En e а о ee te 2 и ан gE Bae дети PR ARS pS Е NER Besen he dee jagen PERE ini iS no re DT ET BEE ET WETTE HALS ENGE TE NOT N TOR HARVARD UNIVERSITY LIBRARY OF THE Museum of Comparative Zoology ra h | Py re | PA ALALO LE. LL. | LIBRARY COME, 200100% NAT: Mr el DETTE м Mhls: abat 4 +n AN CHAR NE0I00S COMM SEIN EROIREN AD VOL. 6 1967-68 MALACOLOGIA International Journal of Malacology Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift DATES OF PUBLICATION At least 50 copies of MALACOLOGIA were mailed to subscribers (including the у Library of Congress, Washington, D. С.) оп the following dates: Vol. VI, No. 1-2 December 31, 1967 Vol. VI, No. 3 June 30, 1968 iv MALACOLOGIA, VOL. 6 CONTENTS D. 5. BROWN The anatomy and relationships of a South African Ferrissia (Basonimatophora:.. Ancylidae) eM HAB SGA A A oa 155 D. S. BROWN and J. B. BURCH Distribution of cytologically different populations of the genus Bulinus (Basommatophora: Planorbidae) in Ethiopia........... 189 D. S. BROWN, С. H. J. SCHUTTE, J. В. BURCH and В. NATARAJAN Chromosome numbers in relation to other morphological characters of some southern African Bulinus (Basommatophora:., Planorbidae) Mana ROME AOU OEE ROA Oe 175 R. BURN Revision of the genus Herviella (Opisthobranchia: вона ея SHORT SE ETES. INT UM I SD 223 J.-K. CHIU A Susceptibility of Oncomelania hupensis chiui to infection with Schistosoma japonicum NE. QUE, CARA ous 145 G. M. DAVIS The systematic relationship of Pomatiopsis lapidaria and Oncomelania hupensis formosana (Prosobranchia: OE A massa Bobet GS” AES Rh eae 1 J. D. DAVIS Ervilia concentrica and Mesodesma concentrica: СООО ae ee лы ао ее и es wee 231 р. J. HUNTER Studies on slugs of arable ground PaSamplang methods sur ne ee с doen a Be 369 P. J. HUNTER Studies on slugs of arable ground A В: ео: er oe ed" ox septs) AE ne PA aS ROR as 379 P. J. HUNTER Studies on slugs of arable ground MTS PG COME habits 2. satan go AS a, ain aos e SR age ee ehe 391 E. MARCUS and E. MARCUS Some opisthobranchs from Sapelo Island, Georgia, U.S. A........ 199 MALACOLOGIA, VOL. 6 CONTENTS (cont.) . B. MILLER Planorbula campestris (Gastropoda: Planorbidae) from the Cudahy fauna (Kansan) of Meade County, Kansas, with notes on the status of the subgeneric categories of Planorbula... ici. ee ORE RC OS 253 . A. STEKLOV Development stages of Neogene terrestrial mollusks of Ciscaucaside. осо. le de Todo Ed ies dl AN: 243 . В. TRUEMAN The locomotion of the freshwater clam Margaritifera margaritifera (Unionacea: Margaritanidae) ....... 401 . VAN DER SCHALIE and G. M. DAVIS Culturing Oncomelania snails (Prosobranchia: Hydrobiidae) for studies of oriental schistosomiasis ......... 0... а 321 . WOOD Physiological and ecological aspects of prey selection by the marine gastropod Urosalpinx cinerea (Prosobranchia: Murieidae)..... 0. 20.0 о. es aa 267 vi TOM 6 МАЛАКОЛЕНИЕ 1968 ОГЛАВЛЕНИЕ И: С: БРОУН Анатомия и родственные взаимоотношения южно- африканских Ferrissia (Basommatophora: Ancylidae) . . . . . 155 ИО: BPOYH И ЛЖ. Б. БЕРУ Распростарнение цитологически-различных популяций рода Bulinus (Basommatophora: Planorbidae) ВОИ o AE pit ол EE MS heat ar LEI ieee BEOYH, К. Х. MOTTE, ДЖ. В. БЕРЧ И P. НАТАРДЖАН Отношение числа хромосом к другим морфологическим признакам некоторых южно-африканских Bulinus (Basommatophora: Planorbidae)ht 3%. mi. m. ni er ть Р. БЕРН Ревизия рода Herviella (Opisthobranchia: Eolidacea) . . . . . 223 ДЖУ-КУАНТ-ШИУ Восприимчивость Oncomelania hupensis chiui ‘заражению Schistosoma japonicum U. EEE a о 145 Г. M. ДЕВИС Систематические взаимоотношения между Pomatiopsis lapidaria и Oncomelania hupensis formosana (Prosobranchia: нубов ее TE ai e a te 1 ДЖ. I. ЛЭВИС Ervilia concentrica и Mesodesma concentrica: (LO MOB OMY иж синоним a. a! ks Vals ria Bee Se, eo ee eo ana eee П. ДЖ. ХАНТЕР Изучение слизней на пахотных землях 5 MERO обо» MSO o о ob 6 66 mo 6 об оо 6 до 959 П. ДЖ. ХАНТЕР Изучение слизней пахотных землях ШО Жизненный: MR ое мое ue eee а eS П. ДЖ. ХАНТЕР Изучение слизней на пахотных землях 1005. REES. CN MOTO dG) ом ae an оса, ка со SEL 9. МАРКУС И 3. МАРКУС О некоторых opisthobranchs из района о. Seno. (Maco DAME, AC Mls: MAGI). ce a о cee ie OG vit IE Б. МИЛЛЕР А. СТЕКЛОВ EY VIBE МАЛАКОЛЕНИЕ Planorbula campestris (Gastropoda: Planorbidae) фауны кьюдехи (канзан) из района MAL каунти, канзас и заметки о положении подродовых Fa Pe ropa. нок Aulas a ere LME в Этапы развития неогеновых наземных моллюсков Предкавказья Движение пресноводных моллюсков Margaritifera margaritifera (Unionacea: Margaritanidae) ВАН ДЕР ШАЛИ И ДЖ. М. ДЕВИС ВУУД Культивирование улиток Oncomelania (Prosobranchia: Hydrobiidae) для изучения восточного UCERO ONECARE TR а о Физиологический и экологический аспекты выбора жертвы морскими гастроподами Urosalpinx cinerea (Prosobranchia: Muricidae) viii . 253 . 243 401 «321 A MALACOLOGIA, VOL. 6 NEW NAMES GASTROPODA burchi (Doridella), Marcus € Marcus, 1967, 205 burchi (Herviella), Burn, 1967, 226 Marciella, Burn, 1967, 227 misa (Tritonia), Marcus & Marcus, 1967, 211 sapeloña (Okenia), Marcus & Marcus, 1967, 203 wattla (Armina), Marcus € Marcus, 1967, 213 Me 0 e у MES ‚ SO LA 5 у ey th Un 7 7 Г pore ie =" M mre Ru AU | | X a e MINA] M Met)! LME: FUN ee RME SAM À (+, | AA ы f 1 | Prager ana CETTE pu aa ag un Rob ad poe! Wish CHAN ; Hin u ee ВИ RP | AIME see CAN | Ser! ra + WA wry hs DS DES A Rowe | fr? dds ИОВ 4 | ie a И NE 4 $ ЧЕ: UAE PY TA с УРНЫ A de Nam) AL vi ae Ñ i; “uy a у y Mes y Yu A i, » «A Им B Г ’ er q п mid В d т ue EN ь acl Fan a + ARTE ‚МУ „вм ale oi PEA 4 ake, MA WAR, 2 2 AQOAORTIAD e PT. E A caja tata a E ка Dey ite di or R IRA re Tr sos : 09: avo TAM > mont | Han АЗИЯ ALT LE trs: PQ лам ao LOS. TOC! ee de CIS | PARE ame вам, Nr ay, | nen Ir; +, M ah M (o | MN 6 No q- -2 MUS. COMP. ZOOL. DECEMBER 1967 LIBRARY FFR 6 1968 HARVARD UNIVERSITY MALACOLOGIA International Journal of Malacology Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift MALACOLOGIA ANNE GISMANN, General Editor 19, Road 12 J. М. HUBER, Managing Editor Museum of Zoology J. B. BURCH, Associate Editor Museum of Zoology The University of Michigan Ann Arbor, Mich. 48104, U.S.A. Maadi, Egypt The University of Michigan UA В: Ann Arbor, Mich. 48104, U.S.A. EDITORIAL BOARD SCHRIFTLEITUNGSRAT РЕДАКЦИОННАЯ КОЛЛЕГИЯ Р. O. AGOCSY К. НАТА! Magyar Nemzeti Múzeum Inst. Geology & Paleontology Baross U. 13 Tohoku University Budapest, VIII., Hungary Sendai, Japan H. B. BAKER N. A. HOLME 11 Chelten Road Marine Biological Assoc. U.K. Havertown The Laboratory, Citadel Hill Pennsylvania 19038, U.S.A. C. R. BOETTGER Technische Hochschule Pockelsstrasse 10a Braunschweig, Germany A. H. CLARKE, JR. National Museum of Canada Ottawa, Ontario Canada C. J. DUNCAN Department of Zoology University of Durham South Rd., Durham, England Z. A. FILATOVA Institute of Oceanology U.S.S.R. Academy of Sciences Moscow, U.S.S.R. E. FISCHER-PIETTE Mus. Nat. d’Hist. Natur. 55, rue de Buffon Paris VE, France A. FRANC Faculté des Sciences 55, rue de Buffon Paris V®, France P. GALTSOFF P. O. Box 167 Woods Hole, Mass. 02543 US TA T. HABE National Science Museum Ueno Park, Daito-ku Tokyo, Japan A. D. HARRISON Department of Zoology University of Natal Pietermaritzburg, S. Africa Plymouth, Devon, England G. P. KANAKOFF Los Angeles County,Museum 900 Exposition Boulevard Los Angeles, Calif. 90007, U.S.A. A. M. KEEN Department of Geology Stanford University Stanford, Calif. 94305, U.S.A. M. A. KLAPPENBACH Museo Nacional Historia Natural Casilla de Correo 399 Montevido, Uruguay Y. KONDO Bernice P. Bishop Museum Honolulu, Hawaii 96819 U.S.A. H. LEMCHE Universitetets Zool. Museum Universitetsparken 15 Copenhagen @, Denmark AKLILU LEMMA Faculty of Science Haile Sellassie I University Addis Ababa, Ethiopia N. MACAROVICI Laboratoire de Géologie Université “Al. I. Cuza” Iasi, Romania D. F. McMICHAEL Australian Conservation Found. Macquarie University, Eastwood №. 5. W. 2122, Australia J. Е. MORTON Department of Zoology The University of Auckland Auckland, New Zealand W. K. OCKELMANN Marine Biological Laboratory Grgnnehave, Helsinggr Denmark CONSEJO EDITORIAL CONSEIL DE REDACTION W. L. PARAENSE Centro Nacional de Pesquisas Malacolôgicas, C. P. 2113 Belo Horizonte, Brazil J. J. PARODIZ Carnegie Museum Pittsburg, Penn. 15213 U.S.A. A. W. B. POWELL Auckland Institute and Museum Auckland, New Zealand R. D. PURCHON Chelsea College of Science and Technology London, S. W. 3, England S. G. SEGERSTRALE Zool. Mus. Helsinki University P. -Rautatiekatu 13 Helsinki, Finland F. STARMUHLNER Zool. Inst. der Universitat Wien Wien 1, Luegerring 1 Austria J. STUARDO Instituto Central de Biologia Universidad de Concepcion Cas. 301, Concepcion, Chile W.S.S. VAN BENTHEM JUTTING Noordweg 10 Domburg The Netherlands J. A. VAN EEDEN Inst. for Zoological Research Potchefstroom Univ. for C. H. E. Potchefstroom, South Africa C. M. YONGE Department of Zoology The University Glasgow, Scotland A. ZILCH Senckenberg-Anlage 25 6 Frankfurt am Main 1 Germany MALACOLOGIA, 1967, 6(1-2): 1-143 THE SYSTEMATIC RELATIONSHIP OF POMATIOPSIS LAPIDARIA AND ONCOMELANIA HUPENSIS FORMOSANA (PROSOBRANCHIA: HYDROBIIDAE)!,2 George Morgan Davis 3 ABSTRACT The North American Pomatiopsis lapidaria (Say) and the Oriental Oncomelania hupensis formosana (Pilsbry & Hirase) were chosen as representatives for 2 related hydrobiid genera. Their comparative anatomy, potential for hybridi- zation, electrophoretic properties and laboratory ecology were studied to de- termine to what extent differences of value to systematics could be found. On the basis of their anatomy Pomatiopsis and Oncomelania are judged to be distinct genera within the same subfamily, the Pomatiopsinae. In the genus Oncomelania (considered tohave 1 species with 4 subspecies) the shell is smooth (except in the ribbed form of O. hupensis hupensis), with moderately deep sutures and moderately convex whorls. The outer lip of the shell has a tendency to form a varix which is usually quite pronounced. The umbilicus is narrow, as is the apical whorl. The parietal callus is elongate. There are at least 35 gill filaments, usually 45 or more. The verge is muscu- lar, the tip has short strips of actively beating cilia and a distinct protrudable papilla. The pleuro-supraesophageal connective is comparatively short; in consequence the osphradio-mantle nerve arising from the tip of the supra- esophageal ganglion is relatively long; it usually does not bifurcate until within the cephalic wall. The supravisceral connective also arises from the tip of the ganglion. The sperm duct and spermathecal duct arise in a common sheath from the right, anterolateral surface of the bursa copulatrix. The female gonad is multibranched and the collecting duct relatively slender. The oviduct en- circles the seminal receptacle in a characteristic manner. The seminal vesicle is a characteristically knotted slender tube. The verge has a single glandular type (studied in O. h. formosana and O. h. quadrasi). The cerebral commis- sure is short. Thetentacles are elongate compared tothe length of the rostrum. Compared with Oncomelania, the shell of Pomatiopsis has a roughened microsculpture, the lip is sharp and there is no tendency to form a varix. In all 4 species the apical whorls are wide. The umbilicus is wide and pro- nounced, sutures are deeply impressed, and the whorls very convex (except in P. binneyi). In P. lapidaria and P. cincinnatiensis, there are less than 30 gill filaments. The verge does not have a pronounced musculature or papilla in the 4 species; penial cilia are lacking in 2 species (P. lapidaria and P. cincin- natiensis); when cilia occur they are bushy and generally inactive (P. califor- nica and P. binneyi). Two species (P. cincinnatiensis and P. californica) have penial filaments, a condition not found in Oncomelania. The verge has 3 lAdapted from a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the University of Michigan, May, 1965. 2This investigation was sponsored (in part) by the Commission on Parasitic Diseases of the Armed Forces Epidemiological Board and was supported (in part) by the U. S. Army Medical Research and Development Command, and (in part) by a research grant (5 T1 AI 41) from the National Institute of Allergy and Infectious Diseases, U. S. Public Health Service. 3Current address: 406 Medical Laboratory, U. $. Army Medical Command, Japan, APO San Francisco, California 96343. (1) 2 G. M. DAVIS glandular types (known for P. lapidaria). The pleuro-supraesophageal con- nective is elongate, the supraesophageal ganglion lies close to the lateral cephalic wall and the osphradial and mantle nerves, which usually bifurcate right after leaving the tip of the ganglion, are correspondingly quite short. The supravisceral connective, in P. lapidaria, arises from the lateral, posterior border of the supraesophageal ganglion, not from the tip. The oviduct does not encircle the seminal receptacle. The spermathecal duct arises from the an- terior end of the bursa copulatrix (P. lapidaria and P. cincinnatiensis), and the sperm duct from the spermathecal duct. The female gonad is little branched, the collecting duct is quite wide. The seminal vesicle is a thick, regularly coiled tube. The tentacles are short, relative to the length of the rostrum. Hybridization does not occur between Pomatiopsis lapidaria and Oncomelania. Dise electrophoretic studies on fresh foot muscle protein of the 2 repre- sentative taxa showed that each taxon has a specific pattern. All subspecies of Oncomelania have 1 or more characteristic dense protein components with Rf values (ratio of the distance from the origin to the center of each band and from the origin to the front) greater than 0.75. Pomatiopsis lapidaria lacks dense, fast moving proteins beyond an Rf of 0.75. All 4 subspecies of Oncomelania are characterized by adaptability to the laboratory culture conditions provided. In 12 months, under conditions less than optimal, the finite rate of mortality (field snails about 1 year old) was 12% per month. Young grew at about 0.65 mm per week with low mortality. Young were produced at rates as high as 2.12 per female per month continuously for more than 2 years. The 4 species of Pomatiopsis investigated did not adapt well to laboratory conditions. Pomatiopsis californica and P. binneyi diedrapidly without pro- ducing young. P. lapidaria and P. cincinnatiensis (field snails about 1 year old) had a finite rate of mortality of 16%per month in “optimal” conditions over a 10-month period for the former and a 3-month period for the latter, after which rates of mortality increased rapidly, in part because of the shorter life span of these snails. Young grew at less than 0.14 mm per week with mor- talities exceeding 30% in 2 months. Young were produced at less than 0.51 per female per month for very short periods of time. CONTENTS 6. Buccal Mass ee 56 Page 7. Musculature. .. = gear 66 8. Nervous System . . . ... 71 INTRODUCTION: 2... at Shen AR 3 D. Oncomelania hupensis SYSTEMATIC DISCUSSION ...... 5 formosana.. В MATERIALS AND METHODS ..... 15 1. Shell ссор ав 80 COMPARATIVE ANATOMY ...... 15 2. External Morphology ОиСТ es 15 and ‘Topography ase 82 B. Materials and Methods..... 15 3: Mantle: Cavity BE C. Pomatiopsis lapidaria ..... 19 4. Female Reproductive ДВЕ OUR eae a O 19 System ... . RER SE 2. External Morphology 5. Male Reproductiv and Topography ...... 21 System... AA suMantle" Cavity 10e. 26 6. Buccal Mass > о 4. Female Reproductive 7::Muscuülature . . eens Systemische chen ar: 28 8. Nervous System ....... 98 5. Male Reproductive E. Summary and Discussion... .109 SV SCCM Ms alte 69. 39 HYBRIDIZATION STUDIES ...... 110 POMATIOPSIS AND ONCOMELANIA 3 ELECTROPHORETIC STUDIES . . ..112 AC introduction... 4.0.5... 04 lle В. Materials and Methods ..... 112 CSUN 0.) 58 ee 115 POT SCS SION iii. Sauce a keiten le LT TABORATORY ECOLOGYS. 1.117 ANTE Че Го) a ara «ste Lt В. Materials and Methods ..... 118 BAIERDETIMENES N... Het 119 Ae CSE MEV OLS DID de pox eyes Léna fe 119 Ze ETOdUCEVIEV EU arr 126 3. Growth and Survivorship ОР ТОНИ К: mern LE 131 сию аи 131 CONCLUDING: DISCUSSION) алан 132 ACKNOWLEDGEMENTS.... 00.0... .139 ОВ Oa Gl Wt Of Bee See nues 135 INTRODUCTION This paper is concerned with the systematic relationship of the American genus Pomatiopsis (Say, 1817) and the Oriental genus Oncomelania Gredler, 1881. Interest in determining to what extent these prosobranch snails were related to each other was first generated when Stunkard (1946), Ward, Travis & Rue (1947) and Berry & Rue (1948) demonstrated that Pomatiopsis lapidavia was capable of serving as an intermediate host of the human blood fluke, Schistosoma japonicum. Ac- cording to Abbott (1948a), malacological studies showed P. lapidaria to be “strikingly similar” to species of On- comelania. Dundee (1957) stated that anatomically P. lapidaria was “quite similar” to Oncomelania and that differ - ences appeared “to be minor.” Van der Schalie, Getz & Dazo (1962) reported success in hybridizing 2 species of On- comelania with P. lapidaria which “strengthened” their “contention that these genera are rather closely related.” Burch (1960a) considered the genera to be synonymous but later, (Burch, 1964), after а detailed cytological study, separated them as _ closely related genera. In this paper Pomatiopsis lapidaria and Oncomelania formosana were chosen as representatives ofthe 2 genera. Their comparative anatomy, potential for hybridization, electrophoretic proper- ties and laboratory ecology were studied. The investigation was under- taken to determine to what extent differ - ences coupled with existing knowledge of the other species of each genus would serve to establish the degree of relation- ship between Pomatiopsis and Oncomel- ania, i.e., whether or not Pomatiopsis and Oncomelania are congeneric, closely related but distinct genera, or genera more widely separated than has previ- ously been considered. Dundee (1957) discussed the major papers pertaining to Pomatiopsis, i.e., ecology, systematics, general distri- bution and anatomy. Aside from a few anatomical details, including a dis- cussion of variability in the radula presented by Abbott (1948a), Dundee pub- lished the only anatomical study on P. lapidaria giving details which can be used in a general comparative manner on the generic level. Van der Schalie & Dundee (1956) presented the basic anatomy of P. cincinnatiensis making possible certain specific comparisons with P. lapidaria, and with the so- called species of Oncomelania. Oncomelania formosana was chosen as a representative of its genus because it appeared to be a form intermediate in the Oncomelania species complex. Burch (1964) studied cytological aspects of the 4 currently recognized species of Oncomelania [Oncomelania hupensis Gredler, 1881; O. quadrasi (Môllen- dorff) 1895; O. nosophora (Robson) 1915; О. formosana (Pilsbry € Hirase) 1905] and their “various hybrids” and, in the light of his findings, coupled with the fact that there was no reduced viability in the F1 and Fo hybrids, he stated that the 4 species were no more than geographic populations or races of the same species. Davis et al. (1965) were of the same opinion, due to successful hybridization: only 1 ab- normal snail was found amongthousands of hybrids of all 4 “species” of Oncome- 4 G. M. DAVIS TABLE 1. The family and subfamily status of Pomatiopsis within the Rissoacea as indicated by various authors Amnicolidae Hydrobiidae Authors Pomatio-| Hydro- |* |psinae | biinae Tryon 1862 Gill 1863 х Stimpson 1865 Binney 1865 Gill 1871 Fischer 1885 x Tryon 1883 Call 1900 x Baker 1902 x Pilsbry & Ferris 1906 Hannibal 1912 x Walker 1918 х Annandale 1924 Baker 1926 Baker 1928 Thiele 1928 x Thiele 1931 x Wenz 1938 Berry 1943 Abbott 1948 Dundee 1957 Davis * =no subfamily mentioned. lania observed. They felt that the genetic compatibility involved was of a con- specific or sub-specific nature. In shell shape, there is a north-south cline, with O. nosophora from Japan having a long, slender shell; O. quadrasi from the Philippines having a relatively more short and broad shell; and O. formo- sana from Taiwan being intermediate. This intergradation was noted by Abbott (19485), who stated that 5% of the Formo- san specimens could not be distinguished from O. quadrasi (Philippines) while 10% had a shape and size similar to many O. nosophora from Japan and China. Studies by Kuo & Mao (1957) indicated that O. hupensis and O. nosophora from China are all О. hupensis and that there was “no clear-cut line of demarcation between them,” i.e., between the smooth shelled O. nosophora type and ribbed O. hupensis type. In all following dis- cussions in this thesis, the so-called Oncomelania “species” of most pre- Pomatio- | Truncat- Pomatio- Pomatio- sinae ellinae psinae psinae Pomatiopsidae Rissoidae |Truncatellidae Pomatio- psinae vious authors will be considered sub- species of O. hupensis, the first named of the “species.” There are a number of anatomical papers describing various features of the subspecies of Oncomelania. Heude (1880), Li (1934) and Kuo & Mao (1957) discussed various aspects of the anatomy of O. hupensis hupensis; Robson (1921), Nakamoto (1923), Itagaki (1955), Wil- liams (Fide Ritchie, 1955, in referring to the 1954 and 1955 Professional Reports of the 406 General Medical Laboratory) and Roth & Wagner (1957) presented material on O. hupensis noso- phora. As far as is known, no anatomi- cal studies have been published on O. hupensis formosana aside from those of Roth (1960), who described the female reproductive anatomy and of Davis (1964) who depicted the structure of the female gonad. Abbott (1945) provided some anatomical notes оп О. hupensis quadrasi. POMATIOPSIS AND ONCOMELANIA 5 SYSTEMATIC DISCUSSION Pomatiopsis and Oncomelania are representatives of the mesogastropod, rissoacean family Hydrobiidae (Tros- chel, 1857) subfamily Pomatiopsinae (Stimpson, 1865). As shown in Table 1, there has been a considerable difference of opinion concerning the proper family and subfamily designation for Pomatiop- sis since 1862. Oncomelania was not named until 1881; this genus generally has been placed in the same subfamily with Pomatiopsis. Pomatiopsis lapidaria was described by Say (1817) as a species of Cyclostoma, a genus of the Pomatiasidae currently split into several genera (Wenz, 1938- 1944). Tryon (1862) recognized the basic differences between the viviparid and hydrobiid types of snails and subse- quently created the family Amnicolidae. In the same paper he separated Ротай- opsis as a subgenus of Amnicola be- cause he felt that “A. lapidaria” differed from the small globose shells of Amni- cola by “shell elongate, spire (of about 6 whorls) much exceeding the length of the aperture. . .”. Tryon, however, did not give a family diagnosis for the Amni- colidae. Gill (1863), in a noteworthy paper, defined the family, but stated that Pomatiopsis (i.e., P. lapidaria) was possibly an aciculid snail4. Heasserted that the “validity” of Pomatiopsis as defined by Tryon was doubtful and that the Amnicolidae contained 3 genera: Amnicola, Chilocyclus and Somatogyrus. Chilocyclus, proposed by Gill (1863), is a synonym for Pomatiopsis, asthe genus is currently understood, and was used to separate Cyclostoma cincinnatiensis (now P. cincinnatiensis) from species of Amnicola. Stimpson (1865) rejected Gill’s family 4Wenz (1938-1944) used Acmeidae for Acicul- idae, but according to Opinion 344 of the International Commission on Zoological Nomenclature (1955), the correct designation is Aciculidae. definition for the following reasons: the definition was almost an exact translation of Moquin-Tandon’s (1855) definition of “Bythinia”, and the definition did not apply to the American forms of the group, founded, at that time, onthe genus Amnicola, e.g., the verge is not bifid in all species, the tenacles of Amnicola proper are not setacious, etc. Aside from the fact that the family was poorly diagnosed, laws of priority exclude the name Amnicolidae from being used in place of the Hydrobiidae as the latter are currently understood. H. B. Baker (1960) reviewed this situation and pointed out that if one accepts Bithyniidae and Truncatellidae as separate rissoacean families, Hydrobiidae is the legal name for the rissoacean family under dis- cussion. Bithyniidae are considered dis- tinct for several reasons. They possess a calcareous operculum which is not found in the Hydrobiidae. Members of the Bithyniidae have a verge with a characteristic prong on the concave side. Within the prong and running back into the body is a “flagellum” described (and figured) by Baker (1928) as “very long, a blind diverticulum,... semi-independent, having no internal connection with the vas deferens.” While some hydrobiids have penial appendages, they are structurally different from those found in the Bithyniidae and the “flagellum” is not present. Animals of the Bithyniidae are characterized by the yellow or orange pigment spots (Baker, 1928; Abbott, 1948b; Taylor, personal communication, 1965) whichare not found in the Hydrobiidae. A number of features described for Truncatella indicate separate family status. The great reduction in the ctenidium con- trasts with the well-developed ctenidium of the Hydrobiidae; concentrated cere- bral, pleural and parietal ganglia in Truncatella are in contrast to the widely separated cerebral and parietal ganglia of the Hydrobiidae; shortened tentacles with eyes at their bases on the mid- line, or displaced medially, in the former, are in contrast with eyes at the 6 G. M. DAVIS outer base of generally elongate tentacles in the latter. The elongated gonoperi- cardial duct and the connection of the bursa copulatrix with the left kidney are unknown in the Hydrobiidae. The modi- fied, small, pedestal-like foot of Trun- catella differs from the elongate, broad foot of the Hydrobiidae. In the Trun- catellidae the corneous operculum is sub-spiral and has characteristically, though not infallibly, a thick, outer, calcarious layer. The central tooth of the radula is triangular and supports a single, anterior, triangular cusp (drawings in Fischer, 1880-1887; Binney, 1865; Clench & Turner, 1948). The hydrobiid operculum is also corneous, usually paucispiral, sometimes multi- spiral, but not subspiral to the degree shown in Truncatella. The central tooth of the hydrobiid radula is trapezoidal or rectangular, the, anterior edge supporting more than 1 cusp. As dis- cussed below, the mode of progression found in the Truncatellidae is distinctly different from that inthe Pomatiopsinae. Stimpson (1865), in an outstanding paper, presented a broad diagnosis for the family Rissoidae so that it included the currently recognized families Ris- soidae, Hydrobiidae and Bithyniidae. The Truncatellidae were excluded onthe basis of radula, eye position and the nature of the “breathing organ.” The Rissoidae are currently characterized as follows: marine, a filament arising from the operculigerous lobe and/or the presence of a pallial tentacle; foot more narrow and agile than that of the Hydrobiidae; the shell may be smooth but is more characteristically ribbed, with spiral cords, or cancellate. The aperture below may be bent outwards (Fretter & Graham, 1962). The Hydrobiidae are a separate family of freshwater snails with a few marine, brackish water, amphibious and ter- restrial forms. The shell is charac- teristically smooth but not infallibly so; the aperture is not bent out below. No filament arises from the operculigerous lobe and a pallial tentacle is known only in Hydrobia ulvae. The difference in the foot has already been mentioned. Fretter & Graham (1962) state that the pleuroparietal connectives of the nervous system are long in the Hydrobiidae and comparatively shortened in the Risso- idae. In summary, the family Amnicolidae is a synonym of the Hydrobiidae as the latter are currently understood. For the anatomical reasons given above, Pomatiopsis cannot be included in the distinct rissoacean families Rissoidae, Truncatellidae, or Bithyniidae. The question arises: Does Pomati- opsis deserve to be separated from the Hydrobiidae in a separate family, Pomatiopsidae? Stimpson (1865) gave excellent reasons for establishing a separate subfamily Pomatiopsinae within the wide group he considered to be Ris- soidae and equal to other subfamilies such as the Hydrobiinae, “Bithiniinae”, Rissoinae, Rissoininae, and Skeneinae. The principal character used was what he called “lateral sinuses” which separated the foot into anterior and posterior parts. Other distinguishing characteristics were: (1)theterrestrial habitat, (2) the peculiar mode of pro- gression, (3) the central tooth of the radula, with basal denticles, not lateral or basolateral cusps as in the Hydrobi- inae. Due to Stimpson’s influence, most authors have considered Pomatiopsis distinctly separate from other hydrobiid snails and have placed it either in a separate family or ina distinct subfamily (Table 1). Gill (1871) raised Pomatiopsis to family rank without a diagnosis; Pilsbry & Ferris (1906) did likewise. F. C. Baker (1902) recognized the subfamily Pomatiopsinae, but later (1926, 1928) used Stimpson’s subfamily charac- teristics to justify family status. Berry (1943) considered family status justified, especially on radular characteristics. A reanalysis of Stimpson’s subfamily characteristics in the light of current studies on morphology and mode of pro- gression is necessary in order to de- POMATIOPSIS AND ONCOMELANIA 7 termine whether or not Stimpson was correct in establishing the subfamily, and, if so, to determine whether or not full family status is warrented for Pomatiopsis. 1) Lateral Sinuses of the Foot. The following discussion pertains to Plate 1, Figs. 2 and 5, in which the right lateral view of the head-foot region of P. lapi- daria is shown with the animal expanding its foot in preparation for advancing (Fig. 2) and with the completion of a “step” or forward advancement (Fig. 5). The head-foot region is characterized by a number of folds, grooves andcreases. The ventral edge of the operculigerous lobe (Op) continues anteriorly as the suprapedal fold (P) and sweeps upwards towards the posterior part of the ros- trum (В). Anteriorly the fold is in- terrupted by the anterior termination of the omniphoric groove (Om). As shown, the groove arises under the mantle on the right side of the “neck” and runs obliquely down towards the an- terior border of the suprapedal fold. The groove is highly ciliated and serves to move fecal pellets along its path as well as particles and mucoid strings from the mantle cavity. The terminal end of the groove juts out over the antero-lateral portion of the foot onto which the fecal pellets and particles fall or are swept by ciliary currents. The groove un- doubtedly serves to transport eggs to the anterior foot prior to their final en- casement in a soil capsule. An identical grove is found on the left side and from time to time a particle may be seen mov- ing anteroventrally in it. More dynamic, however, are the ciliary currents sweep- ing into the left side of the mantle cavity. These currents are created, in part, by cilia on the lateral surfaces of the head. As the animal moves about it is evi- dent that the omniphoric groove is plastic and, due to stretching of the lateral skin, may disappear. Certain move- ments accentuate the groove and the anterior edges of the groove appear lobed, the suprapedal lobe (Su) ventrally and the subocular lobe (S) dorsally. Another groove, the subocular groove (So), is evident below the eye, arising from the suprapedal fold and terminating just posterior to the eye. With the exception of the suprapedal fold, which is always distinct, the other structures discussed are pliable; they appear and disappear as the animal moves about, stretching its head and body in various positions. In relaxed or preserved specimens only the suprapedal fold is evident. Dundee’s figure (1957, Pl. 3) is of a narcotized specimen of P. lapi- daria, as evidenced by the swollen and disproportionate head. The suprapedal fold is directly continuous with the ros- trum, with no sign of the omniphoric or subocular grooves. The pedal crease (Pc) shown in Fig. 5 is equivalent to the lateral sinus of the foot described by Stimpson (1865) andthe vertical fold mentioned by Baker (1928). This crease is evident only upon com- pletion of a “step” (Fig. 5) where the full weight of the animal presses down on the fully contracted foot. With re- duced contraction the pedal fold becomes less evident or non-existent. Muscular contraction bringing the posterior foot forward causes a bunching of muscles which in turn creates the crease or fold. Pigmented epithelium concentrated by this contraction accentuates the crease. Stimpson (1865) was correct in stating that the pedal crease was dependent upon “the peculiar mode of movement.” Abbott (1948a) stated that both Oncomelania and Pomatiopsis have the same mode of progression and pro- duce “folds in the flesh of the foot due to the weight of shell and body.” The folds and grooves described above were illustrated by Stimpson (1865: 33, Fig. 255-1 34, Fig 126081 Piet): Stimpson’s figures were partially correct but failed to consider the plasticity of the organism. His Fig. 26 is quite misleading, as current studies show that the pedal crease is not evi- dent when the fore-foot is expanding or expanded. Table 2 shows the extent to which Stimpson’s figures were copied FIG. FIG. FIG. FIG. FIG. FIG. FIG. Se APT IDEE G. M. DAVIS PLATE 1. Pomatiopsis lapidaria. Dorsal view of the head. Head, foot and locomotion. Right lateral view of the head-foot region; the fore-foot is-extended in the first movement of a “step. ” Sole of the foot without pronounced lateral indentation; see text, p 7. Sole of the foot showing pronounced lateral indentations. Right lateral view; the foot is contracted showing completion of a “step. ” Sole of the foot when the fore-foot is beginning to expand forward. The sole of the foot in various stages of “stepping.” A. The foot is contracted; B, the fore-foot expands while the hind-foot remains firmly attached to the substrate; C, the hind-foot is drawn up to the contracted state. FIGS. 8,9. Variations found in the shape of the flexible foot. At s G L M M O Om Op P anterior foot Be glandular units Ps lateral indentation R point where the mantle covers the “neck” 5 mucous slit So operculum Su omniphoric groove x operculigerous lobe suprapedal fold pedal crease posterior foot rostrum subocular lobe subocular groove suprapedal lobe mid-region of the foot under the pedal haemocoel POMATIOPSIS AND ONCOMELANIA 9 10 G. M. DAVIS TABLE 2. References to reproductions of Stimpson’s (1865) figures of Po- matiopsis lapidaria a Author E Page Plate | Figure Binney, 1865 93 Calls 1900 16 Baker, 1902 345 127 Walker, 1918 34 120 Annandale, 1924 2083 1 Baker, 1928 166, 167 IS and his influence felt. The figures de- picted the crease as a most prominent and rigid structure, which in his words was a truly “distinct fold separating the foot into an anterior and posterior part... ” Abbott (1948a) correctly stated that “previous accounts of the divided foot... are misleading.” Pelseneer (1906) was led to assert that a transverse furrow which crosses the anterior half of the foot was found in Pomatiopsis, a statment which is unfounded, as the sole of the foot is simple and undivided. Annandale (1924) felt that the external anatomy of Pomatiopsis might differ from that of Blanfordia, Stimpson’s drawing of the subocular and supra- pedal lobes leading him to State that the former genus had a “triangular process behind the true tentacle” which did not apply to the latter genus. Actually, the same lobes are found in both genera with about the same degree of develop- ment. Further influenced by Stimpson’s drawings and description of the lateral sinus, Annandale separated Oncomel- ania, creating a new subfamily, Tri- culinae, that did not have a divided foot, as distinguished from the Pomatiopsinae, where the foot was divided by a “trans- verse furrow.” 2) Mode of Progression (Pl. 1). The underside of the foot was examined by placing specimens on a glass slide ina drop of water, waiting until the animal began moving over the slide, inverting the slide, and supporting it above the stage of a dissecting scope. Mucoid secretions of the foot and supportive action of the surface film of water served to keep the snail from dropping from the slide. The snails moved freely across the slide allowing a study of foot shape, structure and mode of pro- gression, When the animal is at rest the foot is slightly contracted and a lateral in- dention is noted on either side of the foot (L, Pl. 1, Fig. 7). This indention represents the ventrolateral end of the pedal crease (Pc, Fig. 5) and corresponds to an area separating anterior and pos- terior portions of the foot. The latter, as Stimpson (1865) observed, is about twice the length of the former. The degree of indentation varies considerably from individual to individual and depends a great deal upon the amount of muscular contraction. In Fig. 3, for instance, only a slight bend of the lateral margins indicated the position of the indentation. At this point the foot is capable of greater contraction. In Fig. 7, the foot at position A is fully contracted. To advance, the anterior portion of the foot (At) expands forward, during which pro- cess the lateral indentation becomes progressively less evident, as this area also expands forward. When the fore- foot is fully extended (position B) the point corresponding to the lateral in- dentation (L, position A) has moved forward from 0.36 to 0.42 mm. In this expansion of the fore-foot, the posterior portion (Ps) remains firmly attached to the substrate (position B). The extended fore-foot then becomes attached to the substrate and the foot musculature con- tracts in the area marked x (Fig. 7), drawing the posterior foot forward until the foot again assumes the shape shown in position A. These movements con- stitute a “step.” While the snail is engaged in a series of “steps” the foot does not become fully contracted (po- sition C, Fig. 7). Stimpson (1865) stated that in pro- gression, and as part of the stepping process, the “snout” was thrust forward and its “disc-like extremity affixed to POMATIOPSIS AND ONCOMELANIA 11 the ground as far ahead as possible.” With the “snout” and the posterior por- tion of the foot solidly in place, “the anterior part of the foot becomes free and is thrust forward to the disk of the rostrum where it is again planted.” He considered 3 points of support to be used in progression, the 2 parts of the foot and the rostral tip. In studying numerous specimens of this speciesitis evident that the rostrum is not used for support nor does it ever become “affixed” to the substrate. The rostrum is only used for feeding as is described more fully below. The animal is sup- ported only by the foot with the points of greatest support associated with the mid- and posterior foot. The snail is quite capable of stretching the anterior foot forward and lifting it off the sub- strate while solidly supported by the mid- and hind foot. The various movements of Pomatiopsis lapidaria over rough and smooth substrates are classified and described below. a. Movement and feeding on moist filter paper. Filter paper serves asa roughened but unyielding surface on which the movement of the snail is normal and unstrained. When the animal is not actually feeding, the basic move- ments of the foot described above operate smoothly. Here the extending fore-foot expands beneath the tip of the rostrum and beyond it often as faras0.6-0.8 mm. The contraction drawing up the hind-foot is most often followed by a contraction of the columellar muscle which tends to raise the spire of the shell as well as pull it forward. On a smooth level substrate the last movement may not be pronounced but its occurrence gives the snail the appearance of hunching forward. With completion of the hunch- ing movement the rostral tip is again brought before the front edge of the foot. Often the rostrum sweeps from one side to the other without touching the substrate. While feeding, the animal may rasp the substrate within an arc limited by the extensibility of the rostrum. In initiating a step while the animal is browsing, the fore-foot is extended to the tip of the rostrum but not beneath it. Completion of a“step” automatically causes an advance of the rostrum. With the advance, the rostrum may remain near the substrate, the radula rasping here and there, or, with the hunching motion, it may stretch straight out and, subsequently, be drawn straight back to the body while rasping the substrate. Occasionally, with the hunching move- ment, the rostrum is raised andthe whole head stretched upward at an angle of about 60°. The rostrum is fully ex- tended and then lowered to the substrate as far ahead of the body as possible. In all these movements the foot is coordinated with the actively probing rostrum but the animal is not dependent upon the rostrum for support. b. Movements on soil. On loose soil, especially on a sharp incline, the step- like movement may appear less evident due to back sliding and the crumbling of soil beneath the foot. Where greater exertion is necessary the hunching move- ment becomes quite pronounced. c. Movement in water. While moving across glass and submerged under water, this species does not simply glide as stated by Abbott (1948a). Water greatly reduces resistance to movement caused by the weight of shell and visceral mass. Glass affords so smooth a substrate that ciliary activity facilitates a slight glide. The main movement is achieved, however, by what appears to be analmost effortless short extension of the fore- foot and subsequent contraction at mid- foot as described above. The “step” is not at all pronounced due to the re- duced pressures on the foot. Under water the hunching movement is greatly reduced while the ciliary action, as well as the undulations of the margin of the fore-foot are more pronounced. Pomatiopsis lapidaria was observed moving upside down, adhering to the surface film of water. Thefore-foot was extended and cupped, thus forming a deep concavity. Ciliary currents swept 12 G. M. DAVIS into the concavity and the animal glided over the surface of the water. Only occasionally did the mid-foot contract or the snail make a hunching movement. The point to be made here is that the foot of P. lapidaria is like that of the Hydrobiidae while the mode of pro- gression is evidently an adaptive change enabling movement on land where the buoyant effect of water is lost and the increased weight of the foot makes ciliary movement impossible. The pedal crease is a result of accentuation of certain muscles enabling the step- like mode of progression and is accen- tuated in Pomatiopsis by pigmentation on the lateral foot surface below the suprapedal fold. Weight of body and shell also serve to accentuate the crease. The basic difference between the mode of progression of members of the Pomatiopsinae and that of members ofthe Truncatellidae was pointed out by Stimpson (1865) who stated that in Trun- catella there were only 2 points of support and that progression in Trunca- tella should be called “looping” as opposed to “stepping” in Pomatiopsis. In the former the rostrum and the whole foot were described by Stimpson as the 2 points of support while in the later the 2 “sections” of the foot and the ros- trum were considered as 3 points of support. In reality, however, as dis- cussed above, Pomatiopsis does not so use the rostrum. Fretter € Graham (1962) gave the following description for the mode of progression in Truncatella. It “extends the snout, which is very extensible, and grips the substratum TABLE 3. Cusp formulae for the teeth of Pomatiopsis lapidaria previously presented or dis- cussed in the literature* Central Author Anter. cusps Lateral Zu Ant Basal cusps Marginal arginal 1. Stimpson, 1865 5 2. Binney, 1865 copied from Stimpson 3. Baker, 1902 copied from Stimpson 4. Walker, 1918 copied from Stimpson 5. Annandale, 1924 10 6. Thiele, 1928 5 7. Baker, 1928 9 8. Thiele, PEN RES 3-5 9. Abbott, 1948a 2(3)-(3)2 2-1-2 (3-4) | 5 1-1-1 10. Dundee, 1957 2-1-3 = 2-2 * Compare with Table 8 and Plate 19. ** Description given for the genus Pomatiopsis. POMATIOPSIS AND ONCOMELANIA 13 with its tip (p 598, Fig. 315); it then pulls the foot up to grasp the ground just behind the snout, dragging the shell in its rear, releases the snout and starts the process once again. Sometimes the foot slides along the surface of the ground as it is drawn forward, sometimes it is lifted clear.” They state that the small rounded foot and the expanded tip of