipsuoja sisdossug ° wpunfosd Slippiz0uayn¢ ° pIuaInjg DUOSOUWLLOY ] * UNADIJUUD DULAPHIPOISA] J 184a1ad DULOSOASA FT vaprlouryo UTR (sure [[) g “ds pyjanosy * (sue 9) y “ds nyjahasy ° SISUAPDUIAS 1aISDULDAA suasjnf 4a1SD0107 * snipuasw saispyquia Ny ROpIOIIsy proinrydo pay “+ q/v ‘ds unisnuoiyd¢ °€/Z soday vanunsadrylog * | eoprom1ydg Wm aNmntuno QMaant *-UW QQOST tod suaumtzads Jo Joquinu se passoidx9 o1e SatIsusqd “E0/ WAI WIAly sulinp poarasqo (nyjaiajdngq ‘asuods ay] pue) sutapouryse Jo satisusp uonendog °7 3]qk 1, 134 DAVID L. PAWSON Gephyrothuria glauca (Clark). The three holothurian species are apparently all burrowing forms. Molpadia species are | known to be active burrowers (Clark, 1907; Rhoads and Young, 1971). It is surprising that} Gephyrothuria glauca, with its dorsal whip-like papillae, is also apparently an infaunal species. Ir} might have been expected that this species uses its papillae as sensory devices in a manner similar to that of the elasipodids, and that the animal is an epifaunal dweller. Dead tests of the spatangoid echinoid Brissopsis elongata were observed during some’ transects (see Table 2); live specimens of this species were undoubtedly burrowed into the © substratum in the same areas, but no trace of their burrows was visible from the Alvin. It is evident from the above discussions that ecological studies of larger deep-sea | invertebrates based entirely upon observations from submersibles or upon trawled samples | might not necessarily reflect the true situation on the seafloor, and some of the most important “consumers”’ will be unwittingly omitted from consideration. It is important to remember that some of the most effective reworkers of sediments might not be represented in trawls or in photographs taken by submersibles. 4. INTESTINE CONTENTS OF SELECTED SPECIES The intestine contents of several species of echinoderms collected in the Tongue of the Ocean by the University of Miami were examined, in order to determine which species might be using as food the fragments of seaweed and turtle grass that are scattered over the seafloor in that area. Vegetarian feeding habits are commonplace among the echinoids (Lawrence, 1975) but other echinoderms seem to be less inclined towards such a diet. In the case of two species of echinoids, comparisons were made with specimens collected from further north, where plant material is not nearly so abundant, but is nonetheless present, as Menzies et al. (1967) and Menzies and Rowe (1969) have shown for Thalassia and as Schoener and Rowe (1970) have shown for Sargassum spp. a. Hygrosoma petersi: In all Tongue of the Ocean specimens examined, intestine contents consisted almost exclusively of fragments of Sargassum and fragments of Thalassia. Presence of occasional pteropod shells and foraminiferal skeletons indicate that some sediment is also ingested, but this may be accidental. Usually Sargassum dominated in intestines examined, Many Sargassum fragments carried colonies of encrusting bryozoans; apart from these, and the skeletal remains mentioned above, no other animal materials were found in the intestines. Specimens collected from further north (see Table 3), between Georgia and New York, appeared to have a more ‘“‘mixed”’ diet, indicating that when abundant plant material is not available locally, this species can subsist on organic material extracted from sediments. Mortensen (1935) studied specimens of this species from South Africa, West Indies and southwest Ireland, and found that the intestines contained ‘‘. . . only mud, formed into small balls about the size of peas.” (p. 205). In a later paper, Mortensen (1938) noted that the Pacific species Hygrosoma luculentum (Agassiz) had its intestines filled with “‘bits of plants” (p. 226). b. Phormosoma placenta: Intestine contents of this species in the Tongue of the Ocean consisted exclusively of small mud balls 1-3 mm in diameter, bound together by mucus. Several specimens from elsewhere in the Eastern Atlantic had similar mud balls in their intestines. This observation concurs with that of Mortensen (1935), c. Plesiodiadema antillarum: Intestine contents poorly defined mud balls, not strongly bound by mucus. No plant material. Mortensen (1938, 1940) notes that P. indicum from the Indo-Pacific eats pieces of plants almost exclusively. Despite the abundance of plant material in the area, P. antillarum does not appear to ingest it, “IOJOWRIP UI WU $-¢ s]Teqpnur May AJOA YIM S]USUISRIT wWnsspsin¢ ATISOUI Sst OH SYST SH\ OYE INGE EE 1692 SJUSUISPIF PIdM a[qelynusprun ‘sazis snowed Jo syjeqpnyw OST UI 8CZ‘T SAN, 9b oI Z ‘N67 06€ [81d sJusUIseIy p9oM o[qQeIyNUSprun My ‘JOISUeIp UT WIL ¢ s[[eqpnur Ajsout SST W EOP TM ,OE,IT o%L “N ,9So8E GEC slusuIsely poo o]qetnusprun ssl W LTST SMA SPPE obL ‘N OE ,6b0S€ STIC suvoZOoAIq SUNSNIOUS YIM IUIOS ‘S]USUISPI] Winssps.iD > OSI pnw pue s}usWIsely wnsspsing = SHI ‘SET susUIseIy [eUTUR pue Jue d]eUTULIO1OpuUT OLI SIUIUISeIT wnssvd.1v> WW STET *MA,,0€,0F 9L “NOPE MOF B Osye “JO]OWIVIP UT WU ¢€ ‘sT[eqpnu ATIsOU 56 8/97 Ee) $]U9]UO0D JUT)SIUT 1S] JO JOIOWIRIG, uoneig ssoneqry “YIOK MAN puP RISIOIDH UIIMI0q WI] Isuajag VIUMOSOLSAF] JO S1UIIUOD IUTISIIUT pUk IZIS “¢ JIqQR I]. 136 DAVID L. PAWSON d. Salenocidaris profundi: Intestine contents poorly defined mud balls; no plant material, e. Ophiomusium spp.: The two species of Ophiomusium apparently consume nothing but: sediment, perhaps on a non-selective basis. In most specimens examined the stomachs weret virtually empty; in a few, the stomach was filled with sediment. f, Bathypectinura heros: Of approximately 20 specimens examined, all but two had empty + stomachs. The two exceptions contained exclusively fragments of Sargassum. Some of the» “empty” stomachs contained a dark brown finely divided material that may once have been | pieces of Sargassum. Madsen (1973) examined “‘a few stomach contents” (p. 142) of this species ; and found them to include mainly unidentifiable organic detritus, foraminiferans, etc. This then | is the first record of a vegetarian diet in this species. Fell (1952) made the remarkable observation | that the related shallow water New Zealand species Pectinura maculata (Verrill) can feed on_ anthers of the southern beech tree, Nothofagus sp., which fall into the water from overhanging trees. Schoener and Rowe (1970) found that another deep-sea ophiuroid, Amphiophiura bullata (Lyman), is capable of ingesting and presumably using as food bladders of Sargassum. 5. SOME ASPECTS OF BEHAVIOUR AND MORPHOLOGY OF SELECTED SPECIES Hygrosoma petersi. This large, conspicuous and fast-moving epibenthic echinoid is common in several areas of the Atlantic Ocean in depths of 200-2,870 m (Mortensen, 1935). Grassle et al, (1975) note that the species is very active, and include an excellent photograph of a specimen in situ. Tongue of the Ocean representatives of this species differ in some respects from their conspecifics further to the north, which were observed by the author during Alvin Dive 592, ata depth of 1930-1988 m, in the area of Deepwater Dumpsite 106, 39°09.9N, 71°54.8°W. The Tongue of the Ocean form has much smaller and less conspicuous hooves on the oral spines (see figs 2a, 2b), although in all other respects this form conforms to the traditional concept of the species, These specimens also showed behavioural differences from the specimens observed during Alvin Dive 592. When being approached by the submersible, the northern specimens, upon sensing the bow wave (or the lights) would immediately “gallop” away from the source of the disturbance. The Tongue of the Ocean specimens were less inclined to move away rapidly; in many cases they remained in position, but directed their aboral spines away from the source of the disturbance. Ellipinion delagei reached great population densities in some areas, especially during transect 7. This species is almost completely transparent when alive, and the body is virtually colourless. Through the body wall the coiled sediment- filled intestine can be clearly seen, Barham et al. (1967) and Pawson (1976) have found that some elasipodids tend to be oriented so that their anterior ends face into the prevailing current. In the case of Ellipinion, this tendency was exhibited to some extent, but was not rigidly followed. In some areas all specimens appeared to be facing in the same direction (fig. 2c), while in others a more random pattern of orientation was observed. In the study area there was almost no detectable current activity, and it is possible that in the absence of a significant current, the animals have a rather random orientation. Benthodytes typica in the Tongue of the Ocean is generally light brownish to cream, with little evidence of shades of red, The anterior end is dark brown. This species was not observed to swim, not even when violently disturbed by the submersible’s bow wave. A dense population of small (approximately 50 mm long) specimens was found in one area during transect 5. This might have been a year-class of specimens, leading to the suggestion that this species might have an annual reproductive cycle. Benthodytes sanguinolenta is much more reddish than the preceding species, and usually ECHINODERMS FROM THE TONGUE OF THE OCEAN, BAHAMAS Fig. 2. Echinoids and holothurians. 2a, Hygrosoma petersi from Alvin Dive 592 (see text), with conspicuous hooves on oral spines. 2b, Hygrosoma petersi from Alvin Dive 703. Note that hooves are not visible, and that aboral spines are directed away from source of light (or from bow wave of submersible). 2c, six specimens of Ellipinion delagei, one of Benthodytes lingua (far right) and one of Psychropotes depressa (top left). Note that in this photo all specimens of Ellipinion are facing in the same direction. 138 DAVID L. PAWSON considerably larger. Several specimens were observed to be swimming (figs 4a, b). Swimming — movements were similar to those described for “Euphronides sp.”’ (=Psychropotes depressa (Théel) according to Hansen (1975) ) by Pawson (1976). ___ Benthodytes lingua is a large species, usually more than 30 cm long, uniformly light to dark violet (figs le, 2c), Specimens were relatively common in all transects and were among the most conspicuous invertebrates encountered during Dive 703. This species is apparently incapable of swimming. Enypniastes extmia is common in the Tongue of the Ocean. The observations of Pawson (1976) can be enlarged upon here. Living specimens are light brown, translucent, and fragile. All but one of the specimens seen were swimming; some appeared to be drifting down towards the bottom passively, and others were ascending, gently undulating the anterodorsal veil. The mouth is apparently always directed upwards (figs 3a, b). All specimens observed swimming or floating past the viewing port were seen to have light coloured material of the same colour as the bottom sediments in their intestines. Eventually, a single specimen was found to be feeding on the seafloor (fig. 3c). The tentacles were very actively sweeping material in towards the mouth. We were unfortunately not able to determine the length of time that this specimen spent on the bottom, but judging by its active feeding rate, it would not need to remain there for more than a few minutes in order to fill its intestine. It seems likely that this species does rely on the seafloor for at least some of its food supply, and that itis nota permanent member of the nekton. Paelopatides sp. was very common in some transects, especially numbers 5 through 8. Further identification of this species is impossible, regrettably, because the University of Miami collections did not include specimens of this fragile light pink species, which can reach a length of approximately 50 mm. Pseudostichopus species A and B, and Mesothuria verrilli are often virtually indistinguishable through the viewing port of the submersible, because of their tendency to cover themselves with a layer of sediment (figs 4c, 5a). They were nowhere extremely common, but were conspicuous where they occurred because of their size (up to approximately 30 cm) and their conspicuous tracks. Systematic problems have necessitated a delay in making final identification of the two species of Pseudostichopus. Freyella sp. is a six-armed brisingid asteroid which lies mouth down on the seafloor, with arms resting on the seafloor (Fig. Sc). This is an unusual feeding position for members of this family, for they usually raise their arms into the water column (Pequegnat et al., 1972, fig. W-1; Pawson, 1976, Pl. 3 figs. A-B). Maureen Downey of this Institution (personal communication) believes that this may be a new species of Freyella. A single specimen was collected by the Alvin, and it is now in the collections of the U.S. National Museum. The 11]-armed Freyella sp. B was usually found on rocks, with its arms raised into the water column. “Luidia” is a single specimen of five-armed asteroid which superficially resembled the genus Luidia. The arms were flattened, strap-like, and tapered gently from the small central disc to the bluntly pointed extremities. No specimens similar to this were found in the University of Miami collection. Bathypectinura heros is a highly active uniformly orange ophiuroid, common in transects 1-4. It has been well described in two recent papers (Schoener, 1967; Madsen, 1973). Dr Daniel M. Cohen and I each saw a single individual of this species making active swimming movements in response to the approach of the submersible. In one case a specimen travelled a distance of approximately one metre by vigorously thrashing its arms. Maximum height above the seafloor was approximately one metre. ECHINODERMS FROM THE TONGUE OF THE OCEAN, BAHAMAS. 