the snout are related to this move- ment. 3. Radula. Stimpson (1965) figured the radula of Pomatiopsis lapidaria. As shown in Table 3, this drawing was ex- tensively copied. Additional figures were prepared by Annandale (1924), Baker (1928), Thiele (1928) and Abbott (1948a). The radula of P. cincinnatiensis was figured by Troschel (1863) under the synonym of Amnicola sayana and sub- sequently copied by Stimpson (1865) and Binney (1865). Baker (1928) and Berry (1943) provided new figures, those of the last being the best. Stimpson (1865) states that the radula of the Pomatiopsinae is distinct from that of the Hydrobiinae in that, in the former, the basal cusps (denticles) of the central tooth are placed at or near the base. In this view Stimpson is correct. In the Hydrobiidae the basal cusps are often attached to a thickened ridge along the lateral angle of the central tooth (La, Pl. 19, indicates that lateral angle, unthickened inthe Pomatiopsinae). This arrangement is particularly noted in genera such as Amnicola and Hydrobia. In the Pomatiopsinae the supports for the cutting edge of the basal cusps do not arise from a thickened ridge along the lateral angle but from moulded thicken- ings of the tooth running posteriorly from points situated anteriorly on the lateral angle (Slb, central 6, Pl. 19). By changing focus on the lateral angle it is evident that each thickened support causes a Slight undulation to the lateral angle where it arises. The most promi- nent of the supports arises just lateral to the outer anterior cusp (or cusps) flanking the large central cusp (Slb, central 6, Pl. 19). Baker (1926) in the first diagnosis of the Pomatiopsidae used the following radula characteristics as having major importance: “radula with its few cusps of large size and the large denticles on the base of the central tooth.” Baker (1928) stated inthe family diagnosis: “central tooth of the radula with but one large basal denticle; denti- cles of the lateral and marginal teeth very large and few in number, pro- portionally much larger than in the Amnicolidae.” Berry (1943) states that the radula is “very different from the Amnicolidae. The central tooth has the basal wing terminating as a cusp and a single basal denticle down from the lateral ridge. The few large cusps on the central, lateral and marginal teeth are distinct characters of Pomatiopsis, and not Amnicolidae.” The radulae of Pomatiopsis and On- comelania are described in detail in the section on anatomy. A few comments must be made here, however, as they relate radular structure to the family or subfamily taxon characters. Baker (1926, 1928) was mistaken in using a single pair of basal cusps as part of the criterion for defining the family Pomatiopsidae. P. lapidaria has 2 or 3 pairs of basal denticles (Thiele, 1928; Abbott, 1948a). P. cincinnatiensis has 2 pairs of basal denticles (Berry, 1943), as do the radulae of P. californica and P. binneyi which I have observed. In studying the radula collection of P. cincinnatiensis at the University of Michigan Museum of Zoology (UMMZ), it was evident that the basal portion of the lateral angle did not terminate as a cusp, although the outer basal cusp had a definitely more external position than that of P. lapidaria. It was possible to discern, basolaterally to the outer basal cusp, a well-defined termination of the lateral angle. The support for the basal cusp arose quite anteriorly, on the edge of the lateral angle. The radulae found in the Pomatiop- sinae are clearly included in the range of types found in the Hydrobiidae through- out the world. They are similar in that 14 G. M. DAVIS the central tooth is wider than long, and has basal cusps, that its anterior edge is narrower than the posterior (basal) edge, and in that there is more than one anterior cusp. The shape and denticul- ation of the laterals and marginals are compatible with those found inthe family Hydrobiidae. The pomatiopsid radula is distinctive in that the cusps of the laterals and marginals are generally fewer and larger than in most hydrobiids. P. cincin- natiensis is extreme in having only 3-4 large cusps on the marginals. P. lapi- daria has up to 9 cusps on the inner marginal while P. binneyi has up to 11. In the last species, the cusps of the marginals are small and needle-like, as in many hydrobiid species. The cen- tral is distinctive in having only 1 or 2 cusps on either side of the central cusp on the anterior edge of the tooth. Other hydrobiids generally have 3 or more cusps on either side of the central cusp. The distinctive nature of the supports for the basal cusps has been mentioned. When the radulae of other taenio- glossid prosobranch families are compared with those of the Hydrobi- idae, it is evident that the radula of Pomatiopsis is a hydrobiid type. While members of the Bithyniidae, e.g., Bithynia tentaculata, have a radula seemingly more similar to some hydro- biids than that of Pomatiopsis, Bithynia does show numerous other major mor- phological differences which clearly separate the groupfrom the Hydrobiidae. In the Truncatellidae the unusual tri- angular central tooth with the single anterior cusp is distinctive. In the assimineid radulae the central and outer marginals have shapes which are dis- tinctly different from those found in the Hydrobiidae. The length of the central is greater than its width. The central is without the lateral angle character- istic of the hydrobiid radula and is often without basal cusps. The outer marginal is exceptionally wide com- pared with the relatively slender type found in the Hydrobiidae. In the littorinacean Littorinidae and the cerithiacean Pleuroceridae the central lacks distinctive basal cusps. Radular characteristics are clearly not suf- ficient to separate Pomatiopsis from members of the family Hydrobiidae. In conclusion, the Pomatiopsinae represent a Subfamily including several genera outside the United States; e.g., Tomichia from South Africa, Blanfordia from Japan, as well as Oncomelania from the Western Pacific area. In these genera the shell is elongated and turreted in contrast to the globose type of Amnicola, the bulimoid shape of Littoridina, or the planispiral shell of Horatia. The tendency in the group is towards an amphibious to terrestrial habitat. Correlated with the amphibious habitat is a step-like mode of pro- gression. A crease develops in the anterolateral foot upon full contraction of the foot. Eggs, laid singly, are covered with a mud capsule (not known for Tomichia). The radula has fewer and larger cusps on the marginal teeth than other hydrobiid snails. The anterior cusps on either side of the central cusp of the central tooth are 1 or 2in number. The basal cusps of the central teeth have a distinctive type of support and are generally 2 ог 3 in number. The eye is in a pronounced swelling at the base of the tentacle, differing from the species of the genus Hydrobia, where the eye is in a slight swelling of the tentacular base. Characters in common with other hydrobiids are the simple verge (hydro- biids have verges either simple or with various appendages), a paucispiral corneous operculum, a broad simple foot which is truncate in front and gently rounded behind. The foot has an anterior transverse mucous slit (See Ms, Pl. 1, Fig. 7A). The rostral shape and tip, and basic internal anatomical features are hydrobiid. POMATIOPSIS AND ONCOMELANIA 15 MATERIALS AND METHODS The body of this paper is divided into 4 major sections: anatomy, hybridization studies, electrophoretic studies and laboratory ecology. Methods pertaining to each of these sections will be dis- cussed under those 4 headings. The snails used throughout these studies were all fully mature adults as indicated by shell size in Pomatiopsis and varix formation in Oncomelania. The only exceptions were the newly hatched young used in the growth experiments. Field collected Pomatiopsis lapidaria were utilized for anatomical studies and electrophoretic experiments. These were obtained from the Barton and Hog Back stations described by Dundee (1957) as well as the Parker Mill Station dis- cussed by van der Schalie & Dundee (1959). These stations are within 5 miles of Ann Arbor, Michigan, U. $. А. Oncomelania subspecies used were Е] or Fa laboratory reared snails of field collected parental stock. O. hupensis formosana came from Pu Yen village, a small farming community a few miles south of the city of Changhua, Taiwan (Formosa). О. hupensis noso- phora were sent from the Kofu Valley in the Yamanashi Perfecture of Japan. O. hupensis quadrasi were sent from Palo, Leyte, in the Philippines. COMPARATIVE ANATOMY A. Introduction One encounters several major problems in attempting to make detailed comparisons from published anatomical material. The material is often stylized and portrays organ systems in a general manner omitting exact contours, di- mensions, variations and positional re- lationships with other organs. Homolo- gous organs are presented in different views by the various authors, making comparisons difficult or impossible. In this study the gross anatomy of the muscular, nervous, reproductive systems, parts of the alimentary sys- tem and the external morphology are discussed. The systems and organs of both species are presented in the same manner and orientation, thereby facili- tating comparisons. The systems and organs were studied in order to de- termine in a comparative manner (1) the presence or absence of a structure, (2) qualitative differences inthe structure of homologous organs and (3) quantitative differences or Similarities in organs or structure. This is not a complete anatomical de- scription, as details of the excretory and circulatory systems are not covered. The systems investigated were chosen because of their potential in providing characters which could be readily used in a systematic discussion. Pomatiopsis lapidaria will be dis- cussed first, followed by a similar treatment of Oncomelania hupensis for- mosana. Comparisons between the 2 species will be discussed with each anatomical section presented for O. hupensis formosana. B. Materials and Methods Dissections were carried out under magnifications of 40X and 60X using a Nippon Kogaku dissecting microscope. Measurements of all structures were made using a standard ocular micro- meter. Proportions and structural di- mensions in all drawings were checked against the specimen, using this micro- meter. A 9 cm Petri dish filled with paraffin and blackened with norite served as the container and substrate for all dissections. Tools used for dissections were “Minutien-Nadeln” (insect pins) em- bedded in solid glass rods, iridectome scissors of the finest grade, jeweler’s forceps with extra fine points, andpliers for cracking the shell. Well over 200 snails were used for the anatomical studies of each species. The snails were studied while living or just freshly preserved. Living animals were most suitable in studying the organs and ducts of the reproductive systems. 16 G. M. DAVIS Aqueous neutral red was very useful in accentuating the tubes of the reproductive systems as well as nerves and ganglia. Aqueous methylene blue aided in staining the visceral ganglion and associated nerves in the living snail. In studying freshly killed snails the animals were removed from the shells, pinned out in the desired position under water, the water was poured off, and Bouin’s fixative was added full strength. Structures in the head were more readily studied in the freshly killed snail, as mucoid secretions were a hindrance in the living animals. Studies on nerves were facilitated by dissecting under Bouin’s solution as minute nerves stood out prominently in that fluid under direct illumination. Muscles were studied in the contracted state where the smaller muscles were more prominent and where a fairly stable configuration of muscles was assured when numbers of speci- mens were studied. Radula. Radulae were dissected from the buccal mass and placed in a 10% solution of KOH for varying amounts of time (about 24 hours). Upon dis- solution of attendant membranes the cleaned radula was placed on a slide with a drop of 4% acetic acid. The acid loosened the lingual membrane (radular shield) within an hour, so that the radula could be readily flattened out; it also facilitated removal of separate teeth or groups of teeth from the mem- brane. With the radula flattened out, measurements were made of the length and width of the radula and the number of rows of teeth were counted. Measure- ments and counting were carried out under a magnification of 150X, using a Nippon Kogaku compound microscope. At this point, the radular ribbon was mounted whole or teeth were stripped from the membrane to facilitate a study of each tooth. The acid was allowed to dry on the slide and then a drop of CMC-10, a non-resinous mounting medium, was added and a coverslip applied. In 24 hours the CMC-10 dried and the edge of the coverslip was ringed with clear fingernail polish assuring the permanency ofthe slide. Detailed studies of the teeth were readily made without need for stain as the smallest cusp readily stood out. Drawings of the radular teeth were made using oil im- mersion (1000X) and camera lucida. Jaws. The buccal mass was removed and the area of the outer lips cut free with iridectome scissors. This separated anterior end of the buccal mass was placed on a slide and opened from the dorsal surface, thus exposing the oral tube and the anterior part of the buccal cavity. A drop of CMC-10 was placed on the tissue, the tube was opened, and a coverslip was applied. The jaws, fully exposed, were studied under the compound microscope. They were drawn using a camera lucida. Shell. Shells were boiled in sodium hypochlorite (5.25%) (commercial Clorox) to remove the periostracum and all occluding matter such as algae, dirt, etc. Cleaned in this manner, the sculpture, shell surface, apical whorls, and sutures were readily studied. Anatomical orientation. There is no problem in discussing features of the head as the foot is ventral and the rostrum points anteriorly. However, the remainder of the organism is coiled within the shell and orientation becomes a problem when discussing anatomical features. As Fretter & Graham (1962) state for Littorina, “as the animal lies in its shell the outer part of each whorl corresponds to the dorsal surface of the body, and the inner to the ventral.” The inside of the coil or ventral surface is that which is appressed to the colu- mella. All visceral anatomy presented here was described from the uncoiled snail, with the head lying to the right, the body surface presented in the illustrations being the columellar (ventral) side. Left lateral is towards the bottom of the drawings wherever the term is used, and right lateral is towards the top of any given figure; “posterior” means towards the apex, “anterior” towards the head. 17 POMATIOPSIS AND ONCOMELANIA 39715 eAljeredui0og "(18 d) 6 pue (61 9) у sojqeL ur uaaı ose "рирзошло{ Sısuadny *O "Y ‘P140p140] “JT “Y "Dunsowmxof sisuadny DVIUD]IMOIUN pue 22410р140] sisdoymuod jo 5ПЭЧ$ £ < ALV'Id 18 G. M. DAVIS Gr ER, 0.55 тт 9 PLATE 3. Shell features of Pomatiopsis lapidaria and Oncomelania hupensis formosana. FIGS. 1, 2, 3. Apertural view of the shell of P. lapidaria showing the wide umbilicus and the shortened parietal callus (x). FIG. 4. Shell of O. hupensis formosana showing the sinuate outer lip and the varix. IDEs) BY Shell of P. lapidaria showing the straight outer lip without the varix. FIG. 6. Apical whorls of P. lapidaria. The short arrow points out the width of the tip of the apical whorl; the long arrow points out the width of the first whorl. HIGS 7, 8 Apertural view of O. hupensis formosana showing the relatively narrow um- bilicus and elongate parietal callus. FIG. 9. Apical whorls of O. hupensis formosana. The scale is the same as in Fig. 6. POMATIOPSIS AND ONCOMELANIA 19 TABLE 4. Conchological measurements of Pomatiopsis lapidaria from Ann Arbor, Michigan | Length in mm Width in mm Number of Snails Structures Measured Shell 6. 0 whorls Shell 6..5 whorls Shell 7. 0 whorls Aperture Apical Whorl Tip of apical whorl X = The mean S = Standard deviation Se = Standard error of the mean All illustrations were made by the author. C. Pomatiopsis lapidaria 1. Shell Say (1817) described the shell in the type descriptionas “turreted, subumbili- cate, with 6 volutions, which are obso- letely wrinkled across. Suture im- pressed. Aperture longitudinally ovate- orbicular, rather more than 1/3 of length of shell. Length about 1/5 of an inch.” Although later authors, in particular F. C. Baker (1928), have elaborated on Say’s original description, more de- tail and discussion of the shell is necessary. The shell (Pl. 2, A)isindeed elongate and turreted. Adult shells have 6.5-7.0 whorls. Shells of 7.5 whorls are rare in non-fossil material (see p 20). The nuclear whorls are 2.0-2.75 in number, glassy, and in cleaned material may appear amber, thereby set off from the brownish or yellow-brown horn color of the remainder of the shell. In uncleaned material the nuclear whorls may appear glistening or dull white. The first nuclear whorl, as Baker pointed out, is usually not emergent and is partially “embraced by the second whorl.” This often gives the apex a flattened appearance. The sutures of the whorls are deeply impressed and the whorls correspond- Number of Snails The aperture is somewhat narrowed and angled above” (Baker, 1928). The inner lip is connected with the outer lip ingly very convex. “elongate, ovate, by a parietal callus. In some speci- mens the callus is so thickened that it looks as if the inner lip continued into the outer lip. Occasionally a speci- men is found where the inner lip is not adnate to the parietal wall. The inner lip is slightly reflected over the um- bilicus. The parietal callus varies in length, the greater the length, the more occluded the umbilicus. The outer lip is sharp, strong and does not form a varix. Observing the edge of the outer lip with the aperture rotated 90° to the left of apertural view, one observes that it is straight, not sinuate (Pl. 3, Fig. 5). The apical part of the outer lip may have a Slight sinuation in some cases but this is not nearly as plain as the sinuation found in Oncomelania hupensis formosana (Pl. 3, Fig. 4). The umbilicus is very pronounced and deep (Pl. 3, Figs. 1-3). As shownin these same figures, the base of the shell is rounded. Without magnification the whorls of the cleaned shells appear smooth and glistening. Under 6-16 magnifications it is evident that the surface of the whorls are wrinkled by growth lines which are irregular in diameter, vary in prominence, and are closely packed. The overall effect is to 20 G. M. DAVIS give the shell a roughened micro- sculpture. The coarse growth lines start immediately after the nuclear whorls. A series of measurements, which are felt to be of use in specific comparisons, are discussed below. Others could be made, but those presented are adequate for the comparisons intended. A series of 62 specimens in all, collected from the Barton, Parker Mill and Hog Back stations were studied conchologically without reference to sexual dimorphism. Length, width, aperture length, parietal callus length, width of the first nuclear whorl and width of the tip of the nuclear whorl (Pl. 3, Fig. 6) were measured. These were recorded with the cor- responding whorl count. The results are shown in Table 4. Shells of 6 whorls had an average length of 5.5 mm and an average width of 2.9 mm; the cor- responding measurements for shells of 6.5 whorls were 6.2 mm and 3.1 mm, and for shells of 7.0 whorls 6.7 and 3.2 mm, respectively. No shells with 7.5 whorls were found among several hundred additional snails observed (see below). The average length of the aper- ture for snails of 6.0-7.0 whorls was 2.1 mm. These dimensions were com- pared with those gathered from several lots of this species housed at the UMMZ (Table 5). These lots represented popu- lations scattered over the extensive range of this species (distribution map in Abbott, 1948a). No significant differ- ence was found for the parameters of shell length, width, or aperture length among these populations. The first nuclear whorl (Pl. 3, Fig. 6) varied but little (Table 4) with a width of 0.53 mm. The tip of the first whorl averaged 0.19 mm and this width like- wise was extremely constant. These features did not deviate significantly among the populations studied. The length of the parietal callus, however, did vary quite a bit in length and signifi- cantly so between populations (Table 5). The Ann Arbor snails of the current studies had an average callus length of 0.6 mm, the shortest of all the popu- lations studied. Correlated with this feature was an unusually pronounced umbilicus. As shown in Table 5, the average callus length of various popu- lations ranged from 0.66 mm to 0.96 mm. These differences are possibly related to growth patterns which deviate under different environmental con- ditions. Variation in length of parietal callus is shown in Plate 3, Figs. 1-3. Hubricht (1960) studied the shells of Pomatiopsis lapidaria from a number of localities and stated that the species appeared modified by different eco- logical conditions; “slender, thick shells occur in dry habitats, obese thinner Shells in wet ones. This is especially true in the South.” Inthat paper Hubricht considers P. praelonga Brooks & Mac- Millan and P. hinkleyi Pilsbry syn- onymous of P. lapidaria. I agree with Hubricht in synonymizing these forms. Certainly P. hinkleyi (Table 5) showed no shell characteristics significantly different from P. lapidaria from many localities, except that in lengthof parietal callus, a character which is shown to be variable. A lot of fossils (UMMZ) fromthe same locality as Pomatiopsis scalaris (Baker, 1927) a Pleistocene fossil, was also studied (Table 5). P. scalaris has been described as “strikingly” different from P. lapidaria, being longer, having 1 more whorl (8 whorls) and very deep sutures. The lot of fossils here studied formed a series which graded from the P. lapidaria observed in Ann Arbor to typical P. scalaris, i.e., the present fossil material had shells of 7.0-7.5 whorls. (P. scalaris was described as having 8 whorls). Many of the recent specimens of P. lapidaria had the very deeply impressed sutures and convex whorls attributed to P. scalaris. Speci- mens of recent P. lapidaria, from various parts of the country, with 7.5 whorls and the characteristics of P. scalaris, can occasionally be found, although they are comparatively rare. The umbilicus of P. scalaris as well as of the fossil series POMATIOPSIS AND ONCOMELANIA 21 TABLE 5. The average callus length for lots of Pomatiopsis lapidaria from various wide- spread localities in the U. S. A. Locality Number of Average callus specimens length in mm ln Alabama, Florence; Lauderdale Co. Lot UMMZ 69912* topotypes of P. hinkleyi 16 0.72 Florence; Bolder Falls; Lauderdale Co. Lot UMMZ 91487 paratypes of P. hinkleyi 4 0.96 Indiana, New Harmony; Posey Co. Lot UMMZ 69915 Fossil Material 11 0.72 Iowa, Marion; Linn Co. Lot UMMZ 132464 11 0.84 Michigan, Ann Arbor; Washtenaw Co. Lot UMMZ 91559 Barton Station 9 0.90 Lot UMMZ 183219 Barton Station, 1952 12 0. 66 Lot UMMZ 183217 Hog Back Station, 1952 0. 84 North Carolina; Broad River, Point Rock Lot UMMZ 91495 0.72 Ohio; Miami Co. Lot UMMZ 59325 Drift off Big Miami River 0.84 Wisconsin; Baraboo; Sauk Co. Lot UMMZ 143721 0. 84 *University of Michigan, Museum of Zoology catalog numbers. from the UMMZ appeared more rounded, as that found in recent P. lapidaria. As wider, and deeper than that met in recent a result of these studies, P. scalaris is P. lapidaria from many localities, a considered an early extreme of P. lapi- feature that is correlated with a pro- daria and synonymous with it. nounced tendency for a short parietal callus and an inner lip which is barely 2. External morphology and topography reflected. However, these same charac- ters are quite pronounced in current The folds and grooves of the head have populations of P. lapidaria from the Ann been mentioned. Arbor area. The width of the apical Pigmentation. The head, dorsally and whorl in the fossil material was the same laterally, is black to grey-black due to 22 G. M. DAVIS PLATE 4. Head, foot, and mantle region of Pomatiopsis lapidaria. anus bursa copulatrix columellar muscle ctenidium esophagus fecal pellet glandular units intestine cut edge of the kidney ventral surface of the kidney edge of the mantle cut edge of the mantle operculum osphradial ganglion omniphoric groove operculigerous lobe Opi Opo osphradial pit opening of the pallial oviduct oviduct portion of oviduct entering pallial ovi- duct suprapedal fold pedal crease pericardium pallial oviduct rostrum spermathecal duct sperm duct subintestinal sinus stomach blood vessel POMATIOPSIS AND ONCOMELANIA 23 heavy pigmentation (Pl. 1, Figs. 1, 2, 5). The edges of the foot below the supra- pedal fold are dusted with pigment, al- though more lightly than the dorsal sur- face. Pigmentation continues along the neck into the mantle cavity but fades out towards the base of the “neck.” Оп the anterodorsal edge of the foot pigmented patterns appear to outline channels in the foot (Pl. 4) which probably coincide with the mucous ducts figured by Abbott (1948a). The sole of the foot (Pl. 1, Figs. 3, 4, 6-9) appeared to have 2 color patterns when studied under direct illumination. The periphery of the sole was a light slate grey to blue-grey while the central area appeared opaque white to yellow- white. The central area, onthe average, was 0.50 mm from the front edge of the foot, 0.96 mm from the posterior end, and about 0.40 mm in from each side. This area corresponds to aposition over the pedal haemocoel. White granular units of about 25y dia- meter were especially crowded in the posterior part of the foot, becoming sparse along the sides. Granules were Sparse in the central area. In addition to the relatively large granules, the sole was covered with small whitish bodies appearing as tiny rods all perpendicular to the sole (observed at 40X). These tiny rods were closely packed at the anterior edge of the foot, and less dense posteriorly. Viewing the living animal through the shell (apertural view with snail re- tracted) the pattern of pigmentation on the outer or dorsal side of the body tube is readily observed. The intestine filled with fecal pellets is clearly discerned crossing the body whorl, underlined bya thin band of pigment 0.24-0.40 mm wide. This band continues along the dorsal surface, widening above the body whorl (0.48-0.72 mm). In the apical whorls this band of pigment becomes more slender again (0.25-0.40 mm). Dundee (1957, Pl. 6) shows this pattern in the apical whorls. The band is generally positioned on the periphery of the coil or displaced towards the aperture on each whorl. The band is not neatly delineated with paralled sides but ir- regularly scalloped, flammulate at the edges, especially the edge towards the apex. This pattern is observable inboth males and females. From the ventral aspect some of the pigment is observed curling over from the dorsal surface (Pls. 5, 6). The exterior epithelium covering the ctenidial area (see Pl. 4) is pigmented and the pigmentation tends to outline the position of each gill fila- ment. Abbott (1948a) states that in Pomati- opsis lapidaria “the most distinguishing color markings are the bright splotching of yellow or yellowish-white granular dots over each eye forming false ‘eye- brows’.” These glandular units (Pl. 1, Fig. 1, 2, 5% BPiy4: Ранг. ра tially surround the eye and occlude the medial, posterior edge of the eye. Col- lectively they are a mass about 0.36 mm long with the greatest width of 0.17 mm. Coloration varies from white to y ellow-white. Tentacles and Eyes. The eyes are in large, distinct swellings at the outer base of each tentacle (Pl. 1, Fig. 1). Viewed ventrally these swellings appeared con- tinuous with the tentacles but from the dorsal surface they appear as units fused with the tentacles and set off from them by a slight crease. In any event, the ocular units are not the simple swellings in the outer tentacular bases seen in Hydrobia and others of the Hydrobiinae and Rissoidae. The characteristic shape of the tentacles is shown in Plate 1, Fig. 1. They are highly contractible and pliable but can most often be seen with a swelling at their base anterior to the ocular units. General Topography. In the uncoiled snail with the columellar side exposed, one can observe, in addition to the head, 3 general areas: (1) the region of the mantle cavity (Pl. 4) which extends from the mantle edge (M) where it encircles the “neck” (Pls. 5, 6) back to a point just posterior to the tip of the visceral 24 G. M. DAVIS PLATE 5. Uncoiled female Pomatiopsis lapidaria showing parts of the female reproductive system. FIG. 1. Uncoiled female P. lapidaria. The oviduct (Ov)is broken due to the stress of uncoiling the snail. FIG. 2. The portion of the reproductive system uncovered by peeling away the connective tissue and kidney tissue between the bursa copulatrix (B) and the edge of the mantle cavity (x). Ast anterior chamber of the stomach Ova portion of the oviduct passing ventral to B bursa copulatrix the spermathecal duct to enter the pallial Cl columellar muscle oviduct D digestive gland Pe pericardium Es esophagus Pi pigment band showing the flammulate Go gonad pattern at the edge Gp gonopericardial duct Po pallial oviduct K, ventral surface of the kidney Pst posterior chamber of the stomach M edge of the mantle Sd spermathecal duct Mc ventral wall of the mantle cavity Sdu sperm duct Oo oocyte Vg visceral ganglion Ov oviduct x the posterior end of the mantle cavity POMATIOPSIS AND ONCOMELANIA 25 PLATE 6. Uncoiledmale Pomatiopsis lapidaria showing parts of the male reproductive system. FIG. 1. The uncoiled snail. FIG. 2. The prostate pressed against the intestine and turned over to expose the point of en- trance of the posterior portion of the vas deferens in relationship to the point of exit of the anterior portion of the vas deferens. FIG. 3. The prostate as viewed in Fig. 1; the connective tissue was cleared away to show the posterior vas deferens (Vd;) passing under the edge of the prostate. F fecal pellet Mc ventral wall of the mantle cavity Cl columellar muscle Pi pigment D digestive gland Pr prostate Es esophagus Sh shell fragment G, large, white “granular” units St stomach Go gonad Vd, posterior portion of the vas deferens Gu a single glandular unit from the prostate Vda anterior portion of the vas deferens In intestine Ver verge covered by mantle wall K, _ kidney Vg visceral ganglion M edge of the mantle 26 G. M. DAVIS ganglion (Vg); (2) the mid-body region extending from the end of the mantle cavity to the posterior portion of the stomach and in which are observed the stomach (St; Pst, Ast), kidney (Kj), part of the intestine (In), a segment of the esophagus (Es), and a portion of the oviduct or vas deferns (Ov or Vdy, respectively); (3) the digestive gland (D) which continues posteriorly from the stomach and contains the gonad (Go) in its left ventral surface beneath the epithelium just posterior tothe stomach. In the female (Pl. 5), the pronounced pallial oviduct (Po) traverses the an- terior portion of the mid-body and the length of the mantle cavity. In the male (Pl. 6), the prostate (Pr) lies over the 2 areas but does not extend anteriorly the whole length of the mantle. The bursa copulatrix (B) iS prominent in the mid-body of the female. The colu- mellar muscle (Cl) emerges from the “neck” area of the head (Pl. 4) on the ventral surface and is pressed against the ventral exterior mantle cavity wall which is exposed in Plates 5 and 6. Connective tissue usually binds the colu- mellar muscle in place but it has been torn and the muscle pulled away from the mantle cavity wall to expose that area and the associated structures, i.e., the visceral ganglion (Vg), anterior section of the vas deferens in the male (Vdo) and the spermathecal duct in the female (Sa). The various organs which were pointed out in the plates discussed above are readily observed, although no connective tissue has been removed, because of their position just beneath the connective tis- sues, their size and bulk, or color and texture. The gonadsinthe living animals stand out bright yellow, as does the bursa copulatrix (В, Pl. 5). The kidney (Ky)is visible because the ventral tissues of this organ are rather transparent. The organ is sac-like and filled with fluid. №- merous white granules are in constant agitation and can be observed through the membranes. The total effect is that the kidney appears quite white compared with the surrounding tissues of other organs. The mantle cavity and pallial oviduct are flecked with pigment. Imbedded within the connective tissue all over the ventral surface are what appear to be white granules (some appear under the compound microscope to be glandular as shown in Pl. 11, Fig. 6). These are particularly concentrated in several areas (Pl. 5): the triangular area from the mantle edge to the point where the anterior portion of the pallial oviduct disappears into the mantle cavity, the connective tissue sheet between the bursa copulatrix and the posterior edge of the mantle cavity. The space over the style sac between the left edge of the antero- ventral arm of the kidney and the eso- phagus is frequently crowded with granules as is the V-shaped area be- tween the anterior (Ast) and posterior (Pst) chambers ofthe stomach. Granules are scattered over the ventral tissue of the digestive gland. In the male the reduced size of the prostate, as compared with the female pallial oviduct, permits a clearer ex- ternal view of the intestine (In), which crosses the mantle cavity and is often filled with fecal pellets (F) in a charac- teristic manner (Pl. 6). Likewise, characteristic for the male, from an external view, is the dorsal swelling of the mantle cavity corresponding to the large verge coiled within. 3. The Mantle Cavity. The mantle cavity was opened by cutting posteriorly along the right lateral margin of the mantle wall where itfuses with the “neck” (Pl. 4) or posteriorly just to the right of the mid-dorsal mantle wall (Pl. 11, Fig. 1). In Plate 4 the columellar muscle is shown pulled away from the mantle cavity wall and down- ward; the mantle edge (M) is pulled forward to stretch the left wall, enabling a clear view of the ctenidium (Ct) and osphradium (Opi, osphradial pit and Og, osphradial ganglion within). Organs and structures associated with POMATIOPSIS AND ONCOMELANIA 27 the mantle cavity are the ctenidium, osphradium, anterior wall of the peri- cardium (Pe), opening of the kidney (Or, Pl. 8, Fig. 1), openings of the anus (A), pallial oviduct (Opo) and spermathecal duct (Osd, Pl. 9, Figs. 4, 7). In males, the verge is housed within the cavity. The reproductive structures and their association with the mantle cavity willbe discussed in sections dealing with the reproductive systems. Ctenidium. The ctenidium is com- posed of triangular gill filaments charac- teristically of а hydrobiid nature (Pl. 11, Fig. 1; Pl. 13, Fig. 3). Dundee (1957) stated that there were 15-20 filaments while Abbott (1948a) stated that there were 27-29 lamellae. In a study of over 50 mature specimens without regard to sex the number found varied between 20-28 with an average of 24. Males characteristically had fewer lamellae than females, an average of 22 + 2 for the former and 25 + 3 for the latter. This difference is correlated with sexual dimorphism, the males being smaller than the females. A blood channel (V, Pl. 4) is noted in the mantle collar connected with a blood sinus (Si) running along the anterior intestine. Near the anus the tissues of the sinus make a collar around the intestine and send a tubular passage to the blood channel in the mantle edge. The gill lamellae connect with the sinus (Si, subintestinal sinus). The base of each lamella con- nects with a vessel (V) which runs tothe auricle (Au) (Pl. 11, Fig. 1; Pl. 8, Fig. 1). Osphradium. The osphradium is an elliptical groove or pit (Opi) located at the base of the gills near the anterior end of the mantle cavity as shown in Plate 4 and Plate 11, Fig. 1. The edges of the groove are swollen and lip-like, packed with small white granules (possibly glands). Swelling up within the grooveis the osphradial ganglion covered with an epithelium which itself appears swollen and filled with fluid (Og, Pl. 4; Pl. 11, Fig. 1). The osphradial nerve (On, Pl. 11) enters the ganglion just anterior to the mid-ventral line. The osphradium is 0.59 + 0.12 mm long and 0.33 + 0.02 mm wide. Visceral ganglion. The visceral ganglion (Vg) is observed imbedded with- in the tissues of the floor of the mantle cavity at the base of the “neck” (Pl. 11, Fig. 1). As mentioned previously, the ganglion is just as readily observed from the external surface of the ventral mantle wall where it is imbedded in the con- nective tissues (Pls. 5, 6). The sub- visceral and supravisceral connectives (Sbv, Suv, Pl. 11, Fig. 1) are seen running anteriorly on either side of the “neck” to disappear into the epithelium covering the anterolateral portions of the “neck.” Mucoid glands. There is no distinct hypobranchial gland, such as is figured by Fretter & Graham (1962) for Littorina littorea, which serves in producing a copious supply of mucus. There are, however, posterior to the gills at the posterior mantle cavity, under the area hidden by the spermathecal duct (Sd, Pl. 4), numerous, individual, large, spheroidal, glandular units within which one can observe, under the compound microscope, numerous tiny granules all in high agitation. Mucus is liberated upon disrupting these units. In many cases these glandular units are so thick that they appear coalesced into a large glandular sheet covering the epithelium of the intestine and right wall of the mantle within the posterior recess of the cavity. Posterior Mantle Cavity. The mantle cavity narrows posteriorly and its ter- minal epithelium is appressed against 2 organs, the pericardium and the kidney. In Plate 11, Fig. 1, the pericardium (Pe) appears at the posterior end of the “neck.” The opening of the kidney (not shown) is, in this figure, above the pericardium. In Plate 8, Fig. 1, one can see that these organs are ap- pressed against the dorsal surface of the body tube and that the pericardium lies on the left dorsolateral curvature while the anterior kidney wall arises 28 G. M. DAVIS from the right dorsolateral curvature. The wall of the kidney abuts on the pericardium. The opening of the kidney (Or) is a slit-like aperture bounded by a pair of “whitish tumid lips” (Dundee, 1957). The lips are provided with a sphinctor muscle. 4. Female Reproductive System. Dundee (1957) provides a useful table of terms used by various authors for the different organs of the female re- productive system as found in proso- branch snails. Further comparative material is found in Fretter & Graham (1962) where the schematic diagrams of reproductive systems from various prosobranch genera are presented. In the overall scheme, oocytes pass from the gonad and travel along the oviduct past the entrance of the seminal re- ceptacle and sperm duct to enter the posterior end of the pallial oviduct. In theory the eggs travel down the pallial oviduct to emerge over the omniphoric groove. Actually, no one has recorded the passage of an egg through the pallial oviduct. Sperm enter the spermathecal duct which opens into the mantle cavity near the anterior end of the pallial oviduct. They traveltothe bursa copula- trix or pass into the sperm duct at the entrance of the bursa and move into the oviduct and then into the seminal receptacle. The spermathecal duct and the pallial oviduct are not fused, but separate structures. Gonad (Pls. 5, 10). The female gonad is a plastic tubular structure about 1.2- 1.5 mm long and 0.57-0.59 mm wide. It is characterized by few branches, the style of branching being plastic and variable. The outer epithelium was removed from one of the ovaries (Pl. 10) to demonstrate how the gonad can be gorged with oocytes. The epithelium of the gonad is often extremely stretched by gonadal products. The anterior edge of the ovary is generally not further than 0.2-0.3 mm from the posterior edge of the stomach. The oviduct runs an- teriorly from the gonad as shown (Pl. 5) and crosses the edge of the stomach at a point beneath which the digestive gland opens into the stomach. Passing over the posterior portion of the esophagus (Es) the oviduct runs below the left ventro- lateral edge of the kidney and seems to disappear at the posterior end of the mantle cavity. Up to this point, all along the oviduct, oocytes can be fre- quently observed characteristically squeezed and flattened into elongate spheres (Pl. 5, Fig. 2; Pl. 7, Fig. 2). Along this length (1.4-1.8 mm) the oviduct is about 0.06-0.10 mm wide. The mid- region of the body, the anterior portion of which seemingly engulfs the oviduct, is complex in its interrelationships of organs. The juxtaposition of kidney, pericardium, nerves, reproductive tubes and connective tissue layers is of such a complex nature that some space is devoted to describing this region. Mid-Region of the Body (Pls. 5, 7, 8). The posterior end of the mantle cavity (x, Pl. 5) is readily defined by a marked crease where the mantle cavity wall folds dorsally along with the distinctive kidney tissue of the ventral anterior arm of that organ. At the edge of the pallial oviduct between the bursa copula- trix (B) and the mantle cavity (Mc) one observes a thick layer of connective tissue which occludes the posterior por- tion of the spermathecal duct (Sd). This tissue is generally full of large white granules; it runs as a sheet to the left ventral curvature of the body tube and folds under the area traversed by the esophagus (Es). It is into this con- nective tissue that the oviduct turns dor- sally just at the edge of the mantle cavity. The ventral surface of the kidney on the right side stretches between the edge of the anterior portion of the stomach and the posterior end of the pallial oviduct. It surrounds the pos- terior end of:the bursa copulatrix (В) and sends an arm anteriorly betweenthe left edge of the bursa and the oviduct. This ventral anterior arm, like the oviduct, turns dorsally at the posterior edge of the mantle. POMATIOPSIS AND ONCOMELANIA 29 In Plate 7, Fig. 1, the connective tissue Sheet between the bursa and the mantle was peeled away. Staining the living organism with neutral red aided in revealing more clearly the underlying structures. The spermathecal duct (Sd) runs dorsal to the point where the oviduct (Ov2) enters the pallial oviduct (Po). At the junction of the spermathecal duct (Sd) and the bursa copulatrix (B), there arises a tube from the left side, which first turns left, then turns to the right over the spermathecal duct, then left again to enter the oviduct. This is the sperm duct (Sdu). As previously described, the kidney (K) is a thin walled, fluid filled sac. Itis molded around and between organs from the posterior edge of the mantle to the anterior chamber of the stomach (Ast, Pl. 5). As the kidney fills the residual space within the bounds described, it is like a second body cavity, the ventral wall of which is shown in Plates 4-7. Opening the ventral wall and peeling it away (Pl. 7, Fig. 2) exposes the cavity of the kidney. The bursa copulatrix (B) is shown pulled slightly outward and rotated about 45° to the right. The posteroventral and all of the dorsal surfaces of the bursa are covered with kidney wall. Pulling the bursa outwards exposes the oviduct lying coiled dorsal to the bursa, likewise wrapped in kidney tissue. The dorsal wall of the kidney (Ko, Pl. 7, Fig. 2) covers the style sac (Sts), which is better shown in Plate 8, Fig. 1. The style sac arises from the anterior chamber of the stomach, runs anteriorly for about 1.44 mm, not quite reaching the end of the mantle cavity. This structure is about 0.96 mm wide at its posterior end. From its left ventro- lateral surface at a point (Oi) about 0.45 mm from the edge of the anterior cham- ber of the stomach, there arises the intestine with a width of 0.36 mm. The intestine runs anteriorly (Inj), swings over the rounded ventral and anterior tip of the style sac and turns dorso- posteriorly, still appressed against the style sac. The intestine then makes a sharp turn swinging anteroventrally again (Ing). The fecal pellet com- pressor is located in this sharp turn. The dorsal wall of the kidney is ap- pressed and molded around these struc- tures. A deep crevice is formed to the right of the point where the intestine turns dorsoposteriorly over the style sac. The crevice runs down towards the dorsal surface (Cr, Pl. 7, Fig. 2). This deepened portion of the kidney ex- tends anteriorly up to the anterior wall of the kidney abutting on the rear of the mantle cavity. In Plate 8, Fig. 1 the kidney tissue was cleared from the oviduct (Ovj, Ova), the pallial oviduct (Po) was cut anterior to the point where the oviduct (Ova) entered it and the posterior portion of the pallial oviduct with bursa copulatrix (В) and tubes was lifted up and outward, thereby exposing structures otherwise hidden by that complex. In that same figure the ventral wall of the mantle cavity was removed. The anterior wall of the kidney is shown abutting onthe epithelium of the posterior mantle cavity (W). The left edge of the anterior wall abuts on the pericardium (Pe), the right edge abuts on the intestine (Ing). The opening of the kidney (Or) is shown on the right side of the pericardium. The pericardium lies anterior to the style sac (Sts) and pushes out into the mantle cavity (Dmc). The auricle (Au) is shown and the vessel (V) whichbrings blood from the ctenidium to the auricle. Just posterolateral to the point where the auricle joins the ventricle a thin tube, the gonopericardial duct (Gp), con- nects the pericardium with the oviduct (Ov1). The gonopericardial duct has not been previously mentioned for Pomatiopsis. It arises from the dorsal surface of the oviduct where the latter turns into the body tube under the antero- ventral arm of the kidney (Pls. 7, 8). This area is generally occluded by pigmented connective tissue and the gonadal nerve (Gn, Pl. 7, Figs. 1, 2) which runs posteriorly over the oviduct at this point and is bound to the latter 30 BIG aw FIG. 2. G. M. DAVIS PLATE 7. Mid-body region of Pomatiopsis lapidaria. The ventral surface of the kidney (K) is shown. Connective tissue was removed to re- veal the point where the sperm duct and spermathecal duct connect and their relation- ship to the bursa copulatrix (B). The ventral wall of the kidney was slit open to reveal the cavity of the kidney. The bursa copulatrix (B) was pulled out of the cavity to show how it and the coiled oviduct (Ov) were wrapped in kidney tissue. B Cl Cr bursa copulatrix columellar muscle the deep crevice between the right edge of the style sac and the anteroventrally running intestine. This is the deepest portion of the kidney. external mantle cavity nerve 3 esophagus fecal pellet gonadal nerve white granules found in epithelium intestine kidney ventral surface of the kidney portion of kidney wall adjacent to the style sac cut edge of kidney wall oocyte coiled portion of the oviduct oviduct posterior to the gonopericardial duct portion of the oviduct entering the pallial oviduct pallial oviduct subvisceral connective spermathecal duct sperm duct stomach supravisceral connective visceral ganglion POMATIOPSIS AND ONCOMELANIA PACE 2 AAA Rn 31 32 FIG. FIG. FIG. FIG. FIG. G. M. DAVIS PLATE 8. Female reproductive system of Pomatiopsis lapidaria. The ventral wall of the kidney was removed as well as the ventral wall of the mantle cavity. All kidney tissue was removed from the reproductive structures. The pallial oviduct and spermathecal duct were cut and the posterior portion of the reproductive system was lifted from the body tube to show underlying structures. The bursa copulatrix with the left lateral “crest” showing, and the oviduct between the bursa and gonopericardial duct pulled out like a spring to show the nature of coiling. The ventral surface of the bursa copulatrix exposed with the oviduct pulled out as in iZ The relationship of the seminal receptacle (Sr)to the bursa copulatrix is shown as well as the coiled portion of the oviduct. Note the spatial relationship between the gono- pericardial duct (Gp) and the seminal receptacle (Sr). The relationship of the sperm duct and bursa copulatrix showing the position of the opening of the seminal receptacle (Osr) into the oviduct. Au auricle B bursa copulatrix Cl columellar muscle Dmc dorsal wall of the mantle cavity Es esophagus Gp gonopericardial duct In; portion of the intestine circling over the tip of the style sac In intestine anteroventral to the pellet compressor Ing intestine running along side the pallial oviduct Ke cut edge of the kidney wall Oi the point where the intestine arises from the posterior por- tion of the style sac Or opening of the kidney into the posterior mantle cavity Osr opening of the seminal receptacle into the oviduct Ov coiled portion of oviduct Ov, oviduct posterior to the gonopericardial duct Ov, portion of the oviduct entering the pallial oviduct Pe _— pericardium Po pallial oviduct Sd spermathecal duct Sdu sperm duct sr seminal receptacle Sts style sac (here partly covered by a remnant of the dorsal wall of the kidney) V vein draining the ctenidium and leading to the auricle W posterior wall of the mantle cavity abutting on the kidney POMATIOPSIS AND ONCOMELANIA 33 34 G. M. DAVIS PLATE 9. Female reproductive system of Pomatiopsis lapidaria. FIGS. 1,2. Different views and variations of the bursa copulatrix and associated ducts. FIG. FIG. FIG. FIG. FIG. 3. 