139 Cc Fig. 3. The holothurian, Enypniastes eximia. 3a, specimen swimming, with gently undulating anterodorsal veil. 3b, another swimming specimen, showing mouth and tentacles. Photo, D. M. Cohen. 3c, a specimen on seafloor, actively feeding. 140 DAVID L. PAWSON Fig. 4. Holothurians Benthodytes and Mesothuna. 4a, 4b, Benthodytes sanguinolenta swimming up and away from source of disturbance. 4c, Mesothuria vernilli on seafloor, with thin covering of sediment. ECHINODERMS FROM THE TONGUE OF THE OCEAN, BAHAMAS 141 Ophiomusium species A and B have not yet been positively identified due to systematic problems. Barham et al. (1967), Grassle et al. (1975) and Wigley and Emery (1967) discuss behaviour and population densities of species of Ophiomusium. “Red ophiuroid”: this puzzling ophiuroid (fig. 5b) was seen in fairly large numbers during Alvin dives further north (see Cohen and Pawson, 1977). Regrettably, we were unable to collect any specimens of this brittle star. It is about the same size as Ophiomusium lymani with a light red disc, light orange-red arms and conspicuous red tube feet. Tongue of the Ocean specimens, like those observed further to the north, adopted a great variety of postures and also, presumably, feeding methods. Commonly, the animal lies mouth down on the seafloor with two or three arms raised into the water column. Several specimens were found occupying burrows (of their own making?) with two or three arms extended onto the seafloor surface. Others were more or less completely buried, with two or three arms projecting. A preliminary survey of University of Miami collections revealed no specimens which might represent this species, but a more detailed examination by a specialist is required. 6. FORMATION OF CONICAL MOUNDS ON THE SEAFLOOR Several authors (Heezen and Hollister, 1971; Pequegnat et al., 1972, and others) have illustrated mounds of various kinds on the deep seafloor, and have suggested that the mounds are constructed by various burrowing organisms. It is well-known that some mounds with holes at their summits or with depressions near their bases are constructed by crustaceans or fishes. Other mounds, which are featureless, not associated obviously with other topographic features, have caused some puzzlement. During Alvin Dive 703 and other dives in the Tongue of the Ocean, several such mounds were disturbed mechanically by the movement of the submersible, and we were surprised to observe that in at least two cases the “‘core”’ of the mound consisted of a mass of holothurian faeces, such as may have been deposited by the genera Ellipinion or Pseudostichopus. Apparently the faeces, bound with some kind of mucus, act as a local surface feature upon which drifting and falling sediment can accumulate. In areas with little current activity, such mounds are almost perfectly symmetrical cones. 7. SHORT TRACKS ON THE SEAFLOOR In numerous seafloor photographs, short holothurian tracks have been observed. These may begin and end within the field of the photograph, with no trace of the holothurian which made them. Two explanations may be offered to account for these tracks, and both involve swimming activities: a. A swimming species landed on the seafloor, fed for a short distance, and then died, the dead animal eventually disappearing, leaving the track. Such a track may have been made very recently or perhaps hundreds of years ago, especially in areas were sedimentation rates are slow (Heezen and Hollister, 1971). b. A swimming species landed on the seafloor, fed for a short distance, and then left again, perhaps to seek a more palatable sediment. During Alvin Dive 703, two specimens of Benthodytes sanguinolenta were seen to swim away, leaving short tracks behind them (fig. 1b). It is possible that a type of “‘trial and error” feeding is commonplace among the more mobile holothurians. 8. POPULATION DENSITIES OF ECHINODERMS Counts of echinoderms in each transect (in numbers per 1000 m*) are givenin Table 2. For comparison, counts for another large invertebrate, the Venus flower basket sponge Euplectella suberea Thomson are also given. Some comments on distribution patterns follow. In view of the large scale sized used, it is not considered profitable to submit the figures given here to statistical 142 DAVID L. PAWSON Fig. 5. Holothurian Pseudostichopus and asterozoans. 5a, Pseudostichopus with covering of sediment and pieces of Thalassia. 5b, a specimen of the unidentified “red ophiuroid”’. Sc the “six-armed Freyella’’ in typical position on the seafloor. This specimen is unusual in having seven arms. ECHINODERMS FROM THE TONGUE OF THE OCEAN, BAHAMAS 143 tests for random versus non-random distribution patterns, such as were employed by Grassle et al. (1975) in their study of some deep-sea communities. However, it is believed that in spite of the limitations of the data, some inferences can be drawn which would seem to be reasonable in light of the data on feeding propensities presented elsewhere in this paper. With a few exceptions, population densities were generally rather low. Ophiomusium spp. in transect 2, Benthodytes typica/sanguinolenta in transect 5 and Ellipinion delagei in transect 7 approached densities of 1 per m*, but apparently did not exceed them. By contrast, Grassle et al. (1975) found that Ophiomusium lymani in their study area (Dives 280 and 436) was far more numerous, reaching densities of 1.7 and 2.4 specimens per m? respectively. However, densities of other echinoderm species listed by Grassle et al. were relatively low, and tend to agree well with those given here. a. Distribution patterns of “vegetarians”: Hygrosoma petersi, which feeds on Sargassum and Thalassia, was more or less evenly distributed in all transects, except that there was a notable population increase in transect 7 followed by a sharp decline in transect 8. Bathypectinura heros, by contrast, reached relatively high numbers during transects 1-4, and was virtually absent from all subsequent transects except transect 8. Note that the increase in this last transect is matched by a decline in numbers of Hygrosoma petersi. Although the Sargassum-Thalassia fragments are ubiquitous, it appears that the two species discussed above are not evenly scattered on the seafloor, and that there may be some evidence here of competitive exclusion, involving other species of echinoderms, or other invertebrates. b. Distribution patterns of ‘‘mud-ball swallowers’’: Phormosoma placenta was unevenly distributed over all transects, reaching peaks in transects 1, 2 and 7. Plesiodiadema antillarum and Salenocidaris profundi were common in the first four transects and then dropped out altogether. Perhaps the latter two species are more selective in their feeding and the endpoint of transect 4 might represent the local southern limit of the ‘“‘range” of their desirable food resource. No herds of Phormosoma like those described by Grassle et al. (1975) were found in the Tongue of the Ocean. Specimens were rather widely scattered. c. Distribution patterns of ‘non-selective’? mud swallowers: While most deep-sea holothurians might be classed as non-selective in their feeding habits, undoubtedly they continually move to areas of higher nutrient content for their feeding, in the same manner as do their shallow-water counterparts. Table 2 shows that several species are present in relatively low numbers in several transects, and then over a few transects population counts rise dramatically. The two species of Benthodytes achieved such high densities in transect 5 that it was impossible to count them with any degree of accuracy. Ellipinion delagei reached a peak in transect 7, and by transects 10 and 11 had disappeared completely. Psychropotes depressa and Benthodytes lingua also reached peak populations in transect 7, but in contrast to Ellipinion delagei these species were also well represented in transects 10 and 11. Paelopatides sp. shows a pattern similar in many respects to that of Psychropotes depressa and Benthodytes lingua, except that the peaks for the first species were reached in transects 6 and 7, while for the latter two species there was a sharp increase in numbers from transect 6 to transect 7. Mesothuria verrilli and Pseudostichopus A and B were more or less evenly distributed over all transects, their numbers dropping slowly towards the last transect runs. Examination of intestine contents of several of the species discussed above revealed no obvious differences at the gross level. A study of organic content of sediments in the various transect areas might be useful. Sanders et al. (1965) found that for the Gay Head — Bermuda study, distribution of animals on the seafloor was not obviously correlated with organic content of the sediment. This negative result was ascribed to an inadequacy in the analytical techniques 144 DAVID L. PAWSON used. Clearly, further study of this topic is required. It is evident from Table 2 that two important population peaks for mud-swallowers were: reached in different transects. The first, in transect 7, includes Ellipinion delagei, Psychropotes depressa and Benthodytes lingua, while the second comprises Benthodytes typica and B, sanguinolenta. Differing nutritional requirements, competition, aggregation as a result of social behaviour might be suggested as causes for staggering of population peaks for the various feeding categories discussed above. Contagious distributions have been noted in deep-sea echinoderms on several occasions (see Grassle et al., 1975). The phenomenon of social behaviour in echinoderms has received some attention recently (Pearse and Arch, 1969; Grtinbaum et al., 1978), and it is possible that the obvious “‘clumping” of deep-sea species discussed here is the result of some social interaction that is not yet understood. In some cases, the aggregation is clearly related to distribution of food resources, as shown by Pawson (1976) for a species of Scotoplanes, but for the most part, no obvious explanations are forthcoming. ACKNOWLEDGEMENTS Submersible dives were supported by the U.S. National Oceanic and Atmospheric Administration as part of a study of deepwater benthic animals in the Tongue of the Ocean. Iam especially grateful to Dr Daniel M. Cohen for facilitating my participation in dives of the Alvin | and for passing on to me his observations on behaviour of various echinoderms. Drs Klaus Ruetzler and James Norris of this Institution kindly gave me information on siliceous sponges and Sargassum spp. respectively. Dr Gilbert L. Voss of the University of Miami generously allowed me access to the Tongue of the Ocean echinoderm collections. My former assistant, Miss Cynthia Gust, helped with sorting of Miami collections and with collation of station data, REFERENCES Barham, E. G., N. J. Ayer, Jr. and R. E. Boyce, 1967. Macrobenthos of the San Diego Trough: photographic census and observations from bathyscaphe, Trieste. Deep-Sea Res. 14: 773-784. Clark, H. L., 1907. The apodous holothurians (a monograph of the Synaptidae and Molpadidae). Smithson. Contr. Knowl. 35: 1-231. Cohen, D. M. and D, L, Pawson, 1977. Observations from DSRV Alvin on populations of benthic fishes and selected larger invertebrates in and near DWD-106, Pl. 423-450 in Baseline report of environmental conditions in Deepwater Dumpsite 106, NOAA Dumpsite Evaluation Report 77-1, vol. II, Biological characteristics, U.S. Dept. Commerce, NOAA, Nat. Ocean. Survey Rockville, Maryland. Fell, H. B., 1952. Echinoderms from southern New Zealand. Zool. Publs. Victoria Univ. Wellington 18: 1-37. Grassle, J. F., H. L. Sanders, R. R. Hessler, G. T. Rowe and T. McLellan, 1975. Pattern and zonation: a study of the bathyal megafauna using the research submersible Alvin. Deep-Sea Res. 22: 457-481. Grtinbaum, H., G. Bergman, D. P. Abbott and J. C. Ogden, 1978. Intraspecific agonistic behaviour in the rock-boring sea urchin Echinometra lucunter (L.) (Echinodermata: Echinoidea). Bull. Mar. Sci. 28(1): 181-188. Hansen, B., 1975. Systematics and biology of the deep-sea holothurians. Part 1. Elasipoda. Galathea Rep. 13: 1-262. Heezen, B. C. and C, D, Hollister, 1971. The face of the deep. Oxford University Press, New York, Pp. 659. Lawrence, J. M., 1975. On the relationships between marine plants and sea urchins, Oceanogr. Mar. Biol. Ann, Rev. 13: 213-286. Madsen, F. J., 1973. The Ophiodermatidae (Ophiuroidea). Galathea Rep. 12: 133-143. ECHINODERMS FROM THE TONGUE OF THE OCEAN, BAHAMAS 145 ‘Menzies, R. J.,R. Y. George and G. T. Rowe, 1973. Abyssal environment and ecology of the world oceans. John Wiley, New York, Pp. xxiiit 488. 'Menzies, R. J. and G. T. Rowe, 1969. The distribution and significance of detrital turtle grass, Thalassia testudinata, on the deep-sea floor off North Carolina. Jnt. Revue ges. Hydrobiol. 54 (2): 217-222. ‘Menzies, R. J., J. S. Zaneveld and R. M. Pratt, 1967. Transported turtle grass as a source of organic enrichment of abyssal sediments off North Carolina, Deep-Sea Res. 14: 111-112. NV \Mortensen, T., 1935. A monograph of the Echinoidea. IJ. Bothriocidaroida, Melonechinoida, Lepidocentroida and Stirodonta. Reitzel, Copenhagen. Pp. 647. 1938. On the vegetarian diet of some deep-sea echinoids. Annot. Zool, Japan 17: 255-228. 1940. A monograph of the Echinoidea. III. 1. Aulodonta. Reitzel, Copenhagen. Pp. 370. Pawson, D. L., 1976. Some aspects of the biology of deep-sea echinoderms Thalassia jugosl. 12 (1): 287-293. Pearse, J. S. and S. W. Arch, 1969. The aggregation behaviour of Diadema (Echinodermata, Echinoidea). Micronesica 5 (1): 165-171. Pequegnat, W. E., B. M. James, A. H. Bouma, W. R. Bryant and A. D. Fredericks, 1972. Photographic stydy of deep-sea environments of the Gulf of Mexico. Jn V. J. Henry and R. Rezak (Eds.) Texas A & M University Oceanographic Studies Vol. 3: Contributions on the geological oceanography of the Gulf of Mexico: 67-128. Gulf Publishing Co., Houston. } Rhoads, D. C. and D. K. Young, 1971. Animal-sediment relationships in Cape Cod Bay, Massachusetts, II. Reworking by Molpadia oolitica (Holothuroidea). Mar. Biol. 11: 225-261. Sanders, H. L., R. R. Hessler and G. R. Hampson, 1965. An introduction to the study of deep-sea benthic faunal ’ assemblages along the Gay Head Bermuda transect. Deep-Sea Res. 12: 845-867. Schoener, A., 1967. Occurrence of Bathypectinura (Ophiuroidea) in New Zealand waters. Trans. Roy. Soc. New Zealand 10 (9): 77-80. Schoener, A. and G. T. Rowe, 1970. Pelagic Sargassum and its presence among the deep-sea benthos. Deep-Sea Res. 17: 923-925. Wigley, R. L. and K. O. Emery, 1967. Benthic animals, particularly Hyalinoecia (Annelida) and Ophiomusium (Echinodermata) in sea-bottom photographs from the continental slope. J. B. Hersey (ed.) Deep-sea photography: 235-249. John Hopkins Press, Baltimore. 9, ABIOMETRICAL STUDY OF POPULATIONS OF THE EUROPEAN SEA-URCHIN ECHINUS ESCULENTUS (ECHINODERMATA: ECHINOIDEA) FROM FOUR AREAS OF THE BRITISH ISLES DAVID NICHOLS Department of Biological Sciences, University of Exeter, Devon, England SUMMARY Results submitted by mainly amateur diving groups during Underwater Conservation Year 1977 in the United Kingdom show that there are regional differences in the relationship between both size and shape of specimens of the European sea-urchin Echinus esculentus Linnaeus and the depth at which they occur. Popylations from South-West England are significantly bigger at all depths than those from the other areas surveyed, those from Western Scotland increase in size more rapidly with increasing depth of water, and those from the North Sea decrease in size with increasing depth. Two sites surveyed in South-West Ireland show that exposure may affect the size of urchins inhabiting shallow waters. The results are compared with those of a similar survey by Larsson (1968) on the same species in Swedish waters. INTRODUCTION The European sea-urchin, Echinus esculentus Linnaeus, was the subject of a nationwide survey during 1977 as part of a special project for amateur divers during Underwater Conservation Year (UCY 77) in the United Kingdom. The project was timely, since there has been unsupported evidence over the past few years that populations of the animal have been suffering at the hands of collectors for the curio trade (see, for instance, Natural Environment Research Council, 1973). Itis possible that the animal may also become the subject of additional pressure from the luxury food trade, since the roe is considered a delicacy by some (Southward and Southward, 1975). In addition, there is contradiction in the results of previous studies that have examined the population structure of this animal in European seas: Moore (1935) and Reid (1935), working on dredged material from the Isle of Man and Scotland, both state that the largest urchins inhabit shallow water, while Larsson (1968), who used SCUBA techniques to study populations in the Koster Fjord region of Sweden, found larger specimens in deeper water. Studies of extensive populations, and over a wide geographical area, require larger teams of investigators than are usually available in the normal course of scientific work, and for this reason the opportunity to use the diving expertise of competent amateurs during a year of special effort was welcomed. Before the start of the project, standardised instructions were prepared which outlined in straightforward terms the procedures to be adopted. Several different observational and experimental projects were suggested (Nichols, 1978a), and this paper describes the results of one, an investigation of the size and shape of the urchins relative to the depth of water at which they live. METHODS Details of the instructions sent out to diving groups prior to the start of the project are given in Nichols (1979), Diving groups were advised to construct a simple pair of calipers with which the two dimensions of diameter and height could be taken on the animal while underwater and read off along the side of a recording board. Since this was also a conservation exercise, a more elaborate design of calipers was suggested to some teams which obviated the need to disturb the urchins, even when taking the height measurement. Australian Museum Memoir No. 16, 1982, 147-163 148 DAVID NICHOLS Teams were asked to record the measurements of all urchins encountered within ¢ | convenient area at any depth. Where possible, depth gauges were calibrated or corrected in | pressure chambers, and all depth readings were corrected to Lowest Astronomical Tide. The |} surveys were conducted within four months of each other (Table 1). Results were transferred to} a standard form for return to the project co-ordinator, and data were processed using a desk computer to provide standard statistical treatment. In his study of a Swedish population of the same species of sea-urchin in 1966, Larsson recorded only the diameter of individual urchins and plotted this dimension against depth off water as a histogram (see Larsson, 1968, fig. 15). The time of year that Larsson made his survey is not given. These results have been replotted in the present paper ina comparable form to those: of the British specimens and standard statistical treatment applied to them. RESULTS Of the total results submitted, those from four localities have been selected for the purposes -} of this paper, to provide as wide a geographical spread around the British Isles as possible (Fig. 1). Details of the sites in each area, the survey teams and the numbers of urchins measured in’ each case are given in Table 1. The test dimensions of diameter and height are plotted separately \ | for each site, and the ratio of height to diameter also plotted on a separate axis on the same graph || (figs 2 to 4). Table 1. Summary of the surveys included in this paper. Location Survey team and Leaders Dates No. of urchins measured © 1. W. Scotland, Army Air Corps, 10-24 a. Black Is 103 | Middle Wallop, August b. Crowlin Is 84 Isle of Skye Sub-Aqua Club 1977 c. Eilean Ban 107 © d. Eileanan Dubha 255 Set. R. Perren e. Tulm Is 154 Total 703 2. S. W. England, University of 30 June 151 Exeter Sub- to Lamorna Cove Aqua Club 8 July 1977 Deborah Garner Andrew Smith 3. S. W. Ireland, University of 16 June i. Carrigavaddra 504 Cambridge to ii. Sheelane Is 504 Bantry Bay Sub-Aqua Club 17 July Alasdair Edwards 1977 Total 1008 Alison Morris 4. North Sea, University of 8 May 40 Durham to St. Abb’s Sub-Aqua Club 16 June and 1977 Newton Christine Howson Charles Anderson BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS 149 SWEDEN Koster Fjord W. SCOTLAND Isle of Skye SW. IRELAND S.W. ENGLAND Lamorna Cove aS A Fig. 1. Map of part of Europe, to show the location of the four areas around the British Isles surveyed in this study, and the location of the area in Sweden surveyed by Larsson in 1966. Figure 2 shows a summary of results from five separate sites in Western Scotland; four of the sites were within a radius of 5 km, while the fifth, Tulm Island, was about 30 km away from the main group of sites. Figure 2f plots the mean values for all five sites. Confidence limits (standard errors) are included as vertical bars about the mean values at each" depth, where these limits extend beyond the size of the symbols. In each case linear regressions of best fit are drawn through the points. The graphs in figure 2 show that for all sites in Western Scotland the dimensions of diameter and height increase with depth of water, and there is no significant difference between rate of increase in the two dimensions except in the case of Tulm Island (fig. 2e). So far as the ratios of the two dimensions are concerned, Tulm Island ‘tig. 2e) is the only one to show a negative slope, meaning that at this site alone the animals may become squatter in deeper water. However, it must be added that only two animals were measured at each of the depths 17 and 18 m, and without the results from these small samples the regression for the ratio, like the others, shows a positive slope. In South-West Ireland (fig. 3), two sites were surveyed. The one (Carrigavaddra) was 150 Fig. 2. Graphs showing the relationship between mean dimensions (diameter, open circles; height, open triangles) and depth of water in which they occur of populations of the sea-urchin Echinus esculentu of Skye, Western Scotland (a to e), and the mean values for all sites (f). The mean values for the ratio of height to diameter: for each individual are plotted (solid Squares) on the same graphs, the left-hand axis being the dimensions and the: right-hand axis being the ratio in each case. Calculated linear regressions are drawn in as solid lines, and the confidence - limits (standard error) are drawn as vertical lines about the mean values where these limits extend beyond the symbols. Regression equations are as follows: a. Black Island. b. Crowlin Island. c. Eilean Ban. d. _Eileanan Dubha. e. Tulm Island. f. Mean of all sites. DAVID NICHOLS Diameter: Height: Ratio: Diameter: Height: Ratio: Diameter: Height: Ratio: Diameter: Height: Ratio: Diameter: Height: Ratio: Diameter: Height: Ratio: mate a) Xe < y y 7 = 7.38 + 0.15x y = 4.60 + 0.18x = 0.64 + 0.0073x = 6.47 + 0.264x = 4.55 + 0.269x y = 0.57 + 0.017x = 7.04 + 0.245x = 5.27 + 0.237x = 0.74 + 0.007x = 7.81 + 0.036x = 5.17 + 0.077x = 0.63 + 0.01x = 6.13 + 0.345x = 5.44 + 0.209x = 0.83 — 0.004x = 7.05 + 0.176x = 4.91 + 0.182x = 0.69 + 0.007x s from five separate sites near the Isle4 BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS 115 4c i 9 \ | ray a. Black Is. : ia Ratio HY b. Crowlin Is. 1 me BH 1 z F Ben eS ere Se Iie sd ian + us Ratio 7; ' 7 ae 2 Seid, \ 2 ms 42 ws \ $ Dimensions * ae v 6 of test, cm x 0 $ A ost Cole —Sr Beas x Diameter — Ie ty 8 oftest, om test, cm ——, Ly ns 2 POR 5% -O.f at ke Height —=—_ 1 4 ae \x G | eee SN LA ier Depth of water. ES m : kk “ Depth of water, m , 1 L L 1 0 5 10 15 (0) 5 10 15 20 oF c. Eilean Ban pape ices | Ratio = v 84 d. Eileanan Dubha nen ae Ratio Ht iis i . “a oY Diam| O7 } an a 12 ik . if Dimensions Pe Bee of test, cm ee +” pio 0-0 ai Sr 10 i 0.54 2 2 Dimensions -O Ly my of test.cm. =~ oR: Oo -O g fal b Lg @ Diameter” “2s = SOs 9 / t py Pe 2 i A y eS x ae res eG Se os a a L63- RoR Height “BA A ; 4 ‘A t a +4 War So he A Depth of water, m ss Depth of water, m 0 > z a ital n Se i L 15 o) 5 10 1S 20 e. Tulm Is an 1 5 95 f. MEAN. All Surveys 94 a7 ah os A kati pian kK 5 } ] Bese Si ‘i : ‘ Ratio #7 ( ‘ See Hee ons 3 ‘ ST g- tera if & 4 . vy = Wie ie Sen Lae: ay v t 4 ne Ybiam ele et t qT 1 fe) {/ 2. 12 2/ ° Dimensions [ Ki “ie | Oke of test, cm + var \ | Diametet it H10 3g 2-8 Lt oats Dimensions pee ak ee eee ~: ere tcl oftest.om. 3-2. / > s “9 A | Heat is fo oe ge 6— 4 my ‘Sg sgt ae em a Depth of water, m Depth of water, m 1 1 1 5 0 5 10 5 20 152 DAVID NICHOLS Fig. 3. Graphs showing the relationship between mean values of diameter, height and ratio to depth of water for + specimens of Echinus esculentus from two sites in Bantry Bay, South-West Ireland. Conventions as for Figure 2. Regression equations are as follows: a. Carrigavaddra. Diameter: Height: Ratio: b. Sheelane Island. Diameter: Height: Ratio: 45 + 0.031x 52 + 0.047x .768 + 0.003x 8 6 0 SOL Se nO RON. 6 0 .078 + 0.010x .762 + 0.0025x BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS 1.0 a. Carrigavaddra Ratio ae 10 Dimensions a o~ f of test, ae ; ee ts v Depth of water, m. ) 5 10 15 20 25 30 b. Sheelane Island is SaOUSS oaks Ratio Peed abak irene ae Bie a apna agai "0.8 ati a aa eee RatioHtty 10 07 Dimensions of test, cm. 9 8+ — Bea. ers 2 ; ree. Dame Jace] 7 ; ee ices eat eee J a aig) es I ee tea ete 8 Depth of water, m. | are O 5 10 15 20 25 30 53 154 DAVID NICHOLS Fig. 4. Graphs showing the relationship between mean values of diameter, height and ratio for specimens of Echinus esculentus froma. Lamorna Cove, South-West England, and, b. St. Abb’s and Newton, on the North Sea coast of Britain. Conventions as for Figure 2. Regression equations are as follows: a. South-West England. Diameter: y = 10.88 + 0.033x Height: y = 8.433 + 0.071x Ratio: y = 0.797 + 0.003x b. North Sea. Diameter: y = 9.49 — 0.168x Height: y = 6.768 + 0.104x Ratio: y = 0.712 + 0.0021x 195 BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS ‘WW JajeM JO Ydaq S UOIMEN 8 SIFY IS ‘VAS HLYON O€ Se Od Sl Ol S O SACD eUIOWET “GNVIONS MS 156 DAVID NICHOLS towards the seaward end of a long, narrow inlet (Bantry Bay), while the other (Sheelane Island)» was at a slightly more sheltered site further into the Bay. The results for Carrigavaddra (fig. 3a)! are similar to the mean result from Western Scotland (fig. 2f), while those for Sheelane Island — (fig. 3b) show that here the average diameters of the urchins measured decreased with depth of water, and the average heights increased. At both sites the ratios show a positive slope, that is, on average the urchins are becoming taller with depth. In South-West