4. Variation in the seminal receptacle. The pallial oviduct and the spermathecal duct oriented to show the opening of the spermathecal duct and the dense connective tissue sheets binding the spermathecal duct to the pallial oviduct. Terminal portion of pallial oviduct and the spermathecal dict viewed as in Plate 4, but with connective tissues not removed and in greater detail. The pallial oviduct oriented to show the female orifice as well as the connective tissue “tubes” encircling the lips and running into the mantle tissue. The pallial oviduct and spermathecal duct oriented to show the opening of the sper- mathecal duct. POMATIOPSIS AND ONCOMELANIA 35 bursa copulatrix connective tissue sheets anterior end of the spermathecal duct occluded by heavy strands of the con- nective tissue opening of the pallial oviduct opening of the spermathecal duct opening of the seminal receptacle oviduct Sdu Vas 0.5mm portion of the oviduct passing ventral to the spermathecal duct to enter the pallial oviduct pallial oviduct spermathecal duct sperm duct seminal receptacle vascular channels inthe connective tissue sheets 36 G. M. DAVIS by tenacious connective tissue. Lifting the bursa and pallial oviduct and displacing them (Pl. 8, Fig. 1) re- veals the coiled nature of the oviduct between the gonopericardial duct andthe bursa copulatrix. In this plate the re- lationship of the seminal receptacle (Sr) to the bursa copulatrix is shown and the point where the sperm duct (Sdu) enters the oviduct. The coiled portion of the oviduct readily fits into the space between the end of the style sac and the posterior slope of the pericardium. As can be seen in Plate 5, Fig. 2, all of the coils of the oviduct are packed dorso- laterally to the point where the sperma- thecal duct (Sd) enters the bursa copula- 1х (В). Gonopericardial Duct to Pallial Oviduct (Pls. 5, 7-9). Gonopericardial Region and Coiled Oviduct. The oviduct narrows just pos- terior to the gonopericardial duct (Pl. 8, Fig. 4). Oocytes have not been seen past this point althoughthey may be found lined up, one behind the other, pos- teriorly, right back to the gonad. The small section of the oviduct from which the gonopericardial duct arises is dis- tinct in that it is characteristically swollen (Pl. 5, Fig. 2); it is about 0.24 mm long and 0.19 mm wide. The duct penetrates connective tissue layers to enter the pericardium and is open at both ends. It is about 0.096 mm long and 0.048 mm wide. The connective tissues occluding the reproductive tract between the bursa copulatrix (B) and the oviduct (Ov) shown in Plate 5, Fig. 1, were removed. The exposed structures (Pl. 5, Fig. 2) are presented as ob- served, with one exception. The gono- pericardial duct arises more dorsally than is shown and would, therefore, be barely visible. The coiled portion of the oviduct forms a very compact cylinder some 0.31 mm in diameter and 0.48-0.36 mm in length depending upon whether there are 4 or 3 coils in the tube. The tube in the coil is up to 0.17 mm wide. Uncoiled, the length of the oviduct between the gono- pericardial duct and the point of entry of the seminal receptacle into the oviduct is about 2.0 mm. Coiling may be regular or with irregular twists. In Plate 8, Figs. 1-4, the oviduct has been slightly stretched out as one would stretch a spring to demonstrate the nature of the coils and twists commonly found in that section. Seminal Receptacle. The seminal re- ceptacle (Sr) is not observable from the ventral surface (Pl. 5, Fig. 2; Pl. 8, Fig. 5; Pl. 9, Fig. 2) although the point where it enters the oviduct may be seen (Pl. 8, Fig. 5). This small, spherical, sac-like organ is bound to the anterior dorsal surface of the bursa copulatrix by a connective tissue sheath in which numerous white granules are often densely imbedded (Pl. 8, Fig. 4). The Seminal receptacle does not com- municate directly with the bursa copula- trix. By slowly rotating the bursa from its normal position (Pl. 5, Figs. 1, 2), like turning the page in a book, one gradually exposes the seminal receptacle (Sequence in Pl. 8: Figs. 5, 3, 2, 4; but with different specimens). The shape of the seminal receptacle varies greatly depending upon the extent to which it is gorged with sperm and fluid. It may be elliptical or circular with gradations between (Sr, Pl. 8, Figs. 4, 1, respectively). It often appears to have a dense hard core (Fig. 1). The duct leading to the oviduct varies in length (Pl. 8, Figs. 1-4; Pl. 9, Fig. 3) from 0.12 to 0.25 mm, while the width varies from 0.06 to 0.15mm. The longer Slender duct is most commonly en- countered. The spherical portion of the organ varies in length from 0.20 to 0.24 mm and the width from 0.17 to 0.24mm. The duct enters the oviduct about 0.48 mm from the opening of the sperm duct (Pl. 8, Figs. 21-3, 5; © POSER SER Sperm Duct. The sperm duct (Sdu, Pl. 8, Figs. 1-3, 5) connects the sperma- thecal duct and the oviduct. It may arise abruptly from the spermathecal duct (Sd) at its juncture with the bursa copulatrix (B, Pl. 9, Figs. 1, 2) or out along the POMATIOPSIS AND ONCOMELANIA 37 PLATE 10. Variations in the gonad of female Pomatiopsis lapidaria. The ovarian follicles are gorged with oocytes. The membrane has been pulled away in the upper right-hand drawing to show the nature of the oocytes. 38 G. M. DAVIS spermathecal duct as far as 0.5 mm from the anterior portion of the bursa (Pl. 8, Fig. 2). It varies in length from 1.2 to 0.4 mm, but is, on the average and most commonly, 0.70 mm long. The width varies between 0.14 and 0.09 mm. The degree of con- volution of the sperm duct (Sdu) varies between the sinuous condition shown in Plate 8, Fig. 1, and the almost straight (Pl. 5, Fig. 2), the latter condition being rather rare. The oviduct, beyond the entry of the sperm duct (Ova, Pl. 8, Fig. 1), is short, some 0.17 mm long and 0.12 mm wide. It passes into the pallial oviduct (Po) 1.0-1.7 mm from the posterior end of the latter, not in the mid-length of that organ as stated by Dundee (1957). This point is 0.65+0.16 mm from the pos- terior end of the bursa copulatrix (B) and is covered by the sheet of con- nective tissue discussed above (Pl. 5, Fig. 1). The oviduct (Ova) passes ventral to the spermathecal duct (Sd) and does not communicate with it, although both are closely bound together by connective tissue. Bursa Copulatrix. The bursa copula- trix (B), as viewed from the ventral side, lies over the anterior tip of the style sac (Sts) with part of the curvature often within the kidney space onthe right of the style sac (Pl. 5, Fig. 1; Pl. 7, Figs. 1, 2; Pl. 8, Fig. 1). As shown in Plate 5, Fig. 2, the ventral surface is evenly rounded, the medio-lateral sur- face is not rounded but narrows to a crest (21.48, Figs: 1, "РЕ 9, Fig” 1) This organ is characteristically ap- pressed against the posterior end of the pallial oviduct as shownin Plate 8, Fig. 1. At times the end of the bursa projects beyond the edge of the pallial oviduct (Pl. 7, Fig. 1) but this is rare. In another variation the tip of the pallial oviduct swings slightly away from the bursa, (Pl. 5 М. 1). "Theibursarhas never been found posterior to the tip of the pallial oviduct as shown by Dundee (1957, Pl. 11). The general positions of the bursa and the pallial oviduct in relationship to the organs in the mid- body region shown in Plates 5, 7 and 8 were found to be invariable. The length of the bursa along its ventral surface averaged 0.72 + 0.14 mm and the width averaged 0.48 + 0,03 mm. Pallial Oviduct and Spermathecal Duct (PlS 305 7, 870) Spermathecal duct. The spermathecal duct (Sd) arises from the anterior tip of the bursa copulatrix (B) as a tube 0.096- 0.168 mm wide. It runs anteriorly, closely appressed to the ventromedial edge of the pallial oviduct (Po) and passes dorsal to the oviduct (Ova, Pl. 8, Fig. 1) where the latter enters the pallial oviduct. Passing over the end of the mantle cavity, the spermathecal duct narrows to 0.07-0.10 mm and runs separately at a distance of 0.02-0.07 mm from the pallial oviduct. As shown in Plate 5, the spermathecal duct (Sd) is readily observed following the curve of the pallial oviduct (Po) imbedded within the ventral superficial tissues of the mantle cavity wall. Anteriorly, with the sharp curve of the mantle cavity and pal- lial oviduct, the spermathecal duct turns towards the pallial oviduct, becomes more slender, and disappears under the ventromedial edge of the pallial oviduct. The area where the spermathecal duct disappears from view is about 0.48 mm from the point where the pallial oviduct is observed to disappear withinthe man- tle cavity (Pl. 5, Fig. 1). The gross aspects of the relation- ship between the termination of the spermathecal duct and pallial oviduct within the mantle cavity are shown in Plate 4. It would seem, from this gross aspect, that the spermathecal duct enters the tip of the pallial oviduct. About 0.7 mm posterior to the opening of the pallial oviduct (Opo) the spermathecal duct becomes closely bound to the pallial ovi- duct by sheets of connective tissue; however, the opening of the spermathecal duct is not evident. A number of whole mount slides were prepared of the anterior 1.0 mm portion of the sperma- thecal duct, pallial oviduct and intestine, POMATIOPSIS AND ONCOMELANIA 39 using CMC-10 (Michelson, 1960). Water mounts were also made. Under the compound microscope it was found that the spermathecal duct did not open into the pallial oviduct as stated by Dundee (1957), but that the stout connective tissue layers continuing to the tip of the pallial oviduct obscured the point where the spermathecal duct does open into the mantle cavity. In Plate 9, Fig. 5, the structures under discussion are shown as they are oriented in Plate 4. In Plate 4 the opening of the pallial oviduct (Opo) has been indicated only to demonstrate the point where it opens. In reality what would be seen is the lip- like tip (outer lips) of the pallial ovi- duct (Pl. 9, Fig. 5) appressed to the inner mantle wall by muscular con- traction, thereby sealing off the opening of the pallial oviduct. Both lips around the opening (Pl. 9, Fig. 6) are muscular, and thickened by connective tissue strands which encircle the edges of the lips. The connective tissue anterior to the opening of the spermathecal duct is heavily pigmented, full of whitish gran- ules, and forms the tubular channel (Vas) for the flow of blood to the tip of the pallial oviduct and around the lips (Pl. 9, Fig. 4). This vascular tube is thick and so dense that one can perceive only with the greatest difficulty where the spermathecal duct ends and the con- nective tissue tube begins (Pl. 9, Fig. 5). The vascular channels running down each side of the lips of the pallial oviduct end in sinuses within the mantle edge. By rotating the connected sperma- thecal duct and pallial oviduct shown in Plate 4, and Plate 9, Fig. 5, as turning a page in a book from left to right, one can observe with the aid of a bright light, that the spermathecal duct ter- minates about 0.58 mm from the tip of the pallial oviduct. The opening of the spermathecal duct (Osd) is permanent (Pl. 9, Figs. 4,7) but, due to the rather thin edge of the tube at this point, it is not distinct. Note that the opening of the spermathecal duct is not oriented in the same direction as that of the pallial ovi- duct but rotated 90°. In the intact ani- mal, where the mantle is in its normal position, the opening of the pallial oviduct is appressed against the right ventrolateral mantle wall, while the Spermathecal duct opens towards the “neck.” The opening of the spermathecal duct is about 71u long and 25u wide. In some specimens it appears slit-like measuring about 60 x 12y; in others it was round with a diameter of 35y. In Plate 9, Fig. 6, the tips of the organs are rotated still further, exposing the thickened lips of the opening of the pallial oviduct (Opo) but causing the opening of the sperma- thecal duct to disappear from sight. Pallial Oviduct. The pallial oviduct is the largest organ of the body witha length of 4.3-5.0 mm and a width of 0.72 mm towards the posterior regions. In the living animal the organ appears divisible into 2 parts of equal length; the posterior half appears greyish, thick and quite glandular, and the anterior portion is whitish, narrower and more smooth. These observations correspond to Dundee’s (1957) histological findings. 5. Male Reproductive System A male snail is shown uncoiled in Plate 6, as was the female previously de- scribed. Sperm from the gonad (Go) travel via the vas deferens (Уа1) to the prostate (Pr), enter the prostate, and leave through the anterior portion of the vas deferens (Vd2) which leads into the base of the verge, along the length ofthe verge (Ver) to exit at its tip (Pl. 11, Fig. 1). Gonad (Go, Pls. 6, 12). The testis, observed through the ventral epithelium of the digestive gland (P1.6, Fig. 1), does not appear as distinct as did the gonad of the female. The reason for this vagueness is that the lobes of the gonad are small and that it is finely branched. The main collecting duct, the vas deferens, is dorsal to the gonad and therefore hidden from sight until it turns ventrally near the edge of the stomach. Peeling off the ventral epithelium reveals the structure of thetestis (Pl. 12, Fig. 1). 40 FIG. FIG. FIG. FIG. FIG. FIG. G. M. DAVIS PLATE 11. Male reproductive system of Pomatiopsis lapidaria. Head, verge and mantle cavity. The verge is shown uncoiled and partially extended. Head of the gland shown in Fig. 3. Longitudinal view of the glands which pack the area between the vas deferens and the concave curvature of the verge (Gl of Fig. 5). Gland type commonly found near the tip of the verge, appearing dense or black under direct illumination. They are oriented with the circular muscle fibers (Gl, of Fig. 5). Verge showing the vas deferens and glandular areas. Gland type (Glo) clustered about the vas deferens. anus ctenidium fecal pellet glandular units around the dorsomedial surface of the eye glandular types shown in Figs. 2, 3 gland type shown in Figs. 4, 5 gland type shown in Figs. 5, 6 cut edge of the mantle thick layers of circular muscles encircling the vas deferens at the base of the verge osphradial nerve osphradial ganglion osphradial pit pericardium prostate rostrum subvisceral connective supravisceral connective blood vessel at the base of the ctenidium anterior portion of the vas deferens verge visceral ganglion POMATIOPSIS AND ONCOMELANIA AR RAA PA UV? ds но MAUR: Ay a. KA 41 42 G. M. DAVIS PLATE 12. Male reproductive system of Pomatiopsis lapidaria. FIG. 1. Gonad showing 9 multibranched units and the coiled “seminal vesicle” behind. FIG. 2. A single multibranched unit arising from the vas efferens. FIGS.3,4,5. Variation in the coiling of the “seminal vesicle. ” FIG. 6. An enlarged view of a single multibranched unit supporting testicular lobes. FIG. 7. Variation in the structure of the testicular lobes seen laterally. FIG. 8. A multibranched unit showing testicular lobes and how the lobes are vascularized. Ar artery to the testicular lobes Br branch arising from the vas efferens O] point where the vas deferens arises from the vas efferens and leads to the “seminal vesicle” Sv seminal vesicle E testicle lobe Ve vas efferens 43 POMATIOPSIS AND ONCOMELANIA 44 G. M. DAVIS With a length of about 2.4-2.8 mm, it is longer than the female gonad. The width of the gonad at its widest part, whichis usually in its anterior third, is about 0.7 mm. From a narrow collecting duct (0.05-0.09 mm wide), the vas efferens (Ve), there arise 8-9 multi- branched units, whose nature is shown in Plate 12, Figs. 2, 6 and 8. Atthe end of each branch are lobes of varying size. The lengths and widths of the lobes depend upon the degree to which they are filled with gonadal products. A single tubular branch may support several interconnecting lobes (Fig. 8). In that figure vascularization of 3 groups of testicular lobes is shown. The whole gonad appears very bright yellow when all of the testicular lobes are productive. Posterior Vas Deferens (Pls. 6, 12). A narrow tube (0.048 mm diameter), which is extremely convoluted in most cases, arises just anterior to the mid- length of the vas efferens (Ve). This is the beginning of the section of the pos- terior vas deferens called the “seminal vesicle” by Dundee (1957). As shown in Plate 12, Fig. 1, the tube arises ata place (01) which would otherwise be filled by a multibranched unit. The “Seminal vesicle” is coiled in a characteristic manner (Sv, Pl. 12, Figs. 3, 4,5). The initial slender convoluted portion thick- ens within 0.6 mm of its origin, forming a series of coils which are regular like those of a spring when the tube is not overloaded with gonadal products (Pl. 12, Figs. 3, 4), but bulge out of position when fully loaded, so as to disrupt the neatly coiled pattern (P1.12, Fig. 5). Thelength of the coil is 1.4-1.9 mm; its width about the same as that of the gonad. Fully uncoiled, the tubing making up the “seminal vesicle” measures up to 6.0 mm in length. The tube may have a diameter of 0.24 mm. At the anterior end of the testis the “seminal vesicle” narrows to about 0.07 mm and turns ventrally to emerge from the dorsal side of the gonad, as shown in Plate 6. The vas deferens (Vdy), as the oviduct, takes an anterior course toward the rear of the mantle cavity. At the place where the oviduct turned dorsally, the vas deferens continues along the ventral sur- face across the body tube, posterior to the mantle cavity, without either coiling, connecting withthe pericardium, or great involvement with kidney tissue. The vas deferens moves to the prostate (Pr) to enter that organ just posterior to the place where the anterior portion of the vas deferens (Vds) is seen leaving (Pl. 6, Fig. 1). The tube is difficult to see because it becomes more slender as it approaches the prostate and is sur- rounded in connective tissue. In Plate 6, Fig. 2, the prostate was turned over to reveal the point of entry of the vas deferens (Vd) into its dorsal surface. At this point the vas deferens measures about 0.024 mm in diameter. Prostate (Pls. 6, 11). As seen from the ventral surface, the prostate is a kidney shaped organ, white in color, situated in the same position as the female pallial oviduct but not as long as that organ. The prostate is 1.68- 1.75 mm long and the greatest widthis about 0.70 mm. The posterior end pro- jects about 0.48 mm beyond the mantle cavity and is encircled by kidney tissue, as was the female bursa copulatrix and pallial oviduct. In contrast to the pallial oviduct, the prostate surface is quite irregular, due to discrete glandular units which were readily separated (Pl. 6, Fig. 3). Viewing the opposite side of the prostate, it is evident that the glandular units drain into a narrow tubular collecting tube (Pl. 6, Fig. 2). The posterior portion of the vas deferens (Vd.) enters the col- lecting tube of the prostate 0.60 mm from its posterior end, while the larger an- terior portion of the vas deferens (Vda) leaves the prostate 0.17 mm anterior to the entry of Vd). Anterior Vas Deferens (Pls. 6, 11). The anterior vas deferens (Vds) emerges from the dorsal side of the prostate (Pl. 6) as a tube 0.96-0.12 mm wide. It usually follows along the edge of the prostate until the anterior end of that _ ri a 96 POMATIOPSIS AND ONCOMELANIA 45 organ before running obliquely over the mantle cavity towards the columellar muscle, but occasionally it divergesim- mediately upon leaving the prostate as Shown in Plate 6, Fig. 1. The vas deferens enters the mantle cavity behind the columellar muscle, 1.4 mm fromthe edge of the mantle. Plate 11, Fig. 1, shows the anterior tip of the prostate (Pr) and the vas deferens (Vdo) running anteriorly on the floor of the mantle cavity along the right side of the “neck” to a point slightly anterior to the base of the verge (Ver); it then turns back and enters the “neck” under the base of the verge. Verge (Pl. 11). The verge of Pomati- opsis lapidaria is very characteristic for the species. The verge would be called simple in much of the mala- cological literature, as it has no ap- pendages and since the vas deferens terminates at the tip, which has no papilla. The verge has, however, a number of significant details of im- portance to a systematic discussion. As Dundee (1957) stated, the verge is very flattened; it is relatively thin bladed. It arises from the “neck” to the right of the mid-line and is carried, coiled counter-clock-wise over the “neck.” It measures up to 3.4 mm in length in the extended condition,but it can possibly be expanded further. The animal shown in Plate 11, Fig. 1 is a relatively small snail; the verge shown in Fig. 5 is from a larger male. Abbott (1948a) described the verge as varying considerably, “ranging from a simple, flattened, tapering cylinder to a prong with a meat-chopper blade on one side.” I found the “simple” verge only in immature males. Inadults the verge is characteristically of the Shape shown in Figs. 1 and 5. In very rare cases (less than 1%) the inner curvature near the anterior end has a fleshy lobe (Pel, Pl. 13, Fig. 2) reminis- cent of a penial appendage common in some other hydrobiid snails. The inner margin of the mid-verge is swollen out into a convex curve which appears scal- loped. The scalloped effect (Pl. 11, Fig. 5) is due to the rounded ends of blunt, finger-like, lobes with parallel sides. The verge was cut from the body deep at its base so as to include the vas deferens where it entered. This organ was Studied in a drop of saline, gently placing a coverslip on it, and viewing it through the compound micro- scope. The vas deferens (Vd,) runs along the base of the verge before turning up into the mid-portion of that organ. The basal portion of the vas deferens is greatly thickened because of a pronounced layer of dense circular muscles (Mvd, Fig. 5). Beyond the base, the vas deferens takes an anterior course either along the mid-line of the verge or Slightly displaced towards the outer curvature. The tube, loosly un- dulating, begins to narrow noticeably about mid-verge. At the tip the vas is only about 0.024 mm wide. The area of the verge between the crenulated edge and the vas deferens is highly glandular with ducts and fissures running beneath the epithelium, along the edges of the “lobes,” towards the edge of the verge. At150X distinct glandular units were noted pushing against the epithelium so that the whole area seemed pustulate. The glands (Gl, Pl. 11, Figs. 2, 3, 5) were club-shaped, the diameter of the heads (Fig. 2) measuring 10-304. The head of such a gland appears full of vesicles (8-20 per head) measuring 2-10u in diameter. The length of the glands was about 160y; the basal portions were without vesicles (Fig. 3). The entire area was full of these glands, the whole length of which could often be seen lying under the epithelium. Two other glandular types can be found in the verge. Along the vas deferens are numerous crowded groups of vesicles (Glo, Pl. 11, Figs. 5, 6). Also near the anterior end of the verge one Sometimes finds dense spheres of minute vesicles which look like black spots under the dissecting microscope (С11, Pl. 11, Figs. 4, 5). The longitudinal axes of these dense spheres are oriented lengthwise 46 PLATE FIG. FIG. FIG. FIG. FIG. a À wo G. M. DAVIS 13. Structures of the muscular, respiratory, reproductive, digestive and nervous systems in Oncomelania and Pomatiopsis. Muscles arising in the pedal haemocoel at a level below the pedal ganglion. The draw- ing was made from O. hupensis formosana and shows a cross section through the pedal haemocoel. The pedal ganglion on the right is transected. The right propodial and metapodial ganglia are shown. The main structures to be pointed out are the mid- ventral protractors arising from the anterior wall of the pedal haemocoel. Verge of P. lapidaria showing a rarely found penial lobe (Pel). The structure of 2 gill filaments from P. lapidaria. Variation in the salivary glands of P. lapidaria. Variation in the nerves associated with the buccal ganglion. a anterior wall of pedal haemocoel by dorsal buccal nerve bo esophageal nerve bg central buccal nerve ba anterior buccal nerve bs odontophoral nerve bg posterior buccal nerve Be buccal commissure Bg buccal ganglion Cb cerebro-buccal connective Cl columellar muscle Mg metapodial ganglion m27 mid-ventral protractor Pel rare penial lobe Pg pedal ganglion Prg propodial ganglion POMATIOPSIS AND ONCOMELANIA Cb № 0-33mm 47 48 G. M. DAVIS PLATE 14. Dorsal buccal mass and associated organs in Pomatiopsis lapidaria. Bg buccal ganglion Bm buccal mass Cc cerebral commissure Cg cerebral ganglion Cl columellar muscle Es esophagus ms buccal protractor muscle Mg preventral dilator muscles m,7 suspensors of the buccal mass moo lateral cephalic retractor muscle Ml, median labial nerve 1 М5 median labial nerve 2 Mn,* mantle nerve 2 Оп* osphradial nerve Opt optic nerve ррз lateral nerve 3 Psc pleuro-supraesophageal connective Sa salivary gland SI supralabial nerve Sug supraesophageal ganglion Suv supravisceral connective Tn tentacular nerve * As a rule a single large trunk, the joint osphradio-mantle nerve, emerges from the supra- esophageal ganglion, which then bifurcates to form the osphradial nerve (On)and mantle nerve 2 (Mng) (see p 79). 49 POMATIOPSIS AND ONCOMELANIA 50 PLATE 15. FIG. FIG. G. M. DAVIS Muscular, lapidaria. nervous 1. The lateral aspect of the buccal mass with associated neural structures. 2. The posterodorsal portion of the pedal haemocoel showing the relationship of the colu- mid-columellar supportive (mag), and pedal ganglia in anterior mellar muscle (Cl), view (compare with Pl. 20, Fig. 2). nerve characteristically arising from P6 nerve characteristically arising from Pg ventral to “a” dorsal buccal nerve esophageal nerve central buccal nerve anterior buccal nerve buccal ganglion cerebro-buccal connective cerebral ganglion columellar muscle cerebro-pedal connective cerebro-tensor nerve external odontophore membrane esophagus buccal protractor muscle preventral protractors anterior jugalis buccal constrictor dorsolateral buccal protractor buccal retractor membranous jugalis preventral dilators suspensors of the buccal mass rostral retractors lateral cephalic retractors mid-columellar supportive and digestive systems in the cephalic region of Pomatiopsis dorsal pedal tensor dorsal propodial retractor metapodial ganglion medial labial nerve 1 median labial nerve 2 optic nerve lateral retractor nerve pedal nerve to the anteroventral wall of pedal haemocoel major lateral nerve propodial connective metapodial connective dorsolateral pedal nerve pedal ganglion pleuropedal connective lateral nerve 1 penial nerve lateral nerve 3 lateral nerve 4 propodial ganglion right pleural ganglion salivary gland supralabial nerve statolith statocyst sublabial nerve tentacular nerve POMATIOPSIS AND ONCOMELANIA D À 23 <>» A АС Se à LS 51 52 а. М. DAVIS PLATE 16. Musculature and pharyngeal structures in Pomatiopsis lapidaria. FIG. 1. Posterior portion of the pharyngeal tube showing the dorsal aspect of the odontophore with the radula removed. The odontophore divaricator muscle (mg) is not shown as it is vaguely defined when the dorsal surface of the odontophore is viewed, due to its ventro-lateral position (fig. 4). FIG. 2. Left buccal cartilage in dorsal view wrapped in intrinsic muscles. Muscle my is not shown. FIG. 3. Lateral aspect of the left buccal cartilage showing muscles arising or inserting on the cartilage. The position of the cartilage is as in Fig. 4. FIG. 4. Schematic representation of the odontophore within the posterior portion of the pharyn- geal tube. The left side of the odontophore is shown. FIG. 5. The fused cup-like muscles making up the mediolateral cartilage tensor (mg) and the subradular membrane which continues from this muscle. The medial radular retractor (m4) arises from the postero-ventral curvature of this muscle. a a thickened portion of the subradular 117 Suspensor of buccal mass membrane 1120 rostral retractor Bca _ buccal cavity mo oral sphinctor Bf floor of pharyngeal tube Mo mouth Bpl bending plane of the radula (=tip of OI outer lip odontophore) Ot oral tube Ca cartilage Ra radula Es esophagus Ras radular shield Ev esophageal valve Rb pharyngeal roof, or roof of the buccal Fg food groove mass т] lateral cartilage tensor Rf roof of the rostrum mo mediolateral cartilage tensor Rfl floor of the rostrum mg odontophore divaricator Rpe rostral portion of cephalic haemocoel m4 medial radular retractor Rs radular sac ms buccal protractor Sp sublingual space mg preventral protractor Sur subradular membrane 115 sSuspensor of radular sac Vf ventral fold POMATIOPSIS AND ONCOMELANIA 53 54 FIG. FIG. FIG. FIG. FIG: G. M. DAVIS PLATE 17. Buccal and rostral musculature of Pomatiopsis lapidaria. Muscles arising from the mid-columellar supportive muscle (тоз). Dorsal view of the odontophore before removing the radula. Dorsal view of the odontophore with the radula removed and the medial radular retrac- tor pulled backwards from the odontcphore. Dorsal aspect of the opened pharyngeal tube showing the odontophore. The cartilages are shown separated and folded out to show the medial surface of the cartilages wrapped in the medio-lateral cartilage tensor muscle. thickened anterior portion of the subradular membrane bending plane of the radula (=tip of the odontophore) cartilage collostyle tip esophagus esophageal valve food groove central groove in the ventral fold inner lip jaw lateral cartilage tensor mediolateral cartilage tensor odontophore divaricator medial radular retractor radular protractor buccal retractor rostral retractor mid-columellar supportive outer lip radula radular sac rostral portion of the cephalic haemocoel rostral wall subradular membrane ventral fold 55 POMATIOPSIS AND ONCOMELANIA 56 а. М. DAVIS with the circular muscle fibers encasing the verge. Anteriorly, the verge has a charac- teristic indention on the inner curvature (Pl. 11, Figs. 1, 5). The edge of the verge appears to have no cilia when viewed at 600X. 6. Buccal Mass In Plate 14 the rostrum is shown opened along the mid-dorsal line. The epithelium of the “neck” was slit open revealing the wide, powerful columellar muscle (Cl). Structures are shown as they appeared upon opening this area with 2 exceptions. First the salivary glands (Sa) are usually folded and tucked ventro- laterally around and about the cerebral commissure (Cc), esophagus (Es) and cerebral ganglia (Cg) in a variable manner (Pl. 15, Fig. 1). They were pulled upward and straight back to per- mit at least a limited view of the re- lationships of the buccal mass (Bm) with the dorsal area above the cerebral ganglia (Cg), esophagus (Es) and under- lying musculature. Second, the area dorsal to the cerebral ganglia (Cg), eso- phagus (Es) and pleuro-supraesophageal connective (Psc) is hidden by trans- verse strands of connective tissue just beneath the roof of the cephalic area. These strands support the cephalic aorta which travels along the esophagus dor- sally, crossing from left to right, to descend, near the right cerebral ganglion, into the pedal haemocoel. These structures were removed to per- mit observation of the underlying organs. Dorsally viewed the buccal mass (Bm) is somewhat pyriform. Inthe contracted state it varies in length from 0.85-1.08 mm; in width from 0.79-0.84 mm. It fills the rostral portion of the cephalic haemocoel. The posterior end turns sharply ventrally at the beginning of the esophagus, the mid-posterior curve of which is pressed against the postero- ventral part of the buccal mass (Pl. 15) by the cerebral commissure (Cc). The position of the cerebral commissure cor- responds to the mid-dorsoventral axis of the posterior buccal mass. The esophagus again turns dorsally, under the pleuro-supraesophageal connective (Psc) and swings left to follow the left edge of the columellar muscle (Cl) posteriorly, lying on that muscle. The esophagus is a large tube 0.31-0.24 mm in diameter. Muscles arising from the dorsal and dorsolateral surface of the buccal mass (1117) are irregularly placed and serve to suspend the buccal mass from the rostral roof. The dorsal posterior por- tion of the buccal mass appears fleshy and laced with only a few muscle fibers. The mid-dorsal portion appears as if 2 fleshy folds were pressed together and bound in position by superficial layers of connective tissue. Each lip-like edge is the dorsal termination of a hemisphere of tissue arising from the lateral and ventrolateral faces of the buccal mass. Studying the lateral aspect of the buccal mass (Pl. 15, Fig. 1) one observes that the anterior, lateral and dorsolateral regions are highly mus- cular (mg, mg, mg). The buccal mass may be considered a composite of 2 main parts; (1) the pharyngeal tube and attendant muscles which swell between the mouth and the esophagus; (2)the odontophore apparatus which projects into the ventral posterior portion of the pharyngeal tube (Pl. 16, Fig. 4). The odontophore is a neatly delineated structure made up of the buccal cartilages, and the musculatures binding together the cartilages as well as the odontophore to the pharyngeal tube, radula, and radular sac. Connected to the posterior dorsal region of the buccal mass are the paired salivary glands which drain into the pharyngeal tube above the odontophore apparatus. Salivary Glands. The salivary glands are characteristically simple, flattened structures (Pl. 14). They are delicate and easily damaged. As shown in Plate 15, Fig. 1, they arise from the dorso- lateral surface of the buccal mass 0.24- 0.36 mm from the mid-dorsal line. The duct leading from the main expanded POMATIOPSIS AND ONCOMELANIA 97 glandular blade varies in length and is closely appressed to the tissue of the buccal mass. The anterior 0.36 mmare generally covered by a sheet of con- nective tissue. The total length of gland and duct may reach 1.37 mm but is generally less, about 1.0 mm. The width of the gland, which varies considerably, averages about 0.20 mm. Variation in the shape of these glands is shown in Plate 13, Fig. 4; Plate 14; Plate 15, Fig. 1. Interior Buccal Mass. Slitting the epithelium between the dorsal lips of the buccal mass and pulling aside the hemispheres of the pharyngeal tube ex- poses the inner regions of the tube and the odontophore apparatus (Pl. 17, Fig. 4). The large triangularly shaped odontophore, upon full contraction of the buccal mass, causes the dorsal lips of the buccal mass to stretch apart (Pl. 14). The odontophore imparts shape to the buccal mass and gives support to the pharyngeal tube and the initial portions of the esophagus. A longitudinal section of the pharyngeal tube is shown in Plate 16, Fig. 4. This figure differs from the other illustrations in that the mouth lies to the left. The unit at the posterior end of the buccal cavity (Bca) isthe odontophore apparatus which comprises the cartilages (not visi- ble) covered by the medial radular re- tractor (m4), subradular membrane (Sur), radular shield (Ras) and radula (Ra). The long axis of the cartilages runs obliquely from posteroventral to anterodorsal. The radula runs along the mid-dorsal surface of the odontophore (Pl. 17, Fig. 4) and bends over the an- terior odontophore tip called the bending plane of the radula (Bpl, Pl. 16, Fig. 4). The radula teeth are formed on a chitinous membrane, the radular shield (Ras) which tightly caps the anterior, anterolateral, and dorsolateral portion of the odontophore (Pl. 16, Fig. 4; Pl. bf; Fig. 4). Posteriorly, the radula emerges from the radular sac (Rs), which bends ven- trally between the cartilages (Pl. 16, Fig. 4) and turns dorsally again at its tip. The ventral aspects of the radular sac and odontophore are shown in Plate 18. There is a fleshy papilla, the collo- style tip (Carriker, 1946a) at the point where the radula leaves the sac (Cos, Pl. 17, Figs. 2, 4). The collostyle is an elongate plug of dense white tissue between the radular teeth and the posterodorsal curvature of the radular sac epithelium. Only the tip protrudes beyond the opening of the radular sac. A sheet of membranes from both pos- terolateral sides of the inner buccal mass meet, coalesce, and form the radular sac. The posterolateral con- tinuation of this sheet thickens noticeably and forms the esophageal valve (Ev, Pl. 16, Figs. 1, 4; Pl. 17, Figs. 2, 4). This structure, quite variable inform between individuals, is extensively discussed by Fretter & Graham (1962). The antero- dorsal edge of the valve is pressed against the radular sac. Together with the collostyle tip it serves to keep materal from dropping into the radular sac or other ventral spaces behind the odontophore. The lateral portions of the valve send membrano-tendonous folds of tissue anteroventrally along either side of the odontophore onthe sides of the pharynx (Pl. 16, Fig. 1; Pl. 17, Fig. 4). These cross over the odonto- phore divaricator muscles (m3) to run into the loosened membranes arising from the pharynx floor in front of the odontophore under the sublingual space (Sp, Pl. 16, Fig. 4). Part of this same membrane system sweeps over the di- varicator muscles to join the radular sac. Dorsally, the esophageal valve may be folded with an internal anterior lip (an- terior portion of Ev., Pl. 17, Fig. 2). The dorsal valvular structure may be folded in various ways (Pl. 16, Fig. 1; Pl. 17, Figs. 2, 4). The esophageal valve also serves as the floor for the initial portion of the esophagus. The mouth is bounded by the outer lips (Ol, Pl. 16, Fig. 4; Pl. 17, Fig. 4). It opens into a short oral tube (Ot) 58 G. M. DAVIS PLATE 18. Musculature ventral to the buccal mass of Pomatiopsis lapidaria. The buccal mass was pulled out of the rostrum (opened along the mid-dorsal line) to expose the muscles and nerve endings underneath. a area where the floor of the rostrum slopes downward as the anterior wall of the pedal haemocoel b posterodorsal gap of the pedal haemocoel bg posterior buccal nerve Be buccal commissure Bg buccal ganglion Cb cerebrobuccal connective EL columellar muscle Eo external odontophore membrane ms buccal protractor mg preventral protractor m7 radular protractor 111 dorsolateral buccal protractor 112 buccal retractor 1120 rostral retractor mg9 lateral cephalic retractor 03 mid-columellar supportive Moy tensor magnus Mb membrane around the radular sac Ml, median labial nerve 1 Mls median labial nerve 2 pı lateral retractor nerve from the pedal ganglion Rs radular sac Sul sublabial nerve 59 POMATIOPSIS AND ONCOMELANIA TEXT FIG. 1. Marginal teethand jaw of Pomatiopsis lapidaria. Variation inthe outer marginals is illustrated in 1-6, that in the inner marginals in 7-9. The jaws are shown in 10. a thickened central core of the peduncle b thin, wing-like expanded portion of the peduncle POMATIOPSIS AND ONCOMELANIA 61 TABLE 6. Radular statistics Feature Radular length (mm) Radular width (mm) Total no. of rows of teeth No. of rows in the formative stage Mean Standard deviation Standard error of the mean x 5 Se which terminates at the 2 lateral out- swellings of the pharynx wall, the inner lips (Il, Pl. 17, Fig. 4), and a pro- nounced transverse ventral fold (Vf). The inner lips bear the paired jaws (J). The ventral fold is centrally grooved (Gr) and is the threshold to the buccal cavity, i.e., the entire space of the pharyngeal tube. The portion of the buccal cavity beneath the anterior tip of the odonto- phore is the sublingual space (Sp, Pl. 16, Fig. 