England the site chosen was at Lamorna Cove, Cornwall, a fairly exposed site. The results (fig. 4a) show that here again diameter and height, and the tallness of the urchins, increase with depth of water. The last area consisted of two sites near the English-Scottish border on the North Sea coast of the United Kingdom. This area yielded the least satisfactory set of results, because the North Sea between the British Isles and Scandinavia is more turbid than other waters from which surveys were taken, and few populations of the sea-urchin occur here. The two sites are about 30 km apart; the one at Newton-By-Sea, in the English county of Northumberland, has populations of urchins at depths of between 1 and 5 m, while the other, at St. Abb’s, in the Scottish county of Berwickshire, has populations at 11 m depth. The results (fig. 4b) are combined, despite their slightly separated provenance. The graph shows that this area may be different from the others here described, in that both diameter and height decrease with depth of water; the tallness, however, increases. However, it must be noted that the sample size in this case is somewhat smaller than the others. The calculated mean linear regressions for diameter, height and ratio for the four British Isles areas are summarised in figure 5. Where more than one site in each area has been surveyed, asin Western Scotland (5 sites) and South West Ireland (2 sites), the means of all sites in that area are plotted here, to show regional differences, if any. The lower part of the figures shows the diameters (upper line of each quadrilateral) and height (lower line) and the depth range for each of the surveyed areas; the upper part of the figure shows the mean ratios for the four areas. The figure shows that the urchin populations from South-West England are significantly different from those of the other areas, in that both the overall size at all depths is larger, and they have taller tests at all depths. Other differences that are revealed by this figure are that the rate of increase in size with depth and the increase in tallness with depth for urchins from Western Scotland are both greater than for other areas, and that the small and shallow sample from the North Sea shows that here the urchins decrease in size with depth of water. Notall survey teams that contributed to this project were able to dive to the depths to which Echinus extends. Indeed, in some areas, such as the North Sea, the urchins themselves do not extend to any great depth, at least in the area surveyed. At some sites too, rather few specimens were encountered in deeper water, so the inclusion of the deepest results may be to some extent unjustified. To make a fairer comparison, the linear regressions for all sites down to a depth of 15 m only have been summarised in figure 6. In fact, the general statements above about the separation of the various areas hold true for these restricted results too. But there are minor differences. For instance, for populations in South-West England, omission of the deeper specimens shows that the regressions for both diameter and height (top and bottom lines for the ‘South-West England’ quadrilateral in figure 6) now show negative slopes, that is, the specimens become marginally smaller with depth. The ratio (tallness) for this area (upper part of the figure) also now shows a negative slope, though there is no significant difference between this line and a constant ratio (horizontal line) (P>0.5). Larsson’s (1968) paper included only diameters of the specimens he measured from the Koster Fjord, Sweden, in 1966. These results have been plotted on the same axes as the results here described from the British Isles (fig. 7). This graph shows that the Swedish population BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS — 157 structure for this urchin is very different from that from all the British sites; in particular, the urchins in shallow water have a much smaller mean size, though the ranges overlap somewhat in _ deeper populations. It should be mentioned that Larsson’s results and those reported here are —_— —_ separated by 11 years, but variation in individual sizes with time to the extent seen in these sets of results is unlikely, and in any case cannot be tested until the results from either country are repeated in the future. DISCUSSION In this paper, trends in the size and shape of the components of the populations with changing depth of water are indicated by plotting linear regressions of the mean values of each depth for the two dimensions of diameter and height of the animal’s body and also that of mean values at each depth of the ratio between these two dimensions for each individual. It is insufficient to rely solely on the regressions for diameter and height to express shape. For instance, in cases where the regressions for diameter and height lie approximately parallel to one another, as for Crowlin Island, Western Scotland (fig. 2b), this does not necessarily mean that the tallness of the animal remains constant throughout the depths surveyed; indeed, in this example the regression for the ratio of height to diameter has a markedly positive slope, showing that the average tallness of the animals at this site increases significantly with depth. There is little substantial difference between results for the sites surveyed in Western Scotland, so far as the calculated linear regressions reveal the trends. Black Island, Crowlin Island, Eilean Ban and Eileanan Dubha (fig. 2, a to d) are all strikingly similar. The somewhat different results obtained from Tulm Island (fig. 2e) are more apparent than real, in that the regressions are skewed because of the inclusion of results from small samples (2 urchins only at each) from 17 and 18 m depth. In particular, the line for the ratio of height to diameter shows a negative slope. If results from these two small samples are omitted, the linear regressions show slopes that are similar to those for all the other Scottish sites. The two sites surveyed in South-West Ireland differ from each other in that the one (Carrigavaddra) is nearer the open sea, and therefore more exposed, than the other (Sheelane Island), which is about half way up the elongated bay. This may account to some extent for the differences between the regressions for diameter and height, the urchins being relatively smaller inshore at Carrigavaddra, but larger inshore at Sheelane. Perhaps this reflects the exposure of the area in which they live, since a larger urchin is more likely to be displaced in the rougher waters of the exposed site. The site in South-West England, Lamorna Cove (fig. 4a) is remarkable for a drop in the average size of urchins between about 5 and 10 m depth of water. This could be explained by some factor, such as the substratum, affecting the general success of the urchins at these depths, though nothing was reported by the diving team; or alternatively it could be a factor related to predation. It happens that this site is a favourite one for the collection of urchins by amateur divers for the curio trade (Nichols, 1978b). Such divers normally descend to between 5 and 10 m depth so that the dive can be recorded in their log-books, and it seems quite likely that the activities of these people have denuded the populations of the larger urchins at these depths. . The results from the North Sea sites, combined on one graph (fig. 4b), are the least satisfactory in this study; although they show an unequivocal trend towards a reduction in size in deeper water, this cannot be supported with confidence on these small and separated populations. The shape of the urchins from both sites appears to remain almost unchanged in the 10 m depth through which the animals occur. A comparison of all four areas surveyed around the British Isles, as shown on the summary 158 DAVID NICHOLS Fig. 5. Summary graph of means of all results from each of the four areas of the British Isles surveyed in this investigation. In the lower part of the figure the top line of each quadrilateral is the regression for the mean values of the diameters of sea-urchins over the depth range surveyed, and the lower line is the regression for the mean values of the heights. In the upper part of the figure, the regressions for the mean values of the ratio of height to diameter against depth of water are plotted. Left-hand axis represents the dimensions and right-hand axis the ratios. Regression equations are as follows: South-West England. Diameter: Height: Ratio: Western Scotland. Diameter: Height: Ratio: South-West Ireland. Diameter: Height: Ratio: North Sea. Diameter: Height: Ratio: Rebeca | + 0.033x + 0.071x + 0.0033x + 0.176x + 0.182x + 0.007x + 0.009x + 0.029x + 0.0025x = (Wiser — 0.104x + 0.0021x BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS — 159 Dimensions of test, cm. SW. England SW. Ireland Depth of water. m. 160 DAVID NICHOLS Fig. 6. Summary graph of means of results from each of the four areas of the British Isles surveyed in this investigation, but ignoring all results from below 15 m depth of water. Conventions as for Figure 5. Regression equations are as follows: South-West England. Diameter: = Height: Ratio: IDL ALS WONG: | = 9.159 — 0.038x | = 0.843 — 0.004x = 6.75 + 0.219x Height: 5.64 + 0.227x Ratio: 0.673 + 0.0095x y yy, y Western Scotland. Diameter: y yy, MW South-West Ireland. Diameter: y = 8.33 — 0.005x W Vf M MW Mf Height: = 6.20 + 0.04x Ratio: = 0.748 + 0.0051x = 9.49 — 0.168x = 6.768 — 0.104x = 0.712 + 0.0021x North Sea. Diameter: Height: Ratio: BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS 161 09 S.W. England he ee SW. Ireland i 08 North Sea W.Scotlan 12. 06 Dimensions of test, cm. 11 10 S.W. England 162 DAVID NICHOLS 15 Diameter of test, cm. Sw oD: are SW En 10K yy a Leg SW Ireland a a ato Ww. 1 15 20 25 30 35 Depth of water, m. Fig. 7. Graph showing the relationship between mean values for test diameter and depth of water in which they occur * (closed triangles) for the sea-urchin Echinus esculentus in the Koster Fjord, Sweden. Data extracted from Larsson (1968). Other conventions as in Figure 2. Regressions for the mean values for test diameters of urchins from the four British areas investigated in the present work are also included within the same axes for comparison. Regression equations are as follows: Koster Fjord. y = 1.433 + 0.242x South-West England. y = 10.88 + 0.035x North Sea. y= 9:49 —0.168x South-West Ireland. y = 8.234 + 0.009x Western Scotland. vy = 7,051 + 0,176x graph (fig. 5), reveals the following features: first, South-West England has populations of urchins that are significantly larger at all depths than elsewhere; secondly, in Western Scotland the size of urchins increases to a greater extent with increasing depth than in the other localities; thirdly, the North Sea urchins alone, so far as meaningful conclusions can be drawn from these results, decrease in size with increasing depth of water; and, fourthly, the populations in South-West England and South-West Ireland show little significant change in size with increasing depth. That these general conclusions are not merely a factor of the different depths to which the various diving teams surveyed is borne out by the similarity in appearance of the graph if the results are recalculated down to only 15 m depth of water (fig. 6). This depth was chosen principally because it is within that to which all the separate sites were surveyed, except those in the North Sea, and also because it roughly corresponds to the depth to which algae, a principal food of Echinus, penetrate. Insufficient data are available on the physical conditions in the areas and sites surveyed to suggest reasons for the features that have emerged. However, the following suggestions are made: perhaps the slightly warmer waters in South-West England induce a larger overall size of urchins in that area; perhaps a difference in the availability of food in shallow waters helps explain the small size of urchins in the shallowest populations of Western Scotland; perhaps there is a more rapid fall-off in the density of algae with depth in the turbid water of the North Sea causing a reduction in mean size of urchins with increasing depth in that area. The study BIOMETRICAL STUDY OF BRITISH ECHINUS ESCULENTUS 163 underlines the need for detailed physical, faunistic and floristic information to be considered alongside the biometrical data when surveys like this are undertaken if the biological significance of the trends uncovered is to be suggested. Without similar additional information from the Swedish waters surveyed by Larsson in 1966 (fig. 7), it is just as difficult to suggest reasons for the marked differences between populations there and from the British Isles. The average size of the Swedish urchins is small in shallow waters, but increases at a greater rate with depth of water than those from any of the British areas, coming to oyerlap the British size ranges in deeper waters. Although the time of year of Larsson’s survey is not known, it is unlikely that this, or a possible difference in the time of spawning of the Swedish population, could account for the difference between the Swedish and British populations. For all the shortcomings in interpretation of these results, owing to the lack of ecological data taken with the initial measurements, the survey conducted by separate groups of divers, many of them amateurs, during Underwater Conservation Year in the United Kingdom, took the study of the population structure of Echinus esculentus much further than had previously been possible. It has underlined that there are regional differences in the overall size of individuals, a fact that could be significant to the curio industry now based on the dried test of the animal, and to any proposed industry based on its roe; it has shown that the population structure apparently can be affected by the exposure of the area; and it has suggested that the effects of human predation can be detected by surveys of this sort. More than this, however, it has demonstrated that the collection of scientific data can be aided by the activities of amateur divers, suitably briefed, and hopefully the experience gained by such people in taking part in such a programme will help them encourage others to conserve species like Echinus that are now under threat. ACKNOWLEDGEMENTS The Author thanks the Projects Officer of UCY 77, Dr Charles Sheppard, and his successor, Dr Bob Earll, for their help in involving diving teams in this survey; he also thanks Dr D. Stradling and Mr C. J. Lawrence for advice on statistical techniques and Dr A. B. Smith for helpful suggestions on the manuscript; finally he thanks the many individual divers and their organisers for all their hard work in collecting the data on which this paper is based. References Larsson, B. A. S., 1968. SCUBA-studies on vertical distribution of Swedish rocky-bottom echinoderms. A methodological study. Ophelia, 5: 137-156, 16 figs. Natural Environment Research Council, 1973. Marine wildlife conservation. N.E.R.C. Publications, series B, 5. pp. 40, 5 tables, 3 figs, 19 pl. Moore, H. B., 1935. A comparison of the biology of Echinus esculentus in different habitats, Pt. II, 7. mar. biol. Ass. U.K., 20: 109-128, 7 tables, 10 figs. Nichols, D., 1978a. Using amateur divers in scientific research. Fedn. Aust. Underw. Instr. Tech. Bull. 2: 35-53, 6 figs. Nichols, D., 1978b. Can we save the sea-urchin? Diver, 23: 490-491, 2 figs, 3 pl. Nichols, D., 1979. A nationwide survey of the British sea-urchin Echinus esculentus. Progr. Underwater Sci., 4: 161-187. Reid, D. M., 1935. The range of the sea-urchin Echinus esculentus. ¥. anim. Ecol. 4: 7-16, 2 tables, 1 fig. Southward, A. J. and E. C. Southward, 1975. Endangered urchins. New Sci. 66: 70-72, 3 figs. 10. CHANGES IN THE ECHINODERM FAUNAIN A POLLUTED AREA ON THE COAST OF BRAZIL TANIA MARIA DE CAMARGO Institute of Oceanography, University of Sao Paulo, Brazil SUMMARY The purpose of this research was to compare the changes in echinoderm fauna in a region under increasing eutrophication. A preliminary survey of the fauna was conducted by Tommasi in 1964; such results here served as a baseline for comparison with the data obtained during a more thorough investigation made from 1974 to 1976 by a team from the Institute of Oceanography. The animals were identified to species and their distributions correlated to environmental parameters like type of sediment, water temperature, salinity, depth and dissolved oxygen, as well as parameters indicative of pollution, mostly nutrients and turbidity. The results for 1974 to 1976 period indicate a lower diversity as well as the disappearance of a number of species when compared to the 1964 period. INTRODUCTION The echinoderm fauna of the coast of the State of Sao Paulo, Brazil, has been studied by Bernasconi (1956) and by Tommasi (1957, 1958, 1965 and 1966). In 1964 Tommasi studied the distribution of these animals as part of a more thorough investigation of the benthic assemblages of the Bay of Santos (24°00’S, 46°20’W), his results were published in 1967, at which time considerable concern had already been aroused by drastic changes which had occurred as the result of sewage, dredgings as well as the installation of smelters and other industrial plants. As pointed out by Isaacs (1973) the discharge of sewage as well as of industrial effluents in estuarine regions can affect the communities of benthonic invertebrates in two ways: (a) a great influx of organic matter tends to have a deleterious effect on filter feeders, detritus feeders and also on their predators; (b) the change in consistency of sediments, the increase in the level of heavy metals and toxic organic compounds, the reduction in dissolved oxygen as well as the increase in sulphides as a result of the deposition and decomposition of organic matter, may inhibit larval attachment or may have toxic effects directly on already attached larvae or on adults which may be sensitive to such pollutants. From 1964 onwards the Santos region has suffered a considerable increase in eutrophication levels as a result of its development as a summer resort, accommodating during the summer months more than one million people, as well as the increase of its docking facilities and installation of new industrial plants. In 1974 a joint programme was established between the Institute of Oceanography of the University of Sao Paulo and the State Centre for Basic Monitoring of Environment (CETESB) in order to investigate the present status of the animal assemblages and environmental conditions in the region. As part of this programme, a study was carried out on the distribution of the echinoderm fauna, similar to that carried out by Tommasi (1967). As much as possible, the echinoderm distribution and abundance has been considered in relation to environmental parameters such as nutrient content, temperature, salinity, type of sediment, dissolved oxygen and depth. Australian Museum Memoir No. 16, 1982, 165-173 166 TANIA MARIA DE CAMARGO DESCRIPTION OF AREA STUDIED The bay and the estuary of Santos (24°00’S, 46°20’ W) on the coast of the State of Sao Paula | (fig. 1) receive a very heavy load of organic pollutants as a result of the outflow of sewage, effluents of the city of Santos and nearby villages. Aside from the several kinds of detritus and, substances originating from mangroves and nearby rivers, a number of industrial wastes, sewage and oil residues are introduced in the Santos region. The final picture in the estuary and bay area is one of high water turbidity, high level of suspended material and considerable eutrophicatiom| of marine environment. The area under study can be divided into two distinct regions, namely one in the west sides and one in the east side of the bay of Santos. The more saline water which comes from the south |} penetrates under the more diluted water which leaves the estuarine area and the bay. This highi'} salinity wedge which penetrates the east side of the bay is more pronounced than the one on thei} west side where there is a much greater mixture of deep and surface water. Previous studies (Emilson, 1955) have shown that the west side of the bay is under heavier) influence of the tidal currents, while the east side is characterized by waters originating from the estuarine area. Such a pattern of water circulation produces two distinct areas, one on the west| side where there is a predominance of waters from the continental platform and the other on the east side where waters are mostly of low salinity, originating from the estuary of Cubatao River. Also such a pattern exerts influence on the types of sediments of the bay; on the west side’ predominate medium and fine sand, while on the east side, finer sediments (silt) are to be found. The granulometric analysis carried out by Tommasi (1967) indicated a predominance of | very fine sediments in several stations of the area under study. On the southern part of the bayas | well as along the shores predominated sandy bottoms while on the eastern side some of the | stations provided samples with a number of dead shells, coarse sediments and detritic material. The sample obtained in the last years showed only the presence of sediments ranging from | 0.063 to 0.250 mm in diameter, namely from clay and silt to fine sand; silt and clay were found mostly near the estuary, very fine sand on the east side and fine sand along the beaches and west side of the bay. Prevailing winds are from the coast during spring and summer, southerly winds predominating during fall and winter (Oliveira-Santos, 1965). Annual precipitation in the area reaches values of 2000 mm or more; highest levels are measured from January to March and the lowest ones in July and August (Camargo, 1960). Mean values for salinity increase from 29°/00 in December (rainy season) to July 34°/00 (dry season). The curves for salinity and temperature (fig. 2) show an inverse relationship. Values for dissolved oxygen and salinity are extremely variable in the channel, and | estuarine region; salinity values as low as 5.5°/00 have been found in surface waters on some occasions, while that at the bottom remains fairly constant (27 to 32°/00). Temperature | fluctuates from values as low as 16.6°C in the winter to 29°C in the summer, mean values ranging 17 to 27°C. METHODS The bottom samples were obtained with a Van-Veen bottom grab, which sampled an area of 0.10 m* and had a capacity of 10-12 litres. Periodic collections were made every season during the period December 1974 to November 1976 in 31 stations ecompassing all the estuary and Bay of Santos (fig. 3) to a maximum depth of 15 metres. 167 CHANGES IN ECHINODERM FAUNA ON COAST OF BRAZIL 62 82 BLO 92 Re 7 ASG 22 “]IZeIg JO 1sB0d dY1 UO soJURS Jo AID ‘| “BI @PUBIN PLUEW “P4S ep qd sijodouelso}4 Jefes| nBeuesed bi - bf pau =e Sep otues OelSegaS'S pue}s| ap oy apues9 alan Pers oP 42, ate t t | = Same?! M,07 A) 27 se? 77 .G7 97 L7 87 .67 168 TANIA MARIA DE CAMARGO Tic) Dez Jan. Fev. Mar. Abr. Mai. Jun, Jul. Time (months) o= O (zero) m x= 5m de aaa = S%oo ( Average Salinity) = T°C (Average Temperature) Fig. 2. Curves of salinity and temperature from the region studied. CHANGES IN ECHINODERM FAUNA ON COAST OF BRAZIL 169 Morro da o Barra 24° 22° 46° 20° 18° BS = Bay of Santos BSV= Bay of Sao Vicente Fig. 3. Sampling positions in region studied. 170 TANIA MARIA DE CAMARGO In each station two bottom samples were obtained; on board the ship all the macrofauna\ was immediately separated by sieving the sample through 2 mm and 1 mm mesh. The animals, were preserved for further examination under a stereoscopic microscope. The echinoderm) specimens were then separated and identified down to species level. Samples of water were also taken near the bottom for analysis of salinity, dissolved oxygen and nutrients (phosphate and nitrate); such analyses were conducted according to the methods recommended by Environmental Protection Agency (EPA) 1972 and Standard Methods for the Examination of Water (1971). Also, a part of the bottom sample was separated to be used in the granulometric analysis. RESULTS Before presenting the results obtained in the present survey, it seems advisable to list those found by Tommasi in 1964. At that time 14 species of echinoderms were found, the ophiuroids being the most abundant group. The species included Amphipholis januarn Ljungman, Hemipholis elongata (Say), Micropholis aira (Stimpson), Micropholis gracillima (Stimpson), Ophiactis lyman’ Ljungman, Ophioderma januari Lutken and Ophiothrix angulata (Say). The asteroids were represented by Astropecten brasiliensis Miiller and Troschel, Astropecten marginatus Miller and Troschel, Echinaster brasiliensis Miiller and Troschel, Coscinasterias tenuispina (Lamarck) and Luidia senegalensis (Lamarck). The occurrence of both crinoids and echinoids was restricted te Tropiometra carinata (Lamarck) and Mellita quinquiesperforata (Leske) respectively. In the more recent samples (1974 to 1976), the animals collected included the ophiuroidsA. januarii, H. elongata, M. atra, M. subtilis (Ljungman), O. lymani and Ophiophragmus liitkem (Ljungman), the asteroids A. marginatus and L. senegalensis; the only echinoid found was M. guinguiesperforata, which occurred in fairly dense populations. The quantitative aspects of the samples are presented in Table 1, which shows that the more abundant species were the ophiurans H. elongata and M. aira. The crinoid species 7’. carinata, the asteroids A. brasiliensis, E. brasiliensis, C. tenuispina, and the ophiuroids M. gracillima, O. januari and O. angulata were not found in any of the samples. In the eastern position of the bay, next to the Morro da Barra, mostly in station 19 and 31% high densities of the ophiuroids M. atra and H. elongata were found. Tommasi (1967) had already pointed out that this region presented a much higher number of echinoderms as compared to other areas of the Bay. It is to be noticed that in station 19 and 31 the bottom deposits are constituted mostly of silt and fine sand, In the region next to the beaches there was a predominance of sandy bottoms, the asteroid A. marginatus being the dominant species; on the west side of the bay the echinoid M. quinquiesperforata presented densities higher than any other species. Such a general pattern of distribution had already been observed in the 1964 samples. During all the sampling period no echinoderms were found in the estuarine region; the only records are those for station 18, near the entrance to the channel. The nutrient levels in station 1, fairly close to the iron smelter plant, were about 10 mes CHANGES IN ECHINODERM FAUNA ON COAST OF BRAZIL 171 higher than those generally found in the Bay. In station 1, the levels for POs and NOs were respectively 0.44 mg/l and 0.21 mg/l, dissolved oxygen being of order of 1.8 mg/l; transparency of water is reduced to about 0.85 m, the bottom being constituted mostly of silt and clay. In station 16, typical