4). The ventral fold is traversed on either side by ventrolateral ciliated grooves, the food grooves (Fg, Pl. 16, Fig. 1; Pl. 17, Fig. 4). They traverse the ventrolateral edge of the pharyngeal tube, pass up and over the large odontophore divaricator muscles, and over the lateral edges of the esophageal valve into the dorsal portion of the anterior section of the esophagus. Jaw. F. C. Baker (1928) showed a small section of the jaw patterned like bricks in a wall. Dundee (1957) stated that the jaw consisted of 25-30 cuticular plates. She derived this number from a longitudinal, histological section ofthe jaw area which included the inner lips. A camera lucida drawing of each jaw is shown in Text Figure 1 (10). The edge of each jaw alone is composed of at least 27 cuticular plates andthe whole jaw of many more, arranged in anarrow sheet. The length of this sheet is 0.13- Pomatiopsis lapidaria (fr. 9 radulae) Oncomelania hupensis formosana (fr. 16 radulae) 0.16 mm and its width varies around 0.06 mm. Radula. The radula is of the typical taenioglossate type, i.e., with numerous rows of teeth, each row consisting of 7 teeth: a central, flanked on either side by a lateral, an inner and an outer marginal. Nine radula ribbons were straightened out on slides and studied. The statistics on radular length, width, total rows of teeth and rows of teeth in the formative state are presented in Table 6. Rows of teeth within the radular sac are in vari- ous stages of formation, those closer to the end of the sac being the least formed. In this posterior area the central may not yet be formed, although the peduncles of the laterals could be counted. The total number of rows was determined by counting all the centrals until they became too indistinct to count and then all the remaining rows of peduncles. Rows of teeth were con- sidered to be in the formative stages unless all the cusps could be discerned on each of the 7 teeth. Each of the 7 teeth showed considerable variation in the number of cusps not only between specimens from a Single population but also along the same radular ribbon: this variability is not one due to wear and tear, which is readily observed for what it is. The formulas for the cusp arrangments have 62 G. M. DAVIS TABLE 7. A general formula for the most common cusp arrangment in Po- matiopsis lapidaria ——— —— Snails* in which ar- Cus rangement occurred Tooth EE in at least 90% of individual teeth % Central (anter. & т 62 basal cusps) Lateral 2-1-2(3) 75 Inner ur a 84 Marginal el Outer Marginal ee a * 50 snails from 3 populations. been considered important throughout systematic molluscan literature. Re- viewing the literature it becomes evi- dent that, on the one hand, this vari- ability is not generally appreciated, and, on the other hand, many who have ob- served variability often discredit the use of cusp arrangement as a major characteristic of either genera or Species. Knowledge of variability is essential and the presence of variation is not a problem if all the classes of variation can be adequately accounted for. It is, therefore, not sufficient to describe 1, 2, or even a dozen radulae, until the full range of variation is documented. Inthe study of this species it was necessary to study 50 radulae from 3 distinct populations in order to adquately encom- pass variation in cusp number (Tables 7 and 8). In addition to cusp number, basic tooth morphology also may vary between taenioglossate radulae of various genera, subfamilies and families. Some of the variation in the morphology of the central tooth has already been discussed above in the section on systematics. In Plate 19, the top row of teeth de- TABLE 8. The various types of cusp arrange- ment for the different teeth among 50 radulae of Pomatiopsis lapi- daria and the percentage of radu- lae showing that arrangement at least once Central Lateral Arrangement Arrangement of cusps: of cusps: anterior basal 3-3 Peal 0 etl, re ot bP 2-1-2 one side 2-1-3 other side Outer Marginal Number of cusps Number of cusps picts the relationship between teeth as seen on the lingual ribbon, i.e., shows tooth folded over tooth. All other teeth are shown disarticulated from the mem- brane and were drawn from various planes and views. Theteethwere chosen to represent variations observed among the centrals and laterals. A series of inner and outer marginals is shown in Text Fig. 1. For an accurate count of POMATIOPSIS AND ONCOMELANIA 63 the cusps on the laterals and marginals it is advisable to fold the teeth back far enough to expose all the cusps (Pl. 19, lateral 5; Text Fig. 1, teeth 2-4, 7-9). Each cusp is composed of a thickened, rounded supporting piece (Sup) anda thin, blade-like cutting edge (Cu). In radular preparations the cutting edges are often torn away leaving only the supports (Text Fig. 1, tooth 6). Where the cutting edge is drawn, the edge of the underlying support is represented by a dashed line (Pl. 19; Text Fig. 1). The body of the central tooth supports both anterior and basal (posterior) cusps. The sup- ports for the basal cusps of the central have been discussed in the section of systematics (р 13). In Plate 19, central 6, note the characteristic shape of the posterior edge of the support (Slb). It can be discerned by focusing down through the large medial basal cusps. In central 6, the large cusps are repre- sented only as dashed lines to make the shape of the supports clear. The plane of focus on centrals 3-5 is on the upper surface of the basal cusps so that the underlying supports are not evi- dent. The supports for the large medial basal cusps give the face (Fa) of the tooth a square appearance. There is no thickened tongue-like pro- jection from the face ofthe central tooth, as figured for some hydrobiids (e.g., by Berry, 1943, for “Amnicola integra”, Pl. 3, Fig. 4). As shown in central 6, the posterior contour of the tooth is Slightly concave towards the lateral angle (La); it is bowed out posteriorly at the center as a tongue-shaped structure (Bp) which is neither thickened nor obvious. This structure is the mem- branous attachment of the basal edge of the central tooth to the lingual mem- brane. Viewing a row of central teeth on the radula, the attachment is often not noted, although it is there, e.g, centrals 1-5. The laterals and marginals are attached to the membrane by the base of the slender peduncle (Pd, Pl. 19; Text Fig. 1, tooth 1). The lateral tooth has a distinctive thickening (Th, Pl. 19, lateral 3), which arises from the com- paratively massive, swollen support for the innermost cusps, i.e., in a lateral tooth with the cusps formula 2-1-3 (See Table 8) the “2-1” cusps. Thethickened ridge runs posteriorly over the medial face of the lateral toothandturns sharply to run along the peducle of the tooth. Posterior to this ridge the peduncle is very thin and membranous. In Table 7, the arrangement of cusps most commonly found on the various teeth is shown along with the percentage of radulae on whichthat arrangement was found to occur in at least 90% of the teeth. Only 61% of the snails had the representative formula as a whole, i.e., that shown in column 2 of Table 7; but inevitably there were some teethinthese radulae which were not representative types. In Table 8, every cusp arrange- ment found for each of the teeth is tabulated, with the percentage ofradulae on which it was found, regardless of the frequency with which it occurred. The only significant difference between popu- lations was found in the central tooth. Centrals from the radulae of snailsfrom the Parker Mill population had a formu- la of 1-1-1/2-2 in 61%ofthe populations, while in the Hog Back and Barton popu- lations 80% of the population of snails had a central witha formula of 1-1-1/3-3. A rare case was found which had no centrals at all. An unusually high per- centage (12%) had a lateral toothformula of 2-1-3 on 1 side of the central and 2-1-2 on the other side. This arrange- ment would continue the length of the ribbon. Variation within the same lingual rib- bon was greatest among the cusps of the marginal teeth. In 60% of radulae, vari- ation within a ribbon was limited to the marginals. Whether a marginal had 7 or 8 cusps would often depend on the presence or absence of a tiny lateral spur as shown in Text Fig. 1, tooth 7. In only 20% of the radulae was vari- ation but minor, i.e., did a few outer 64 G. M. DAVIS PLATE 19. Variation in the radular teeth of Pomatiopsis lapidaria. The horizontal top row of teeth shows 1 row of teeth on the radula in natural position“ In some of the teeth the cutting edges of the cusps (Cu) have been torn off and only their thickened supports (Sup) are visible; when the cutting edges are shown, the supports are indicated by broken lines. The central column shows variation in 6 central teeth. Focus on centrals 1-5 is on the upper surfaces of the large, medial basal cusps (Lb). On this plane of focus the supports (Sup) for the basal cusps are not clearly discerned (centrals 1, 2) or not seen (centrals 3-5). The posterior, basal, edge of the tooth (Be) appears straight. The plane of focus is lowered on central 6 so that the characteristics of the supports for the basal cusps (Slb)can be observed. The large, medial basal cutting edges are shown as dashed lines to make more clear the shape of the basal sup- ports. At this level the basal process (Bp), a tongue-shaped attachment of the posterior tooth to the lingual membrane, can be seen. Central 7 (to the bottom, left) is shown with its posterior edge lifted up, showing how the basal supports are thickened and knob-like. To help interpret the cusp formulae quoted in the text, the numbers are marked on central 7 near the cusps in- volved. The 2 columns of lateral teeth numbered 1-5 (on the left) and 6-10 (right) show variations in cusp number and shape of the cutting edges. Lateral tooth 5 is shown with the anterior edge de- pressed so that each cusp is clearly observed for counts. Ae anterior edge of the tooth Be basal (posterior) edge of the tooth Вр basal process attaching central tooth to the lingual membrane Cu cutting edge of the cusp Fa face of the tooth La lateral angle Lb large medial basal cusp of central tooth Pd peduncle Slb supportive for large medial basal cusp Sup supportive for cutting edge (dashed when shown under cusp) Th molded thickening which gives the anterior medial part of the tooth a distinctive curvature *When the plate is turned sideways so that the labels are horizontal. ы 65 POMATIOPSIS AND ONCOMELANIA 19mo STRUIS IIA [810387] ТелдиэЭ Телэзет STRUI3ABIAL 66 G. M. DAVIS marginals vary between 5 or 6 cusps. In the 40% of the snails where the cen- tral or lateral tooth varied along the ribbon, a small stretch of about 10 cen- trals might have one formula while the next stretch of 15 or so would have another. In other cases only 5 or 6 centrals along the whole ribbon would vary. In Table 3 (p12) are listed the authors who have previously figured or discussed the radula of Pomatiopsis lapidaria. Also shown are the formulas they have figured or presented. The variations discussed in this paper en- compass the formulas presented by Stimpson (1865), Thiele (1928), Abbott (1948a), and Dundee (1957). Baker (1902) studying the work of Stimpson (1865), discussed the 2 incurved basal denticles on each side of the central tooth. In 1926, however, in his family diagnosis for the Pomatiopsidae, he mentions only 2 large basal denticles for the central and it is this arrange- ment he figures in 1928, i.e., 1 large basal cusp on either side of the central. In the light of the weight of evidence of the other workers, it is evident that Baker (1926, 1928) either did not observe the other cusps which were present or that he examined an extremely rare variant. Further, he probably reversed the inner and outer marginals; it is more likely that the inner marginal had 9 and the outer marginal 6 cusps in the specimen that he studied. Annandale (1924) reports the highly improbable number of 10 cusps for the inner and outer marginals. These can only be considered as misinterpre- tations. The central tooth formula for the anterior cusps, i.e., 2-1-2, could only represent a rare variant. Thiele (1931) gives a formula intended for the genus, hence including species other than P. lapidaria, with a lower limit of the number of cusps in the marginals. Inner and outer marginals with 4 and 3 cusps, respectively, are characteristic of P. cincinnatiensis. 7. Musculature Muscles of the Odontophore and Buccal Mass It is beyond the scope of this paper to attempt the detailed study on buccal mass musculature and function such as presented by Carriker (1946a) for Lymnaea stagnalis appressa (Say) or by Nisbet (1953), as presented by Fretter & Graham (1962), on the functional re- lationships in Monodonta lineata. For a full review of the literature pertaining to the buccal mass musculature one should consult Herrick (1906), Carriker (1946a), and Fretter & Graham (1962). There is not much literature on the buccal mass musculature of taenio- glossate snails. Johansson (1939) pre- sented a noteworthy paper onthe muscu- lature of Littorina littorea; Krause (1949) discussed the anatomy of Litho- glyphus naticoides. Fretter & Graham (1962) show several figures ofthebuccal mass musculature of Viviparus vivi- parus. The last authors review other literature which pertains to this group. As much of the musculature shown by these authors apparently does not per- tain to the muscles described in this study, muscle nomenclature becomes a problem. Johansson (1939) avoided this problem by numbering the muscles. I have named all the muscles discussed on the basis of position or inferred function. Intrinsic Muscles of the Odontophore The intrinsic muscles of the odonto- phore are those intimately associated with the cartilages. Removing the radula, radular shield and radular sac fromthe odontophore (Pl. 16, Fig. 1) and viewing its dorsal aspect, one observes 2 muscles (m4) each running from a posterior, ventrolateral position towards the mid- anterodorsal line of the odontophore. They fuse and form a trough which supports the radular sac. This muscle, the medial radular retractor, is readily lifted from the odontophore and pulled backward exposing the tips of the buccal POMATIOPSIS AND ONCOMELANIA 67 cariilages (Ca, Pl. 17, Fig. 3), which overlap with either the right over the left tip, or the reverse. Each cartilage is about 0.6 mm long, with a height of about 0.35 mm, and swells posteriorly, reaching a width of 0.26 mm. The shape of the medial surface of the carti- lages is shown in Plate 17, Fig. 5. The dorsal outline of the (left) cartilage, seen throughthe enveloping musculature, is shown in Plate 16, Fig. 2. The pos- terior ends of the cartilages are widely separated (Pl. 17, Fig. 3) leaving adequte room for the radular sac to nestle between. 1) Lateral Cartilage Tensor (my, Pl. ONE SP. 17.) Figs) 3.9) "The lateral cartilage tensor passes as an arc-like band about 0.36 mm wide about the ventroanterior end of the 2 carti- lages, binding the cartilages together. The muscle fastens along the exterior lateral edges of the cartilages (Pl. 16, Fig. 3). 2) Mediolateral Cartilage Tensor (mo, Pato ries: 11,289 93 "Pl: 17. “FHigs: 3, 5). The mediolateral cartilage tensor is a pronounced muscle seen arising from the ventral edge of each cartilage, as well as the ventral and postero- lateral edge, thus forming a cup-like structure encasing the posterior end of each cartilage (Pl. 16, Fig. 5; Pl. 17, Fig. 5). The muscle runs dorsally and anteriorly over the medial surface of each cartilage and appears thickened at the dorsal edge of the cartilages. From this thickening appears to arise a membranous continuation (Sur), called the subradular membrane, which con- tinues as a wide, thin band down laterally, and around the anterior edge of the cartilages, covering the lateral cartilage tensor (Pl: 165. Pio 17, Figs. 13:15). The mid-ventral crest of the membrane is thickened (a, Pl. 17, Figs. 3, 5) and serves as the place for the insertion of the radular protractor muscles (тн, РЕ 11, bigs 5): ¿The radular “shield (Ras) is tightly appressed to the sub- radular membrane (Pl. 16, Fig. 4). Ven- trally the subradular membrane is con- tinuous with the floor of the pharynx (Bf) to which it becomes tightly and inseparably appressed (Pl. 16, Fig. 4). On either side of the ventral crest of the subradular membrane a thin muscle (my 8, not figured) inserts: the lateral membrane protractor. Its origin is on the posteroventral edge of the cartilages. 3) Odontophore Divaricator (m3, Pl. 16 EES. NM CPI 17) Fist) AE he odontophore divaricator arises from the posterolateral edge of each cartilage. It is a band which runs laterally to the pharynx wall and serves, in part, as the major source of support and attach- ment of the pharynx wall to the odonto- phore. 4) Medial Radular Retractor (ma, Pl. 16, Bigs. №245; РЕ BES NS M5); The origin of this muscle is from the posterior, ventrolateral portion of the mediolateral cartilage tensor (mg). As mentioned above, this muscle runs dor- sally between the cartilages, over the medial surface of the mediolateral car- tilage tensor, fuses with its counter- part from the other cartilage, and rests with its anterodorsal tip on the crossed anterior ends of the buccal cartilages. The paired muscles insert on the an- terior, ventrolateral portion of the radular sac at a point 1/5 the way pos- teroventrally from the tip of the muscle. The muscle is trough-like and serves to support the radular sac as well as to retract it. Not figured is a pair of thin muscles (m 9) which run from the midventro- lateral edge of the style sac, betweenthe cartilages, under the anteroventral edge of the lateral cartilage tensor (my), where they insert on the subradular membrane to either side of the radular protractor (my). This tiny muscle is the ventral membrane protractor. Extrinsic Muscles of the Odontophore 5) Buccal Protractors (ms, Pl. 14; Pl. 15) tie: 134 PI UN а в 18): The origin of these paired muscles is shown in Plate 18. There are charac- teristically 2 slips for each of the pro- tractors; one from the rostral re- 68 G. M. DAVIS tractor (1120) and the other from the anterior, ventrolateral rostral wall. The Slips arising from the rostral retractors are usually fused as shown. Occasionally all the slips are fused in a single arc of muscle, but this occurs rarely. Each muscle inserts partially on the postero- ventral surface of the medial radular retractor (my) and partially on the mediolateral cartilage tensor (mo, Pl. 16, Figs. 4, 5). These broad, heavy muscles serve to pull the odontophore an- teriorly as well as to depress the pos- terior end of the odontophore. 6) Preventral Protractors (mg, Pl. 15, Fig. 1; Pl. 16, Fig. 3; Pl. 18). These muscles originate at the tip of the ros- trum immediately adjacent to the oral aperture (Pl. 15, Fig. 1). They insert on the posterior ventrolateral edge of each cartilage (Pl. 16, Fig. 3). These bands of muscle are most readily ob- served from the lateral, external view of the buccal mass (Pl. 15). They serve to protract the odontophore as well as the pharyngeal walls. The number of muscle bands is variable. 1) Radular Protractor (my, Pl. 17, Figs. 1, 5; Pl. 18). The paired radular protractors arise from the base of the rostral retractors (mag) as shown in Plate 17, Fig. 1. They run anteriorly side by side, bound in a connective tissue sheath with the main vascular supply to the buccal mass. They pass over the point where the fused medial bands of the buccal protractors (m5) separate (Pl. 18), pass into the con- nective tissue sheet (Mb) which sur- rounds the radular sac,and run between the cartilages to insert on the central crest of the subradular membrane (Sur, Pl. 17, Fig. 5). These muscles serve to depress the tip of the odontophore while protracting the radula slightly over the bending plane of the odonto- phore. Extrinsic Muscles of the Buccal Mass 8) Anterior Jugalis (mg, Pl. 15, Fig. 1). Arising from the anterior dorsal crest of the buccal mass, the muscle band, about 0.40 mm wide, runs obliquely posteroventrally covering the posterior portion of the buccal constrictor (mg). It inserts on the odontophore divari- cator (mg) and the ventrolateral edge of the cartilage. The cerebrobuccal con- nective (Cb) passes between this muscle and the buccal constrictor (mg). 9) Buccal Constrictor (mg, Pl. 15, Fig. 1). The buccal constrictor surrounds the anterior end of the buccal mass, en- casing the fleshy walls of the oral tube in a sheath of muscles. This muscular sheath extends from the anterior oral tube posteriorly to the rear of the pharyngeal tube to a point corresponding to the place where the subradular mem- brane (Sur, Pl. 16, Fig. 4) fuses with the floor of the pharyngeal tube. The posterior portion of this muscle is covered by the anterior jugalis (mg). 10) Odontophore Levator (not figured). The odontophore levator runs obliquely from the odontophore to the dorsal por- tion of the buccal constrictor (mg) be- tween the buccal constrictor and the anterior jugalis (mg). The insertion of this slender muscle is with the an- terior jugalis. 11) Dorsolateral Buccal Protractor (m44,, PL. 15, Е Борна muscle arises from each side of the anterolateral rostral wall (as shown in Pl. 18) as a single or as 2 thin parallel strands, which run to an insertion on the odontophore divaricator (mg) or the buccal retractor (m2). The insertion of this muscle is hidden by the buccal retractor (п112, Pl. 15, Fig. 1). Com- monly the protractor bifurcates, the an- terior slip fusing with the anterior slip of the retractor, both inserting on the divaricator muscle. The posterior pro- tractor slip runs posteroventrally over the exterior odontophore membrane to fuse with fibers from the tensor of the odontophore membrane mentioned below (under 14). 12) Buccal Retractor (m9, Pl. 15, Figs... 1... 23.9) Pli1%,) Fig, ВВ These pronounced paired muscles arise from the basal part of the mid-colu- mellar supportive (поз, Pl. 17, Fig. 1; POMATIOPSIS AND ONCOMELANIA 69 Pl. 18). They run anteriorly passing the medial surfaces of the cerebral ganglia (Cg), the pleuro-pedal (Pp) and the cerebro-pedal (Cp) connectives to in- sert on the odontophore divaricator mus- cle (mg) or the lateral cartilage surface. Before inserting, this broad muscle bifurcates, sending the anterior slip to an insertion on the odontophore; the posterior slip sends fibers into the membranous jugalis (m3). 13) Membranous Jugalis (my 3, Pl. 15, Fig. 1; Pl. 14). Posterior tothe anterior jugalis (mg) and dorsal to the external odontophore membrane (Eo), the buccal mass is quite fleshy; its surface is laced with thin and irregularly oriented muscle strands suggesting a thin mem- branous network rather than a distinct, stout muscular layer. The salivary glands enter this tissue; the esophagus arises from its posterior continuations. 14) Tensor of the Odontophore Mem- brane (not figured). The lateral muscula- ture of the odontophore bulges outward making the contour of the odontophore evident to one observing the lateral buccal mass (Pl. 15, Fig. 1). This muscu- lature is hidden from view as it is wrapped in a membrane, the external odontophore membrane (Eo, Pl. 18), which is continuous between the 2 halves of the odontophore. Posteroventrally this membrane passes between the ven- tral, protruding, recurved end of the radular sac and the odontophore muscul- ature. A slender muscle, the tensor of the odontophore membrane, runs from side to side across the external odontophore membrane through the angle formed by the emerging esophagus and the mem- brane. This muscle sends branches over the ventrolateral surface of the mem- brane. 15) Suspensor of the Radular Sac (m15, Pl. 16, Fig. 4). The origin of this muscle is the thickened tensor of the odontophore membrane at the point where the latter passes over the recurved ventral tip of the radular sac. Itinserts on the membranes of the tip of the radular sac. The muscle may be forked, i.e., its origins on the tensor of the odonto- phore membrane are slightly separated while it has a common point of insertion on the tip of the radular sac. 16) Preventral Dilator Muscles (тб, Plaid PL dbs Figs): Numerous thin muscle strands run from the buccal constrictor muscle (mg) and the pre- ventral protractors (mg to the antero- lateral rostral wall. 17) Suspensors of the Buccal Mass (m7; Pls (4s ee PL, (Bigs 11216: Fig. 4). Irregularly placed muscle strands run from the dorsal and dorso- lateral surface of the buccal mass to the rostral roof. Anteriorly, from the dorsal crest of the buccal constrictor (mg) and the anterior jugalis (mg), these muscles are more dense and some of them undoubtedly serve to protract the buccal mass. 18) Lateral Membrane Protractor (1118, discussed under 2). 19) Ventral Membrane Protractor (1119, discussed under 4). Body Musculature The principal muscle of the body is the columellar muscle (Cl). In Plates 5 and 6 this muscle is observed emer- ging from the ventral mantle tissue behind the collar (M). Only a portion of this wide muscle is shown. As pre- viously mentioned, the muscle is nor- mally pressed and bound to the ventral wall of the mantle cavity. It is fused with the columella of the shell at a level corresponding to the posterior end of the mantle cavity. In Plate 4, the muscle is shown emerging from the epithelium covering the “neck.” In Plate 14, the rostral and “neck” epithelium are slit and folded back to reveal this broad muscle, which is the basis of support for the head- foot region. To show the underlying features (Pl. 18), the esophagus was cut at the level of the pleuro-supra- esophageal connective (Psc), the cerebral commissure (Cc) was cut, and also the posterior region of the buccal retractors (mj9). The posterior end of 70 G. M. DAVIS the buccal mass was then pulled upward out of the rostral portion of the cephalic haemocoel and forward, using the origins of the buccal protractors (m5) asa hinge. The radular protractors (m7) were then cut and the buccal mass pulled forward completely. Finally, removing the dorsal nervous system, the musculature (Pl. 18) underlying the organs shown in Plate 14 can be observed. The columellar muscle, beneath the point where the pleuro-supraesophageal connective (Psc) crosses the esophagus (Pl. 14) sends 3 pronounced bands an- teriorly, while it sweeps ventrally in an arc as shown in Plate 15, Fig. 2. The 3 bands are the 2 lateral cephalic retractors (mg9) and the centrally positioned mid-columellar supportive (mg3, Pl. 15, Fig. 2; Pl. 18). The mid-columellar supportive mus- cle (mg3) serves as the origin for a number of important muscles. On either side of the origin of this muscle is noted a cavity (b, Pl. 18), leading into the posterior portion ofthe pedal haemo- coel, whose posterior wall and roof are formed by the ventrally curved colu- mellar muscle and the mid-columellar supportive, respectively. Laterally there is a space between the mid-colu- mellar supportive and the lateral cephalic retractors (m99) which marks the dorsolateral edges of the pedal haemocoel. The mid-columellar sup- portive bifurcates anteriorly into 2 band-like muscles, the rostral re- tractors (mag), which continue across the dorsal pedal haemocoel and run an- teriorly over the floor of the rostrum (PISA Riess Pls 18): Johansson (1939) has an excellent photograph of a gross dissection of the rostral area of Littorina littorea show- ing these paired muscles running toward the oral aperture as they do in Pomati- opsis lapidaria. Anteriorly the re- tractors pass beneath the point where the medial slips of the buccal protractors (m5) take their origin; they send numer - ous inserting slips around the oral aperture on the rostral floor. The origin of the radular protractors (m7) is at the dorsomedial base of the rostral retractors (mag, Pl. 17, Fig. 1). They are characteristically slightly swollen at their base. The buccal re- tractors (mj9) arise at the dorsal, pos- terolateral base of the mid-columellar supportive muscle (mg3, Pl. 17, Fig. 1; Pl. 18). Around the oral aperture, at the ros- tral tip, is a thin circular band of muscles best observed by clearing the rostral floor of the buccal protractors (m5). This band is the labial sphincter (m21, Pl. 16, Fig. 4). Anterior to the origin of the buccal retractors isa sheet of muscles running from side to side between the lateral cephalic retractors (1122). Characteristically, this sheet, the tensor magnus (mg 4), is split upinto 3-5 discrete bands (Pl. 18). The sheet is 0.36 mm wide and about 0.96 mm long. It rests upon the mid-columellar sup- portive muscles (m,3) as well as the posterior portion of the rostral re- tractors (1120) and supports the pos- terior portion of the contracted buccal mass, providing a supportive frame- work for the cerebral ganglia as well. The tensor magnus and the mid-colu- mellar supportive form a sort of roof over the pedal haemocoel separating it in a loose manner from the cephalic haemo- coel. The paired pleuropedal connectives (Pp), one on either side of the mid- columellar supportive, pass from the cerebral area between the posterior and mid-slips of the tensor magnus down into the pedal haemocoel to connect with the pedal ganglia. The cerebropedal connectives (Cp) do likewise, passing between the mid and anterior slips of that muscle (Pl. 18). The origin of the buccal retractor (m2) is sometimes split so that an anterior slip appears to arise from the posterior slip of the tensor magnus (Pl. 18). The lateral cephalic retractors are powerful bands (п122) giving support to the posteroventral wall of the rostrum. They terminate at about the point where the rostral floor turns ventrally forming POMATIOPSIS AND ONCOMELANIA 71 the anterior wall of the pedal haemo- coel (Pl. 18). The relationship of the ventrally curving columellar muscle (Cl), mid-columellar supportive (mag) and lateral cephalic retractor (mg9) is shown with regard to the posterior pedal haemocoel and the pedal ganglia (Pl. 15, Fig. 2). The columellar muscle sweeps ventroposteriorly beneath the oper- culum. Where the rostral floor slopes down to form the anterior wall of the pedal haemocoel some transverse mus- cle fibers are noted passing between the lateral cephalic retractors (1122). These muscles, the dorsal pedal tensors (m25), lie against the anterior wall of the pedal haemocoel and pass across the mid- anterior surface of the pedal ganglia or a little dorsal to the mid-length of these ganglia. A single muscle arises from the ven- tral origin of the rostral retractors (m20), the dorsal propodial retractor (1126, Pl. 15, Fig. 2). This muscle passes anteroventrally between the pedal ganglia over the pedal commissure, forks, and sends a Slip ventrolaterally to the anterior haemocoel wall beneath the dorsal pedal tensor (mp5). At the level of the propodial ganglia (Prg), from the mid-anterior haemocoel wall, arises a muscle about 0.15 mm wide which bifurcates and sends a slip laterally, right and left respectively, under the point where the propodial ganglia enter the anterior musculature, back under the metapodial connective, to the posterolateral wall of the de- scending columellar muscle. This is the mid-ventral protractor (mg7) shown for Oncomelania hupensis formosana, which has the same arrangement (Pl. 13, Fig. 1). 8. Nervous System In the study of neural anatomy for comparative, systematic use, con- siderable attention was directed towards the position of ganglia and their di- mensions, the number of nerves and their respective points of origin on a given ganglion, the lengths of the major com- missures and connectives, and especially the amount of variation encounteredinall of the above. Secondary and especially tertiary branches of nerves were found to be highly variable and were not generally considered for this com- parative study. There are major discrepancies be- tween the previous work on the neural anatomy of Pomatiopsis lapidaria (Dun- dee, 1957) and my findings. The present work is derived entirely from my own observations made on over 100 snails dissected especially for neural struc- ture. A classical, extensive study on com- parative prosobranch neural anatomy is that of Bouvier (1887). General proso- branch neural anatomy is reviewed by Fretter € Graham (1962). Johansson (1939) presented some unique photo- graphs of the gross dissections of the central nervous system of Littorina littorea. Krull (1935) gave data on prosobranch nervous systems and their relevance to prosobranch phylogeny. Krause (1949) published excellent draw- ings on the nervous system of Litho- glyphus naticoides. Further references are reviewed by these authors. There are 6 major ganglionic com- plexes, the cerebral, buccal, pedal, pleural, parietal, and visceral. Cerebral Complex a. Dorsal Aspect. Opening the rostral cavity from the dorsal side one exposes the buccal mass and associated struc- tures. Particularly noticeable is the heavy pigment dusted over the ganglia and nerves seen fromthe dorsal surface. The paired cerebral ganglia (Cg, Pl. 14; Pl. 20, Fig. 1) connected by the cere- bral commissure (Cc) are pressed against the mid-posterior curvature of the esophagus where the latter bends ventrally (Pl. 15, Fig. 1). The anterior tips of the ganglia press against the mid- or ventral, posterolateral wall of the buccal mass, in particular against the anterior end of the buccal retractor muscle (m12, Pl. 15, Fig. 1). The cerebral commissure (Cc) is 0.14 72 FIG. FIG. FIG. FIG. FIG. On* PLATE 20. G. M. DAVIS Nervous system of Pomatiopsis lapidaria 1. Dorsal aspect of the “brain” or central portion of the nervous system lifted out of the rostrum. 2. Anterior aspect of the pedal ganglia. 3. A rare variant in that a distinct pleuro-subesophageal connective is found between the left pleural ganglion and the subesophageal ganglion. 4. A variant in the arrangement of nerves leaving the supraesophageal ganglion (same scale as Fig. 3). 5. Medial surface of the right cerebral ganglion, showing the exact positions where the nerves from that ganglion arise. nerve from pg nerve from pg pedal commissure cerebro-buccal connective cerebral commissure cerebral ganglion cerebro-pedal connective cerebro-tensor nerve external mantle cavity nerve 1 external mantle cavity nerve 2 external mantle cavity nerve 3 gonadal nerve left lateral cephalic wall left pleural ganglion mid-columellar nerve metapodial ganglion median labial nerve 1 median labial nerve 2 mantle nerve 1, ganglion mantle nerve 2, from the supraesopha- geal ganglion mantle nerve 3, from the subesophageal ganglion osphradial nerve Opt Py Pa from the left pleural Sg . 6. Visceral ganglion viewed from the ventral side. optic nerve lateral retractor nerve from the pedal ganglion nerve to the anteroventral wall of the pedal haemocoel major lateral nerve of pedal ganglion propodial connective mid-propodial nerve metapodial connective pedal ganglion pericardial nerve pleuro-pedal connective propodial ganglion pleuro-supraesophageal connective pleuro-subesophageal connective renal nerve right pleural ganglion supralabial nerve subesophageal ganglion subvisceral connective supraesophageal ganglion sublabial nerve supravisceral connective tentacular nerve visceral ganglion variant branch of Mno * Usuallya single large trunk emerges fromthe supraesophageal ganglion the common osphradio- mantle nerve, which then bifurcates to form the osphradial nerve (On) and mantle nerve 2 (Mng) (See p 79). POMATIOPSIS AND ONCOMELANIA 73 74 G. M. DAVIS mm long, but in some specimens, a length of 0.19 mm was found. The width is 0.05-0.06 mm. In dorsal view the cerebral ganglia are 0.36-0.29 mm long. Eight nerves arise from each cerebral ganglion. Seven of these nerves are pronounced and run anteriorly, while 1 is quite thin and has a ventral course. Upon opening the dorsal mid-line, 5 of the most conspicuous nerves appear, jumbled, intermixed, or appressed to the buccal mass (Pl. 14) while the 2 remaining ones are hidden from view beneath it. In all cases the nerves are easily untangled and separated from the buccal mass. In a rare case the right salivary gland was found tucked down between the cerebral ganglion and buccal mass, intermixed in the nerves arising from the ganglion. With the buccal mass removed, one can observe all 7 of the pronounced nerves (Pl. 20, Fig. 1) although, for most of them, their origin on the ganglion cannot be observed. 1) Tentacular Nerve. The tenta- cular nerve (Tn) is the most prominent nerve rising from the cerebral ganglion (РР 12; 71. 15% Fig: 15 Pl. 20, "Figs, 5). It arises from a marked swelling, the tentacular bulb, on the anterodorsal end of each cerebral ganglion and runs anterolaterally into the lateral rostral wall, through the wall and into the tentacle. Arising from the tentacular nerve about half way out towards the rostral wall is a slender nerve which runs into the rostral wall apart from the entry of the tentacular nerve. Oc- casionally 1 or 2 other nerves are seen emerging from the tentacular nerve at a point 0.24 mm beyond the tentacular bulb. These last mentioned nerves are rarely found. When they occur they enter the rostral wall posterior to the tentacular nerve. 2) Optic Nerve. This nerve (Opt) arises from the mid-, ventrolateral surface of each ganglion and runs an- teroventrally to enter the rostral wall about 0.12 mm posterior to the ten- tacular nerve. The nerve is quite slender. It innervates the eyes (Pl. 14; Pl, 15, ‘Figs № РР 20; Pie7 3) Supralabial Nerve. The origins of the remaining nerves that arise from the cerebral ganglia are displayed in Plate 20, Fig. 5, which shows the medial surface of the right cerebral ganglion. The supralabial nerve (51) sweeps dor- sally from the anteromedial surface of the cerebral ganglion crosses over the other nerves from the ganglion, and runs dorsolaterally to the rostral wall. It becomes bound to the dorsolateral rostral wall about 0.12 mm anterior to the point where the tentacular nerve (Tn) enters the wall. Travelling an- teriorly, dorsal to the other labial nerves, it becomes more Slender and, in the dissection shown, undulates near its anterior end due to the con- tracted state of the rostrum (Pl. 14, 51). The nerve terminates more pos- teriorly than the other labial nerves. The supralabial often sends off a small branch near the point where it fuses with the rostral wall (Pl. 20, Fig. 1). 4) Median Labial Nerve 1. Arising Slightly anteroventrally of the previous nerve (Pl. 20, Fig. 5), the median labial (Ml) runs anteriorly unfused with the rostrum but usually lying against the rostral wall. Anteriorly it passes above the origin of the dorsolateral buccal protractor muscle (m 11) and runs to the tip of the rostrum beneath the pre- ventral dilators (mıg, Pl. 15, Fig. 1; P1. 18; Pl. 20. Figs 175). 5) Median Labial Nerve 2. Median labial 2 (Ml) arises from the anterior edge of the cerebral ganglion beneath the tentacular bulb (Pl. 20, Fig. 5). It runs anteroventrally to median labial 1, likewise unfused with either the rostral wall or buccal mass. It charac- teristically forks anteriorly sending a root to either side of the origin of the dorsolateral buccal protractor (m1); one root innervates this muscle. 6) Sublabial Nerve. This nerve (Sul) arises closely appressed to the cerebro-buccal connective (Cb) and sometimes these 2 nerves arise as a POMATIOPSIS AND ONCOMELANIA 75 fused, inseparable trunk before branch- ing. The sublabial travels freely along the floor of the rostrum. In the con- tracted rostrum, the nerve coils. From the coiled part a branch arises to in- nervate the anterior rostral floor (Pl. 18). Anteriorly the nerve forks; one branch passes to the origin of the buccal protractor (m5) while the other passes beneath that muscle to travel anteriorly to the rostral tip (Pl. 15, Fig. 1; Pl. 18; Pl:20, Figs: 1.45). 7) Cerebro-Buccal Connective. This nerve (Cb) runs anteriorly over the lateral edge of the rostral retractor muscle (mag, Pl. 18), up under and behind the buccal protractors (m5) before the latter disappear beneath the external odontophore membrane. As shown in Plate 15, Fig. 1, the cerebro-buccal connective travels between the anterior jugalis (mg) and the buccal con- strictor (mg) over the anterolateral edge of the odontophore, emerges over the surface of the membranous jugalis (1113) to run posteriorly into the an- terior tip of the buccal ganglion (Bg). Before reaching the buccal ganglion it passes beneath the inserting fibers of the posterior slip of the buccal re- tractor (m1). 8) Cerebro-Tensor Nerve. This nerve (Cg) arises from the anteroventral edge of the cerebral ganglion (Pl. 20, Fig. 5). It is a slender nerve which runs directly ventrally to innervate the anterior slip of the tensor magnus mus- cle (Mo4). b. Right Lateral Aspect of the Cere- bral Complex. Two connectivesleavethe ventral surface of the cerebral ganglia (Pl: 15,. Fie:wl: Pl. 20, Fig:,5), drop ventrally and converge on the dorsal surface of the pedal ganglia (Pg). The posterior connective is the pleuro-pedal (Pp), the anterior one is the cerebro- pedal (Cp). The former is about 0.19- 0.22 mm long and 0.048-0.050 mm wide. The latter is about 0.19 mm long with a width of about 0.03 mm. No nerves arise from along the length of the cerebro-pedal connective. From the juncture of the pleuro-pedal connective and the cerebral ganglion arises the pleural ganglion (Rp, Pl. 15, Fig. 1; Pl. 20, Figs. 1, 5). A few nerves, usually 4, labeled ppı_4, arise from the pleuro-pedal connective. 1) Lateral Nerve 1. At the ventral juncture of the pleural ganglion with the connective a nerve arises (PP: PIS; Fig. 1) which runs ventrolaterally under the penial nerve (pp) to enter the mus- culature of the rostral wall just above the lateral insertion of the posterior slip of the tensor magnus muscle (m9,) in the lateral cephalic retractor (mao). This nerve may be absent; when present its thickness is found to vary con- siderably. It is 0.024 mm in diameter or thinner. 2) Penial Nerve. In males this nerve (pps, Pl. 15, Fig. 1) is greatly thickened (0.036 mm wide). It arises from the mid-length of the connective (in some cases slightly lower, e.g., Pl. 15, Fig. 1). Passing posteriorly as well as dorsolaterally to the cephalic wall above pp,, it becomes thinner and proceeds to the mid-cephalic roof and then into the basal musculature of the verge. In females the nerve is slender, about 0.024 mm wide, and passes dorso- laterally to the cephalic wall. 3) Lateral Nerve 3. Slightly ven- tral and lateral to the penial nerve a nerve arises from the pleuro-pedal connective (pps, Pl. 15, Fig. 1) which initially travels anteriorly and ventro- laterally, then dorsolaterally. Near its end it forks and sends both branches into the lateral rostral wall just pos- terior to the point where the tentacular nerve enters the wall. The optic nerve occasionally passes between the bi- furcation. Thepoints where the branches of this nerve enter the wall are variable and may be ventral or posterior to the positions described. The nerve itself is markedly variable. Instead of a single nerve arising from the con- nective, 2 closely associated nerves often arise and run directlytothe lateral rostral wall. In rare instances 2 minute 76 G. M. DAVIS nerves just ventral to ppg arise from the connective and run laterally to the rostral wall. 4) Lateral Nerve 4. When dis- secting under Bouin’s solution with the aid of a very bright light, I often found arising from the ventral surface of the cerebral ganglion between the pleuro- pedal connective and the cerebro-pedal connective a thin strand (pp4, Pl. 15, Fig. 1) appressed to or fused with the pleuro-pedal connective, or distinctly separate so as to traverse the space between the connectives. It is suspected that this nerve, when not observed, is incorporated in the pleuro-pedal con- nective. The nerve passed into the area where the ppg and pp3 nerves arose, but its final destination was not ascertained. c. Left Lateral Aspect of the Cerebral Complex. The counterpart of PP» found on the right side, was rarely present on the left. When encountered, it was a thin strand running ventrolaterally into the lateral cephalic retractor muscle (mo) near the posterior edge of the attachment of the tensor magnus (m24). The counterpart of the penial nerve is regularly present. It runs dorsolaterally to enter the cephalic wall just behind the entrance of the optic nerve where the latter enters the wall. The counter- part to рр. runs laterally and only slightly dorsally, bifurcates, eachbranch entering the ventrolateral cephalic wall opposite the pleuro-pedal connective or posterior to that point (pp3, Pl. 14). Again, variation is commonly en- countered in the arrangement of these lateral nerves from the pleuro-pedal connective. In unusual instances one finds, instead of pp, a nerve arising from the posterior surface of the pleuro- pedal connective below the point where the pleural ganglion arises. The nerve runs ventrally and innervates the pos- terior slip of the tensor magnus. In contrast to the right side, ppg and pp3 arise from the ventral portion of the connective. Pedal Complex The pedal complex consists of 3 pairs of ganglia. Each of the large dorsal pedal ganglia connects ventrally with an antero-medial propodial ganglion and a postero-lateral metapodial ganglion. The statocyst is part of the pedal com- plex. It is appressed to the dorso- posterior surface of each pedal ganglion at the base of the pleuro-pedal con- nective. The pedal ganglia are quite large and fill the pedal haemocoel beneath the tensor magnus (m94). Each ganglion seen from the anterior face (Pg, Pl. 15, Fig. 2; Pl. 20, Fig. 2) is about 0.31 mm long and 0.24 mm wide. They are connected, as shown, by a commissure (C) which varies in length from 0.08- 0.04 mm. The right and left ganglia are Similar with regard to the number of nerves which arise, their position, and variability. The description below pertains to either ganglion. Seven nerves, labeled pı_7 (including the pro- podial and metapodial connectives) arise from the pedal ganglia. 