for most of the channel, values for nutrients were around 0.17 mg/l for POs, 0.16 mg/l for NOs and 4.9 mg/l dissolved oxygen; transparency of the water was usually close to 1.75 m. Station 19, which presented the greatest concentration of echinoderms, is located on the east side of the bay; levels for nutrients were of the order of 0.01 mg/l POs and NOs, 6.4 mg/l dissolved oxygen and transparency of the water around 1.10 m. The bottom is constituted mostly of very fine sand and silt, similarly to station 31. In the western part of the Bay, the bottoms are mostly of fine sand; stations 26 and 27 presented fairly large population of Mellita quinquiesperforata. Measurements of POs and NO3 showed values of 0.07 mg/l and 0.15 mg/l respectively, dissolved oxygen being around 6.7 mg/l. DISCUSSION The analysis of the results obtained for the two different sampling periods shows a fairly definite change in the echinoderm fauna in the area under consideration, suggesting different tolerance limits for different echinoderm species in relation to the environmental parameters under consideration. Due to the lack of information in relation to some of the environmental parameters (like nutrients) for the 1964 sampling, it is not possible to carry out a more detailed comparison with the more recent information. It would seem that the disappearance of some of the species is to be correlated with change in the nature of the bottom sediments, due to continued dredging operations near the docking facilities in the channel and the removal of the new material to the vicinity of Itaipu Point. The construction of new roads in the same general area has also helped produce the present picture of a substrate consisting mostly of very fine material. The importance of the type of sediments in the distribution of benthic species has already been discussed by Beanland (1940), Holme (1949) and Sanders (1958). As already mentioned, Tropiometra carinata, Coscinasterias tenuispina, Echinaster brasiliensis and Ophiothrix angulata, all characteristic of hard bottoms, were collected in 1964, but were not found again during the 1974-1976 period. They had been found associated with bottoms constituted of ‘‘detritic’’ material and dead shells, in areas which are now covered by sediments in the range 0.062 mm to 0.088 mm. Special reference should be made to Hemipholis elongata and Micropholis atra, the two ophiuroid species which have predominated in soft bottoms on different occasions and which can occur in fairly high concentrations at some stations. The changing environmental conditions seem to have had little or no effect on the number and distribution of these two species. REFERENCES Beanland, F. L., 1940. Sand and mud communities in the Dover estuary. 7. mar. biol. Ass. U.K. 24: 589-611. Bernasconi, I., 1956, Equinoideos y asteroideos de la coleccion del Instituto Oceanografico de la Universidad de San Pablo. Segunda contribuicion. Bolm. Inst. Oceanog. S, Paulo 7 (1/2): 119-148, 4 est. Camargo, A. P., 1960. Balango hidrico no Estado de Sao Paulo Servigo Nacional de Pesquisas Agronomicas, Bolm. 12, 172 TANIA MARIA DE CAMARGO Rio de Janeiro, 634 pp. Emilsson, I., 1955. Pesquisas hidrograficas ao longo da costa da Ponta de Itaipu. Jn Cunha, A. & Neto, J. M. de A. — Lancamento de esgotos sanitarios de Santos e Sao Vicente, cap. 16. Environment Protection Agency (EPA), 1972. Water Anality Criteria D.C. Holme, N. A., 1949. The fauna of sand and mud banks near the mouth of the Exe estuary. 7. mar. biol. Ass. U.K. 28: 189-237. Isaacs, J. D., 1973. The ecology of the southern California Bight: Implications for Water Quality Management. SCCWRP: 531 pp. Oliveira Santos, E., 1965. Characteristics climaticas de Baixada Santista. Jn Azevedo, A. e col. — A Baixada Santista. Aspectos geograficos, vol. 1, pp. 96-150. Sanders, H. L., 1958. Benthic studies in Buzzard Bay. I. Animal — sediment relationships. Limnol. Oceanog. 3 (3): 245-258. Standard Methods for the Examination of Water and Waste Water, 1971. American Public Health Association Washington, D.C., 13th ed. Tommasi, L. R., 1957. Os equinodermes do literal de Sao Paulo. I. Papeis Dep. Zool. S. Paulo 13 (2): 19-44, 30 figs. 1958. Idem, II. Contrgoes Inst. oceanogr. Univ. S. Paulo, ser. Ocean. Biol., no. 2: 27pp., 6 est. 1965. Lista dos crinoides recentes do Brasil. Contrcoes Inst. Oceanog. Univ. S. Paulo, ser. Ocean. Biol., no. 9:33 pp., 30 figs. 1966. Distribuicao geografica de alguns equinodermes do Brasil. Revta. bras. Biol. 26 (3): 239-246. 1967. Observacoes preliminares sobre a fauna bentica de se dimentos moles da Baia de Santos e regides vizinhas. Bolm. Inst. Oceanog. Univ. S. Paulo 16: 43-66, 6 figs. W733 TudyIN] snuise1ydorydg, TuewA] snoerydo syjqns styoydoss1w ene stoydoss1pw BiesUuola syoydrua yy Tenuel syoydrydury elelopiodsamnburnb BUTT W stsuayedauas EYpIngy snjeulsseul uajedonsy ‘(0% ) UdIINIIO Jo a3eIUII9d pur (u) paiaTjoo suaumtsads ye101 ‘uoneis Yyoea Woy susutToads e101 “satoads Jo Ist] “sajdures ay} Jo sioadse aaneimuend “| 2qe 1 11. A STABLE SYSTEM OF PREDATION ON A HOLOTHURIAN BY FOUR ASTEROIDS AND THEIR TOP PREDATOR CHARLES BIRKELAND*, PAUL K. DAYTON** AND NORMAN A. ENGSTROM*#** Friday Harbor Laboratories, Friday Harbor, Washington, U.S.A. SUMMARY Seven species of asteroids feed on Cucumaria lubrica, but together they harvest only 3% of the population or 10% of the standing crop biomass per year at the locality of the study. The rates of predation by the asteroid Solaster dawsoni on the predators of C. lubrica are high enough and the rates of growth and successful recruitment into the area by the predators of C. lubrica are low enough to indicate that the predators of C. lubrica are possibly kept low in abundance by the higher predator Solaster dawsoni. Solaster stimpsont, the most abundant predator of C. /ubrica in the area, has a behavioural escape mechanism which becomes increasingly effective as S. stimpsoni grows large and when it is on vertical rock surfaces. While S. dawsoni removes about 24-32% of the S. stimpsoni population each year, probably preventing a buildup in numbers, the refuge in size of a reproductive stock allows the persistence of the long-lived, slow-growing, S. stmpsoni. Dermasterias, a predator of C. lubrica with a refuge in size but with no behavioural escape mechanism to S. dawsoni, is 0.07 times as common as S. stimpsoni with a size-frequency distribution represented predominantly by large adults. Solaster endeca and Leptasterias, predators of C. lubrica with no known refuge to S. dawsoni, are 0.004 and 0.008 times as common as S,. stimpsoni and may be considered strays from other habitats. No significant changes in abundance were observed in the 3 trophic levels of the association from 1965 to 1976: C. lubrica, 4.4 x 103 m-?; S. stimpsoni, 0.5 m2; S. dawsoni, 0.007 m~*. The stability of the system results from different control mechanisms and refuges at each trophic level. INTRODUCTION In basic ecological theory, predator-prey systems have an inherent tendency to oscillate or to become extinct (Lotka 1920; Volterra 1926; Gause 1934; May 1973). In natural systems, populations usually fluctuate to a much lesser degree than would be expected (Murdoch and Oaten 1975). The factors preventing over-exploitation of a prey by its predators fit into two general categories (MacArthur 1972:31): (1) a refuge for the prey or (2) a factor limiting the predators to numbers low enough to prevent annihilation of prey (e.g., a higher level predator, a limiting resource other than the prey in short supply, cannibalism, territoriality, etc.). Prey refuges stabilize a community by providing protection for a reproductive stock but allowing relatively easy access of the predator to the “surplus” (Errington 1946) or excess “product” (Connell 1970; Smith 1972) of the prey population. For the simpler organisms in a heterogeneous environment, the susceptibility of prey to predation is usually inversely related to their abundance. The prey with the weakest escape or defence responses or those in the least safe location are caught first; so as prey become more scarce, only the less available are present. Also, scarcity and unpredictability can become refuges in themselves, even if site selection is disregarded (Smith 1968; Birkeland 1974). The effects of refuges are generally inversely related to prey population size and are thereby a stabilizing factor in predator-prey systems. As pointed out by Elton (1927), species size distribution has a major influence on community organization. The “‘size of the prey of carnivorous animals is limited in the upward * Present address: Marine Laboratory, UOG Station, Mangilao, Guam 96913. ** Present address: Scripps Institution of Oceanography, LaJolla, California 92093. *** Present address: Department of Biological Sciences, Northern Illinois University, DeKalb, Illinois 60115. Australian Museum Memoir No. 16, 1982, 175-189 176 C. BIRKELAND, P. K. DAYTON AND N. A. ENGSTROM direction by the carnivore’s strength and ability to catch the prey, and in the downward direction | by the feasibility of getting enough of the smaller food to satisfy the carnivore’s needs” (Elton \ 1927:60). A number of predators, even those quite generalized in their diets, will take only : earlier age classes from the prey populations. Prey individuals that survive long enough will gain | a refuge in size (Paine 1965, 1976; Dayton 1971). Many benthic marine invertebrates are : characterized by the type III survivorship curve implying “extremely heavy mortality beginning _ early in life, but the few individuals which survive to advanced ages have a relatively high | expectation of further life’ (Deevey 1947:286). A refuge in size is prevalent among marine invertebrates (Thorson 1955, 1958, 1961; Paine 1965, 1976; Dayton 1971; Connell 1972), However, in cases in which effective refuges do not exist, populations which sustain mortality in all age classes could be composed largely of the young stages (Grassle and Sanders 1973), Anage | structure composed preponderantly of long-lived adults also could imply a crowded habitat — governed by competitive interactions (Pianka 1970) or unpredictable reproductive success (Murphy 1968). Therefore, similar age structures could be the manifestations of very different processes. To be used as evidence for the effects of a process, age structures must be presented along with natural history information and ample data on the rates of the critical process. In this paper we examine a predator-prey system which displays remarkable constancy despite intensive predation at two trophic levels. Our goal is to elucidate the factors promoting stability. In the San Juan Islands of Washington State, the holothurian Cucumaria lubrica H. L. Clark attains great population densities (4 to 6 x 10° m™~) over extensive areas of rock or cobble substrata in shallow water (about 8 to 20 m depth). Within these areas, this holothurian species is preyed upon by 7 species of asteroids (3 of which are congeners) and forms the major portion of the diet of 4 of these predators (Mauzey, Birkeland and Dayton 1968). These predators are consistently common as a group and we have observed no major fluctuations in either prey or predator abundance over a period of eleven years. With this apparently opulent food supply, one would expect that total predator density might increase until the predators begin to eliminate their prey in local areas or at least seriously reduce its abundance. OBSERVATIONS Cucumaria lubrica is a small (up to 0.5g dry weight) dendrochirote holothurian which is numerically predominant over large areas of nearshore subtidal rock or cobble substrata in the San Juan Islands. The population density from November 1968 through May 1969 at Eagle Point (cf. Fig. 1, a map, in Mauzey, Birkeland and Dayton 1968) was estimated as 4420 + 400 (+)m~° from counts in eight 0.01 m* and four 0.06 m* quadrats. In the four quadrats of May 1969, an average recruitment of 7.7 x 10’ m® additional tiny C. Jubrica was observed. Cucumaria lubrica is thus numerically prevalent and, in fact, occupies up to 43% of the primary substrata in such areas as Black Rock (Fig. 1; Mauzey, Birkeland and Dayton 1968) and much of the area along the west shore of San Juan Island (e.g., Eagle Point and Edward’s Reef). In 1972 the mean abundance from all samples from Eagle Point was 4380 m7’, not significantly different from the 1968-1969 samples. Our observations over an eleven year period (December 1965 through August 1976) indicated that this is a stable condition; C. lubrica continued to occupy a major portion of the substrata at Eagle Point and Edward’s Reef throughout this period. Like most dendrochirote holothurians, C. lubrica is a passive suspension-feeder, and a position with adequate access to the water current is potentially a limiting factor. The size distribution of C. lubrica from dry weights taken on 239 specimens collected in November 1968 and February 1969 is shown in Figure 1. The size distributions from these two collections did not differ significantly, so the data from the two collections were combined. Five normal curyes were extracted from the size distribution in Figure 1 by using the method of Cassie (1954). Assuming the different normal curves represent year classes, the longevity of C. Ee was estimated as 5 years, Estimates of size classes derived from the method are given in Table 1. PREDATION ON A HOLOTHURIAN 177 IN SIZE CATEGORY NUMBER 5 10 15 20 25 30 35 40 DRY WEIGHT IN 0.01 GM_ UNITS Fig. 1. Size distribution of Cucumaria lubrica at Eagle Point (at 10 m depth) from collections taken November 1968 and February 1969. In areas in which it is predominant, this diminutive holothurian is preyed upon by seven species of asteroids (Mauzey, Birkeland and Dayton 1968): Luidia foliolata Grube, Solaster stimpsont Verrill, S. endeca (Linnaeus), S. dawsoni Verrill, Dermasterias imbricata (Grube), Leptasterias hexactis (Stimpson) and Pycnopodia helianthoides (Brandt). While four of these species are generalists (L. foliolata, D. imbricata, L. hexactis and P. helianthoides) and prey upon organisms other than holothurians in areas where holothurians are not predominant, the abundances of the Solaster species appear to be related to the abundance of dendrochirote holothurians (Mauzey, Birkeland and Dayton 1968). Luidia foliolata and Pycnopodia helianthoides only occasionally feed on C. lubrica. Forty-four percent of the observed prey items (N=61) of L. foliolata were holothurians and of these only about 47% (N=8) were C. lubrica (Mauzey, Birkeland and Dayton 1968). Luidia fohiolata is characteristic of gently sloping or flat sand bottoms where it feeds on ophiuroids, holothurians and bivalves and is rare in current swept areas on solid substrata characterized by C. lubrica. Although large numbers of C. lubrica (17, 17 and 35) were found in Pycnopodia stomachs on 3 occasions, less than 2% of the feeding observations for Pycnopodia in the San Juan Islands included C. lubrica (Mauzey, Birkeland and Dayton 1968). Since Luidia and Pycnopodia only rarely eat C. lubrica, they will not be discussed further. CEC = 178 6 °C 6b + O'Ih 9 OVE Of 8E a, Is oY OTE 91 0°77 CV SC € Sick 80°C. OG Cale 8Z 91 c € vl vLC OREO ES 6L°6 I v9 r8'1 09 810 831 0 4 UONDIT[OD, uonelAasd asury urow sseD TRIO T, JO % prepuels : (suun Sut QT) 14sIay\ Aq Jeo pounssy ‘($S6]) dISseD JO Spoylour ay] 01 BUIPIOe paaliap vIUGn] DUDIUNIND JO SUOTINGINSIP ssB] 921§ “T IGP L PREDATION ON A HOLOTHURIAN 179 Leptasterias hexactis and Dermasterias imbricata tend to specialize on C. lubrica in areas where C. /ubrica is predominant, although they have very different diets in other areas. Small holothurians make up 82% of the diet of subtidal L. hexactis in the San Juan Islands (Mauzey, Birkeland and Dayton 1968), and C. /ubrica comprises 85% of the holothurian prey or 69.7% of the total diet. In intertidal areas, however, L. hexactis feeds mainly on barnacles and molluscs (Menge 1972). In C. lubrica beds, 96.7% of the diet of D. imbricata (N=123) is made up of dendrochirote holothurians and of these, 92.4% are C. lubrica (89.4% of the total diet). In other areas, however, over 95% of the diet of D. imbricata may be made up of anemones, sponges, or pennatulaceans (Mauzey, Birkeland and Dayton 1968; Birkeland 1974) or 45% echinoids (Rosenthal and Chess 1972). Twenty-three D. imbricata contained from 1 to 30+ C. lubrica in their stomachs, with a mean of 7.7 C. /ubrica per stomach. Solaster stimpsoni and S, endeca feed mainly on dendrochirote holothurians (Mauzey, Birkeland and Dayton 1968). In the San Juan Islands, we recorded 424 feeding observations for S. stimpsoni (64% of the 656 S. stimpsoni examined were feeding). Holothurians made up 96% (or 408) of the feeding observations and of these, Cucumaria lubrica accounted for 93% (89.3% of the total diet), Eupentacta sp. and C. miniata (Brandt) 2% each and Psolus chitinoides H. L. Clark 1%. The S. stimpsoni preying upon C. lubrica were found with 1 to 8 specimens of C. lubrica per seastar or with a mean of 1.6 + 0.9s. Solaster stimpsoni spends 64% of its time feeding and 89% of its diet consists of C. lubrica. The average time required for digestion of a meal of C. Jubrica was found to be 1.5 days for four observations in aquaria. Therefore, each S. stimpsont eats about 222 C. lubrica per year. (This and similar calculations for the rate of predation on C, lubrica by the other species of asteroids are given in Table 2. The proportion of the C. lubrica population and biomass consumed by the combined activities of all asteroids is calculated in Table 3.) One hundred thirty-eight S. stimpsont were individually tagged with FD-67 Floy Tags. Within a year, most of the tagged seastars had disappeared, ‘although many of these apparent disappearances could be simply a loss of tags with no sign of damage to the seastar. The S. Stimpsoni recovered were individuals which had disappeared for several months, then reappeared. Other solasterids, S. dawsoni and Crossaster papposus (Linnaeus), are characteristically very motile (Mauzey, Birkeland and Dayton 1968; Birkeland 1974). Although Solaster stimpsoni and S. dawsont appeared to wander in and out of the area, they were relatively scarce in other areas and appeared not to be feeding as well in these other areas. We have examined many different habitats and localities between 1965 and 1976 (cf. also Mauzey, Birkeland and Dayton 1968). The percent of the Solaster that were feeding was lower in these other habitats (Birkeland 1974) and some of the prey items were of rather doubtful benefit to the Solaster (Mauzey, Birkeland and Dayton 1968; Birkeland 1974). For instance, Solaster stimpsoni feeds mainly on holothurians which were more scarce in these other habitats so the percent of the S. stimpsom found feeding was lower. Sometimes S, stampsoni was found with its tays wrapped around and its stomach extruded upon the surface of a sea pen, Palosarcus gurneyi (Gray), or over a tunicate, Pyura haustor (Stimpson). In these cases there was no sign of damage to the prey. Although individual S. stimpsoni switched from “wall” to “wall” (areas in our study site) or disappeared while new individuals arrived in the area, the abundance of S. sumpsoni did not change significantly during the 9 year period (Table 4). The numbers of S. stimpsoni were counted on specific rock walls and on certain sections of horizontal cobble substrata in our study areas on 10 dates from 14 December 1967 to 13 August 1976. If we take dates as treatments or conditions and the six larger sample areas as subjects or replicates, we can use the nonparametric Friedman two-way anova by ranks (Siegel 1956) to test if the number of S. sampsoni in the area varied significantly with time. Since the areas over which the counts were made must be complete matched sets for given dates, we tested the counts from the upper four rows (areas) in C. BIRKELAND, P. K. DAYTON AND N. A. ENGSTROM 180 -W 3°, url uaiea ou Aq paleusa jwyusea “p T S198. L Woy ssepo Tea Yea Jo 1Yy8IaM AIp UkdU dy) saw & UUINJOS SuTATdyNUI Aq PaAlep -W/N *5 “ENGL SS; JOBL] SIaJoId mMoswus * g asNedAq Sasse]d IedIK III] oy] Suoure Z a]qe J, Woy [e101 9y Burprarp ATuaAd Aq posnboe 7W/U91vI “ON -q ‘T 9198, L Woy sasse]o IkIA UT SJOgUINU ay] JO ONI IY} 0} SuTpIOSOe - UI/O~pPp JO doURPUNeR ay) sutuonsodde Aq poururaiap -W/"ON -e O'ee 6le v Obl OOF TWLOL Or b 6P 9°9€ 61 +Y €°6 79S Ske IC ie IGS 8°86 9°98 68S G Is e19 D895 979 (dures) | =i Stas i SE8T LAS 7-UWI Ualea 3 z-W 3 2W ualed ‘ou 7 ‘ou SSEID ody P 2 q B “s1olepaid proraise si Jo satoads xIs [Te Aq patunsuod sseurorg HIUgN] DUDUNIND Jo UONIOdosd ay Jo a1LUINSd pareynoyeo y *¢ IQR], “S AGRI, UE RIep Woy parelNs[eo., “Spuaispulia(] 1OJ A10e1OQR] aY2 UT sUOTeAJASGO aay puR imosduns “S$ TOF ATOLIOGE] OY UT SUONBAIASGO INO} Jo aSBIDAK IY] UO paseq Prosaise Jo saroads yoea 10J paumsse st Shep ¢* | JO DOUgN] ‘2 A0J aw UONsasip puke uondunsuod y,, Se9rl IVLOL L6 cE 0°97 I nS saploylunyay vipodourt J Of SZ O'T OL br suapxay spiuaisvidaT CCSE L901 OuL 68 +9 DIDIUQUL SDLUAISDULAACT el 81 07 ji €S lMOsmDp 4218D]09 8I 88 OT 8S 79 poapua 4a1spjos Ss60l (aE oul 68 +9 mosquijs 1a1Sv]05 ex TBOA xS'T (QD) paugn] *D (q) vouQgn) “OD (¥) Surpeaz punojy satsadg Jad -w 001 TOAV SOE) TVA sutute]uoo jo sunstsuos STENPIAIPUT JO % Iolepalg Jad uaiea Jod yenptarpur yoeurols Jad 1QIP Jo % pouqny “D Jed uajea pougny “dD JO pougn} “dD “ON aSRIDAY “Splosaise Jo satdads XIs woly UONepaid Aq voUgn] DLDWNIND JO JeAOUIAI JO AVY *7 IQR I. PREDATION ON A HOLOTHURIAN 181 Table 4 for all dates except December 1967 and January 1969 for which counts were lacking in one of the areas. Although there appears to be an increase in the numbers of Solaster stimpsont over the nine year period (Table 4), the Friedman two-way anova by ranks test indicates that there was a seventy to eighty percent probability that the differences could have been greater by chance. On five dates, counts were made in all six areas. We compared October 1967, November 1968, February 1969, January 1972 and August 1976 for all six areas and found that there was a ten to twenty percent ’ probability that the differences between dates would have been greater by chance alone. Therefore, the S. stimpsoni population seems to be characterized by a stability in numbers during a constant wandering of individuals in and out of a given local area, This is characteristic of several seastar populations (Menge 1974), including S. dawsoni (Birkeland Bae: mean abundances of S. stimpsoni and the seastars relevant to our discussion are given in Table Seven tagged S. stimpsoni were recovered after one year, three of these after 2.6 years, but no predictable growth patterns could be observed. After reaching a diameter of about 25 cm, S. stimpsoni may increase or decrease in size. Those seastars which decrease in total diameter (tip of ray 1 to tip of ray 6) look “unhealthy”; the rays are particularly thin and occasionally even have concave grooves. From these few data (Table 6) we estimated the growth of adult S. simpsoni (excluding from the analysis those decreasing in size) at 2 cm total diameter per year. Presumably, small individuals would grow at a greater rate. About 21% of the S. stimpsoni at Eagle Point during 1968 and 1969 (N= 74) were infected with a parasitic green alga Diogenes sp. Two of the six tagged S. simpson: that decreased in size were heavily infected; none of the six that grew were infected. Both of the S. stmpsoni had been infected by Diogenes sp. during the entire observation period, so S. stimpsoni is able to survive with Diogenes sp. for at least 32 months. One untagged S. stimpsoni was observed to be very green and very near death, motionless with concave grooves in its thin rays. Of a total of 42 field observations on Solaster endeca, 27 were feeding: 16 on Cucumaria lubrica, 3 on C. miniata, 2 on Psolus chitinoides, and 6 on bryozoa, tunicates and unidentified organisms. Solaster endeca was very rare in our study areas with 0.004 times the abundance of S. stimpsoni (Table 5). Of 138 field observations on Solaster dawsoni, 65 (47.1%) were feeding: 39 (60% of those feeding) were feeding on S. stimpsoni, 5 on C. lubrica, 4 on Crossaster, 3 on Dermasterias, 3 on Leptasterias, 3 on Mediaster, 3 on arms autotomized from Evasterias, 2 on arms autotomized from Pycnopodia, and 1 each on Henricia, Balanus crenatus Brugiere and Solaster dawsom. Solaster dawson is cannibalistic and will defend itself against predation by others of its own species with a response similar to that used by S. stimpsoni (cf. Fig. 2in Mauzey, Birkeland and Dayton 1968). The recorded diameter ratios of prey S. stimpsomi to predator S. dawsoni which had successfully captured them were 0,38, 0.72, 0.90, 0.94, 1.00, 1.00, 1.02, 1.03, 1.08, 1.09, 1.10 and 1.14; the prey-predator dianteter ratios recorded for attacks i in which. stimpsont ultimately escaped were 1.18, 2.18, 2.23 and 2.23. The probability of the 4 escapes of S. stimpsoni being only by those greater than 1,15 times the diameter of their predator while the 12 successful captures of S. stimpsoni were only by those less than 1.15 times the diameter of their predator was due only by chance would be P=0.0011 by a two-tailed Fisher exact probability test (Siegel 1956). From this we conclude that the defence response of S. sfimpsoni is effective only in combination with a refuge in size of approximately 1.15 times the diameter of the S. dawsoni attacking it. Since S. stimpsoni and Dermasterias have refuges in relative size, the number of asteroids available as prey toS. dawsoni increases as S. dazwsoni grows larger. TheS. dawsoni feeding onC. Oe v0 ane 0 v 6b my) L802 EOS 9°€ 9°0 Ik $e v0 OLY ial COI (jeuozI10y) Iooyy 219909 ee 70 DE 0 O°CS 0 $86 (Teon494) [TPM yoy pipodoutT spuarspidaT spuaispuuacy poapua*S wmosduys sg wosmop -s (UI ur) Aydessodo 7, pojdures ( [="4 re) 0) \ w \ 5% SOLASTER ‘ UV STIMPSONI \ wi \ N \ n ‘ 20 2” [24 lu $ 5 10 z i el 10 20 30 40 DIAMETER IN CM Fig. 2. Size distributions of Solaster dawsoni and Solaster stimpsoni at Eagle Point. Those Solaster stimpsoni larger than the size at the intersection with the broken line are safe from 50% of the S. dawsoni population by a refuge in size. Those to the right of the solid line have attained an essentially complete refuge in size from predation. In contrast to S. stimpsoni, S. endeca does not have a defence response to predation by S. dawson (cf. tig. 2 in Mauzey, Birkeland and Dayton 1968 and fig. 3 in this paper). Probably due to this in part, S. endeca is the rarest member of the association (Table 5); S. stimpsoni is about 250 times as abundant. Similarly, Leptasterias is too small to escape S. dawsoni once attacked, although its small size may allow it to forage into crevices or cavities under cobble that S. dawsoni would pass over. The Leptasterias in the subtidal C. lubrica meadows have probably strayed from the intertidal part of their range where S. dawsoni very rarely wanders. Dermasterias