1) Lateral Retractor Nerve. The lateral retractor nerve (р1) arises ап- teroventrally with respect to the cerebro-pedal connective (Cp) asa Slender nerve (Pl. 15, Fig. 2; Pl. 20, Fig. 2). It runs anteriorly (Pl. 15, Fig. 1) to enter the lateral cephalic retractor (mp) at its ventral edge near the area where the rostral retractors emerge from beneath the tensor magnus (mgq, Pl. 18). 2) Nerve to the Anteroventral Wall. A slender nerve (ро) leaves the anterior face of the pedal ganglion ventral to the above nerve, and just dorsal to the pro- podial connective (pa). It runs an- teriorly to the anterior wall of the pedal haemocoel beneath the dorsal pedal ten- sor (m95). 3) Major Lateral Nerve. This stout nerve (p3), often 0.048 mm wide, arises from the ventrolateral edge of the pedal ganglion and runs posterolaterally back through the musculature to an area be- POMATIOPSIS AND ONCOMELANIA 77 neath the operculum. The minor lateral nerve (not figured) was found only a few times. It branches off between p3 and the metapodial connective (pg) on the lateral edge of the ganglion and follows pg into the lateral musculature of the foot. 4) Propodial Connective. This short, stout connective (py) joining the pedal ganglion (Pg) to the propodial ganglion (Prg) arises from the ventromedial edge of the former. It is devoid of emergent nerves. It usually swings dorsally from the long axis of the pedal ganglion and enters the anteroventral foot muscula- ture (Pl. 13, Fig. 1). The Propodial Ganglion (Prg). Within about 0.07 mm from the pedal ganglion the propodial connective swells into the propodial ganglion which is about 0.12 mm in diameter or slightly less. The foot musculature forms a spheroidal cavity encasing the propodial ganglion. From the distal tip of that ganglion there arise 3 nerves of about equal size. These nerves are, however, variable and at times only 1 of them is particu- larly pronounced. 5) Mid-Propodial Nerve. A thin nerve (ps, Pl. 20, Fig. 2) emerges from the anterior, mid-surface of the pro- podial ganglion. It runs anteroventrally into the foot musculature. 6) Metapodial Connective. This stout connective (pg) between the pedal ganglion (Pg) and the metapodial gan- glion (Mg) takes off lateral to the pro- podial connective and runs ventro- laterally to it. Between the pedal and metapodial ganglia 2 slender nerves arise from the connective (a, b, Pl. 20, Fig. 2). Commonly they do so next to each other at the dorsal end of the metapodial connective, although the exact point of emergence is quite variable. The dorsal strand (a) may actually, at times, originate at the point where the pedal ganglion gives rise to the con- nective and at the posterior side. The ventral strand (b) may arise from the mid-anterior connective near the pedal ganglion. These 2 nerves are tightly appressed to each other and against the connective, the metapodial ganglion, and nerves leaving the distal end of the metapodial ganglion. The metapodial ganglion (Mg) is not spherical as the propodial ganglion, but more like an elongate swelling of the connective. It is about 0.19 mm long and 0.10 mm wide. Nerves leaving the distal tip vary as to number and thick- ness. Generally, the ganglion narrows to a central wide and thin band flanked on either side by a thinner nerve. Often the medial of the thinner nerves is re- placed by 2 nerve strands. The meta- podial ganglion (Mg) and distal nerves enter a well-defined, roomy channel in the ventrolateral wall of the pedal haemocoel below the level of the more medially placed propodial ganglion (Pl. 1.35) Pig: A). 7) Dorsolateral Pedal Nerve. View- ing the pedal ganglion laterally (Pl. 15, Fig. 1) a nerve (p,) is seen to arise just ventral to the point where the pleuro-pedal connective joins the pedal ganglion. This point of origin is just anterior to the mid-point of the stato- cyst (Stc) where the latter is pressed against the dorsoposterior wall of the pedal ganglion. The nerve soon bi- furcates and each branch runs laterally to the haemocoel wall. The nerve is variable and 2 nerves might arise from the ganglion instead of 1. In one case p, arose from the pleuro-pedal con- nective just dorsal to the pedal ganglion as a thick, single nerve. It ranpostero- ventrally for 0.07 mm and then branch- ed into 3 distinct nerves each of which traveled laterally to the haemocoel wall. The statocyst (Stc, Pl. 15, Fig. 1) is 0.12 mm in diameter and contains a single statolith (Sta) 0.07 mm in dia- meter. Upon full contraction of the rostrum these spheres slide up within the openings (b) shown in Plate 18, lateral to the base of the mid-colu- mellar supportive (mag). Buccal Complex From the dorsal aspect, the buccal ganglia (Bg) are just visible in each 78 G. M. DAVIS angle of the anterodorsal cerebral ganglion and the outer edge of the salivary glands (Pl. 14). It is evident from the lateral view (Pl. 15, Fig. 1) that these ganglia lie in the depression where the esophagus presses against the posterior buccal mass just after its origin. The ganglia are slightly elongate, the long axis generally projecting slightly anterodorsally when the buccal mass is horizontal. Each ganglion is about 0.19 mm long, with a width of 0.14 mm. The 2 buccal ganglia are connected by the buccal commissure which is about 0.24 mm long and 0.048 mm wide. The com- missure passes between the esophagus where it leaves the buccal mass and the posterior buccal mass. The com- missure is shown ventrally in Plate 18 (Вс). The cerebro-buccal connective (Cb) has been discussed (p 75). А number of nerves, labeled b1_6 arise from these ganglia and from the cerebro-buccal connective. 1) Dorsal Buccal Nerve. Thisnerve (by) arises from the mid-dorsal surface oí the buccal ganglion as a stout strand which runs dorsally and bifurcates be- neath the emerging salivary glands. Each branch runs to the dorsal crest of the buccal mass over the area of the esophageal valve (Pl. 15, Fig. 1; Pl. 13, Во 5) 2) Esophageal Nerve. The eso- phageal nerve (65) arises from the dorsal buccal nerve before the latter passes be- neath the salivary gland. It runs pos- teriorly along the mid-esophageal sur- face (PIs 7105, PL A lo ME AT): 3) Central Buccal Nerve. This nerve (bg, Pl. 13, Fig. 5) is variable in position. It may arise from the antero- dorsal surface of the ganglion, from the posterior base of by, or as a branch of by. The nerve runs dorsally to enter the buccal mass just posterior to the root of the salivary gland (Pl. 15, Fig. 1). 4) Anterior Buccal Nerve. This nerve (by, Pl. 13, Fig. 5; Pl.15, Fig. 1) arises from the dorsal surface of the emerging cerebro-buccal connective. It runs dorsally to disappear beneath the root of the salivary gland. 5) Odontophore Nerve. About 0.19 mm anterior to the buccal ganglion a nerve (bs) emerges from the ventral surface of the cerebro-buccal con- nective. This nerve curves over the external lateral odontophore mus- culature beneath the external odonto- phore membrane (Pl. 13, Fig. 5). 6) Posterior Buccal Nerve. At the point where each buccal ganglion (Bg) gives rise to the buccal commissure (Вс) a nerve (bg) emerges from the commissure to run ventrally over the external odontophore membrane (Pl. 13, Pis 5: РЕ 8). Pleural Complex The pleural ganglia (Rp, Lp, Pl. 20, Figs. 1, 2) arise from the pleuro-pedal connective (Pp) immediately at the juncture of the cerebral ganglia and the connective (Pl. 15, Fig. 1) and are, therefore, partially pressed beneath the posteroventral curvature of the cerebral ganglia and intimately associated with them (Pl. 20). As aresult ofthe charac- teristic prosobranch streptoneurous condition the posterior tip of the right pleural ganglion (Rp) is drawn up over the edge of the esophagus (Pl. 15, Fig. 1) where it gives rise to the pleuro- supraesophageal connective (Psc, Pl. 20, Fig. 1). The connective crosses the esophagus to the left side of the body. The shape of the ganglion, cor- responding to this stretching, is not round like its counterpart, but drawn out, with a length of 0.24 mm and width of 0.12 mm. The pleuro-supraesophageal connective (Psc) often lies in a crease in the esophagus, crossing at a distance of 0.24 mm or less behind the cerebral commissure (Pl. 14). The connective has a length of 0.34 + 0.048 mm. It enlarges at the. left lateral extremity into the supraesophageal ganglion (Sug, Pl. 14; Pl. 20, Figs. 1, 4). The left pleural ganglion (Lp), resting on the mid-columellar supportive muscle is round with a diameter of 0.17 mm. Posteriorly it gives rise to a pro- nounced nerve, mantle nerve 1 (Mn, POMATIOPSIS AND ONCOMELANIA 79 Pl. 20, Figs. 1, 3). This nerve runs laterally tothe left ventrolateral cephalic wall and enters it at a point just anterior to the point where mantle nerve 2 (Mno) from the supraesophageal ganglion (Sug) enters the wall. The left pleural ganglion is tightly appressed to and connected with the subesophageal ganglion (Sg). The latter is characteristically not separated from the former by more than 0.04 mm. In only one instance was a short, pro- nounced pleuro-subesophageal con- nective noted (Psb, Pl. 20, Fig. 3). In that Same instance an unusual second nerve arose from the left pleural gan- glion just anterolateral to mantle nerve 1 and ran to the left ventrolateral cephalic wall. No corresponding mantle nerves arise from the right pleural ganglion. Parietal Complex The parietal ganglionic complex is composed of 3 ganglia, the supra- and sub-esophageal ganglia and the os- phradial ganglion. The latter has already been discussed in the section on the mantle cavity (p 27). The supraesophageal ganglion (Sug, Pl. 14; Pl. 20, Figs. 1, 4) is about 0.24 mm long and 0.12 mm wide. The tip of the ganglion is very close to the ventro- lateral cephalic wall (Lew, Pl. 20, Fig. 4 and also Pl. 14), generally only 0.12 mm or less. From the area of the ganglion tip a few nerves arise ina variable fashion. The supravisceral connective (Suv) leaves the postero- lateral surface of the ganglion. It con- sistently leaves the ganglion at this point, travels posteriorly along the left edge of the columellar muscle against the base of the “neck” wall, emerges from the left side of the “neck,” and runs posteriorly to join the visceral ganglion (Vg, Pl. 11), thereby com- pleting one side of the “visceral loop.” The osphradiomantle nerve arises from the tip of the supraesophageal ganglion (Sug) and bifurcates into the osphradial nerve (On) and mantle nerve (Mng). The point of bifurcation, however, is quite variable. In Plate 20, Fig. 1, the mantle nerve 2 and osphradial nerve are shown to arise separately, the for- mer emerging from the ganglion as a very thin fiber. This condition is rare. Commonly the 2 nerves emerge, equally thick, as separate but closely associated trunks. Equally common, a single large trunk emerges from the ganglion and bifurcates immediately before entering the cephalic wall. Occasionally bifur- cation occurs within the wall. The osphradial nerve (On, Pl. 14; Pl. 20, Figs. 1, 4) runs laterally to the osphradium to enter the mid-ventral surface of the osphradial ganglion (Og, Pl. 11). The mantle nerve. (Мпо) runs anteriorly towards the mantle edge after it enters the lateral cephalic wall. This nerve sends a branch to connect in a dialyneury with mantle nerve 1 from the left pleural ganglion. An extreme vari- ant as regards the manner in which nerves arise from the supraesophageal ganglion is shown in Pl. 20, Fig. 4. A branch (x) arose from mantle nerve 2 at the point where the latter emerged from the supraesophageal ganglion. In this same rare variant specimen, anerve also arose from the base of the supra- visceral connective (Suv) and ran laterally to the cephalic wall. The subesophageal ganglion (Sg, Pl. 20 Figs. 1, 3) is of the same dimension as the left pleural ganglion (Lp). It lies beneath the esophagus closely ap- pressed against the left pleural ganglion as previously mentioned. Three nerves characteristically emerge from this ganglion. 1) The subvisceral connective (Sbv) arises from the posterolateral curva- ture. It forms a loop when the rostral area is contracted and then runs pos- teriorly along the mid-line of the colu- mellar muscle (Cl, Pl. 7, Fig. 2) or slightly to the right of the mid-line. It enters that muscle in the posterior “neck” region and appears emerging from the “neck” on the right side (Pl. 11, Fig. 1). The connective runs to the visceral ganglion (Vg) thereby com- pleting the other arm of the “visceral loop.” 80 G. M. DAVIS 2) Mantle nerve 3 (Mng) arises from the anterolateral curvature of the subesophageal ganglion asa strong nerve running laterally across the columellar muscle just as the columellar muscle turns ventrally to form the posterior wall of the pedal haemocoel. This nerve tends to slope downward at the right edge of the columellar muscle and then turns to enter the right wall. 3) The mid-columellar nerve (Mcn) arises from the left, ventral curvature of the subesophageal ganglion as a slender fiber which slips around the left edge of the mid-columellar sup- portive (m3), turns ventrally and enters the columellar muscle. Visceral Complex The visceral ganglion (Vg, Pl. 5; Pl. GE bei PLT. ress 1.) 25 Pdo 11. Fig 01:10 Pl: 20, Fig... 6) 19. aısingle structure about 0.24 mm long and 0.096 mm wide. It is exposed by pulling back the columellar muscle of the uncoiled snail, as shown in Plates 5, 6 and 7. Also exposed is the subvisceral con- nective (Sbv) where it emerges dorsal to that muscle and runs into the gan- glion. Two nerves are seen arising from the subvisceral connective (Pl. 20, Fig. 6) at points usually covered by the columellar muscle. The anterior nerve, exterior mantle cavity nerve’ 1 (Es), is stouter. It runs directly over the mantle cavity, often crossing the spermathecal duct, towards its right edge, to the pallial oviduct in the fe- males or to the intestine in the males. The posterior nerve, external mantle cavity nerve 2 (Es), is variable in size and at times is absent. It takes off about 0.32 mm posterior to Ey, runs to the mid-ventral external wall of the mantle cavity, and turns posteriorly, becoming slender and finely branched. Between these 2 nerves the subvisceral connective is characteristically kinked or undulating. The supravisceral connective (Suv) arises from the dorsal surface of the visceral ganglion near its anterior end (Pl. 20, Fig. 6). About 0.07 mm from the ganglion, the pericardial nerve (Pn) leaves the supravisceral connective, runs posteriorly over the pericardium, forks and sends 2 branches over the pericardial surface. In Plate 20, Fig. 6, the visceral ganglion is shown as observed in Plates 5-7. From the posterior tip of the ganglion, arise 2 major nerves and one minor nerve. Most readily observed is the ventrally located gonadal nerve (Gn) whichtravels along the vas deferens or oviduct to- wards the gonadal area (Pl. 7, Fig. 1). Arising alongside of the gonadal nerve or as an early branch of the gonadal nerve is a minor nerve, the external mantle cavity nerve 3 (Es). It runs over the area covering the end of the mantle cavity towards the right. The renal nerve (Rn) can only be ob- served from the ventral surface by cutting the gonadal nerve and lifting the posterior tip of the visceral ganglion up. From the dorsal surface of the ganglion near the posterior tip this stout nerve arises and runs into the body tube between the posterior mantle cavity epithelium and the kidney epi- thelium pressed against the posterior wall of the mantle cavity. The nerve passes to the dorsal surface of the body tube and bifurcates, sending roots to the kidney. Occasionally very fine fibers were observed emerging from the right lateral surface of the ganglion which ran onto the external mantle cavity epithelium. D. Oncomelania hupensis formosana 1. Shell Pilsbry & Hirase (1905) described this species as Blanfordia formosana. The Shell was described as: “perforate, light brown, rather solid, turrite-conic, the outline of the spire straight, apex rather acute. Whorls 6 3/4, quite convex and parted by well-impressed sutures, smooth except for faint growth lines. The last whorl has a rounded and rather strong crest or varix behind the outer lip. The aperture is ovate, POMATIOPSIS AND ONCOMELANIA 81 TABLE 9. Conchological measurements of Oncomelania hupensis formosana Number of snails Structures measured Shell 7.0 whorls Shell 7.5 whorls Aperture Callus Apical whorl Tip of apical whorl X = Mean S = Standard deviation Se = Standar error of the mean. brown within; peristome brown-edged, the columella concave and somewhat thickened, whitish. Length 7, diam. 3.25 mm, length of aperture with peristome 2.8 mm.” Others who described the shell of this species were primarily Bartsch (1936), Annandale (1924) and Abbott (1948b). Annandale included Blanfordia for- mosana, Katayama nosophora and On- comelania hupensis within the single genus Oncomelania. Abbott also included the genus Schistosomophora (S. quadrasi) as discussed by Bartsch (1936). The de- Scriptions presented by these authors need to be expanded and in some cases modified. Adult shells, i.e., those with a varix, have 7.0-7.5 whorls (Pl. 2B). The nuclear whorls are 2.5, white, glossy, set off from the yellow-horn of the remaining whorls. The first nuclear whorl is emergent. The sutures are moderately impressed and the whorls slightly convex. Pilsbry & Hirase (1905) correctly stated thatthe outline of the spire was straight; the comparatively flattened whorls aid in giving this impression. The aperture is ovate, elongate, narrowed apically. The inner lip is slightly reflected over the narrow umbilicus and is connected with Length in mm Number Width in mm of snails the outer lip by a long parietal callus (Pl. 2; Pl. 3, Figs. 7, 8). The outer lip is thin and strong. Observing the outer lip with the shell rotated 90° to the left of the apertural view, one ob- serves that it is sinuate as is the thick- ened varix behind the lip (Pl. 3, Fig. 4). The base of the shell, as shown in Plate 3, Figs. 7, 8, is not rounded but appears truncate. The surface of the Shell has fine growth lines in contrast to the roughened microsculpture of Pomatiopsis lapidaria. Only occasion- ally is a line here and there more pro- nounced. Cleaning the shell with “Clorox” removes the periostracum and the “brown-edged” peristome. Conchological measurements for 31 adults are presented in Table 9. Shells with 7.5 whorls had an average length of 6.3 mm and a width of 3.0 mm; those of 7.0 whorls measured 5.76 mm and 2.82 mm, respectively. The average length of the aperture for 7.5 mm shells was 2.4 mm andthe average callus length was 1.08 mm. The width of the first nuclear whorl was 0.34 mm and the tip of the first nuclear whorls measured 0.12 mm in diameter. The width of the tip of the first nuclear whorls was almost constant. Several features of the shell of 82 G. M. DAVIS Oncomelania hupensis formosana sepa- rate this species from Pomatiopsis lapi- daria: (1) The apical whorls of P. lapidaria are considerably larger than those of O. hupensis formosana (Pl. 3, Figs. 6, 9). (2) The straight lip of the former con- trasts with the sinuate lip of the latter. (3) The lack of varix formation in the former differs considerably from the pronounced varix in the latter. (4) The wide umbilicus and short parietal callus in P. lapidaria con- trasts with the narrow umbilicus and long parietal callus in О. hupensis formosana. (5) The roughened microsculpture, deeply impressed sutures, and pro- nounced convex whorls of theformer are in contrast to the comparatively smooth shell, moderately impressed sutures and moderately convex whorls of the latter. (6) When both species are fully mature and have the same shell length, O. h. formosana has 1 more whorl than P. lapidavia (compare Tables 4 and 9 for the number of whorls at a shell length of 6.2-6.3 mm). Few P. lapidaria reach 7 whorls while many O. h. formosana have 7.5 whorls. | 2. External Morphology and Topography The folds and grooves of the head as well as the mode of progression are as described for Pomatiopsis lapidaria. Pigmentation. The pattern of pigmen- tation is somewhat different. Oncome- lania hupensis formosana shows sexual dimorphism in that the integument ofthe apical whorls of the male has an intense black pigment which is lacking in the females. This was initially evident, through the shell; the apical whorls of the male appeared black while those of the female appeared light brown and peppered with the usual white granular bodies. In the males the pronounced band of pigment starts at the beginning of the digestive gland (Pl. 24, Fig. 2) with a width of 0.48 mm. The apical edge of the band is smooth and regular in con- trast to the flammulate pattern found in Pomatiopsis lapidaria. The lower edge of the band is irregular but not deeply lobed or flammulate. Anteriorly the dorsal surface of the stomach is lightly dusted with pigment. No pigmented strip underlines the in- testine crossing the body whorl in either sex. Viewing the animals of both sexes through the shell in apertural view one notes a very dark pigmented area to the left. This area corresponds tothe dorsal surface of the mantle which is densely pigmented all the way back to the rear of the ctenidium. The dorsal rostrum and head are gray- black, the intensity of the pigment varying between individuals. The pigment is evenly dusted over the epithelial sur- faces. Below the suprapedal fold on the sides of the foot there is very light pigmentation in contrast to the darker pigment found in P. lapidaria. Lack of dark pigment in this area is one reason for the decreased prominence of the pedal crease. The sole of the foot differs from P. lapidaria: the entire surface appears white due to the densely packed large, white, glandular units, which were found mainly at the posterior part of the foot in P. lapidaria. The lateral indentation is slightly less pronounced. Tentacles and Eyes. The tentacles of Oncomelania hupensis formosana are more elongate than those of Pomatiopsis lapidaria. A series of adult specimens was placed under water with adult P. lapidaria, observed as they moved about, and the tentacular length beyond the eye measured. Thetentacles of both species, although capable of great expansion and contraction, are carried in a character- istic fashion, usually just short of full extension. In O. h. formosana the length measured between 0.96 and 1.20 mm; in P. lapidaria the length varied between 0.60 and 0.90 mm. The width of the tentacle at the base was about 0.12 mm in both species. The “glandular units” partly covering the dorsomedial surface of the eyes formed units about the same length and POMATIOPSIS AND ONCOMELANIA 83 PLATE 21. Uncoiled female Oncomelania hupensis formosana FIG. 1. The uncoiled female. Numerous organs are evident through the epithelium. FIG. 2. A portion of the female reproductive system exposed by removing connective tissue and kidney tissue visible in Fig. 1 anterior to and to the left of the bursa copulatrix (B). A portion of the pallial oviduct is cut away (dashed line) to show the tubes of the sperma- thecal and sperm ducts which are overgrown by tissue of the pallial oviduct. bursa copulatrix columellar muscle digestive gland esophagus fecal pellet gonad gonadal nerve intestine kidney edge of the mantle ventral wall of mantle cavity oviduct coiled portion of oviduct portion of oviduct passing into pallial oviduct pallial oviduct posterior chamber of stomach subvisceral connective spermathecal duct sperm duct shell fragments supravisceral connective visceral artery vascular element running laterally over the digestive gland visceral ganglion end of the mantle cavity 84 G. M. DAVIS PLATE 22. Female reproductive system in Oncomelania hupensis formosana FIGS. 1,4,5. Mature gonads. Membrane was removed from Fig. 5 to show oocytes. PIGS sere Underdeveloped gonads from adult sized snails. FIG. 6. The relationship of the spermathecal duct, pallial oviduct and intestine in the area where the ducts open into the mantle cavity. A anus F fecal pellet In intestine Opo opening of the pallial duct Osd opening of the spermathecal duct Sd spermathecal duct Si subintestinal sinus POMATIOPSIS AND ONCOMELANIA 85 86 G. M. DAVIS width in both species. The glandular patches vary in size. They may continue anteriorly by as muchas 0.09 mm beyond the eye. Abbott (1948b) states these granules to be bright yellow in this species, as in Oncomelania hupensis quadrasi. In reality, the color varies from light, pale yellow to a white- yellow. Very few individuals have been found which have the bright yellow colored glandular units found in O. h. quadrasi. While the yellow tinge is not common in Pomatiopsis lapidaria, pure white granules are very uncommon in O. h. formosana. General Topography. The position and arrangement of organs in both species are the same. The detailed descriptions of the relationships of organs in Po- matiopsis lapidaria likewise pertain to this species. Differences between the 2 snails are mainly ones of structural or dimensional modifications of homologous organs. The descriptions which follow deal mainly with those features which are different as compared with P. lapi- daria. The organ systems of Oncome- lania hupensis formosana are presented in Plates 21-31 in an analogous manner, so that a general comparison with P. lapidaria can be readily made. One feature of the digestive gland of this species deserves attention: the way in which the vascular elements stand out beneath the external epithelium (vas, V, Pls. 21, 24). The blood vessels are so evident because they are outlined by pigment as well as by the white “gran- ules” imbedded in the ventral surface of the digestive gland. The main artery supplying this portion of the vascular system runs under the epithelium above the left ventral edge of the digestive gland. It runs over the edge of the gonad and can be traced to an area on the stomach where the anterior and posterior chambers of the stomach join (Pls. 21, 24). At this point there is a slight depression covered by dense epithelium studded with numerous “granules.” From here the artery can be traced anteriorly to the right where it parallels the esophagus and gonoduct until the mid-region of the right anterior arm of the kidney beneath which it disappears. It passes over the style sac under the kidney tissue, beneath the intestine where the intestine loops over the tip of the style sac and connects with the aorta soon after the aorta sends a main branch anteriorly towards the head. This artery is present in Pomatiopsis lapi- daria but never stood out prominently, as it did in this species. The vessels from the artery running along the gonad pass from left to right, perhaps branching once before passing over the right edge of the digestive gland. These same vessels are not pro- nounced in P. lapidaria and are discerned with comparative difficulty, since the pronounced pigmented outlines are lacking. 3. Mantle Cavity The gill filaments are numerous: 46 + 4. Abbott (1948b) stated that 40-60 ctenidial filaments are characteristic for the genus Oncomelania. The osphradium is 0.5 + 0.09 mm long and its width is about 0.14 + 0.024 mm. Theosphradium (Pl. 24, Fig. 1) of this species is definitely more narrow than that of Pomatiopsis lapidaria. 4. Female Reproductive System The uncoiled female is shown in Plate 21, Fig. 1. The general arrangement of organs and tubes of the reproductive system is similar to that described for Pomatiopsis lapidaria. There are, how- ever, distinct differences to be found in the structure of the gonad, the coiling of the oviduct posterior to the bursa copula- trix, and the arrangement of the ducts leaving the bursa. Gonad. The. ovary is multibranched (Pl. 21, Fig. 1). The average length of this organ is about 1.39 + 0.19 mm and the width is 0.53 + 0.096 mm at the widest portion. The lobed nature of the gonad of Oncomelania hupensis quadrasi was shown schematically by Abbott (1948b). Roth & Wagner (1957) published POMATIOPSIS AND ONCOMELANIA 87 a schematic drawing of the gonad from О. h. nosophora. Roth (1960) presented a schematic diagram for the female герго- ductive tract of O. h. formosana depicting a gonad composed of lobe-like struc- tures. Variation in the size and Shape of the gonad of O. h. formosana is shown in Plate 22. Gonads in Figures 2 and 3 show rudimentary development although they were taken from otherwise adult specimens. The gonads of both of the species investigated showed no variation corresponding to the time of the year, i.e., no Swelling or decrease in pro- ductivity throughout the year. The gonad is rather delicate compared with that of Pomatiopsis lapidaria (Pl. 10). The branches, in the latter species, arise from a large swollen tube while those in the former arise from a slender collecting duct. The fully matured ovaries of Oncomelania hupensis for- mosana have 5 or 6 branched units, each supporting several terminal lobes. The posterior end of the gonad ends in one of the multibranched units. The oocytes of Pomatiopsis lapidaria are distinctly larger than those of this Species, the larger oocytes measuring about 0.17 mm in the former against about 0.11 mm in the latter. Coiled section of the Oviduct. The length of the oviduct between the gono- pericardial duct and the opening of the seminal receptacle into the oviduct measures about 1.7-2.1 mm. The tube is narrower than that of Pomatiopsis lapidaria; the diameter measures up to 0.12 mm. The convoluted oviduct in this region does not form the relatively compact cylinder found in P. lapidaria. It forms 1 or 2 irregular loops (Ov) as shown in Plate 23, Fig. 3. In Figures 3 and 4 the oviduct is shown just beyond the gonopericardial duct. After making 2 irregular coils the oviduct circles under the bursa (B) and encircles the seminal receptacle (Sr) which is appressed against thebursa. The oviduct does not encircle the Seminal receptacle in) Pr \lapidarıa (Pl. 8, Figs. 2, 3). Roth & Wagner (1957) show the coil encirciling the seminal receptacle in a similar manner in Oncomelania hupensis nosophora. Seminal Receptacle. The seminal re- ceptacle (Sr, Pl. 23, Figs. 3, 4) varies in shape and dimension as it does in Pomatiopsis lapidaria. The spherical or elliptical portion of the organ varies in length from 0.17-0.31 mm and the width from 0.12-0.24 mm. A dense central core is often observed within the central portion of the organ. The duct leading to the oviduct varies in length from 0.14-0.19 mm. It is slender, witha width of 0.048 mm. The duct enters the ovi- duct about 0.14 mm from the opening of the sperm duct. In one case (Pl. 23, Fig. 3) the duct entered the oviduct right at the base of the entrance of the sperm duct (Sdu). Bursa Copulatrix. The bursa, viewed from the ventral surface (Pl. 21, Figs. 1, 2; Pl. 23, Fig. 2) measures 0.84 + 0.096 mm in length and 0.38 + 0.048 mm in width. The spermathecal duct (Sd) does not arise from the anterior tip of the bursa as it does in Pomatiopsis lapidaria. It arises anteriorly, together with the sperm duct (Sdu) 0.12 mm from the end of the bursa, or the right side (PE Whigs... 25402. Pl 23. Ш. 2). Study of this area as a whole mount in CMC-10 revealed that the spermathecal duct (Sd) andthe sperm duct (Sdu) emerge from the bursa as 2 distinct tubes bound together in a common sheath of con- nective tissue (z, Pl. 23, Fig. 5), which at magnifications of 60X appeared as 1 tube (z, Pl. 23, Fig. 2). The2 tubes loop to the right and anteriorly. They emerge from the connective tissue sheath just posterior to the point where the oviduct (Ova) passes ventral to the spermathecal duct’ (Sd; PE 21, 2: Pl. 23, Figs. 2, 3, 5). In Plate 23, Fig. 5, the tubes within the connective tissue sheath are shown as dashed lines. Inone specimen the sheath was lacking altogether. Where the tubes loop anteriorly, they are over- grown by tissue of the pallial oviduct (Po, Pl. 21, Fig. 2). The sperm duct 88 G. M. DAVIS PLATE 23. Female reproductive system of Oncomelania hupensis formosana FIG. 1. The terminal portions of the anus, pallial oviduct and the spermathecal duct. FIG. 2. Mid-body region cleared of kidney tissue showing the relationship of the tip of the style sac (Sts) to the pericardium (Pe). Note the point where the gonopericardial duct (Gp) enters the pericardium. Note how the commonly bound spermathecal and sperm ducts leaving the bursa (z) seem to enter the pallial oviduct. FIG. 3. The bursa has been removed to showthe relationship of tubes and structures associated with the bursa. Note the coil about the seminal receptacle (Sr). FIG. 4. The bursa has been rotated to show how the oviduct coils about the seminal receptacle. FIG. 5. The point where the sperm duct and the spermathecal duct leave the bursa in the com- mon connective tissue sheath is shown. The tubes are separate within the sheath (dashed lines). A anus Оу2 portion of the oviduct entering the pal- Ast anterior chamber of the stomach lial oviduct B bursa copulatrix Pe pericardium Cl columellar muscle Pn pericardial nerve Ct ctenidium Po pallial oviduct D digestive gland Pst posterior chamber of the stomach Es esophagus Rn renal nerve Gn gonadal nerve Sbv subvisceral connective Gp gonopericardial duct Sd spermathecal duct In intestine Sdu sperm duct In; _ intestine starting to swing over the tip Si subintestinal sinus of the style sac Sr seminal receptacle Ing _ intestine at the hair-pin turn, the point Sts style sac of the pellet compressor Suv supravisceral connective M edge of the mantle Vg visceral ganglion Mc ventral wall of the mantle cavity y passage from the stomach to the di- Osd opening of the spermathecal duct gestive gland Ov; coiled section of the oviduct beyond the z sheathed tubes leaving the bursa copu- gonopericardial duct latrix POMATIOPSIS AND ONCOMELANIA 0.33mm 89 90 G. M. DAVIS PLATE 24. Male reproductive system of Oncomelania hupensis formosana FIG. 1. Head, verge and mantle cavity. FIG. 2. Stomach and digestive gland as observed without removal of external epithelia. Note the “seminal vesicle” (Sv) irregularly coiled and protruding from behind the gonad (Go). A anus Ast anterior chamber of stomach Ct ctenidium D digestive gland Es esophagus fecal pellet glandular units gonad edge of the mantle the “neck” Ох osphradial ganglion in osphradial pit Pe pericardium Pi pigment Pr prostate Pst posterior chamber of stomach R rostrum Sbv subvisceral connective Suv supravisceral connective Sv seminal vesicle V visceral artery Vd, posterior portion of the vas deferens Vdo anterior portion of vas deferens Ver verge Ve visceral ganglion 2290" POMATIOPSIS AND ONCOMELANIA Swear UNTER 91 92 FIG. FIG. FIG. FIG. FIG. FIG. FIG. ES G. M. DAVIS PLATE 25. Male reproductive system of Oncomelania hupensis formosana The gonad with several of the multibranched units removed to reveal the structure of the seminal vesicle. The gonad as viewed from the ventral surface. The single glandular type found in the verge. The prostate as seen from the ventral surface A and turned over, B,to show the points where the vas deferens enters and leaves the organ. The scale is the same as Fig. 2. The structure of the testicular lobes. Verge. The tip of the verge magnified several times to show the strips of cilia on either side of the papilla. Mvd thick layer of circular muscles encircling the vas deferens at the base of the verge. O¡ initial portion of the vas deferens Pa papilla at the end of the verge Pr prostate Sv seminal vesicle Vd, posterior portion of the vas deferens Vda anterior portion of the vas deferens Ve vas efferens POMATIOPSIS AND ONCOMELANIA 93 94 G. M. DAVIS (Sdu), as it enters the oviduct, measures about 0.12 mm in diameter. Upon entering the common sheath with the spermathecal duct it narrows noticeably to 0.048 mm. Both tubes, emerging from the bursa, have a collective diameter of about 0.036 mm. Roth & Wagner (1957) describe the spermathecal duct as the “vagina” andthe sperm duct as the “duct of the bursa.” They figure the ducts arising from the mid-portion of the bursa in Oncomelania hupensis nosophora. A more thorough investigation of that snail has verified that the arrangement of tubes leaving the bursa is as described for О. h. for- mosana. Pallial Oviduct andSpermathecal Duct. The spermathecal duct (Sd) is a slender tube in this species, with a diameter of about 0.048 mm. Its opening (Osd, Pl. 22, Fig. 6; Pl. 23, Fig. 1) is quite evident at 60X. The complications found in Pomatiopsis lapidaria with the thick, heavily pigmented connective tissue sheets are not encountered here. The opening of the duct is 0.48-0.57 mm pos- terior to the opening of the pallial ovi- duct (Opo). The opening of the sper- mathecal duct is 254 wide and is surrounded by lips 12.54 thick. The pallial oviduct (Po) is about 3.36- 4.00 mm long and up to 0.60 mm wide. This organ is distinctly smaller than that of P. lapidaria. The width of the pallial oviduct at the anterior end of this species is 0.14 mm as against 0.27 mm for P. lapidaria. For a distance of 0.12 mm posterior to the opening, the pallial oviduct is very slender and non- glandular. In both species snails of adult size may possess an underdeveloped pallial oviduct, only 0.28 mm wide for the whole of its length, but this occurs more fre- quently in Oncomelania hupensis for- mosana. There has been some confusion con- cerning the area near the bursa where the oviduct (Ova) passes ventral to the spermathecal duct (Sd). Itagaki (1955) shows the ducts intercommunicating in O. h. nosophora. Current investigations, as well as those of Roth & Wagner (1957), clearly show that these ducts do not intercommunicate. Itagaki (1955) does not mention the sperm duct. He states that the spermathecal duct enters the pallial oviduct near the latters’s ter- minus, as did Dundee (1957) for Pomatio- opsis lapidaria. As Roth & Wagner (1957) showed, this is not the case in O. h. nosophora; the spermathecal duct and the pallial oviduct have separate openings into the mantle cavity. 5. Male Reproductive System Gonad. Thetestis appears as indistinct as that of Pomatiopsis lapidaria when viewed through the ventral epithelium of the uncoiled digestive gland (Pl. 24, Fig. 2). The distinctly outlined vascular pathways further tend to obscure the gonad. The structure of this organ is revealed by removing the ventral epithe- lium from the digestive gland (Pl. 25, Fig. 2). Comparison of the gonad of P. lapidaria (Pl. 12) with that of this species shows several differences. Al- though there are about the same number of multibranched units (7-9) arising from the vas efferens (Ve), the units of On- comelania hupensis formosana lack the many finely branched tubes supporting testicular lobes at their tips. The testicular lobes in this species are thick and elongate, tending to rise from wider, more basal branches very close to the vas efferens (Pl. 25, Fig. 5). The length of the gonad is 1.92 + 0.24 mm and the width is 0.55 + 0.10 mm. The vas deferens arises from the vas efferens at a position (O1) similar to that described for Pomatiopsis lapidaria. The intial tube is not tightly coiled but runs directly into the “seminal vesicle.” The “Seminal vesicle” (Sv) is very characteristic for this species. It never forms the neatly delineated coil described for P. lapidaria (Pl. 12, Figs. 3, 4) but forms a knot, a spherical mass of intertwined tubes bound together by connective tissue strands and laced with vascular elements (Pl. 25, Fig. 1). The POMATIOPSIS AND ONCOMELANIA 95 left edge of this confused mass of tubing projecting out from the dorsal surface of the gonad is generally visible in ventral view (Pls. 24, 25, Fig. 2). The tubes of the “seminal vesicle” are quite narrow, not more than about 0.096 mm wide, but often more slender. The “knot” is about 0.60 mm long and 0.40 mm wide. Uncoiled, the vas deferens, up to the prostate, is about 2.4 mm long. Prostate Gland. The prostate mea- sures 2.20-2.25 mm in length and 0.62- 0.72 mm in width. The relationship of the anterior and posterior portions of the vas deferens to the prostate is the same as that found in Pomatiopsis lapi- daria (compare Pr, Pl. 25, Fig. 4A, B, with Pl. 6, Figs. 2, 3). Verge. Although the verge is charac- terized as “simple” in the literature, it has a number of distinctive features which are of value for comparative pur- poses. The length of the extended verge is 3.36 + 0.12 mm and the greatest width (near the tip) is about 0.80 mm. The verge is thicker and more muscular than that of Pomatiopsis lapidaria. The anterior end is so muscularly thickened that it has a pronounced convex curvature (51: 24. Fig. 1; Pl. 25, Шо. 6). The anterior portion of the verge has а distinct pinkish, salmon color which was also observed by Itagaki (1955) for On- comelania hupensis nosophora. In this area longitudinal muscle strands are readily observed. The tip is blunt and flattened, or in the contracted state, a little concave (Ver, Pl. 24, Fig. 1). When the verge is expanded a dis- tinctly fleshy papilla is pushed beyond the tip (Pa, Pl. 25, Figs. 6, 7). At a magni- fication of 600X, it is evident that the tip of the verge is ciliated on either side of the papilla (Pl. 25, Fig. 7). The ciliary bands extend, invariably, only about 504 on each side of the papilla; they project beyond the epithelium by 3.8 и and actively beat posteriorly. The inner curvature of the verge does not have the expanded glandular area so pronounced in Pomatiopsis lapidaria. Only 1 glandular type was found scattered throughout the verge (Pl. 25, Fig. 3). The verge differs from that of P. lapi- daria in the following: (1) The tip is ciliated; (2) it possesses a distinct protrudable papilla; (3) there is 1 glandular type; (4) there is a pro- nounced musculature which is “pink” in color; (5) it lacks the glandular edge; and (6) the anterior end is not crimped and thereby set off from the rest of the verge (compare with P. lapidaria, Pl. 11, Fig:L5): Roth & Wagner (1957) noted the papilla at the tip of the verge as well as the “several small white lines running parallel to the axis of the penis ... in the tapering portions” for Oncomelania hupensis nosophora. Those “white lines” represent fibers of longitudinal muscle. 6. Buccal Mass The buccal mass and associated struc- tures are similar in the 2 taxa under consideration. The most noticeable difference is one of size. Oncomelania hupensis formosanais uniformly smaller when one compares dimensions of the buccal mass, ganglia, muscles, etc. The length of the buccal mass is 0.79 + 0.07 mm and the width is 0.53 + 0.05 mm. The dorsal view of the buccal mass is shown in Plate 26; the opened pharyngeal tube in Plate 28, Fig. 2. Jaw. The jaws are similar to those of Pomatiopsis lapidaria, except that they are slightly longer. The plates of the jaw are comparatively larger [com- pare Pl. 30, Fig. 7, with Text Fig. 1 (10)]. Radula. The structure of the radula is the same in both species. A series of 16 radulae was straightened out on slides and studied (Table 6). It can be seen that the radula of this species is shorter, less wide and has fewer rows of teeth than P. lapidaria. There was no significant difference in the number of rows of teeth in the formative stages. The cusp formula most commonly en- countered for each tooth in 50 snails investigated is presented in Table 10, along with the percentage of radulae on 96 G. M. DAVIS sx Er NE. FRS DAS A AAA A 4 Ее Rp A 1.) и | Es A н.