has no behavioural defence or escape mechanism, but it can eventually grow too large to be consumed by S. dawsomi. With the lack of a behavioural mechanism that increases the effectiveness of the refuge in predator-prey size ratio, Dermasterias is only about 7% as abundant as. stimpsoni. Once attaining the refuge in size, Dermasterias is susceptible to eventual mortality from infection by Diogenes. When attacked by S. dawsoni, Pycnopodia will autotomize one of its rays, which the S. dawsoni eats as Pycnopodia leaves the area. The most remarkable aspect of this food-web association is its relative constancy in abundance from year to year at all trophic levels. The C. lubrica population maintained an abundance of 4.4 x 103 m7? from 1968 to 1972. Our observations indicated no major differences 187 Fig. 3. Solaster dawsoni preying upon a Solaster endeca with a Solaster stimpsoni about 20 cm away. Note the lack of a behavioural escape or defence response by S. endeca. The dark tufts are tentacles of C. lubrica. Cucumaria lubria covers a major portion of the substratum. from 1965 to 1974. Its most common predator species, Solaster stimpsoni, did not significantly differ in abundance through a nine year period (Table 4). The top predator, S. dawsont, averaged 0.7/100 m* (2087 m? sampled) in 1968-1969 and 0.7/100 m? (648 m? sampled) in 1972. The interactions controlling the populations of these species differ with trophic level. A large portion of the individuals or available biomass of the C. lubrica population is left by predators (Table 3). The prevention of overexploitation of C. lubrica might be explained as follows: Solaster endeca, Leptasterias, Dermasterias and Pycnopodia are all rare in the areas of C. lubrica under study because of predation by S. dawsoni. Solaster endeca and Leptasterias do not maintain a reproductive stock in the area. The few individuals present had strayed in from outside areas or wandered down from the intertidal and were temporarily missed by S. dawsoni. Pycnopodia is rarely killed by S. dawsont, but leaves the area (often losing a ray) when attacked. Dermasterias oe maintain a reproductive stock in the area, but recruitment to this stock may be limited by S. awsoni. Recruitment to the S. stimpsoni population is seriously impaired by predation, but the 188 C. BIRKELAND, P. K. DAYTON AND N. A. ENGSTROM reproductive stock remains consistently common because of a behavioural defence mechanism that becomes increasingly effective with size and works especially effectively on vertical substrata that serve as spatial refuges. Predation by Solaster dawsoni keeps S. endeca, Leptasterias and Pycnopodia from establishing populations in meadows of C. lubrica. Solaster stimpsoni and Dermasterias can maintain a reproductive stock in the area with a refuge in size, but with severe predation on the recruitment to these populations, S. dawsoni can prevent S. stimpsoni and Dermasterias from increasing to population sizes capable of overexploiting their food resource. The slow buildup of populations of seastars having reached this size refuge is most likely prevented by infection from the parasitic green alga Diogenes sp. These processes together contribute to the maintenance of a remarkable constancy in numbers of this association at three trophic levels. ACKNOWLEDGEMENTS The authors were supported on the National Science Foundation Grant No. GB 6518X to the Department of Zoology, University of Washington. We wish to thank Robert L. Fernald for providing facilities at the Friday Harbor Laboratories and Robert T. Paine for his advice and support. Richard E. Norris tentatively identified the parasitic green alga for us as Diogenes sp. Bruce A. Menge made helpful comments on the manuscript. REFERENCES Atwood, D. G., and F, S$. Chia. 1974. Fine structure of an unusual spermatozoan of a brooding sea cucumber, Cucumana lubrica. Can. F. Zool, 52 (4): 510-523. Birkeland, C. 1974. Interactions between a sea pen and seven of its predators. Ecol. Monogr. 44: 211-232. Cassie, R. M. 1954. Some uses of probability paper in the analysis of size frequency distributions. Aust. 7. Mar. Freshw. Res. 5: 513-522. Coe, W. R. 1956. Fluctuations in populations of littoral marine invertebrates. 7. Mar. Res. 15; 212-232. Connell, J. H. 1970. A predator-prey system in the marine intertidal region. I. Balanus glandula and several predatory species of Thais, Ecol. Monogr. 40: 49-78. 1972. Community interactions on marine rocky intertidal shores. Ann. Rev. Ecol. Syst. 3: 169-192. Dayton, P. K, 1971. Competition, disturbance, and community organization: the provision and subsequent utilization of space in a rocky intertidal community. Ecol. Monogr. 41: 351-389, Deevey, E. S. 1947. Life tables for natural populations of animals. Quart. Rev. Biol. 22: 283-314. Elton, C. S. 1927. Animal Ecology. Methuen, London. 207 p. Errington, P. L. 1946. Predation and yertebrate populations. Quart, Rev. Biol. 21: 144-177, 221-245. Feder, H, M.,and A. M. Christensen. 1966. Aspects of asteroid biology, p. 87-127.In R. A. Boolootian (ed.) Physiology of Echinodermata. Interscience Publishers, New York. Feeny, P. P. 1970, Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology 51: 565-581. Flaherty, D. L. 1969. Ecosystem trophic complexity and Willamette mite, Eotetranychus willametter Ewing (Acarina: Tetranychridae), densities. Ecology 50: 911-915. Gause, G. F. 1934. The Struggle for Existence. Williams and Wilkins, Baltimore. 163 p. Grassle, J, F., and H. L. Sanders. 1973, Life histories and the role of disturbance. Deep-Sea Research 20: 643-659. PREDATION ON A HOLOTHURIAN 189 Janzen, D. H. 1970. Herbivores and the number of tree species in wopical forests. Am. Nat. 104: 501-528. 1971. Escape of juvenile Dioclea megacarpa (Leguminosae) vines from predators in a deciduous tropical forest. Am. Nat. 105: 97-112. Lotka, A. J. 1920. Analytical notes on certain rhythmic relations in organic systems. Proc. Nat. Acad. Science 7: 410-415. MacArthur, R. H. 1972. Geographical Ecology. Patterns in the Distribution of Species. Harper & Row, New York. 269 p. Mauzey, K. P., C. Birkeland, and P. K. Dayton. 1968. Feeding behavior of asteroids and escape responses of their prey in the Puget Sound region. Ecology 49: 603-619. May, R. M. 1973. Stability and Complexity in Model Ecosystems. Princeton University Press, Princeton, New Jersey. 256 p. Menge, B. A. 1972. Competition for food between two intertidal starfish species and its effect on body size and feeding. Ecology 53: 635-644. 1974. Effect of wave action and competition on brooding and reproductive effort in the seastar, Leptasterias hexactis. Ecology 55: 84-93. Murdoch, W. W., and A. Oaten. 1975. Predation and Population Stability. Adv. Ecol. Res. 9: 1-131. Murphy, G. I. 1968. Pattern in life history and the environment. Am. Nat. 102: 391-403. Paine, R. T. 1965. Natural history, limiting factors and energetics of the opisthobranch Navanax inermis. Ecology 46: 603-619. 1976. Size-limited predation: an observational and experimental approach with the Myrilus-Pisaster interaction. Ecology 57: 858-873. Pianka, E. R. 1970. On r- and K-selection. Am. Nat. 104: 592-597. Root, R. B. 1967. The niche exploitation pattern of the blue-gray gnatcatcher. Ecol. Monogr. 37: 317-350. Rosenthal, R. J., and J. R. Chess. 1972. A predator-prey relationship between the leather star, Dermasterias imbricata, and the purple urchin, Sztrongylocentrotus purpuratus. Fish. Bull. 70: 205-216. Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. 312 p. Smith, C. C. 1968. The adaptive nature of social organization in the genus of tree squirrels, Tamiasciurus. Ecol. Monogr. 38: 31-63. Smith, F. E. 1972. Spatial heterogeneity, stability, and diversity in ecosystems, p. 309-335. In E. S. Deevey (ed.) Growth by Intussusception: Hutchinson. Transactions published by the Connecticut Academy of Arts and Sciences, New Haven. Thorson, G. 1955. Modern aspects of marine level-bottom animal communities. Sears Found., 7. Mar. Res., p. 387-397. 1958. Parallel level-bottom communities, their temperature adaptation, and their “balance” between predators and food animals, p. 67-86. In A. Buzzati-Traverso (ed.) Perspectives in Marine Biology. University of California Press, Berkeley. 1961. Length of pelagic larval life in marine bottom invertebrates as related to larval transport by ocean currents, p. 455-474. In M. Sears (ed.) Oceanography. A.A.A. S. Publ. 67, Washington, D.C. Volterra, V. 1926. Variazioni e fluttuazioni del numero d’individui in specie animali conviventi. Memo. R. Accad. Naz. det Lincet. Ser. VI, vol. 2: 31-113. 12. THE SHRIMPS ASSOCIATED WITH INDO-WEST PACIFIC ECHINODERMS, WITH THE DESCRIPTION OF A NEW SPECIES IN THE GENUS PERICLIMENES COSTA, 1844 (CRUSTACEA: PONTONTINAE). A. J. BRUCE Heron Island Research Station, Gladstone, Queensland, Australia* SUMMARY At present, fifty one species of shrimp are known to live in association with Indo-West Pacific echinoderms. Of these, only one is a stenopodidean, all others belong to the Caridea, principally to the subfamily Pontoniinae (35 species), with the others in the families Alpheidae (11 species) and the Gnathophyllidae (4 species). The echinoderm hosts may belong to any class but are mainly the Crinoidea (26 species), Echinoidea (18 species) and Asteroidea (18 species), although only a very small number of shrimp species are associated with the latter class. Three ophiuroids, all basket stars, and eight species of holothurians are known to have shrimp associates. The available knowledge of the biology of these associations is outlined. _ Keys for the provisional identification of these shrimps are provided and one new species, Periclimenes ruber, is described and illustrated. The distribution of the shrimps is outlined and the known hosts listed. INTRODUCTION The shrimp fauna of the tropical and subtropical Indo- West Pacific region is dominated, in shallow water, by three groups, the Pontoniinae, the Alpheidae and the Hippolytidae. Numerous species of these groups are now known to live in “commensal” association with other marine animals. The details of these associations are very poorly known, and the use of the term “commensal” is, in general, rather misleading as it implies that something is known about the trophic relationships involved. This is rarely the case, and in the vast majority of examples virtually nothing is known about the feeding methods concerned. The use of the term “associates” is probably preferable in the present state of ignorance, especially as it seems probable that a variety of feeding strategies may be involved. Associations between shrimps and other marine animals are particularly common in the warm tropical waters around coral reefs and are relatively infrequent in colder waters. Only a single example is known from the British Isles, 7'ypton spongicola Costa, a pontoniine shrimp that lives in sponges. Little is known of these associations in deep water but they appear to be less frequent. Species of many shrimp families do occur in depths well over 100 fms and some of these probably are “commensals”’. SYSTEMATIC ACCOUNT Keys are provided below for the provisional identification of the known shrimp associates of Indo-West Pacific echinoderms. Where possible these identifications should be based on ovigerous females, and should be checked in detail with the original or later descriptions for confirmation. The keys will not separate related species found on non-echinoderm hosts. KEY TO THE GENERA OF SHRIMP ASSOCIATED WITH INDO-WEST PACIFIC ECHINODERMS 1. First two pairs of pereiopods with chelae ............. da rib sctesrtaplpsandiaieans Nes ttalet Me 2 Rea IESIMUETCS “alton fapETSIOPOUS ACHELHLe uc. dareueaetns fin oeeteqe esate reget sao, re Odontozona *Present address; The Darwin Museum, P.O. Box 4646, Darwin, N.T., Australia 5794. Australian Museum Memoir No. 16, 1982, 191-216 192 A. J. BRUCE Fig. 1. a. Allopontonia iaini Bruce, b. Araiopontonia odontorhyncha Fujino & Miyake, c. Palaemonella pottsi (Borradaile), d. Parapontonia nudirostris Bruce, e. Periclimenes hirsutus Bruce, f. Pontoniopsis comanthi Borradaile, g. Stegopontonia commensalis Nobili, h. Tuleariocaris zanzibarica Bruce. Scale = 1 mm. SHRIMP ASSOCIATES OF INDO-WEST PACIFIC ECHINODERMS 193 Fig. 2. a. Zenopontonia noverca (Kemp), b. Levicaris mammillata (Edmondson) (after Fujino & Takeda, 1977), c. Gnathophylloides robustus Bruce, d. Pycnocaris chagoae Bruce, e. Conchodytes meleagrinae, f. Athanas indicus (Coutiére), g. Synalpheus fossor (Paulson), h. Odontozona sp. Scale = 1 mm. 194 A. J. BRUCE First pair of pereiopods slender, second more robust ..........2..---sese,0s-ecseeereesese 3 . Second pair of pereiopods slender, first pair more robust .............:sceceeeeeeeeenees 15 Mandible with incisor process; third maxillipeds with ischio-merus slender, not OPPOSiigen TTS, AN I Lees arria peece ream iae Rete rn gu Ge he Recent aeT Ae are ie coer er a 4 Mandible without incisor process; ischio-merus of third maxilliped broadened, OPEECULALES Oe Soak tower eseeA ant tur Dabs Soe aeles]y: ad he See eaten ieee eae tee eee erates 13 Mandible swaithesimalle2=jommtedspalp, wens-esp ste: cate eceey eee ie ease ie ees okra Palaemonella @Man di ble withtomtepalpys wvcdas cua. dN omens tame seat cue shorts Basthige sie setae Smee oon ane 5 Rostrum generally dorsally dentate (except in Periclimenes insolitus Bruce) .......... 6 o> FROSURMENGTOGEMICSS: = nce mene se, re earn ee ee enn Oe ee eens ie ee 10 FTE PATIGsS PITT PRESEN ~