в ее Mn; Omn PLATE 26. Dorsal aspect of the buccal mass and associated structures in Oncomelania hupensis formosana Bg buccal ganglion Ml; median labial nerve 1 Bm buccal mass Mlg median labial nerve 2 Cc cerebral commissure Мп] mantle nerve 1 Cg cerebral ganglion Omn osphradiomantle nerve Cl columellar muscle Opt optic nerve Es esophagus ррз lateral nerve 3 ms buccal protractor Rp right pleural ganglion 1111 dorsolateral buccal protractor Sa salivary gland m6 preventral dilators SI supralabial nerve 1117 suspensors of the buccal mass Suv supravisceral connective Шор lateral cephalic retractor Tn tentacular nerve POMATIOPSIS AND ONCOMELANIA 97 TABLE 10. A general formula for the most common cusp arrangement found in Oncomelania hupensis formo- sana (from 50 radulae). Snails in which arrangement occurred in at Tooth least 90% of individual teeth % Central (anter. 62 & basal cusps) Lateral 2-1-2(3) 74 Inner Marginal 8-9 100 Outer Marginal 6-5 91 which the arrangement was found for at least 90% of the teeth of each category. Only 53% of these 50 radulae had the representative formula shown in column 2 of Table 10. Table 11 shows every cusp arrangement found for each tooth type, with the percentage of radulae on which it occurred at least once. Studies onthe radulae of the subspecies of Oncomelania by Abbott (1948b), Mao & Li (1948) and Kuo & Mao (1957) reveal the fact that variation in radular for- mulas for a single subspecies of On- comelania hupensis includes the cusp arrangments of the other subspecies as well as of Pomatiopsis lapidaria. How- ever, those studies did not generally consider the frequency of occurrence of cusp arrangment. In the radulae (snails) investigated in this study, 70% of O. h. formosana had a central tooth formula of 1-1-1/2-2 (at least once; Table 11) and 62% of the snails had this formula represented in at least 90% of their central teeth. In Pomati- opsis lapidaria only 40% of the radulae had central teeth withthis formula (Table 8) and the formula was notas commonas that of 1-1-1/3-3 (Table 7). Considering the inner marginals, 5-7 cusps were common in P. lapidaria (84%, Table 7) TABLE 11. The various types of cusp ar- rangement for the different teeth in 50 radulae of Oncomelania hupensis formosana and the per- centage of radulae showing that arrangement at least once. Central Lateral Arrangmt. of cusps anterior basal Arrangmt. of cusps 3-1-4 one side, 2-1-2; other side, 20 2-1-3 one side, 2-1-3; other side, 10 2-1-4 Outer marginal Inner marginal No. of cusps while 8-9 were the rule in O. h. formo- sana (100%, Table 10). No radulae of P. lapidavia were observed where any inner marginal had 9 cusps (Table 8). The question arises, whether other popu- lations of Oncomelania conform to the pattern found inthe laboratory population studied here. For the most part that 98 G. M. DAVIS question remains unanswered, except for data on 1 population of O. h.hupensis in China presented by Mao & Li (1948). The central had a formula 1-1-1/2-2 for 75% of the population. Centrals with 3 basal denticles occurred in only 26% of the snails. There were 8-9 cusps on the inner marginals in 87% of the popu- lations, the remainder having 7. In spite of the overlapping variations between the subspecies of Oncomelania and Pomatiopsis lapidaria, it appears that P. lapidaria is separable from On- comelania on the basis of cusp formula in that there is a pronounced tendency in the former to have centrals with 3 basal cusps on each side as compared with 2, and 5-8 cusps on the inner marginal, as compared with 8-11. Aside from the differences in length, wdith, and number of rows of teeth, there are 2 other small differences. The large mediobasal cusps of Oncomelania hupen- sis formosana are more widely separated than those of Pomatiopsis lapidaria (compare Pl. 19 with Pl. 29). Also, there is a pronounced tendency in the outer marginal to have an enlarged and swollen outer cusp (Pls. 29, 30), a con- dition rarely observed in P. lapidaria. The manner in which the teeth in both Species are attached to the membrane is shown in Plate 30, Fig. 6. 7. Musculature The musculature is shown in Plate 13, Fig. 1, and Plates 26-28. The patterns and arrangements of muscles are the same in both species withfew exceptions (compare with Pls. 14-18). There are grades oof variation in muscular structures which reach extremes in Oncomelania hupensis formosana. The preventral protractors (mg, Pl. 28, Fig. 1) tend to be more strongly developed than in Pomatiopsis lapidaria, although this is variable. The suspensors of the buccal mass (mj37, Pl. 26; Pl. 28, Fig. 1) tend to be more stout and fewer innumber than those found in P. lapidaria, although there is a gradation to the exact con- dition found in the latter species. In several cases stout muscle strands from the anterior jugalis (mg, Pl. 28) were observed running anterodorsally to the anterior rostral roof, a condition never observed in P. lapidaria. There are 3 distinct differences be- tween the musculature of these 2 species: (1) The rostral retractors (mag, Pl. 27) terminate shortly after passing over the pedal haemocoel. They do not pass under the buccal protractors (m5) to in- sert about the oral aperture at the floor of the tip of the rostrum as they do in Pomatiopsis lapidaria. (2) The buccal protractors commonly fuse in a single sheet which takes its origin from the rostral floor (m5, Pl. 27), in contrast to the condition normally found in P. lapidaria (ть, Pl. 18). There also are mediolateral slips, not found in P. lapidaria, which have their origin on each side of the oral aperture and run posteriorly, and then dorsally, to unite with the buccal protractor near the rostral floor before the muscle bifur- cates; but as they are anterior to the main sheet of m5, they cannot be seen in Plate 27. The 2 m; labels in Plate 28, Fig. 1, refer to the mediolateral slip and the posterolateral portion of the sheet, respectively. (3) The mediolateral slips are crossed by a circular band of muscles (m, Pl. 28) which arise from the floor of the rostrum at either side of the oral aper- ture and sweep around in an arc over the mediolateral slips of the buccal protractors. This muscle band does not correspond to the oral sphincter muscle. Usually this arc of muscle is posterior to the oral sphincter. 8. Nervous System Several differences are found between the species with regard to the nervous system. (1) Upon opening the rostral cavity it is seen that the ganglia and nerves are not dusted with the heavy pigment, as was the case in Pomatiopsis lapidaria. (2) The cerebral commissure (Cc, Pl. 31) is short in Oncomelania hupensis POMATIOPSIS AND ONCOMELANIA 99 0.33 mm Es PLATE 27. Musculature of the cephalic region of Oncomelania hupensis formosana posterodorsal gap of the pedal haemo- coel buccal commissure buccal ganglion cerebrobuccal connective columellar muscle external odontophore membrane esophagus buccal protractor preventral protractor radular protractor dorsolateral buccal protractor 112 buccal retractor 1120 rostral retractor 1122 lateral cephalic retractor 1124 tensor magnus Mb membrane around the radular sac М1 median labial nerve 1 Mly median labial nerve 2 Py lateral retractor nerve from the pedal ganglion Rs radular sac Sul sublabial nerve 100 G. M. DAVIS Musculature and nerves of the cephalic region of Oncomelania hupensis formosana 1. The lateral view of the buccal mass with the central portion of the nervous system. 2. Dorsal aspect of the opened pharyngeal tube, showing the dorsal odontophore. 4. The overlapping tips of the cartilages are pulled apart to expose their medial surfaces, which are wrapped in the intrinsic muscles of the odontophore. membranous jugalis preventral dilators suspensors of the buccal mass median labial nerve 1 median labial nerve 2 outer lip optic nerve lateral retractor nerve pedal nerve to the anteroventral wall major lateral nerve dorsolateral pedal nerve pedal ganglion pleuro-pedal connective lateral nerve 1 penial nerve lateral nerve 3 lateral nerve 4 radula radular shield right pleural ganglion rostral portion of the cephalic haemocoel rostral wall salivary gland subradular membrane statolith statocyst sublabial nerve tentacular nerve ventral fold PLATE 28. FIG. FIG. FIG. 3. The right buccal cartilage and associated intrinsic muscles. FIG. bi dorsal buccal nerve b4 anterior buccal nerve bs odontophoral nerve Bg _ buccal ganglion Cg cerebro-tensor nerve Ca cartilage Cb cerebro-buccal connective Cg cerebral ganglion Cos collostyle tip Cp cerebro-pedal connective Eo external odontophore membrane Es esophagus Ev esophageal valve Fg food groove Gro central groove in the ventral fold Il inner lip J jaw m circular muscle running over the medial slips of the buccal protractors my lateral cartilage tensor mo mediolateral cartilage tensor mg odontophore divaricator m4 medial radular retractor ms buccal protractor mg preventral protractor m7 radular protractor mg anterior jugalis mg buccal constrictor 111 dorsolateral buccal protractor buccal retractor 101 POMATIOPSIS AND ONCOMELANIA CET LAN NT ref — 102 G. M. DAVIS PLATE 29. Variation in the radular teeth of Oncomelania hupensis formosana The horizontal top row of teeth displays 1 radular row in natural position* except for the outer marginals 1 (left) and 2 (right) which are erected to expose the exact number of cusps. Variation is shown for the other teeth. The description for Pomatiopsis lapidavia given for Plate 19 generally applies and should be consulted. The plane of focus on central teeth 1-5 is on the upper surfaces of the medial basal cusps. The supports for those cusps are not shown but are similar to those of P. lapidaria. The tongue-shaped attachment (basal process) described for central 6 of Plate 19 is shown diagramatically for centrals 1-2. The anterior end of central 3 is raised thereby demonstrating the dagger-like cusps as seen from this orientation. Lateral teeth 1-4 (left) and 5-7 (right) show variation in cusp number and shape. Lateral 6 shows only the cusps of the tooth; the anterior edge of the tooth is lifted upwards. Lateral 7 is shown with peduncle (Pd, Pl. 19) turned to the right 90° from the normal position to show the hook-like nature of the cusps. Inner marginals 1 (left) and 2-4 (right) are all shown in normal position. *When the plate is turned sideways so that the labels are horizontal. 103 POMATIOPSIS AND ONCOMELANIA AQUUI S[CULS IVI [81938] ТелдаеЭ [819387] JQUUI Iayno STRUISIPIA PLATE 30. Radula of Oncomelania hupensis formosana FIGS. 1,2,3. Inner marginals. FIGS. 4,5. FIG. 6. FIG. 7. Outer marginals. The manner in which the teeth are attached to the radular membrane. Jaw. POMATIOPSIS AND ONCOMELANIA 105 TABLE 12. Anatomical differences between Pomatiopsis lapidaria and Oncomelania hupensis Jormosana considered to be of major importance Character Shell 1. umbilicus 2. varix 3. sutures 4, apex be Lip 6. whorls Gill filaments Tentacles Female Reproductive System 1. gonad branches 2. gonad collecting duct 3. oviduct coils encircle the seminal receptacle 4. spermathecal duct leaves bursa copulatrix 5. Sperm duct arises from Male Reproductive System 1. verge with papilla 2. verge ciliated 3. verge with glandular edge pronounced 4. verge with pronounced musculature 5. seminal vesicle Nervous System 1. cerebral commissure 2. pleuro-supraesophageal connective 3. Supravisceral connective arises on supraesophageal ganglion from osphradial and mantle nerves: dis- tance from origin to lateral cephalic wall 4 + = уез - = по formosana. It is 0.07 + 0.024 mm long and about 0.06 mm wide. In P.lapidaria it is at least 0.12 mm long and 0.07 mm wide. This difference in length is not necessarily due to the generally smaller structures in О. h. formosana. (3) The pleuro-supraesophageal con- nective (Psc, Pl. 31, Fig. 1) is short, measuring 0.168 + 0.048 mm as against at least 0.288 mm in Pomatiopsis lapi- - P. lapidaria O. hupensis formosana wide deep wide straight 6.5-7.0 30 or less short at anterior end spermathecal duct thick; neatly coiled long long posterolateral surface short; bifurcating soon after origin to run separately daria. narrow + moderately deep narrow sinuate 7.0-7.5 35 or more long several slender + laterally (in common sheath w. sperm duct). bursa copulatrix + + ES slender, knotted tube short short tip long and ruming jointly As a result of the shortened connective the supraesophageal ganglion (Sug) rests on the dorsolateral surface of the esophagus. The osphradio-mantle nerve (Omn) and the supravisceral con- nective (Suv) arise from the tip of the ganglion and travel about 0.39 mm tothe left lateral body wall (Pl. 26). In P. lapidaria, due to the lengthened pleuro- supraesophageal connective, the tip of the 106 G. M. DAVIS PLATE 31. . 2. Anterior aspect of the pedal ganglia. arises. nerve from pg nerve from pg pedal commissure cerebral commissure cerebro-buccal connective cerebral ganglion cerebro-pedal connective cerebro-tensor nerve external mantle cavity nerve 1 external mantle cavity nerve 2 external mantle cavity nerve 3 external mantle cavity nerve 4 gonadal nerve left pleural ganglion mid-columellar nerve metapodial ganglion median labial nerve 1 median labial nerve 2 mantle nerve 1, from the left pleural ganglion mantle nerve 3, from the subesophageal ganglion osphradio-mantle nerve optic nerve РУ P2 Nervous system of Oncomelania hupensis formosana . 1. Dorsal aspect of the central nervous system or “brain. ” . 3. Medial aspect of the right cerebral ganglion showing the position where each nerve . 4 Ventral aspect of the visceral ganglion and associated nerves. lateral retractor nerve, from the pedal ganglion nerve to the anterioventral wall pedal haemocoel major lateral nerve of the pedal ganglion propodial connective metapodial connective minor lateral nerve of the pedal ganglion pedal ganglion pericardial nerve pleuro-pedal connective propodial ganglion pleuro-supraesophageal connective renal nerve right pleural ganglion supralabial nerve subesophageal ganglion subvisceral connective supraesophageal ganglion sublabial nerve supravisceral connective tentacular nerve visceral ganglion of the POMATIOPSIS AND ONCOMELANIA 107 108 G. M. DAVIS TABLE 13. Anatomical differences between Pomatiopsis lapidaria and Oncomelania hupensis formosana indicative of specific rank Characteristic Pigmentation Glands at edge of eyes Digestive gland with prominent vasculari- zation Radula 1. length 2. width 3. total rows of teeth Size of organs Osphradium Muscles 1. short rostral retractor 2. mediolateral slip from buccal pro- tractor 3. circular muscle over buccal pro- tractor Male Reproductive System 1. gland types in verge 2. testicular lobes Female Reproductive System 1. opening of the spermathecal duct obscure 2. ova Nervous System 1. buccal nerve 3 2. external mantle cavity nerve 4 3. branch of tentacular nerve arises at + = yes supraesophageal ganglion almost touches the lateral body wall (Pl. 14); the supra- visceral connective arises from the pos- terior side of the ganglion near its tip, but not from the tip (Pls. 14, 20). The osphradio-mantle nerve bifurcates to form the osphradial and mantle nerves just after leaving the ganglion and before entering the lateral wall (see p 79). (4) The main branch of the tentacular nerve (Tn, Pl. 31) arises from the basal swelling of the tentacular nerve, not from the mid-length of the nerve as in Po- no dimorphism white-yellow, white short and slender mid nerve Р. lapidaria O. hupensis formosana sexual dimorphism yellow, white-yellow 0.98 mm 0.12 mm 84 smaller narrower elongate and thickened smaller matiopsis lapidaria. (5) Buccal nerve 3 (bg, Pl. 15, Fig. 1) was not found in Oncomelania hupensis formosana. (6) The other differences are mainly ones of size (compare Pls. 31 and 20). A few minor differences might be men- tioned. The minor lateral nerve (pg, Pl. 31, Fig. 2) from the pedal ganglion was frequently encountered in O.h. formosana while it was infrequently found in P. lapidaria. The external mantle cavity nerve 4 (E4, Pl. 31, Fig. 4), not POMATIOPSIS AND ONCOMELANIA found in P. lapidaria, was observed al- though it varied in strength and was sometimes absent. It rananterolaterally over the mantle wall epithelium. E. Summary and Discussion Abbott (1948a) stated that the repro- ductive and nervous systems of Pomati- opsis and Oncomelania showed few differences. Actually, the major differ- ences found between the species were in these systems, especially in the repro- ductive systems. In Table 12 are listed those differences which I feel are important in defining the generic separ- ation of Oncomelania and Pomatiopsis. Differences of a specific nature are listed in Table 13. Because many of the characters listed are unknown for the other species of Pomatiopsis as well as for the subspecies of Oncomelania, only future investigations of these other forms can show whether these characters are specific or representative of the genus as a whole. Differences in the shell clearly sepa- rate Oncomelania from the species of Pomatiopsis. P. binneyi, however, has a shell quite unlike those of either On- comelania or the other species of Pomatiopsis. It is tiny, about 3.0 mm high; imperforate, and the inner and outer lips are continuous, i.e., there is no parietal callus andthe lips are slightly separated from the body whorl. P. binneyi will be discussed in the final summary statements, asthis formis also aberrant in other ways from the other species of Pomatiopsis. The higher number of gill filaments also clearly separates Oncomelania from species of Pomatiopsis. All the species of Pomatiopsis have shorter tentacles, relative to the length of the rostrum, than Oncomelania. None of the 4 species of Pomatiopsis investigated have the terminal papilla found in the verge of Oncomelania, Blanfordia and Tomichia. Two species of Pomatiopsis, P. cincinnatiensis and P. californica, have penial filaments, 109 unknown in the above genera. Oncomelania and Blanfordia have the characteristic ciliation of the tip of the verge described for O. hupensis formo- sana. Tomichia ventricosa has the same active cilia, but in this speciesthe bands of cilia extend half way back along the verge. In these 3 genera the cilia are about 4u in length and beat actively. The verges of Pomatiopsis lapidaria and P. cincinnatiensis lack cilia. In initial studies on P. binneyi and P. californica, 1 found that the tips of the verges have bristle-like cilia. Unlike those of Oncomelania, they are not active, a cilium here or there beating slowly once in a while. They were bushy, irregularly oriented, and 6-8u in length. In P. californica they extended part way out on the penial filament. There are clear-cut differences inthe origin and the relation of the sperma- thecal and sperm ducts. The sper- mathecal duct arises at the anterior end of the bursa copulatrix in P. lapidaria and from its right ventro-lateral surface in Oncomelania and in 2 other members of the Pomatiopsinae I have studied, Blanfordia japonica and Tomichia ventri- cosa. Thespermathecal and sperm ducts originate together from the bursa, inone common sheath, in Oncomelania and Blanfordia, while in Pomatiopsis and Tomichia the sperm duct arises from the spermathecal duct near its junction with the bursa. Thus, Tomichia shows an intermediate position, in that the sperm duct arises from the spermathecal duct, as in Pomatiopsis, but the spermathecal duct arises laterally from the bursa, as in Oncomelania, and not at the anterior end. The bursa in Tomichia is about twice as long asin Pomatiopsis lapidaria and in the subspecies of Oncomelania hupensis. The arrangement of bursa and ducts in Pomatiopsis might possibly be derived from the condition found in Tomichia by a reduction in length of the anterior end of the bursa (the reverse derivation being likewise open to con- sideration). Howthetubes arise from the bursa is unknown for P. californica and 110 G. M. DAVIS P. binneyi. The slender collecting duct ofthe ovary and the numerous branches separate On- comelania from Pomatiopsis lapidaria and P. cincinnatiensis. The condition is unknown for P. californica and P.binneyi. The oviduct coils around the seminal receptacle in a characteristic mannerin the subspecies Oncomelania hupensis quadrasi, O.h. formosana and O.h. nosophora. The condition is unknown in О. h. hupensis. The arrangement does not occur in Pomatiopsis lapidaria and P. cincinnatiensis. The condition is un- known for P. californica and P. binne yi. The pleuro-supraesophageal con- nective in all the subspecies of Oncome- lania is relatively short, as described, a condition also found in Blanfordia japonica. In these snails the common osphradio-mantle nerve leavingthetip of the supraesophageal ganglion is corre- spondingly long and usually does not bi- furcate before entering the wall of the “neck.” The supravisceral connective arises also from the tip of that ganglion. In the species of Pomatiopsis studied, on the other hand, the pleuro-supra- esophageal connective is elongate: van der Schalie & Dundee (1956) show a long connective for P. cincinnatiensis, and I also found it so in P. binneyi and P. lapidaria. In these 3 species of Pomati- opsis the tip of the ganglion is close to the lateral wall, and the mantle and osphradial nerves, which frequently bi- furcate soon after leaving the tip of the ganglion, have a very short lengthbefore entering the lateral wall. In P. binneyi, as in P. lapidaria, the supravisceral connective arose from the posterolateral surface of the supraesophageal ganglion near, but not from, the tip. P. cali- fornica has not been studied. HYBRIDIZATION STUDIES Success in hybridizing the subspecies of Oncomelania was reviewed by Davis et al. (1965). Van der Schalie, Getz & Dazo (1962) reported success in hybridization experiments when male Pomatiopsis lapidaria were placed in culture with virgin female O. hupensis quadrasi and O. hupensis formosana. They did not report success with crosses involving female P. lapidaria and male Oncomelania. Cross cultures were again set up be- tween male Pomatiopsis lapidaria and virgin females of the various subspecies of Oncomelania inorder to obtain hybrids for anatomical studies. Three different sets of experiments were established over a total of 26 months in which adult, male P. lapidaria were maintained under various culture conditions with virgin female Oncomelania. In all the experi- ments males which died were replaced. Females that died were removed from the culture but not replaced. Cultures which deteriorated due to soil erosion, mold, or algal- accumulation were re- placed by fresh cultures. The deteri- orated cultures were maintained for at least 1 month in order to observe if any young had hatched from eggs laid just prior to changing the culture. All cultures were maintained at 240 + 20 С. Experiment 1. (A) Seven plastic tray cultures (see p 118) were extablished with 10-20 females in each, along with 13-40 males, no culture having more than 50 snails. The cultures were maintained in normal room level light during the day and were in the dark at night. One culture was set up with female Oncomelania hupensis nosophora, 2 with O. h. quadrasi and 4 with O. h. formo- sana. (B) Eight cultures were set up using 3 inch diameter clay flower pots, 1 inch deep, partially filled withloam and main- tained exaclty as those described by van der Schalie et al. (1962). Five speci- mens of each sex were maintained inthe cultures. The only females used were O. h. formosana. In both (A) and (B), the cultures were routinely serviced each day along with the 80 other stock cultures on hand. Within 3-7 months 5 of the plastic tray POMATIOPSIS AND ONCOMELANIA 11 cultures contained young. Upon re- determining the sex of all the adults in the cultures (methods of Wong & Wagner, 1954) it was demonstrated that in every case where young were pro- duced, 1 or 2 males of O. h. formosana or O. h. quadrasi were present. Such contamination had occurred in2 cultures in the 7th month, after the sexes of all the snails had been rechecked in the 5th. In the 7th month all the cultures were placed in isolation and serviced with tools set aside for the cross cultures only. Those cultures, in which con- tamination had occurred, were discon- tinued. When these precautions were taken, no further young were produced, although the cultures were maintained and constantly observed until the 20th to the 26th month. Experiment 2 Thirteen cultures were established in medium size clay pots, atype of vivarium found to provide optimal conditions for the production of young Oncomelania (van der Schalie & Davis, 1968, in press. See p 119, e). Five specimens of each sex were placed in each culture. Two of the cultures contained female O. h. quadras?; 11 of the cultures contained female O. h. formosana. Five of the cultures were maintained under constant light provided by cool, white fluorescent bulbs. The intensity of light was 75 ft. candles. All cultures were carefully maintained in isolation. The experiment was discon- tinued after 14 months. No young were produced. All control cultures with On- comelania males produced young within 2 to 3 months of being established. Male Pomatiopsis lapidaria were ob- served in copulation with the Oncome- lania females many times and spot examination of the male gonads showed them to be fully mature and productive. Experiment 3 Another 16 medium clay pot cultures were established and separated into blocks of 4, each block being provided with male Pomatiopsis lapidaria from a different population. These were col- lected and placed in cultureinthe spring, a time of year when the species exhibits pronounced sexualactivity. Two cultures of each block were established with virgin O. h. quadrasi and 2 with virgin О. h. formosana. Twofemales were uni- formly used in all the cultures with alternately 2 or 4 males. The cultures were maintained in isolation. Copulation was noted frequently but no young were produced over a period of 5 months. At the end of 5 months the cultures were discontinued. Females removed fromtheterminated cultures were fixed in Bouin’s solution, sectioned, stained with standard Hema- toxylin and Eosin and studied to de- termine if the ovaries were productive and if the seminal receptacles were storing sperm. In no case was sperm noted in the seminal receptacle or bursa copulatrix. Oocytes posterior to the gonopericardial duct often showed signs of atrophy and deterioration. Summary and Discussion As a result of these experiments it was concluded that Pomatiopsis lapidaria will not produce an hybrid Fy generation when interbred with Oncomelania. It is felt that the results of previous experi- ments reporting success insuchcrosses were possibly due to contamination of the cultures with male Oncomelania during routine maintenance. Burch (1960b) reported that the sper- matagonial cells of P. lapidaria have 33 chromosomes, 16 pairs and “a hetero- chromatic chromosome, presumably a sex chromosome.” The oögonial cells of Oncomelania hupensis quadrasi and O. h. nosophora have 34 chromosomes (17 pairs). The spermatogonial cells of Pomatiopsis cincinnatiensis have 30 chromosomes plus a heteromorphic pair. According to Patterson (1963) the sex determining mechanism in P. lapidaria is of an XO type while in P. cincinnatiensis and O. h. formosana the sex determing mechanism appears to be an XY type. Although a cross between individuals with 2n=32 and 2n=34 is possible, a 112 G. M. DAVIS successful cross between P. cincinnati- ensis (2n = 32), having a heteromorphic pair of sex chromosomes, and Oncome- lania (2n = 34), which lacks a corre- sponding heteromorphic pair, is highly improbable. Crosses between P. lapi- daria and subspecies of O. hupensis are improbable because of the apparent difference in the sex determing mechan- ism. Also, other karyotypic differences in the chromosomes of the 2 species of Pomatiopsis and subspecies of Oncome- lania hupensis indicate that there would probably not be a sufficient number of similar homologues between them to permit successful hybridization (Burch, personal communication). ELECTROPHORETIC STUDIES A. Introduction Little is known about electrophoresis as applied to molluscan systematics. Cheng (1964) and Davis & Lindsay (1967) review previous work pertaining to mol- lusks. Cheng, using membrane electro- phoresis, investigated several species of marine and freshwater gastropods as well as a few marine pelecypods (also 1 sphaeriid). The gastropods were widely separated taxonomically. Using serum, he obtained up to 5 fractions, although there were generally only 1-3 fractions. From his results he concluded that each “species” could be identified by its serum electrophoretic pattern. He states that “undoubtedly much more extensive surveys of the serum proteins of mol- lusks must be made before useful taxo- nomic information will be obtained.” Wright & Ross (1959) found that paper electrophoretic analysis of proteins in snail blood was not satisfactory and began using cellulose-acetate electrophoresis (1959, 1963). Their studies turned to gastropod egg proteins (1963, 1965) to provide characters of use inthe taxonomy of planorbid snails. In 1963 they published data showing that blood proteins and haemoglobin varied considerably both quantitatively and qualitatively with progressive growth and development of sexual maturity in Biomphalaria glabrata (=Australorbis glabratus). As a result they stated that “the results of this work confirm earlier doubts concerning the taxonomic value of molluscan blood proteins....” Davis & Lindsay (1964, 1967) studied proteins from foot muscle and blood of Helix pomatia using polyacrylamide electrophoresis. Proteins fromthe blood yielded 12, from foot muscle extract 20, components, which showed no qualitative changes with snails of different size (age). However, when size of snail was correlated with protein density, there was а significant inverse quantitative change with haemolymph, but not with foot muscle extract. They (1967) also Showed that different populations of Pomatiopsis lapidaria had population- specific protein patterns. Despite significant variation between popu- lations, the species was characterized by a densitometric pattern clearly recog- nized in each population. Michelson (1966) studied the haemo- lymph of Biomphalaria glabrata using polyacrylamide electrophoresis. He reviews literature pertaining especially to mollusks involved in host-parasite relationships. Michelson reported that size did not affect qualitative results but that density in blood proteins in B. glabrata increased with increased size. His results regarding increased density of blood proteins with larger snails are contrary to those of Davis & Lindsay (1964, 1967) with haemolymph of Helix pomatia. The purpose of the present investi- gation was to study the electrophoretic patterns of Pomatiopsis lapidaria and Oncomelania hupensis formosana in order to determine whether distinct taxon characterizing patterns could be ob- tained. This investigation was under- taken as a preliminary step for com- parative studies involving the other sub- species of Oncomelania hupensis and species of Pomatzopsis. B. Materials and Methods Disc electrophoresis. The technique POMATIOPSIS AND ONCOMELANIA 113 B — CATHODE Sample gel- 1/4" space À р CATHODAL BUFFER BATH Protein GLASS TUBE Separating with inside PLASTIC TUBE gel- 2" Diameter of 3/16" CUT AWAY VIEW OF Stand to hold the glass tube SILICONE STOPPER Solid, polymerized polyacrylamide gel GLASS TUBE ANODAL BUFFER BATH + ANODE TEXT FIG. 2. Diagrammatic set-up for disc electrophoresis A. A glass tube is set upright in a supporting base stand. Three gel solutions are poured into the tube, one atatime. Each solution is allowed to polymerize before the next is added. The protein separating gel is the standard 7. 5% acrylamide gel. When the uppermost solution is polymerized, the glass tube with the internal gel column is removed from the base stand. B. The glass tube containing the gel column is held in place in the upper cylindrical cathodal buffer bath by means of a perforated silicone stopper while the lower end hangs freely in the anodal buffer bath. Usually, as many as 12 cathodal units were used in a common anodal bath. 114 G. M. DAVIS FOLIN-CIOCALTEU PROTEIN ESTIMATION X 102 СМ о 100 200 300 400 (Density Units vs. Wet Weight Of Tissue) Ye = 5.400 x 10°* + 3.480 x 10°° X + 5.437 x 10°° x? 500 600 700 800 900 DENSITY UNITS TEXT FIG. 3. Estimation of protein in foot muscle tissue The density units are direct readings from the scale of the Klett colorimeter. Results were not linear with smallest weights of tissue: with X taken at zero the Y intercept (Ус) was 5. 400 x 10-4 grams of wet (but blotted) weight of tissue. of disc electrophoresis was described in detail by Ornstein (1962, 1964) and B. J. Davis (1964). The reader is also referred to bibliographies on work per- taining to disc electrophoresis available through the Canalco Corporation, Rockville, Maryland, U.S. A. The methods used in this study were discussed fully by Davis & Lindsay (1967). The standard 7.5% acrylamide gel was used. In Text Fig. 2 is shown a schematic drawing of the arrangement of gels polymerized in the glass sup- porting tube (A) and a drawing of how one such tube is positioned so as to bridge the 2 buffer baths (B). The buffer was a tris*-glycine mix- *tris = 2-amino- 2 hydroxymethyl -1, 3-рго- panediol ture with a pH of 8.2-8.4. A constant current of 5 milliamperes was passed through each tube and was maintained by hand regulation of a Heathkit power supply. Bromphenol blue dye was added to the cathodal bath prior to starting the current through the gels and served as a tracking dye to indicate the position of the front or leading band moving through the separating gel towards the anode. When the front band had moved 33 mm into the protein separating gel, the current was turned off; the gels were removed from the glass tubes and placed in Amido-Schwartz stain. After 2 hours of staining, the gels were destained in acetic acid. Preparation of sample. Proteinsfrom foot muscle tissue were extracted in Carriker’s (1946b) physiological saline. POMATIOPSIS AND ONCOMELANIA The reasons for using foot muscle are fully discussed by Davis & Lindsay (1964, 1967). The tissue of 20-50 snails was pooled and 0.15 gm (wet weight, blotted tissue) were homogenized in 1.0 ml saline using a Servall microhomoge- nizer (50,000 rpm). The tissue was homogenized for 30 seconds and then examined to see whether all the tissue was “taken.” If some pieces remained the muscle was homogenized for another 30 seconds. All operations were carried out at 20-30 C. The homogenate was centrifuged at 1200 rpm (250X G) for 5 minutes. Super- natant was mixed with the sample gel in a ratio of 1:2. Only 100 lambda of the mixture were polymerized above the spacer gel in each glass tube (A, Text Fig. 2). An estimate was made of the relative amount of protein present in the super- natant fluid after centifugation as wellas how much actually was separated in the electrophoretic runs. The Folin- Ciocalteu reagent test, based on colori- metric procedures, was applied to de- termine the relative weight of protein in the supernatant and in the electro- phoretic runs. Shreds of foot tissue were weighed on a Mettler Microbalance and subsequently submitted to the Folin test. Wet weight (blotted shreds) of tissue was plotted against Klett colori- metric readings (Text Fig. 3). Fromthe resulting curve it was possible to deter- mine the relative amount of protein in the supernatant fluid in relation to the initial wet weight of tissue. The tissue, of course, was not entirely protein, i.e., some weight was contributed by carbohydrates and fat, but the relative approximation of weights gives a useful indication of the fate of homogenized tissue. It was determined that 27% of the initial wet weight of tissue are found in the total volume of supernatant after homogenization and centrifugation. When supernatant was mixed with upper gel and electrophoresis was completed, it was found that 45% of the proteins in the sample gel were too crude to pass into 115 the spacer gel, 21% remained in the spacer gel, and 33% migrated into the separating gel. At least 20 experiments were per- formed for each species. Each experi- ment consisted of 4-5 tubes loaded with aliquots of a single muscle preparation. Analysis of results. Densitometric tracings of the distribution of the pro- tein fractions in the gels were made using the unmodified Canalco Model E microdensitometer. Later studies (Davis & Lindsay, 1967; published before the present account) were made with the densitometer modified to give an expanded tracing with a more clearly defined densitometric pattern. Rf values for the fractions were deter- mined from the densitometric tracings when peaks were pronounced; they were calculated from direct measurement of the gels when a band was diffuse or faint. An Rf value is the ratio of the distance from the origin to a given fraction and the distance from the origin to the front band. It serves to mini- mize differences in band migration due to small differences in the length of the “run” which did not always measure exactly 33 mm. Measurements were made using a ruler accurately cali- brated in 0.5 mm units. Results were analyzed by studying the differences between Rf values and densitometric patterns of thetaxa. More detailed information on the use of Rf values in comparing taxa can be found in Davis & Lindsay (1967). When Rf values of fractions in different taxa varied by 0.014 or less, they were con- sidered analogous because this value represented the greatest error for value determination when different people measured the same component from the gel. C. Results The results of the study are illustrated in Plate 32, Figs. 1, 2 for Pomatiopsis lapidavia (Parker Mill population) and Oncomelania hupensis formosana, re- 116 G. M. DAVIS PLATE 32. Electrophoretic comparison of Pomatiopsis lapidaria and Oncomelania hupensis formosana 1. P. lapidaria 2. O. h. formosana The vertical lines beneath the densitometric tracings (solid black) represent the separated protein components from the foot muscle. The photographs above each tracing were made by using the stained gels as negatives and placing them under the enlarger. Prints were made directly from the gels. POMATIOPSIS AND ONCOMELANIA 117 TABLE 14. Representative Rf values* for the separated protein components illustrated in Plate 32 Band Pomatiopsis Oncomelania lapidaria hupensis formosana 1 0. 014 0. 014 (1)** 2 0. 056 0.042 (2) 3 0. 099 0.070 4 0.141 0.098 (3) 5 0.196 0.133 (4) 6 0. 244 0.198 (5) 7 0. 296 0. 266 8 0. 350, 0. 390 0. 314 9 0. 440 0. 410 10 0. 497 0. 464 11 0. 552 0. 515 12 0.601 0.649 (13) 13 0. 640 0.691 14 0.718 0.765 15 0.818 0.821 (15) 16 0.965 0.896 17 1. 000 1.000 (17) * Averages of data from numerous tubes. ** The number to the right of the Rf value refers to the band of P. lapidaria with which the Rf is analogous, i.e. , differs by 0.014 or less. spectively. The lines beneath the peaks of the 2 typical densitometric graphs represent the linear spacing of 17 pro- tein fractions indicating the components distinguished. Thetubes andtracings are representative and reliably portray the species differences despite the small variations that do occur between different experiments on homologous prepa- rations. Average Rf values for the separated fractions are listed in Table 14. Com- parison shows that only about 44% of the components were analogous in the 2 species. The reader is reminded that analogy does not mean homology, and that homology must be proved by bio- chemical and/or immunological means. As stated by Davis & Lindsay (1967), P. lapidaria is particularly charac- terized by (1) fractions 11-13 with the twin dense peaks at 12 and 13; (2) by the fact that the area from band 13to the front is devoid of high density com- ponents. In P. lapidaria, band 8 was frequently split into 2 bands (Table 14). O. h. formosana is characterized by bands 12-15, bands 14 and 15 being close to each other and 15 being the less dense. Band 13 is always very faint while 12 is dense. Bands 9-11 are wide and diffuse. Generally bands 1 and 3 are very dense and wide so as to hide band 2. Results in Plate 32, Fig. 2, where band 1 is not dense, are the exception. D. Discussion Initial studies with the other sub- species of Oncomelania hupensis re- vealed that in contrast to Pomatiopsis lapidaria all had 1 or 2 dense fast- moving protein fractions in the region beyond Rf 0.75 (after band 13). The group of subspecies includes the so- called Tricula chiui which was referred to the genus Oncomelania by Davis & Chiu (1964). This snail is currently considered to be O. hupensis chiui (see footnote 5, p 133). The subspecies of Oncomelania hupen- sis are further separated from P. lapi- daria electrophoretically by the fact that they have distinctive densitometric patterns in the gel region between Rf 0.601 and 0.850 (includes bands 12-15 in Table 14). The characteristic fractions for P. lapidaria are found in the gel region between Rf 0.338 and 0.656 (includes bands 8-13 in Table 14) while the gel region beyond Rf 0.656 is devoid of dense components. The limits here given for the gel regions include the variations found for various populations of P. lapidaria (Davis & Lindsay, 1967) and subspecies of O. hupensis (Davis, unpublished). LABORATORY ECOLOGY A. Introduction (1948) Modifications of Vogel’s 118 G. M. DAVIS aquaterrarium have been used with varying degrees of success for rearing the subspecies of Oncomelania hupensis. References to such vivaria are found in Stunkard (1946), Ward et al. (1947), DeWitt (1952), Wagner & Wong (1956) and Moose et al. (1962). These and others have noted a marked contrast between the relative ease in rearing Oncomelania and the difficulties in maintaining and rearing species of Pomatiopsis. Stunkard (1946) noted that Pomatiopsis cincinnatiensis did not survive well in the laboratory and that, although P. lapi- daria remained alive many weeks, it did not reproduce. Ward et al. (1947) failed to maintain P. lapidaria on moist mud in shallow pans. They employed large aquaterraria with a sloping mud band covered with dry maple leaves, but found that stocks died after several months. Berry & Rue (1948) stated in anabstract that laboratory breeding of P. lapidaria was successful. They noted egg laying, followed by hatching after 3 weeks. No other data were given. DeWitt (1952) stated that he was successful in main- taining P. lapidaria from 1944 to 1952, using an “aquaterrarium.” He gave no data on reproduction, growth of young, or if an Е] generation was reared to maturity in the laboratory. Dundee (1957) described а large flower pot container (12 inches in diameter) which she used “successfully” as a vivarium. She reported reproduction and egg hatching, but did not mention whether the young were reared to maturity. Van der Schalie & Dundee (1955), van der Schalie et al. (1962) and van der Schalie & Getz (1962, 1963) discuss various aspects relating to the difficulties in maintaining species of Pomatiopsis in the laboratory. Few quantitative data have been pre- sented comparing Oncomelania and Po- matiopsis with regard to survival in the laboratory, natality, growth rates of the young produced, or survival of the young in the laboratory. Van der Schalie & Getz (1963) provided comparative data on temperature and moisture responses between “species” of the 2 genera. In this study the survivorship of field collected Pomatiopsis lapidaria and On- comelania hupensis formosana was in- vestigated when these snails were placed in different standard vivaria. The pro- duction of young was noted and the growth rates of the young recorded. Records were maintained on the sur- vivorship of the young in culture. B. Materials and Methods 1. Vivaria Utilized in the Investigations a. Plastic Tray Container. This vivari- um isa modification of the aquaterrarium used by DeWitt (1952) and was dis- cussed in detail by van der Schalie & Davis (1965). It is briefly described as follows: the measurements of the plastic tray were 28 x 19 x 6.5 cm. At one end was a soil bank andthe other a water reservoir with about 250 сс capacity. The water in the reservoir was con- stantly aerated. The tray was covered by a Sheet of plexiglass bored withmany small holes to permit gaseous exchange. Filter paper was added as food, as prescribed by van der Schalie et al. (1962). b. Tall Clay Pot. The set up in a tall, unglazed, clay flower pot (5 1/2 inches deep and 7 inches in diameter at the top) was described by van der Schalie & Getz (1962). It was designed to decrease or regulate soil moisture, since other vivaria with saturated soils were par- ticularly detrimental to Pomatiopsis cincinnatiensis. This unit was modified slightly for this study: the filter paper wick was replaced by a thick roll of cheesecloth which projected up into the pot about 2 inches. The pot was filled with sand up to within 3 inches of the top. The sand was covered with 1 1/2 inches of loam. The packed loam was dusted with finely ground dried loam to provide a surface of particularly fine particles. c. Battery Jar. Cylindrical glass con- tainers 10 inches deep and 8 inches in POMATIOPSIS AND ONCOMELANIA 119 diameter were utilized. The bottom of each jar was covered with a double thickness of No. 500 filter paper. A glass plate 3 inches by 4 inches was placed in the center of the jar and was used to support a small mound of mud. The filter paper was continually soaked with water so that a residue of 15-20 cc was present. The jar was covered with a plate of glass. d. Petri DishCulture. Nine centimeter Petri Dishes were used as cultures as described by van der Schalie & Davis (1968, in press). A mound of soil was placed in the center of the dish so that a space of 1.5 cm remained between the edge of the soil and the walls of the dish. About 40 ml of pond water were added to the culture. This type of vivarium was used only for rearing the young, newly-hatched snails to maturity. The vivarium was maintained under a 40 watt, white, “cool,” fluorescent tube suspended 10 inches above the cultures. The light, providing 100-150 ft. candles, was cycled 10-12 hrs. per day. e. Medium Clay Pot. Wagner & Wong (1956) used unglazed flower pot 5 inches in diameter in which a slope of “soil” was packed “high on one edge and ter- minated before reaching the other edge.” The “soil” was a mixture containing soil, gravel and sand mixed in the ratio of 2:1:1. Filter paper and dried leaf were added as food. A modification of the Wagner-Wong vivarium was found to provide optimal conditions for the production of young Oncomelania (van der Schalie & Davis, 1968, in press). In this modification the containers were unglazed, shallow Clay Pots with a diameter of 13 cm and a depth of 4 cm. A central mound of mud was placed on a large disc of filter paper. The bottom of the culture was covered with water, so that 5-6 cc were always present. Five males and 5 fe- males were maintained in each unit of this culture type. 2. Conditions of Temperature and Light The cultures, with the exception of the Petri Dish cultures, were maintained at 249 +20 C. As the Petri Dish vivaria were placed closer to the source of illumination, the temperature was generally 250 + 20 С. The cultures were maintained under different lighting conditions. Those maintained in “room level light” (normal daylight + usual overhead lights) were exposed to 60 + 10 ft. candles over an 8-hour period. They were in the darkat night. Cultures under “alternating light” were exposed to 100 + 10 ft. candles for periods of 10-12 hours except for the Petri dishes, which were exposed to fluorescent light as described above. Cultures under constant light were ex- posed to 140+ 20ft. candles for 24 hours. 3. Routine Maintenance and Collection of Data All cultures were routinely checked each day to secure proper water levels, knock down the snails from the sides of the vivaria (these snails show a pro- nounced negative geotropism), and observe general culture conditions. Every 2 weeks each culture (except the Petri Dish cultures) was thoroughly ex- amined. Dead snails were removed and recorded; young were removed and the number at each whorl stage was re- corded. At this time water reservoirs clogged with soil particles were cleaned. The young were placed in Petri Dish cultures where they remained for at least 8 weeks. In some cases 2-5 young were placed in Medium Clay Pots and observed. In the growth experiments young of Oncomelania hupensis formosana were measured every 3 days using a dissecting microscope and a Nippon Kogaku sliding ocular micrometer. The young of Po- matiopsis lapidaria were initially measured every 3 days but it soon became evident that monthly measurements were sufficient. C. Experiments 1. Survivorship a. Oncomelania hupensis formosana. From a shipment of about 2,000 speci- 120 G. M. DAVIS TABLE 15. The arrangement of vivaria and the number of snails used per culture Е Light SES condition Oncomelania Alternating hupensis formosana Room level Pomatiopsis Alternating lapidaria Constant Room level Alternating Constant Room level Alternating Pomatiopsis Alternating cincin- Constant natiensis Alternating Constant Room level Alternating Constant *BJ = Battery Jar MCP = Medium Clay Pot PT = Plastic Tray TCP = Tall Clay Pot mens received from Taiwan, 1,788 adults were sorted out, i.e., animals with shells possessing varices. Shells showing ex- treme erosion were not used. From the adult size, the condition of shell and from information in the literature the age of the snails was estimated at about 1/2-1 year. Although Sugiura (1933) has demonstrated that O. h. nosophora was capable of living about 5 years in the field and McMullen et al. (1951) have stated that this subspecies could live more than 2 years, other evidence in- dicates that the average life expectancy for subspecies of Oncomelania reaching maturity is less than 2 years. While Li (1953) could not indicate how long field O. h. formosana lived, he con- cluded from his data that alarge number of adults of the previous year died during or soon after the most active breeding season, and calculated the life span to be about 1 year for the vast majority of snails. No. of Total no. No. of snails cultures of snails per culture © 170 26 26 100 10 10 10 10 26 26 10 10 10 10 10 1 re NN — I Où O1 OO O1 O1 o Pe Oa oe Oo — Pesigan et al. (1958) reported that the average rate of daily mortality for field females of O. h. quadrasi was 0.76%, and that the average female lived 65.8 days after reaching maturity. From their data on the growth rate of this Subspecies, the average female would succumb in about the 7th-8th month in life. As for the adults of O. h. formosana of the present collection, it was calcu- lated from Li’s (1953) growth rate in the field that none of the adults received were less than 5 months old, while from the condition of the shells, it was thought that the minimal age was most likely 7-8 months. The snails were placed in Battery Jar and Plastic Tray vivaria; theformer were maintained at “room level light,” the latter under alternating light. The total number of Oncomelania snails placed in each type of culture along with the total number of cultures are tabulated POMATIOPSIS AND ONCOMELANIA 20 Per Cent of Adults Surviving 0 2 4 6 EA ee 20 121 VIVARIUM TYPE LIGHT CONDITION © Plastic tray Alternating © Battery jar Room level PP) 2A 26: 2830323363800 Months in Culture TEXT FIG. 4. Survival of Oncomelania hupensis formosana in 2 types of vivaria in Table 15. The initial proportion of females in each culture varied between 47-55%. The percentage of adults sur- viving each monthis shownin Text Fig. 4. It is evident that a constant fraction of the snails in the Plastic Tray vivarium died each month (exponential rate of 0.127), while the snails maintained in the Battery Jars showed an increasing rate of mortality, which indicates an unfavorable environment. The snails were apparently less able to subsist in that environment with advancing age, although the cultures were cleaned every 2 weeks andthe filter paper was regularly replaced. Extreme erosion of the shells in snails maintained in the Battery Jars for 8-9 months also gave evidence that this environment was not optimal. Asa result of the poor environment, 50% of the adults died within 4 months and only 1% survived until the 32nd month. In the Plastic Tray vivaria, 50% survived until the 6th month and 2% were sur- viving at the end of 38 months. b. Pomatiopsis lapidaria. Adults of this species were collected from the stations previously mentioned (p 15). The growth rate of this species in nature has been given as 0.20 mm per week by Dundee (1957). At that rate of growth, adults used in this study were at least 7 1/2 months old. As the snails over-winter and are inactive for at least 4 months, the snails were most likely a minimal 11 1/2 months of age. Dundee also indicated that the life span of this species in the field is about 2 years. As the adults were collected primarily in July and August, they had over-wintered and had presumably grown from young hatched in June or July of the previous year. The snails were thus assumed to be 11-12 months of age. 122 G. M. DAVIS LIGHT CONDITION ® Room level О Alternating X Constant Per Cent of Adults Surviving On ri Ir A TRIER Months in Culture TEXT FIG. 5. Survival of Pomatiopsis lapidaria in Medium Clay Pots under varying conditions of light Experiments with Pomatiopsis were The survivorship of P. lapidaria was not conducted using a larger array of culture tested in the laboratory until 1 year chambers than used for O.h. formosana. after the establishment of Oncomelania. The reasons for this are several. (1) In that period of time van der Schalie POMATIOPSIS AND ONCOMELANIA 123 VIVARIUM TYPE LIGHT CONDITION e Tall clay pot Alternating x Plastic tray Alternating o Plastic tray Constant Per Cent of Adults Surviving On ale A оао 28, ¿PORTO 1] Months in Culture TEXT FIG. 6. Survival of Pomatiopsis lapi- daria in different vivaria & Davis (1968, in press) had found that the various subspecies of Oncomelania survived better and produced more young per female when cultured in Medium Clay Pot vivaria. (2) The Tall Clay Pot had been found more suitable for rearing P. cincinnatiensis. (3) Additional field specimens of O. h. formosana were not immediately available for testing in the new types of culture chambers. The number of cultures used for P. lapidaria under the varying conditions of light as well as the number of snails in each culture are listed in Table 15. The sex ratio was 1 inallthese cultures. The survivorship curves for this species in 6 different environments are pre- sented in Text Figs. 5 and 6. The lowest mortality rate was found in the snails maintained in Medium Clay Pots, at either room level or alternating light. In these environments the exponential rate of mortality was 0.177 over the first 10 months, after which time it progressively increased. In Table 16 are listed the exponential rates and finite rates of mortality for Р. lapidaria inthe 6 environments tested. The rates were calculated for the first 4-7 months in culture. It is evident that optimal survival is correlated with the Medium Clay Pot vivarium type. It also appears that constant light is correlated with increased mortality. Survivorship in the Medium Clay Pot in room level or alternating light was compatible withthe 2 year life expectancy in the field. The increasing rate of mortality after the snails were about 2 years old perhaps reflects the natural consequence of old age rather than deteriorating culture conditions. Allthe other environments are clearly un- suitable for maintaining P. lapidaria. Mortality rates were rapid, the expo- nential rate exceeding 0.40 per month, i.e., a finite rate of over 33%ofthe snails per month. c. Pomatiopsis cincinnatiensis. Data presented by van der Schalie & Dundee (1955) indicated that the number of adults decreases in the field at the end of August and that the vast majority of young hatch early in August. These authors have shown that the life span is 16-18 months. In this study adults were collected in July and August; a few cultures were established with snails collected in May. These snails had presumably hatched in August or October of the previous year and had over-wintered. The snails were considered to be about 10-12 months old. The number of snails per culture typeis listed in Table 15. The sex ratio was 1. Survivorship curves for P. cincin- natiensis in 7 different environments are presented in Text Figs. 7, 8 and 9. Optimal survival was obtained in the Medium Clay Pot cultures under con- stant light (Text Fig. 8) where the ex- ponential rate of mortality was 0.17 124 G. M. DAVIS TABLE 16. Exponential and finite rates of mortality for Pomatiopsis lapidaria, P. cincin- natiensis, and Oncomelania hupensis formosana under varying environmental con- ditions Exponential* death Finite* death ae i Species rate per month rate per MES Tight (first 4-7 months) month (%) type condition CRD ASS 0. 12 PT Altern. formosana P. lapidaria 0. 16 MCP Room 0. 16 MCP Altern. 0. 34 MCP Const. 0. 45 PT Altern. 0. 49 TCP Altern. 1% 66 BAR Const. В cincinnatiensis 0. 16 MCP Const. 0. 33 PT Const. 0. 34 MCP Altern. 0. 45 Ter Const. 0. 45 PT Altern. 0. 45 ТОР Altern. 0. 60 MCP Room * Ix = e-ax MCP = Medium Clay Pot + 1-е-а PT = Plastic Tray ТСР = Tall Clay Pot for the first 3 months and 0.82 there- after. Only 10% of the snails were alive at the end of 6 months. In Table 16 cultures and lighting conditions are listed in order of increasing exponential rates of mortality. It is evident that survival is correlated with lighting conditions and not culture type. Optimal survival oc- curred under constant light. The least favorable environment was that of room level light in a vivarium which provided optimal survival in constant light, i.e., the Medium Clay Pot. In all but the 1 optimal condition, the finite rate of mortality was 33% or greater per month. After 6 months, cultures were down to 1-4 snails with the exception of those maintained under constant light where there were 6-10 snails per culture. The patterns of survival of P. cincin- natiensis in the laboratory indicate that this species is an “annual” as observed by van der Schalie & Dundee (1955) in Altern. = Alternating Const. = Constant Room = Room Level the field. Constant high rates of mor- tality might be expected for adults col- lected at the end of summer. d. Pomatiopsis californica and P. binneyi. Nothing is known of the life history of these snails, which are native to California. Although exact data were not kept for these species, attempts at maintaining them in Plastic Tray or Medium Clay Pot vivaria have not been successful so far. No young were pro- duced nor did survival generally exceed 5 months. Over 200 specimens of each species were involved in attempts to establish these species inthe laboratory. e. Interspecific comparisons. It is of value to compare the survivorship curves of the different species in the same environment. In only 1 case was a comparison possible between the 2 species of Pomatiopsis and Oncomelania hupensis formosana: in the Plastic Tray vivarium under alternating light, a con- POMATIOPSIS AND ONCOMELANIA 125 VIVARIUM TYPE LIGHT CONDITION Constant N о Plastic tray 80 х Tall clay pot e Tall clay pot Alternating Constant Per Cent of Adults Surviving JA AS JE Ye ACTO 99 Months in Culture TEXT FIG. 7. Survival of Pomatiopsis cin- cinnatiensis in different vi- varia dition providing only anintermediate type of survival for the species of Pomatiopsis (Text Figs. 6, 9). The finite rate of mortality for O. h. formosana was 12% per month while the rate was 45% per month for both species of Pomatiopsis. Survivorship curves for P. cincin- natiensis and P. lapidaria are compared for 5 different environments in Text Figs. 10-13. P. lapidaria is distinctly separated from P.cincinnatiensis, onthe basis of superior survivorship, when maintained in the Medium Clay Pot vivarium under alternating or room level light. P. cincinnatiensis survives in that vivarium under constant light (Text Fig. 11) as well as does P. lapidaria at room level light (Text Fig. 10) for 3 LIGHT CONDITION e Constant о Alternating x Room level 20 Per Cent of Adults Surviving CPE NE TO TUE NE ИРИ Months in Culture TEXT FIG. 8. Survival of Pomatiopsis cin- cinnatiensis in Medium Clay Pots under varying conditions of light months; however, its rate of mortality increases markedly in the 4th month. This indicates that the 2 species have a comparable rate of mortality in their respective optimal environments but that the specific differences in longevity account for the difference in rates of mortality after the 4th month. Inthe first few months both species have a finite monthly rate of mortality of 16% com- parable to that of 12%for O. h. formosana in the Plastic Tray vivarium. Pomatiopsis lapidaria has greater rates of mortality than P. cincin- natiensis, when both are maintained in the same vivarium types, under constant light (Text Figs. 11, 12). Constant light 126 G. M. DAVIS Per Cent of Adults Surviving 0’ 2. 4 eerie 12 14 e ONCOMELANIA HUPENSIS FORMOSANA x POMATIOPSIS CINCINNATIENSIS о POMATIOPSIS LAPIDARIA 16 18 20 22 24 26 28 30 32 34 36 38 40 Months in Culture TEXT FIG. 9. Comparative survivorship in the Plastic Tray vivaria under alternating light was as detrimental for P. lapidaria in the Plastic Tray culture (Text Fig. 12), as room level light was for P. cincin- natiensis in the culture condition proving to be optimal for the former, i.e., the Medium Clay Pot (Text Fig. 10). Plastic Tray and Tall Clay Pot vivaria under alternating light provided an en- vironment in which anintermediate level of survival occured for both species, i.e., an exponential rate of mortality of about 0.60, corresponding to a finite rate of mortality of about 45% per month (Text Figs. 9, 13). The survivorship of P. lapidaria under “optimal” laboratory conditions very closely approximate that of O. h. formo- sana in the Battery Jar environment, considered detrimental to the latter species. In the Battery Jar vivaria, P. lapidaria suffered a 50% mortality within 1 month and none survived past 4 months. 2. Productivity The production of young is a function of survivorship of the female, of the capacity of the female to lay a number of eggs per unit time over her lifespan and of a suitable environment which encourages egg laying and _ hatching. Table 17 gives for each species, ex- pressed in percentages, the proportion of producing vivaria for each type of environment provided, the distribution of young in those types and the average number of young per producing aquarium. Hatchlings of Oncomelania hupensis POMATIOPSIS AND ONCOMELANIA 127 100% 10 NN 70 60 NN POMATIOPSIS LAPIDARIA 50 N iz o Alternating light R Sie ® Room level light O 40 O © O 30 A Ô O O O + + 20 LA = a $ O > e = [72] x = = 10 = = 8 Sara | 5 o 6 5 5 х o 8 8 y 4 POMATIOPSIS CINCINNATIENSIS 3 + Alternating light x Room level 2 1 x x x x DEF are Aa 6. Twe Bia обеды de Months in Culture TEXT FIG. 10. Comparative survivorship in Medium Clay Pot vivaria under different lighting conditions 128 G. M. DAVIS e POMATIOPSIS LAPIDARIA Per Cent of Adults Surviving OR aah a Soh Gh err PR A ION Months in Culture TEXT FIG. 11. Comparative survivorship in Medium Clay Pot vivaria under constant light formosana were obtained in each of the 2 vivarium types used: 54% of them in Plastic Tray cultures and 46%in Battery Jar cultures. The average number per culture brought forth was 243 in the former and 148 in the latter, in spite of the fact that initially there were more females inthe latter type of culture. As regards Pomatiopsis lapidaria, only 50% of the Medium Clay Pots in room level light yielded offspring, and 40% of those in alternating light. Only a relatively small percentage of young were bred in the other culture types. Young from producing Medium Clay Pots in room level light accounted for 68% of all the hatchings, with an average of 9.4 per culture. The second highest о POMATIOPSIS CINCINNATIENSIS о POMATIOPSIS CINCINNATIENSIS @ POMATIOPSIS LAPIDARIA o) Per Cent of Adults Surviving 5 0, № 2053574. 30" 16 1 Months in Culture TEXT FIG. 12. Comparative survivorship in Plastic Tray vivaria under constant light percentage was 19, the result of 1 Plastic Tray culture in alternating light, which yielded 27 juveniles. Although survival of P. lapidaria in Medium Clay Pots was about equal in alternating light and in room level light, only 1% of the young were produced under the former con- dition, but 68% under the latter. That the difference in the initial number of females did not account for the higher rate of reproduction is shown by the percentage of young per initial female, which was 0.04% against 0.65%. Among the vivaria housing Pomati- opsis cincinnatiensis, 60% of the Plastic Tray vivaria in constant light, and 67% of all the Tall Clay Pots in alternating light were producers. However, only POMATIOPSIS AND ONCOMELANIA 129 100 зо 70-\ $ e POMATIOPSIS CINCINNATIENSIS 60 o POMATIOPSIS LAPIDARIA Per Cent of Adults Surviving > o DD — © OM 07 29, RE SL 6 780192 001 Months in Culture TEXT FIG. 13. Comparative survivorship in Tall Clay Pot vivaria under alternating light 16% of the progeny originated in the former, while 80% came from the latter, with an average of 8.3 young in each of the producing cultures of the former and an average of 14.2 in the latter. The initial number of females was about equal in each type of culture, although there were about twice as many females per culture in the Plastic Tray vivaria. The greatest multiplication occurred in a vivarium where there was an intermediate rate of mortality. No young were generated in the culture pro- viding the best rate of survival, i.e., Medium Clay Pots in constant light. These comparisons become more meaningful when the number of young per female per unit time is considered. Young per female per month were cal- culated for those cultures yielding 46% or more of the young (Table 18). The calculations are based on the survival of the females. Exact survivorship was not recorded for the females in par- ticular, but spot checks on the cultures indicated a general trend of equal rates of mortality for both sexes. The data presented in Table 18, are therefore, only a rough estimate, but nonetheless serve to show real differences between the 3 species listed. Oncomelania hupen- sis formosana shows sustained pro- duction of offspring at much higher num- bers of young/female/month than found for the other species. In the most pro- ductive month this number was 5 for Pomatiopsis lapidaria and 10 for О. В. formosana. Accurate data are available for each of the 4 subspecies of Oncomelania main- tained in Medium Clay Pots in room level light. With snails at 1 year of age, the greatest number of young/ female/month in the respective optimum month was, for O. h. hupensis, 23; for О. h. nosophora, 22; for O. h. quadrasi, 51; and for O. h. formosana, 44. It appears that the peak multi- plication in the field at a given sea- son does not carryover in the labora- tory (Table 19). In that table, the percentage of the young produced each month is listed for Oncomelania hupensis formosana, Pomatiopsis lapi- daria and P. cincinnatiensis. It was found that hatchlings of P. lapi- davia appeared in culture 4-5 months after the culture was initiated. As most of the cultures were established in July or August, the greatest per- centage of young were found in December and January. Offspring of P. cincin- natiensis were present in cultures 3-4 months after the cultures were es- tablished in August or September. Six cultures were set up in May and young appeared 1-2 months later. Due to rapid rates of mortality in the adults, no young were produced during a number of months. 130 G. M. DAVIS TABLE 17. Vivaria in which young were found 9 A vivarium | O o ad ae : * dition** Species ENDE coneition type pro- vivarium producing ducing young vivarium Oncomelania PT Altern. 100 54 243 hupensis fovmosana BJ Room 100 46 148 Pomatiopsis PT Altern. 17 19 ZA lapidaria Const. 0 0 0 BJ Room 0 0 0 MCP Altern. 40 1 1 Const. 22 U 4.5 Room 50 68 9.4 TCP Altern. 21 5 2 Pomatiopsis PT Altern. 17 2 4.0 cincin- Const. 60 16 8.3 ti y | EAP MCP Altern. 0 0 0 Const. 0 0 0 Room 25 2 2 TCP Altern. 67 80 14 Const. 0 0 0 * BJ = Battery Jar ** Altern. = Alternating MCP = Medium Clay Pot Const. = Constant PT = Plastic Tray Room = Room level TCP = Tall Clay Pot TABLE 18. The production of young per female per month in cultures which yielded 46% or more of the young Highest y/f/m in any month Culture* Light** No. of months condition condition in production Species Oncomelania Altern. hupensis formosana Room 10.0 Pomatiopsis Altern. 0. 22 lapidaria Room 0.64 Altern. 5. 0 Pomatiopsis Const. 0. 48 cincinnatiensis Altern. * BJ = Battery Jar ** Altern. = Alternating MCP = Medium Clay Pot Const. = Constant PT = Plastic Tray Room = Room Level TCP = Tall Clay Pot POMATIOPSIS AND ONCOMELANIA 131 TABLE 19. The percentage of the young pro- duced in the laboratory in each month of the year Oncomelania Pomatiopsis Month hupensis lapi- cincin- formosana daria natiensis a нЕ Jan. 5 16 31 Feb. 8 10 1 Mar. 8 dell 1 April 5 1 0 May 7 11 0 June 11 17 26 July 18 10 9 Aug. 10 0 Sept. 7 0 Oct. 3 0 Nov. 9 9 Dec. 9 24 3. Growth Rates and Survivorship of Young Van der Schalie & Davis (1964, 1965) found that newly hatched young of On- comelania hupensis formosana, main- tained 1 or 2 per Petri Dish culture, grew more rapidly than snails reared in any other fashion. Petri Dish vivaria provided an environment where the snails grew, on the average, 0.65 mm per week for the first 8 weeks with a mortality below 10%. In Text Fig. 14, the growth curves for male and female O. h. for- mosana are presented. The measure- ments were made using 25 snails of each sex. The snails were maintained 1 per dish under constant light. It was later found that the same optimal growth occurred under alternate light. Under the same conditions young Pomatiopsis lapidaria grew at a slow rate with a high mortality. Within 3 months 50% were dead and little growth had taken place (Text Fig. 15). Only 15% were alive at the end of 6 months. At the end of 1 year 5% remained ofthe 40 snails which started. Over a period of6 months the average weekly increment in length was 0.13 mm. This rate is about the same as that given by Dundee (1957) for the Length in mm 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 Days TEXT FIG. 14. Growth curves for male and female Oncomelania hupensis formosana growth of this species in the labora- tory. Comparing the growth of the 2 species between the ages of 1 and 1.5 months, O. h. formosana was found to grow ata rate of 0.95 mm per week as against 0.20 mm for P. lapidaria. It was discovered that survival of the young could be increased by placing 5 young P. lapidaria in a Medium Clay Pot in room level light, or by leaving the young in the parental cultures. Of 15 snails observed under these con- ditions, 50% survived until the 9thmonth and at the end of 13 months 40% were still surviving. The growth rate, how- ever, was somewhat lower, i.e., about 0.10 mm per week for the first 6 months (as against 0.13 mm). P. cincinnatiensis barely lasted 5 months in Petri Dish vivaria. Of the 20 young snails studied, 50% were dead in 3 1/2 months and none survived past 5 months. In that period the average increment in length was 0.12 mm per week. D. Discussion As has been reported by a number of 132 G. M. DAVIS 0. HUPENSIS FORMOSANA P. LAPIDARIA Length in mm Months TEXT FIG. 15. Growth curves for Pomati- opsis lapidaria and Oncome- lania hupensis formosana authors, the subspecies of Oncomelania hupensis can survive and reproduce under a variety of conditions which are not optimal environments. О. h. formo- sana survived well in the vivaria pro- vided and produced young at an adequate rate to assure survival of the species in the laboratory. Van der Schalie & Davis (1968, in press) found that survival and productivity increased when the 4 Subspecies were maintained in Medium Clay Pot vivaria. There was a uniform, large gap be- tween Oncomelania and Pomatiopsis with regard to longevity, production of young, growth rates of young and survival of the young in the laboratory. The 4 species of Pomatiopsis did not survive well in any of the vivaria provided, with 2 exceptions. P. lapidaria and P. cin- cinnatiensis showed survival in “opti- mal” environments, i.e., in Medium Clay Pots, at room level light for the former and at constant light for the latter, corresponding to what might have been predicted from knowledge of their life span in the field. P. lapidaria produced more young per female in the environment which pro- vided optimal conditions for survival. Although P. cincinnatiensis survived better under constant light, this species produced more young per female in the Tall Clay Pot vivarium type under al- ternating light, an environment in which the adults showed an intermediate rate of survival. Data on the production of young per female per month as well asthe greatest production per female for any given month indicate that 1 or several critical factors in the laboratory environments were either absent or detrimental. For instance, in the field P. cincinnatiensis reproduces throughout the summer with pronounced hatches of young in August and October. In August, I found 20 fe- males and 155 young in a quadrat of 2 x 3 feet. The following month I found in the same quadrat 24 females and 525 young. This condition could be duplicated up and down the river banks for 30 feet in either direction and concerned a popu- lation known to be in a steady state for over 10 years. It is obvious that pro- duction of young per female per month in the field was many times that found for this species in the laboratory. The 4 subspecies of Oncomelania hupensis appear more closely allied in their performance in the laboratory than the 2 species of Pomatiopsis here dis- cussed indetail. Data from the labora- tory studies show that P. lapidaria sur- vives best under room level light, pro- duces the greatest number of young in this optimal environment, and had a finite rate of mortality of 16% over 8-9 months. P. cincinnatiensis survives best under constant light, but produces more young in a Tall Clay Pot under alter- nating light, a condition where thefinite mortality rate was 49% per month for 7 months. CONCLUDING DISCUSSION A large number of similarities be- tween Oncomelania and Pomatiopsis have been discussed by many authors. These Similarities have mainly pertained to POMATIOPSIS AND ONCOMELANIA 133 subfamily characteristics such as the general arrangement of organs in the body, an amphibious mode of existence, eggs laid singly out of water and coated with a mud capsule, the distinctive step- like manner of progression, and the structure of the central tooth of the radula where the basal cusps arisefrom the face of the tooth instead of from the lateral angle. Enough data are now available to state that Oncomelania and Pomatiopsis are distinct genera within the Pomatiop- Sinae, a hydrobiid subfamily which also includes Tomichia from South Africa and Blanfordia from Japan. The genera Pomatiopsis and On- comelania will not hybridize. Reference was made to the cytological differences, such as sex determining mechanisms, between Oncomelania and P. lapidaria and P. cincinnatiensis. Major anatomical differences between the 2 genera are listed in Table 12. Of particular importance are the facts that: 1) In Oncomelania the sperm and spermathecal ducts arise together from the right ventrolateral surface of the bursa copulatrix, bound together as Slender tubes in a connective tissue sheath while, in Pomatiopsis, the sperm duct arises from the spermathecal duct near the point where the latter emerges as a broad tube from the anterior end of the bursa copulatrix. 2) Unlike Pomatiopsis, the verge of Oncomelania has a protrudable papilla and the tip of the verge is quite muscular. Some species of Pomatiopsis have a penial filament. In the laboratory Oncomelania is characterized by the comparative ease with which it adapts to laboratory cul- ture and by the rapid growth of young. Pomatiopsis (all species) does not adapt to the laboratory environment in which Oncomelania thrives; the growth rate of the young is extremely slow com- pared with that of young Oncomelania. Van der Schalie & Getz (1963) pointed out that Pomatiopsis lapidaria and EP. cincinnatiensis were more tolerant of low temperatues (-5° to -7° C) and less tolerant of high temperatures (41° to 449 С) than Oncomelania; these 2 species of Pomatiopsis were less resistant to drowning than Oncomelania. Several of the “specific” differences are most probably indicative of charac- ters applicable to the generic level. For instance the electrophoretic differences between O. h. formosanaand Pomatiopsis lapidaria are of no more than specific order; however, preliminary data for the other subspecies of Oncomelania hu- pensis indicate similarities among these latter in the dense, fast moving proteins, which do not occur in P. lapidaria. Davis (1964) indicated a distinct differ - ence between P. lapidaria and 0. h. formosana in the potential for shell regeneration. О. h. formosana rapidly regenerated a Shell in the apical whorls with low mortality, while P. lapidaria had a high mortality and showed no signs of shell regeneration. Oncomelania is a genus with but 1 species; it has 4 well established? sub- species. The subspecies of Oncomelania do not differ greatly in their internal anatomy. External differences are mainly those of size, ribs on the shells of some populations of O. hupensis hu- pensis, and variance in the intensity of external pigmentation and the yellow coloration of the granules surrounding the medial surface of the eye. In Oncomelania the shell is usually smooth (except for the above mentioned ribs), with moderately deep sutures and with moderately convex whorls. The outer lip of the shell has a tendency to form a varix which is usually quite pro- nounced and is sinuate. The umbilicus is narrow and the apical whorl narrow. 5Davis (1968, in press) discussed the sys- tematic position of the so-called Tricula chiui; this taxon is now assigned to On- comelania, as a 5th subspecies of O. hupen- sis, O. hupensis chiui, on the basis of ana- tomical, electrophoretic and _ serological data. 134 G. M. DAVIS The parietal callus is elongate. There are at least 35 gill filaments, usually 45 or more. The verge is muscular, the tip has short strips of actively beating cilia and a distinct protrudable papilla. The pleuro-supraesophageal connective is relatively short. The osphradio- mantle nerve and supravisceral con- nective both arise from the tip of the supraesophageal ganglion. The former nerve is elongate and most commonly bifurcates only within the lateral wall of the “neck” into the osphradial nerve and mantle nerve. The sperm and spermathecal ducts arise together from the ventral, right anterolateral surface of the bursa copulatrix. The female gonad is multibranched and the collecting duct of the gonad relatively slender. The oviduct encircles the seminal receptacle in a characteristic manner. The seminal vesicle is composed of a slender tube which is characteristically knotted. The verge has a single glandular type. The cerebral commissureis short. The tentacles are elongate relative to the length of the rostrum. Pomatiopsis is a genus composed of 4 distinct species of whichonly P. lapidaria and P. cincinnatiensis are known in terms of anatomy and life history. P. californica and P. binneyi are virtually unknown except for general habitat and Shell. A 5th form, P. chacei, is most like P. californica; the type description given by Pilsbry (1936) is not sufficient to separate this form as a distinct species. Another species, P. robusta Walker, was listed by Abbott (1948a) with Pomatiopsis. However, Pilsbry (1933) had removed this species from Pomatiopsis with justification, as the radula is clearly not of the type found in the Pomatiopsinae. He assigned it to Amnicola. Gregg € Taylor (1965) have placed this species in a new genus and subgenus, Fontelicella (Natricola) ro- busta. Pomatiopsis binneyi is different from the other species in the genus on the basis of shell and habitat, althoughthere are similarities on the basis of some subfamily characteristics. The shell is only about 3 mm high, without an umbili- cus; the inner and outer lips meet without forming a parietal callus and are generally slightly separated from the parietal surface of the body whorl. The species lives high on Mt. Tamalpais in Marin County, California. The snails live in dense shade on the surfaces of rocks and leaves, either in the path of trickling water or sprayed by water streaming down steep canyon slopes. This habitat is markedly different from those inhabited by the other species of Pomatiopsis. P. californica lives on shallow mud banks and marshy seepages leading into shallow streams. The area I observed was a lowland habitat on the edge of Bolinas Bay, Marin County, California. P. cincinnatiensis lives ona narrow margin of river bank while P. lapidaria survives in marshy Typha Swamps, grassy seepages leading into rivers, or wooded lowlands seasonally Swampy due to stream overflow. Much of the Swampy areas inhabited by P. lapidaria closely resemble the habitat of Oncomelania hupensis quadrasi ob- served in the Philippines. Because Pomatiopsis binneyi appears so different on the basis of shell and habitat it should be studied thoroughly to determine whether it is, indeed, a species of Pomatiopsis. Pomatiopsis is currently defined as a genus in which the shell has a roughened microsculpture, the lip is sharp, straight, and there is no tendency to form a varix. The apical whorls are wide. The umbilicus is wide and pro- nounced. The sutures are deeply im- pressed and the whorls very convex. There are less than 30 gill filaments. The verge does not have the pronounced musculature or papilla foundin Oncome- lania. Penial cilia are lacking in 2 species (P. lapidavia and P. cincin- natiensis); in the 2 other species they are bushy and generally not active. Two species have penial filaments (P. cincin- natiensis and P. californica), a feature not found in Oncomelania, Blanfordia, POMATIOPSIS AND ONCOMELANIA 135 or Tomichia. The verge has 3 glandular types (known for P. lapidaria). The pleuro-supraesophageal connective is elongate. The distance from the tip of the supraesophageal ganglion to the lateral cephalic wall is very short. The osphradiomantle nerve is quite short and frequently bifurcates, forming the osph- radial and mantle nerves, before enter- ing the lateral wall. The supravisceral connective leaves the posterior edge of the ganglion, not the tip (known only for P. lapidaria and 1 specimen of P. binneyi). The spermathecal duct arises from the anterior end of the bursa copula- trix. The female gonadis little branched and the collecting duct is very wide. The oviduct does not encircle the seminal receptacle. The seminal vesicle is a thick, regularly coiled tube, not a “knot” of tubes. The tentacles are relatively Short, compared to the length of the rostrum. ACKNOWLEDGEMENTS I wish to express my sincere thanks and gratitude to Dr. Henry van der Schalie for his encouragement and support throughout my doctoral program. I am greatly indebted to Drs. Nelson G. Hairston and Warren H. Wagner for their helpful criticism and assistance during the final stages of the program. I also wish to thank Drs. J. B. Burch and Dwight W. Taylor for their special interest and discussions on many aspects of molluscan systematics. Special thanks go to Mrs. Anne Gismann for the long hours spent in editorial work on the manuscript. Grateful acknowledgement is extended to Gene Lindsay for his assistance in various aspects of the electrophoretic studies; to Berton Roffman, Andrew and Jerry Bratton, and Robert Wakefield who, at various times, assisted in the main- tenance of the snail cultures. A note of special gratitude goes to the following persons and laboratories who aided in providing specimens or facilities which made this study possible: Dr. Robert E. Kuntz, Lt. J. F. Bergner, and Lt. Duell E. Wood, formerly of the О. 5. Naval Medical Research Unit, No. 2, Taiwan; LTC J. W. Moose andMv. J. E. Williams of the 406 Medical Laboratory, U. S. Army Medical Command, Japan; and the late Dr. T. P. Pesigan, Director of Medical Bureau, Department of Health, Philippines. Photographs of the shells were done by William Brudon, Department of Anatomy, University of Michigan. LITERATURE CITED ABBOTT, R. T., 1945, The Philippine intermediate snail host (Schistosomo- phora quadrasi) of schistosomiasis. Occ. Pap. Mus. comp. Zool., Harvard, 1(2): 5-16. 1948a, A potential snail host of Oriental schistosomiasis in North America (Pomatiopsis lapidaria). Proc. U. S. Natl. Mus., 93(3222): 57- 67. 1948b, Handbook of medically important mollusks of the Orient and the western Pacific. Bull. Mus. comp. Zool., Harvard, 100(3): 246-328. ANNANDALE, N., 1924, The molluscan hosts of the human blood fluke in China and Japan, and species liable to be confused with them. Amer. J. Hyg., Monogr. Ser., (3): 269-294. BAKER, F. C., 1902, Mollusca of the Chicago Area. II. Gastropoda. Chicago Acad. Sci. Bull. (Nat. Hist. Survey), 3(2): 137-410. 1926, Nomenclatural notes on American freshwater Mollusca. Wisc. Acad. Sci., 22: 193-205. 1927, Description of new forms of Pleistocene land mollusks from Illinois with remarks on other species. Nautilus, 40(3): 114-120. 1928, The freshwater Mol- lusca of Wisconsin. 1. Gastropoda. Bull. Wisconsin Acad. Sci., 70: 1-494. BAKER, H. B., 1960, Hydrobiidae or Truncatellidae? Nautilus, 74(1): 34- 35. BARTSCH, P., 1936, Molluscan inter- mediate hosts of the Asiatic blood fluke, Schistosoma japonicum, and species confused withthem. Smithson. Misc. Coll., 95(5): 1-60. 136 G. M. DAVIS BERRY, E. G., 1943, The Amnicolidae of Michigan: distribution, ecology, and taxonomy. Misc. Publ. Mus. Zool., Univ. Mich., 57: 1-68. BERRY, EG: RUE, В.Е. 1 1948, Pomatiopsis lapidaria (Say), an American intermediate host for Schistosoma japonicum. J. Parasit., 34 (suppl.): 15. BINNEY, W. G., 1865, Land and fresh water shells of North America. Smith- son. Misc. Coll., 143, (2): 1-120. BOUVIER, E. L., 1887, Systeme nerveux, morphologie générale et classification des gastéropodes prosobranches. Ann. Sci. nat. Zool., 3: 1-510. BURCH, J. B., 1960a, Some snails and slugs of quarantine significance to the United States: Agr. Res. Serv., US. "Dept Agra” 182): 1 173; Sterkiana, 2: 13-53. 1960b, Chromosomes of Po- matiopsis and Oncomelania. Amer. malacol. Union ann. Reps., 1960, 26: 15. 1964, Cytotaxonomy of the genus Oncomelania, intermediate hosts of Schistosoma japonica. Amer. malacol. Union ann. Reps., 1964, 31: 28-29. CARE, RES, 11900! A descriptive illustrated catalogue of the Mollusca of Indiana. 24th Ann. Report, Ind. Dept. Geol. and nat. Resources, 1899: 335-535. CARRIKER, M. R., 1946a, Morphology of the alimentary system of the snail Lymnaea _ stagnalis appressa (Say). Trans. Wisc. Acad. Sci., 38: 1-88. 1946b, Observations on the functioning of the alimentary system of the snail Lymnaea stagnalis ap- pressa Say. Biol. Bull., 91(1): 88- 511% ®HENG;'-T...C.,.'1964; Comparative electrophoretic studies on the sera of marine and freshwater mollusks. Taxonomic Biochemistry and Se- rology. Ed., Leone. Ronald Press, N. Y., p 659-666. CLENCH, W. 3. “€ TURNER. RD. 1948, A catalogue of the family Trun- catellidae with notes and descriptions of new species. Occ. Pap. Moll., Mus. Zool., Harvard, 1(13): 157-212. DAVIS, B. J., 1964, Disc electrophoresis - II. Method and application to human serum proteins. Annals N. Y. Acad. Sci. 121: 404-427. DAVIS, G. M., 1964, Shell regeneration in Oncomelania formosana (Gastro- poda: Hydrobiidae). Malacologia, 2(1): 145-159. 1968, A systematic study on Oncomelania hupensis chiui (Proso- branchia: Hydrobiidae). Malacologia (in press). DAVIS, С. М. & CHIU; а. Е 196% The anatomy of Oncomelania chiui, Annual Report of the University of Mich. to U. S. Army Medical and Re- search Development, Department of the Army. p 14. DAVIS, G. M. & LINDSAY, G. K., 1964, Disc electrophoresis in the study of molluscan systematics. Amer. malacol. Union ann. Reps., 1964, 31: 20-21. 1967, Disc electrophoretic analysis of molluscan individuals and populations. Malacologia, 5(2): 311- 334. DAVIS, G.: M:,::.MOOSE ods Wane WILLIAMS, J. E., 1965, Abnormal development in a hybrid Oncomelania (Gastropoda: Hydrobiidae). Zbid., 2(2): 209-217. DeWITT, E. B., 1952, Pomatiopsis lapi- daria, its occurrenceinthe Washing- ton, D. C., area and its laboratory rearing in comparison to that of On- comelania spp. J. Parasit., 38(4): 321- 326. DUNDEE, D. S., 1957, Aspects of the biology of Pomatiopsis lapidaria (Say) (Mollusca: Gastropoda: Prosobran- chia). Misc. Publ., Mus. Zool., Univ. Mich., (100): 1-37. FISCHER, P. H., 1880-1887, Manuel de Conchyliologie, Paris, Savy. xxiv + 1369 p. FRETTER, V. & GRAHAM, A., 1962, British prosobranch mollusks. Ray Society, London. xvi + 755 p. POMATIOPSIS AND ONCOMELANIA 137 GILL, T., 1863, Systematic arrangement of the mollusks of the family Vivipar- idae, and others, inhabiting the United States. Proc. Acad. nat. Sci. Phil., (1): 33-40. 1871, Arrangement of the families of mollusks. Smithson. Misc. Col (227) т. GREDLER, V., 1881, Zur Conchylien - Fauna von China. Jahrb. Deutsch. malakoz. Ges., 8: 110-132. GREGG, W. O. & TAYLOR, D. W., 1965, Fontelicella (Prosobranchia: Hydro- biidae) a new genus of west American freshwater snails. Malacologia, 3(1): 103-110. HANNIBAL, H., 1912, The aquatic mol- luscs of Southern California and adjacent regions, a transitional fauna. Bull. S. Calif. Acad. Sci., 11(1): 18-46. HERRICK, J. C., 1906, Mechanism of the odontophoral apparatus in Syco- typus canaliculatus. Amer. Natur., 40(478): 707-737. HEUDE, P. M., 1882, Notes sur les mollusques terrestres de la vallée du fleuve bleu. Mémoires concernant l’histoire naturelle de l’empire Chinois par les Péres dela Compagnie de Jésus. p 1-188, pls. 12-21. Mis- sion Catholique à l’orphelinat de Tou- sé-we. Chang-Hai. HUBRICHT, L., 1960, Pomatiopsis lapi- daria on the southern Atlantic Coastal Plain, with remarks on the status of P. praelonga and P.hinkleyi. Nautilus, 74(1): 33-34. ITAGAKI, H., 1955, Anatomy of On- comelania nosophora (Robson) (Gas- tropoda). Venus, 18(3): 161-168. JOHANSSON, J., 1939, Anatomische Studien tiber Die Gastropodenfamilien Rissoidae und Littorinidae. Zool. Bidr., Uppsala, 18: 239-396. KRAUSE, H., 1949, Untersuchungen zur Anatomie und Okologie von Litho- glyphus naticoides (C. Peiffer). Arch. Moll.-Kunde, 78(4/6): 103-148. KRULL, H., 1935, Anatomische Unter- suchungen an einheimischen Proso- branchiern und Beitráge zur Phylo- genie der Gastropoden. Zool. Jb., Anat. u. Ontog., 60: 400-464. KUO; ‘У. & МАО, S., 1957, On the taxonomy of Oncomelania-snail hosts of Schistosoma japonicum. Mor- phological studies of Oncomelania snails from different endemic areasin China. Chinese med. J., 75: 824-831. LI, F. C., 1934, Anatomie, Entwick- lungsgeschichte, Oekologie, und Ras- senbestimmung von Oncomelania, des Zwischenwirtes von Schistosoma japonicum (Katsurada, 1904) in China. Trans. Sci. Soc. China, 8(2): 104-145. LI, S. Y., 1953, Studies on schisto- somiasis japonica in Formosa. Ш. The bionomics of Oncomelania formo- sana, a molluscan intermediate host of Schistosoma japonicum, Amer. J. Hyg., 57(1): 30-45. MAO, C. P. & LI, L., 1948, Snail hosts of Schistosoma japonicum in the Soochow-Wusih Area, Kiangsu, China. J. Parasit., 34(5): 380-385. McMULLEN, D. B., KOMIYAMA, S. & ENDO-ITABASHI, T., 1951, Obser- vations on the habitats, ecology, and life cycle of Oncomelania nosophora, the molluscan intermediate host of Schistosoma japonicum in Japan. Amer. J. Hyg., 54(3): 402-415. MICHELSON, E. H., 1960, A rapid method for preparing mounts of snail genitalia. Nautilus, 74(1): 32-33. 1966, Characterization of the haemolymph antigens of Australorbis glabratus by disc electrophoresis and immunoelectrophoresis. Ann. trop. Med. Parasit., 60(3): 280-287. MOLLENDORFF, О. F., 1895, Diagnosis specierum novarum ex insulis Philippinis. Nachrichtsbl. Deutsch. malakoz. Ges., 27: 137-149. MOOSE, J. W., WILLIAMS, J. E: & FLESHMAN, P., 1962, Rice cereal as sustenance for rearing oncome- lanid snails in the laboratory. J. Parasit., 48(1): 68. MOQUIN-TANDON, A., 1855, Histoire naturelle des Mollusques terrestres et fluviatiles de France. Paris, Bailliére: 513-527. NAKAMOTO, M., 1923, Anatomy of Blan- fordia nosophora, anintermediate host of Schistosoma japonica. Kyoto med. Mag., 20: 1059-1066. NISBET, R. H., 1953, The structure and function of the buccal mass in some gastropod mollusks. 1. Monodonta 138 G. M. lineata (da Costa). Ph.D. thesis, Univ. London, England. OPINION 344, 1955, Validation under the plenary powers of the generic name “Truncatella” Risso, 1826, and addition of that name and the names “Acmaea” Eschscholtz, 1883, and “Acicula” Hartman, 1821 (Class Gas- tropoda) to the “Official List of Generic Names in Zoology.” Opinions € Declar. Rend. Internat. Comm. zool. Nomencl., 10(11): 315-351. ORNSTEIN, L. 1962, Theory of disc electrophoresis. Preprint for Canalco Corporation, Bethesda, Maryland. р1- 44. 1964, Disc electrophoresis - I. Background and theory. Annals N. Y. Acad. Sci., 121: 321-349. PATTERSON, C. M., 1963, Cytological studies of pomatiopsid snails. Amer. malacol. Union ann. Reps., 1963, 3113-14) | PELSENEER, P., 1906, A Treatise on Zoology. Ed.: Lankester, Part 5. London, A. & C. Black. PESIGAN; Ez. Pi; HAIRSTON. №.С,, JAUREGUI, J. J., GARCIA, E. G., SANTOS, "A? Т., SANTOS; 'B.-.C. € BESA, A. A., 1958, Studies on Schisto- soma japonicum infection in the Philippines. 2. The molluscan host. Bull. Wld Hlth Org., 18: 481-578. PILSBRY, H. A., 1933, Amnicolidae from Wyoming and Oregon. Nautilus, 47(1): 9-12. 1936, A Californian Po- matiopsis. Nautilus, 50(3): 84-85. PILSBRY, H. A. € FERRISS, J.H., 1906, Mollusca of the Ozarkian fauna. Proc. Acad. nat. Sci. Phil., 58(3): 529-567. PILSBRY, H. A. & HIRASE, Y., 1905, Catalogue of the land and fresh-water mollusca of Taiwan (Formosa), with descriptions of new species. Ibid., 57(3): 720-752. RITCHIE, L. S., 1955, The biology and control of the amphibious snails that serve as intermediate hosts for Schistosoma japonicum. Amer. J. trop. Med. & Hyg., 4(3): 426-441. ROBSON, G. C., 1915, Katayama noso- phora. British. med. J., 1: 203. 1921, On the anatomy and DAVIS affinities of Hypsobia nosophora. Ann. & Mag. nat. Hist., Ser. 9, 8: 401-413. ROTH, A. A., 1960, Aspects of the function of the bursa copulatrix and seminal receptacle in the prosobranch snail Oncomelania formosana Pilsbry and Hirase. Trans. Amer. micr.Soc., 79(4): 412-419. ROTH, A. A. & WAGNER, E. G., 1957, The anatomy of the male and female reproductive systems of Oncomelania nosophora. Ibid., 76(1): 52-69. SAY, T., 1817, Description of seven species of American fresh water and land snails, not noticedinthe systems. Js Philas Acad: Зе. STIMPSON, W., 1865, Researches on the Hydrobiinae and allied forms. Smithson. Misc. Coll., 201: 1-59. STUNKARD, H. W., 1946, Possible snail hosts of human schistosomiasis in the United States. J. Parasit., 32(6): 539-552. SUGIURA, S., 1933, Studies оп biology of Oncomelania nosophora (Robson), an intermediate host of Schistosoma japonicum. Mittlgn. a.d. path. Inst. d. med. Fakult., Niigata, Japan, 31:1-18. THIELE, J., 1928, RevisiondesSystems der Hydrobiiden und Melaniiden. Zool. Jb., Systematik, Geogr. u. Biol. 55: 351-402. 1931, Handbuch der sys- tematischen Weichtierkunde. Vol. 1. Reprintedin 1963 by A. Ascher & Co., Amsterdam, from the 1931 edition by Gustav Fischer, Jena. 778 p. TROSCHEL, Е. H., 1856-1863, Das Gebiss der Schnecken zur Begrtindung einer natürlichen Classification. p 1- 252, pls. 1-20. Nikolaische Verlags- buchhanal., Berlin. TRYON, G. W., 1862, Notes on Ameri- can fresh water shells, with de- Scriptions of two new species. Proc. Acad. nat. Sci. Phil., 451-452. 1883, Structural and system- atic conchology. Vol. 2: 272. VAN DER SCHALIE, H. & DAVIS, G.M., 1964, Stunting of Oncomelania formo- sana inculture. Amer. malacol. Union ann. Reps., 1964, 31: 13-14. 1965, Growth and stunting in POMATIOPSIS AND ONCOMELANIA 139 Oncomelania (Gastropoda: Hydro- biidae). Malacologia, 3(1): 81-102. 1968, Culturing Oncomelania snails for studies of Oriental schisto- somiasis, Zbid., (in press). VAN DER SCHALIE, H. & DUNDEE, D. S., 1955, The distribution, ecology, and life history of Pomatiopsis cincin- natiensis (Lea), an amphibious oper- culate snail. Trans. Amer. micr. Soc., 74(2): 119-133. 1956, The morphology of Pomatiopsis cincinnatiensis (Lea) an amphibious prosobranch snail. Occ. Pap., Mus. Zool.,Univ. Mich., 579: 1-17. 1959, Transect distribution of eggs of Pomatiopsis lapidaria Say, an amphibious prosobranch snail. Trans. Amer. micr. Soc. 78(4): 409- 420. VAN DER SCHALIE, H. & GETZ, L. L., 1962, Distribution and natural history of the snail Pomatiopsis cincin- natiensis (Lea). Amer. Midl. Natur., 68(1): 203-231. 1963, Comparison of temper- ature and moisture responses of the snail genera Pomatiopsis and Oncome- lania. Ecology, 44(1): 73-83. VAN DER SCHALIE, H., GETZ, L. L. & DAZO, B. C., 1963, Hybrids between American Pomatiopsis and Oriental Oncomelania snails. Amer. J. trop. Med. Hyg., 2(3): 418-420. VOGEL, H., 1948, Über eine Dauer- zucht von Oncomelania hupensis und Infektionsversuche mit Bilharzia japonica. Z. f. Parasit.-Kunde, 14: 70-91. WAGNER, E. D. & WONG, L. W., 1956, Some factors influencing egg laying in Oncomelania nosophora and On- comelania quadrasi, intermediate hosts of Schistosoma japonicum. Amer. J. trop. Med. Hyg., 5(3): 544- 558. WALKER, B., 1918, A synopsis of the classification of the freshwater Mol- lusca of North America, north of Mexico. Misc. Publ., Mus. Zool. Univ. Mich., (6): 1-213. WARD, P. A., TRAVIS, D. € RUE,R.E., 1947, Experimental infection with Schistosoma japonicum. Nat. Inst. Hlth Bull., (189): 95-100. WENZ, W., 1938-1944, Gastropoda. Allgemeiner Teil und Prosobranchia, In: О. H. Schindewolf (ed.), Handbuch der Paldozoologie, 6. Berlin. WILLIAMS, J. E., 1954, Professional report of the 406 Medical General Laboratory. p 182-184. 1955, Zbid., 230-231. WONG, L. W. & WAGNER, E. D., 1954, A Rapid method of sexing snails, On- comelania nosophora. Trans. Amer. micr. Soc., 73(1): 66-67. WRIGHT,» EA. €:ROSS, «Gees 1959, Electrophoresis of snailblood. Trans. roy. Soc. trop. Med. & Hyg., 53: 308. 1963, Electrophoretic studies of blood and egg proteinsof Austro- lorbis glabratus (Gastropoda: Planor- bidae). Ann. trop. Med. & Parasit., 57: 47-51. 1965, Electrophoretic studies of some planorbid egg proteins. Bull. Wld Hith Org., 32: 709-712. 140 G. M. DAVIS RESUMEN RELACIONES SISTEMATICAS DE POMATIOPSIS LAPIDARIA Y ONCOMELANIA HUPENSIS FORMOSANA (PROSOBRANCHIA) G. M. Davis Pomatiopsis lapidaria (Say) de Norte América y la Oncomelania hupensis formo- sana (Pilsbry & Hirase) oriental, se eligieron como representantes de dos géneros relacionados de hydróbidos. Se estudió la anatomía comparada, hibridización potencial, propiedades electroforeticas y ecología de laboratorio, para determinar en que alcance pueden encontrarse valores sistematicos. : En base a sus anatomias Pomatiopsis y Oncomelania, se juzgan como generos distintos dentro de la misma subfamilia Pomatiopsinae. En el género Oncomelania (considerado como formado por una especie con 4 sub- especies), la concha es lisa (excepto en la forma costulada O. hupensis hupensis) con suturas de profundidad moderada y anfractos convexos. El labio externo tiene tendencia a formar una varice muy pronunciada. El ombligo es estrecho, asi como el anfracto apical. El callo parietal es alargado. Tiene рог los menos 35 filamentos branquiales y generalmente 45 o más. La verga es muscular, la punta con bandas de activas cilias y una papila sobresaliente. Los conectivos pleuro-supraesofágicos comparativamente cortos; en consecuencia el nervio manto-osfrádico elevándose de la punta del ganglio supraesofágico es relativamente largo; generalmente no se bifurca hasta estar adentro de la pared cefálica. El conectico supravisceral tambien surge de la punta del ganglio. El espermoducto y ducto de la espermateca surgen de la derecha en una sóla vaina, en la superficie antero lateral de la bursa copulatrix. El oviducto rodea el recptaculo seminal en una manera caracteristica. La verga tiene un tipo glandular único (estudiada en O. h. formosana y H. quadrasi). La comisura cerebral es corta. Los tentáculos son alargados comparados con la longitud del rostro. Comparada con Oncomelania, la concha de Pomatiopsis tiene una microescultura rugosa, el labio es agudo y notienetendencia a formar varice. En las cuatro especies las vueltas apicales son anchas. El ombligo es dilatado y pronunciado, suturas pro- fundamente impresas y los anfractos muy convexos (excepto in P. binneyi). In P. lapidaria y P. cincinnatiensis hay menos de 30 filamentos branquiales. La verga no tiene una musculatura pronunciada o papila, en las cuatro especies; cilias peniales faltan en 2 especies (P. lapidaria y cincinnatiensis); cuando aparecen cilias estan aglomeradas y generalmente inactivas (P. californica y binneyi). Dos especies (cin- cinnatiensis y californica) tienen filamentos peniales, una condición que no se encuentra en Oncomelania. La verga (tal como es conocida en P. lapidaria) tiene tres tipos glandulares. EI conectivo pleuro-esofágico es alargado, el ganglio supraesofagico descansa cerca de la pared cefálica lateral y los nervios del manto y osfrádico, el cual generalmente se bifurca a partir de la punta del ganglio, son correspondientemente mas cortos. El conectivo supravisceral, en P. lapidaria, surge del lateral, borde posterior del ganglio supraesofágico, no de la punta. El oviducto no encierra la vesícula seminal. El ducto de la espermateca surge del extremo anterior de la bursa copulatrix (P. lapidaria y P. cincinnatiensis), y el ducto espermático del ducto de la espermateca. La gonada femenina es un poco ramificada, y el ducto colector bastante ancho. La vesícula seminal forma un tubo grueso regularmente arrollado. Los tentáculos son cortos, en relación a la longitud del rostro. No se conocen híbridos entre Pomatiopsis lapidaria y Oncomelania. Estudios disco-electroforéticos de proteína fresca del músculo pedal de los dos taxa representativos mostró que cada taxón tiene un patrón específico. Todas las subespecies de Oncomelania tienen 1 о más componentes protéicos densos con valores Rf (proporción de la distancia desde el orígen a el centro de cada banda y del orígen al frente) mayor que 0.75. Pomatiopsis lapidaria no tiene proteínas densas de movimientos rápidos más allá de un Rf de 0.75. POMATIOPSIS AND ONCOMELANIA Las cuatro subSpecies de Oncomelania se caracterizan por adaptabilidad a las condiciones de cultivo en laboratorio. En 12 meses, bajo condiciones menos que Óptimas, la mortalidad (caracoles de natural habitat cerca de 1 año de edad) fué de 12% por mes. Los jóvenes crecieron 0.65 mm por semana con mortalidad baja. Las hembras produjeron crías en porporción de 2.12 mensualmente por más de 2 años. Las 4 especies de Pomatiopsis investigadas no se adaptaron bien a condiciones de laboratorio. P. californica y P. binneyi murieron rapidamente sin reproducirse. P. lapidaria y P. cincinnatiensis (de habitat natural, 1 año de edad) tuvieron 16% de mortalidad por mes en condiciones “óptimas” sobre un periodo de 10 meses para el primero y 3 meses para el segundo, después de lo cual la mortalidad aumentó rapidamente, en parte por razón del más corto término de vida de esos caracoles. Jóvenes crecieron menos de 0.14 mm por semana con mortalidad excediendo el 30% en 2 meses. Jóvenes fueron producidos en proporción de menos de 0.51 por hembra mensualmente por periodos cortos. ABCTPAKT СИСТЕМАТИЧЕСКИЕ ВЗАИМООТНОШЕНИЯ МЕЖДУ POMATIOPSIS LAPIDARIA И ONCOMELANIA HUPENSIS FORMOSANA (PROSOBRANCHIA: HYDROBIIDAE) Г. М. ДЕВИС Для исследования были выбраны представители двух родствен- ных родов гидробиид: северо-американский вид Pomatiopsis lapidaria (Say) и восточный Oncomelania hupensis formosana (Pilsbry € Hirase). Изучалась сравнительная анатомия этих форм, потенциальные возмзжности для гибридизации, электрофоретиче- ские свойства и лабораторная экология; все это должно было определить, насколько имеющиеся различия имеют ценность для систематики. С точки зрения анатомии Pomatiopsis и Опсотеата считаются вполне самостоятельными родами одного и того же подсемейства Pomatiopsinae. У представителей рода Опсотешта (имеющего 1 вид с 4 подвидами), раковина гладкая (исключая ребристую Форму O. hupensis hupensis), с умеренно-глубокими швами и умеренно- выпуклыми оборотами. Наружная губа раковины имеет тенденцию образовывать поперечный гребень (varix), обычно довольно выдающийся. Пупок (umbilicus) узкий, как и апикальный оборот. Париетальный каллус удлиненный. Имеется по крайней мере 35 жаберных Ффиламентов, обычно 45 или больше. Край мантии мускулистый, на конце имеются короткие ряды активно-рабо- тающих ресничек Y ясно-выдающаяся папилла. Плевро- надглоточная коннектива сравнительно короткая; соответственно, осфрадийно-мантийный нерв, отходящий от верхушки надглоточного ганглия, относительно длинный; он обычно не раздваивается вплоть до внутренней части цефалической стенки. Суправисцеральная коннектива, также отходит от верхушки 141 142 G. M. DAVIS ганглия. Семепровод и семеприемник находятся в общей оболочке и отходят от передне-боковой поверхности совокупи- тельной сумки. Яйцевод характерным образом опоясывает семеприемник. Железа края мантии единого типа (изучена на O. h. formosana и О. h. quadrasi). Церебральная комиссура короткая. Тентакулы, по сравнению с длиной рострума, удлиненные. По сравнению с Oncomelania, раковина Pomatiopsis имеет более грубую микроскульптуру; губа раковины острая и не имеет тенденции образовывать гребень (varix). Пупок широкий и хорошо- развитый; швы глубоко вдавленные, обороты очень выпуклые (кроме Р. binneyi). Y P. lapidaria и Р. cincinnatiensis имеется менее 30 жаберных Филаментов. У 4 видов край мантии не имеет хорошо выраженной мускулатуры и папилл; у 2 видов (Р. lapidaria uP. cincinnatiensis) пениальные реснички отсутствуют; когда реснички имеются, они обычно густые и мало активные (P. cabifornica и P. binneyi). У двух видов (P. cincinnatiensis u P. californica) имеются пениальные Филаменты, отсутствующие у Oncomelania. Край мантии имеет железы трех типов (R lapidaria). Плевро-надглоточная коннектива, удлиненная, надглоточный ганглий расположен близь боковой части цефалической стенки, a осфрадийный и мантийный нервы, которые обычно раздваиваются сразу после отхода их от верхушки ганглия, соответсткенно довольно короткие. Суправисцеральная KOHHEKTUBA у Ps lapidaria отходит от бокового заднего края надглоточного ганглия, а не от его верхушки. Яйцевод не опоясывает семенной пузырек; протока сперматеки отходит от переднего конца совокупительной сумки (Р. lapidaria и Р. cincinnatiensis), a семепровод - от протока сперматеки. Женская половая железа слабо-разветвленная; собирающий проток довольно цщирокий. Семенной пузырек ввиде толстой правильно-извитой трубки. Тентакулы относительно длины рострума короткие. Гибридизация между Pomatiopsis lapidaria и Опсотаата не наблюдается. Исследование белков из свежего HOXHOTO мускула y представителей двух групп моллюсков методом дискового злектрофореза показало, что каждый таксон Имеет в этом отношении свои специфические компоненты. Все подвиды Опсотеата имеют 1 или более характерных плотных протеиновых компонентов с величинами Rf (отношение расстояния от источника, до центра в каждой полосе и от источника до переднего края) более 0.75. Pomatiopsis lapidavia не имеет плотных быстро "двигающихся" белков, со значениями Rf более 0.75. Все 4 подвида Oncomelania характеризуются способностью адап- тироваться к искусственным лабораторным условиям. За 12 месяцев при ‘условиях, далеко не оптимальных, конечная скорость отмирания улиток (взятых из природныз условий в возрасте 1 года), составляло 12% в мксяц. Темп роста молоди был около 0.65 мм в неделю, при малой смертности. Непрерывно втечение более 2 лет продукция молоди составляла 2.12 на 1 POMATIOPSIS AND ONCOMELANIA самку в месяц. Четыре исследованных вида Pomatiopsis не адаптировались достаточно хорошо к лабораторным условиям. Pomatiopsis californica и P. binneyi быстро отмирали, не давая потомства. P. lapidaria и Р. cincinnatiensis (собранные в возрасте около 1 года) имели конечную скорость отмирания 16% в месяц, при "оптимальных" условиях за период более 10 месяцев для первого вида и за 3 месяца для второго; после этого периода темп отмирания быстро возрастал, отчасти благодаря более короткому жизненному циклу этих видов. Прирост молоди составлял по крайней мере 0.14 мм в неделю, a смертность - более 30% за 2 месяца. Продукция молоди была 0.51 на 1 самку в месяц за довольно короткий промежуток времени. 143 Вл. ee AANA NI MGM UBD NASIR, о hein PAR | Ч CEE ae A bi AE | Ih #1 rs hu > ico ÍA Mu | de CE LA Diner or) 4 O re MOTOR PO ES al RE о ot O AA | N 7 cy DN Ny у y M yr à У GA TAB к Pai nr db» Pr A hy | A j fi Ч M 14 : 7 a ; fy! a ew! и eri ab MA Wan! с Ny y | IE И Eh an Pres MELLE RIRES if a RE nr E N НИ Pia ap AN À РИ Wi et PAN TRAE ACCS м. \ pyr ay 24 Diva LO ый Bidet Adi Ala FR ee и RON IA МА Зы cat Зо ААА 1c р de | fi an a BER | LUS Сл" RER at emer LC Bea ae ay A A NR A ravir iat aN у” у $ м р y i 7 N И ы di | ü 4 idly te U 0 Fi en ji x Fo ab L | à hu af к. АД y ii 4 у 'q e 7 1 O | | ук Bei , ao у RT SOC (ute : “al PARA ‘eo Ve ND NORTON ART А wi pre des E AP dike ME ahah A E BER o A A Ce hc N oe SF A he MA A e gio qa Noa ACA y ALO ART ВИЙ TA AN * y Иа ny | AY OT A a IN И ME LA A MC N ee Е чаи. AA а be Rees ai © Pea LEONE MOULE E ВИ ory) и Ft ИИ ми ee | NN i wy iW Cr у а уни el A р BEN. FRE: dr a © у hy № i var DIT so d'u ‘PA yc A LY hs | sr. MALACOLOGIA, 1967, 6(1-2): 145-153 SUSCEPTIBILITY OF ONCOMELANIA HUPENSIS CHIUI TO INFECTION WITH SCHISTOSOMA JAPONICUM 1 Jui-Kuang Chiu Department of Parasitology College of Medicine National Taiwan University Taipei, Taiwan (Formosa) ABSTRACT A small amphibious hydrobiid snail, Oncomelania hupensis chiui (originally described as Tricula chiui) from the northern tip of Taiwan, known to act as the snail host of Paragonimus iloktsuenensis, was found to be able to serve as a good intermediate host for both the zoophilic (Formosan-Changhua) and human (Japanese) strains of Schistosoma japonicum. Since the snail host (O. h. for- mosana) of S. japonicum in Taiwan from different locations is either refractory or only slightly susceptible to human strains of S. japonicum from Japan and the Philippines, this discovery of an efficient potential host brings the establish- ment of human schistosomiasis within the range of possibility in Taiwan. The experiments indicated that this snail showed a varying degree of suscept- ibility to the 2 geographic strains of S. japonicum and is, on the whole, a more suitable host for the Changhua strain than for the Japanese strain. As high an infection rate as 100% was obtained for the Changhua strain of S. japonicum, when cercariae were shed fromthe snails 95 days after exposure to 6 miracidia individually. Interesting results were obtained for the Japanese strain. Snails exposed in pairs to 20 miracidiä were 100% infected but did not produce infect- ive cercariae. Only the snails exposed to 5-7 miracidia individually, and in- fected at the rate of 22.2%, shed cercariae 105 days after infection. INTRODUCTION In the course of a study on the tre- matode genus Paragonimus in Taiwan, a small amphibious hydrobiid snail was in- criminated as the first intermediate host of P. iloktsuenensis (Chiu, 1961, 1965b). This snail was subsequently named Tri- cula chiut by Habe & Miyazaki in 1962. Habe & Miyazaki also stated that the species was allied to Oncomelania for- mosana, the snail host of Schistosoma japonicum in Taiwan. This raised the Submitted for publication July 16, 1966. question of whether each of these snails would prove susceptible tothe trematode parasite of the other. As a result, it was found that the snail host of the lung fluke could indeed also carry the blood fluke (Chiu, 1965a). While investigations were in process, a systematic malaco- logical study was made on this snail by Dr. G. M. Davis. He came to the conclusion that the so-called Tricula was in fact Oncomelania (Davis, 1968). Since he considers all so-called “species” of Oncomelania as _ subspecies of O. lThis study was supported by a research grant from the National Council on Science Develop- ment of China. (145) 146 I. E. CHIU hupensis, this new species is now called О. hupensis chiui. The present paper gives further de- tail on the experimental infection of this snail with the non-human Formosan- Changhua strain of S. japonicum and, in addition, reports susceptibility to the human Japanese strain of schistosome. MATERIALS AND METHODS Oncomelania hupensis chiui (Habe et Miyazaki) were collected from the type locality, Alilao village, Taipei County (northern tip of Taiwan) where schisto- somes have not been found. The eggs of the zoophilic Changhua strain (central part of the western coastal plain of Taiwan) of Schistosoma japonicum, used for infecting the snails, were obtained from the liver of a rabbit and a mouse that had been experimentally infected and kept in the laboratory for 40-50 days after infection. Eggs of the human Japanese strain of S. japonicum were secured from the liver of a mouse with a 39 day old infection. The infected Oncomelania nosophora used for in- fecting the mouse were obtained from the 406th Medical Laboratory in Japan. These snails were experimentally in- fected with a strain of S. japonicum originally from Yamanashi, Japan. Mature Oncomelania hupensis chiui snails collected from the field were exposed to miracidia either individually or in pairs in small glasses, allowing 3-5 hours for penetration at 26° + 20C, After exposure, the snails were kept in clay flower-pots in the laboratory as previously described (Chiu, 1965b) at room temperatures varying between 15° and 32°C. Snails were crushed at different times after exposure to mira- cidia, to determine infection. Cercarial emergence was checked by isolating Snails in a small glass of water for 5 hours in the morning. EXPERIMENTS Experiment 1. A preliminary experiment designed to investigate the susceptibility of Onco- melania hupensis chiui to infection with the Changhua strain of Schistosoma japonicum was made as follows: 25 snails were exposed individually to 2-3 miracidia for 5 hours on April 9, 1964. The miracidia were hatched from eggs obtained from the liver of a rabbit which had been infected in the laboratory with cercariae of S. japonicum shed from naturally infected snails. The results obtained from this experiment have been partly reported before (Chiu, 1965a). Upon ist examination, 79 days after infection (Table 1), schistosome sporo- cysts were found inboth snails dissected. These sporocysts contained embryos in various stages. Immature cercariae were seen within some daughter sporo- cysts. Mature cercariae were found emerging from the snails 95 days after exposure to miracidia, and 2 snails crushed contained sporocysts and cer- cariae. Similar positive findings were made on the 121st day after infection, in 1 of 3 snails dissected and on the 127th as well as on the 153rd days in 2 of 3 snails crushed. Cercarial emer- gence was still observed among the 5 surviving snails on the 243rd day after infection. However, on the 282nd day, the 3 surviving snails failed to shed the cercaria and these snails were found dead a month later. In summary, 9 of 13 snails crushed (69.2%) were found infected with the schistosomes. This experiment showed that О. h. chiui is readily infected with the Changhua strain of Schistosoma japonicum. The infectivity of the cercariae shed from Oncomelania hupensis chiui was confirmed by means of animal infection. INFECTION OF ONCOMELANIA HUPENSIS CHIUI WITH SCHISTOSOMA 147 TABLE 1. Examination of Oncomelania hupensis chiui snails exposed to 2 - 3 miracidia each of the Changhua strain of Schistosoma japonicum Date F nd Days after No. snails No. snails Larval stages 1964-1965 infection examined* infected found June 27 2 Sporocysts July 13 (Shedding) Cercariae 2 Sporocysts & Cercariae Aug. 8 121 1 Sporocysts & Cercariae Aug. 14 127 2 Sporocysts & Cercariae Sept. 9 153 2 Sporocysts & Cercariae Oct. 30 (Shedding only) Cercariae Dec. 8 (Shedding only) Cercariae (Shedding only) (-) Total dissected 13 9 (69. 2%) *Numbers in parentheses designate snails that were not dissected. TABLE 2. Examination of Oncomelania hupensis chiui snails exposed to 6 miracidia each of the Changhua strain of Schistosoma japonicum Date examined 1964 No. snails examined Days after infection Sept. 20 Oct. 4 Nov. 27 Dec. 7 Total dissected 8 Three mice were exposed to an un- determined number of cercariae from 2 positive snails for 2 hours, and at autopsy, 40 days later, adult Schistosoma japonicum were collected. Experiment 2. Another experiment, with heavier ex- posure to miracidia, was made to gain further insight concerning the infectivity of Oncomelania hupensis chiui as regards the Changhua strain of Schistosoma japonicum in the laboratory. Fifteen snails were exposed individually to 6 miracidia for 4 hours on August 24, (Shedding only) No. snails infected Larval stages found 1 Sporocysts i Sporocysts Cercariae Sporocysts & Cercariae 6 8 (100%) 1964. The miracidia were hatched from eggs obtained from the liver of a mouse infected in Experiment 1. Upon dissection of 1 snail each on the 27th day and the 41st day after infection, sporocysts were encountered in both (Table 2). On the 95th day, cercarial shedding was demonstrated in the 6 surviving snails. Ten days later, these were dissected, and all 6 were found positive for sporocysts and cercariae. The infection rate among those snails that had survived in this experiment was 100%. This high rate indicates that Oncomelania hupensis chiui pos- 148 J. K. CHIU TABLE 3. Examination of 2 groups of Oncomelania hupensis chiui snails exposed to different number of miracidia of the Japanese strain of Schistosoma japonicum A exposed to 5-7 miracidia each Date Days after examined infection 1965 March 20 March 22 March 25 April 8 April 15 April 22 May 4 May 11 May 18 May 27 - June 19 June 29 July 5 July 12 Sept. 6 Totals dissected 45 (Shedding) (Shedding) 10 (22. 2%) B exposed to 10 miracidia each on the average Larval Larval stages stages found found (-) (59 Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts Sporocysts (-) (Shedding) (5) Cercariae (Shedding) (-) Sporocysts & 1 Sporocysts Cercariae Sporocysts & Sporocysts Cercariae Sporocysts & de- generating cercariae 11% 11 (100%) *The two snails examined as early as the 4th and 6th days after infection are excluded from the total. sesses a high degree of susceptibility to infection with the Changhua strain of Schistosoma japonicum. The snail should be an excellent host for maintaining the life cycle of the Changhua strain of S. japonicum in the laboratory unless the schistosome should adapt specifically to O. h. chiui. Experiment 3. The susceptibility of Oncomelania hupensis chiui to the Japanese strain of Schistosoma japonicum was tested on 2 groups of snails, A and B, using 2 different doses of miracidial exposure: 75 snails in Group A were exposed individually to 5-7 miracidia, and 42 snails in Group B were exposed in pairs to 20 miracidia, for 3 hours, on March 16, 1965. Sporocysts were first detected in the internal organs of a snail from Group B 9 days after infection (Table 3), none having been encountered in the 2 snails examined earlier. However, these 2 snails were very possibly infected, since INFECTION OF ONCOMELANIA HUPENSIS CHIUI WITH SCHISTOSOMA 149 TABLE 4. Infection percentages of Oncomelania* from different geographic locations with 4 strains of Schistosoma japonicum reported by various workers | Schistosoma japonicum strain Species of Oncomelania % O. h. hupensis (China) 34 (D) O. h. nosophora (Japan) 0 (D) O. h. quadrasi (Philippines) 0 (D) O. h. formasana (Taiwan) - Changhua 0 (D) - Ilan - - Kaohsiung = О. h. chiui (Taiwan) - Alilao a Chinese Japanese Philippine Formosan (Changhua) % % % 13 (D) 20 (HH) 0 (D) 21 (D) 9.6 (HH) 21 (D) 44.4 (HRO) 35. 7-43. 8 (MW) 0 (D) 44-75 (P) 6.4 (D) 28. 7-45. 0 (MW) 0 (D) 0 (HH) 35 (D) 0.8 (HRO) 18. 0-36. 2 (MW) 5.6 (MW) 5 (MW) 1 (MW) 0 (MW) 0 (MW) 0-1.8 (MW) 22. 2-100 (С) - 69. 2-100 (С) *The “species” of Oncomelania are here all considered to be subspecies of О. hupensis. Abbreviations: (D) = DeWitt, 1954; (HRO) = Hunter, Ritchie & Otori, 1952; (MW) = Moose & Williams, 1963, 1964; a 100% infection rate was later shown to prevail in this heavily exposed group. The negative finding suggests that schistosome larvae may stay inthe head- foot muscle for a while after penetration. Lower infection rates (22.2% on the average) were observed in Group A, exposed to fewer cercariae. Sporocysts were discovered in 5 out of 15 snails dissected 23-63 days after infection. On the 95th day, shedding failed to occur in both groups. On the 105th day, cer- cariae were shed by snails of Group A, but not of Group B. Six days later, 2 of 5 snails crushed in Group A harbored sporocysts and cercariae, while 1 snail examined in Group B still harbored nothing but sporocysts. On the 118th day, 3 of 20 snails dissected in Group A were found infected, but only 1 snail harbored cercariae, the other 2 merely sporocysts. In a snail from Group B, again only sporocysts were detected on the 111th and 118thdays. The 7 surviving (P) = Pesigan et al. , 1958; (C) = Chiu, this paper; (HH) = Hsti & Hsü, 1960. snails were dissected on the 174th day. None of 5 snails crushed in Group A were infected with schistosomes. In contrast, the 2 surviving snails from Group B were parasitized with sporo- cysts and a few cercariae, but these were degenerated and apparently non- infective. In summary, 10 of 45 snails dissected (22.2%) were infected with the parasite in Group A. On the other hand, al- though presumably a 100% infection rate obtained in Group B, no snail was capable of producing infective cercaria. It was also noted that the snail death rate was significantly higher in Group B (69%) than in Group A (40%). These obser- vations, as compared with those for the Changhua strain of Schistosoma japoni- cum, suggest that Oncomelania hupensis chiui is a less suitable host for the Japanese than for the Changhua strain of S. japonicum. The infectivity of schistosome cercariae of the Japanese 150 J. K. CHIU strain shed from Oncomelania hupensis chiui was also confirmed by animal in- fections. Two mice were exposed to an undetermined number of cercariae for 2 hours, and at autopsy, 42-49 days after infection, adults of Schistosoma japonicum were recovered. DISCUSSION It is well known that oncomelanid snails from various geographic locations possess a varying degree of suscepti- bility to infection with different strains of Schistosoma japonicum (Hunter et al., 1952; DeWitt, 1954; Pesigan et al., 1958; Hsü & Hsü, 1960; Moose & Williams, 1963, 1964). The knowledge available is summarized in Table 4. It is seen that Oncomelania hupensis hu- pensis from China was susceptible to the Chinese, Japanese and Philippine strains of S. japonicum, but refractory to the Formosan-Changhua strain; O. h. noso- phora from Japan was susceptible to the Japanese, Philippine and Changhua schistosome strains, but not to the Chinese strain; O. h. quadrasi from the Philippines to the Philippine and Changhua strains, but not to the other 2 strains; O. h. formosana from Changhua, Taiwan, was susceptible to the Changhua strain, was faintly infected by, but an unsuitable host for, the Japan- ese schistosome strain, and refractory to the other 2 strains. Recent findings by Moose & Williams (1963, 1964) indicate that O. h. formosana from Ilan, in north eastern Taiwan, another endemic area for Schistosoma japonicum, recently dis- covered by Kuntz (1965), was relatively susceptible to the Japanese and Philippine strains of schistosome, but exceedingly resistant to the Changhua strain; whereas snails from Kaohsiung, in southern Taiwan, were altogether unsuitable as hosts and resistant to the Changhua, Japanese and Philippine strains. The author (1965a) has already reported the fact of susceptibility of “Tyicula (=Oncomelania) chiui” to in- fection with the Changhua strain of Schistosoma japonicum. In the present study, Oncomelania hupensis chiui was not only confirmed as a good potential intermediate host for the Changhua strain but also found capable of transmitting the Japanese strain of S. japonicum. In other words, the snail can serve as an efficient host for both the non-human and human strains of S. japonicum. For the Changhua strain of S. japonicum an infection rate of 100% could be obtained. The results for the Japanese strain were interesting, in that moderately heavy exposure (5-7 miracidia per snail) resulted in functional infection at the rate of 22.2%, whereas, with heavier exposure (an average of 10 miracidia per snail), even though the infection rate amounted to 100%, not a single mature infective cercaria could be found throughout the 5 month duration of the experiment. This interrelationship be- tween the cercaria producing capacity of О. h. chiui and the number of miracidia to which it was exposed remains to be understood. The results of this study have further demonstrated that O.h. chiui also showed a varying degree of susceptibility to 2 geographic strains of S. japonicum, as do the other oncome- lanid snails, i.e., at approximately equal exposure, infection rates were about 100% for the Changhua strain of S. japonicum and 22% for the Japanese schistosome (Tables 2 and 3A). The variety in response now realized to exist in local strains of O. h.formosana - from refractive to the indigenous non- human schistosome to slightly sus- ceptible to the foreign human schisto- somes - and the discovery of the new subspecies O.h. chiui, susceptible to both these schistosomes, already entails a revision of our former views, in particular of the view that Taiwan was safe from the threat of human schisto- somiasis, because there was no potential snail host for the human strain of S. japonicum. No doubt further investi- gation will provide further evidence of variation in the snail-parasite relation- INFECTION OF ONCOMELANIA HUPENSIS CHIUI WITH SCHISTOSOMA 151 ship. Among other related hydrobiid snails, DeWitt (1954) reported that the North American Pomatiopsis lapidaria was capable of infection with the Chinese (1%) and the Changhua (3%) strains of Schistosoma japonicum. It is also of interest to note that Hunter & Abbott (1949) reported on an infection experiment with Tricula minima from Japan and the Japanese strain of Schistosoma japonicum. They found that the miracidia could not develop to the cercarial stage in that snail; only a few degenerating mother sporo- cysts were discovered 7-71 days after infection. ACKNOWLEDGEMENTS The author is indebted to Major J. W. Moose, former Chief of the De- partment of Medical Zoology, 406th Medical Laboratory in Japan, for supply- ing Oncomelania hupensis nosophora in- fected with the Japanese strain of Schistosoma japonicum. Thanks are due to Lt. (j.g.) D. E. Wood, formerly of the U. S. Naval Medical Research Unit No. 2 in Taiwan, for assistance in ob- taining snails. Thanks are due to Dr. G. M. Davis, Malacologist of the Department of Medical Zoology, 406th Medical Labora- tory in Japan, for his continuous interest in this study. LITERATURE CITED CHIU, J. K., 1961, Snail host of Para- gonimus iloktsuenensis in Taiwan. J. Form. med. Assoc., 60(12): 1173. 1965a, Tricula chiui: a new snail host for Formosan strain of Schistosoma japonicum. J. Parasit., 51(2): 206. 1965b, Tricula chiui Habe et Miyazaki, 1962: a snail host for Paragonimus iloktsuenensis Chen, 1940 in Taiwan. Japan. J. Parasit., 14(3): 269-280. DAVIS, G. M., 1968, A systematic study on Oncomelania hupensis chiui (Gastropoda: Hydrobiidae). Mala- cologia (in press). DeWITT, W. B., 1954, Susceptibility of snail vectors to geographic strains of Schistosoma japonicum. J. Parasit., 40(4): 453-446. HABE, T. € MIYAZAKI, I., 1962, Tri- cula chiui sp. nov., a new snail host for the lung fluke Paragonimus ilok- tsuenensis Chen in Formosa. Kyushu J. med. Sci., 13(1): 47-49. HSU, ¿SY LL AG SU, VE. Е, 1960. Infectivity of the Philippine strain of Schistosoma japonicum in Oncomel- ania hupensis, O. formosana and O. nosophora. J. Parasit., 46(6): 793- 796. ; HUNTER, G. W. Ш & ABBOTT, R. T., 1949, Studies on potential snail hosts of Schistosoma japonicum. II. In- fection experiments on amnicolid snails of the genera Blanfordia, Tri- сша and Fukuia. Helm. Soc. Washing- ton, 16(2): 86-89. HUNTER, G. W. Ш, RITCHIE, L. S. & OTORI, Y., 1952, A comparison of the infectivity of Schistosoma japoni- cum occurring in Japan for Oncome- lania formosana. J. Parasit., 38(5): 492. KUNTZ, R. E., 1965, Zoophilic schisto- somiasis with areport of a new locality on Taiwan. J. Form. med. Assoc., 64(10): 649-657. MOOSE, J. W. & WILLIAMS, J. E., 1963, Susceptibility of Oncomelania formosana from three different areas of Taiwan to infection with Formosan strain of Schistosoma japonicum. J. Parasit., 49(4): 702-703. 1964, The susceptibility of geographical races of Oncomelania formosana to infection with human strains of Schistosoma japonicum. 406th Medical Laboratory Research Report, U. S. Army Medical Command, Japan, Presented at First Internat. Congr. of Parasit. in Rome, Italy, September 1964. PESIGAN, T., HAIRSTON, N. G., JAURE- GUI, J. J., GARCIA, E. G., SANTOS, 152 A. T., SANTOS, В. С. € BESA, А. A., 1958, Studies on Schistosoma japoni- cum infection in the Philippines. 2. The molluscan host. Bull. Wld Hlth Org., 18(4): 481-578. ADDENDUM J. К. CHIU that Oncomelania hupensis chiui is also susceptible to infection with the Formosan-Ilan, Japanese-Kurume, Philippine and Chinese strains of Schistosoma japonicum, at the rates of 56.5-100%, 87.5-100%, 43.8-100% and 66.7-87.0% respectively. The exposure per snail ranged from 1 to 15 mira- Further experiments have indicated cidia. RESUMEN SUSCEPTIBILIDAD DE ONCOMELANIA HUPENSIS CHIUI A LA INFECCION POR SCHISTOSOMA JAPONICUM Jui-Kuang Chiu Se ha descubierto que el pequeño caracol hydróbido, anfibio, Oncomelania hupensis chiui (descripto originalmente como Tricula chiui) del extremo norte de Taiwan (Formosa), conocido como huésped de Paragonimus iloktsuenensis, es también capaz de servir de intermediario tanto para la raza zoofílica (Formosa-Changhua), como para la raza que ataca al hombre (japonesa) de Schistosoma japonicum. Desde que el caracol huésped (O. h. formosana) de $. japonicumen Taiwan de diferentes localidades, es refractario o muy ligeramente susceptible a las razas de infección humana de S. japonicum, este descubrimiento de un huésped potencialmente eficiente, trae la esquistosomiasis humana dentro del área de posible establecimiento en Taiwan. Los experimentos indicaron para este caracol un grado variable de susceptibilidad a las dos razas geográficas deS. japonicum, y es, en general, un huésped más adaptable a la raza Changhua que a la japonesa. El linaje Changhua mostró una proporción infecciosa del 100%, y los caracoles libraron cercarias 95 dias después de ser expuestos a 6 miracidios por individuo. Resultados interesantes se obtuvieron del linaje japonés. Caracoles expuestos en parejas a 20 miracidios, fueron 100% infectados pero no produjeron cercarias infecciosas. Sólo aquellos expuestos individualmente a 5-7 miracidios, infectados en una proporción del 22.2% libraron cercarias 105 dias después de la infección. ABCTPAKT ВОСПРИИМЧИВОСТЬ ONCOMELANIA HUPENSIS CHIUI К ЗАРАЖЕНИЮ SCHISTOSOMA JAPONICUM ДЖУ-КУАНГ-ШИУ Мелкая амфибийная гидробия Oncomelania hupensis chiui (первоначально описанная как Tricula chiui) с северной око- нечности Тайваня известна как промежуточный хозяин Parago- nimus iloktsuenensis; было найдено также, что она может служить хорошим промежуточным хозяином как для зоофильного штамма (из формозы-Чангуа), так и для человеческого (японского) Schistosoma japonicum. Поскольку улитка-хозяин (O. h. formosana) паразита 5. japonicumô из различных мест Тайваня являесся устойчивой или лишь слабо-восприимчивой к человеческой Форме 5. japonicum из Японии и Филиппин, потенциального хозяина Schistosoma позволяет ожидать развития человеческого шистозомиазиса на Тайване. открытие весьма эффективного INFECTION OF ONCOMELANIA HUPENSIS CHIUI WITH SCHISTOSOMA Эксперименты показывают, что этот моллюск имеет различную степень восприимчивости в двум географическим штаммам $5. japonicum и, в целом, является более подходящим хозяином для штамма из Чангуа, чем для японского штамма. 100% заражение было получено для штамма 5. japonicum из Чангуа, когда, церкарии выходили из моллюсков через 95 дней после индивидуального заражения их 6 мирацидиями каждый. Интересные результаты были получены для японского штамма. Улитки заражались попарно 20 мирацидиями и давали 100% заражение, но не давали заражающих церкарий. Улитки, зараженные 5-7 мирацидиями индивидуально и давшие только 22.2% заражения, давали церкарии через 105 дней после заражения. 153 SOE) naga ves aye Bar MSR ESA RE AN ERA AD aon VW 0 ‘à NO MS OMAN +. A MART ря year ah DCR "EX PANE AN wi he ie: «$ nu ys PANA CA gr ahi # AS + dur ee seg ne oil 9 LTE F | i ep pee gti ham Shah, ee pu Hk AD, soupe «da MON CAEN ved weer à iy 654 AS sind rol P PALA A sr? Haut VAT NE y А, fit и“ in ET ong Seat Las rai AN TEA) MICHEL RE ci Mt TE TS 05 ¿NEBOT E wa A SRD J AE ay Ol ARS et. Meet ER ET LATE ET N TR RT EA TONER N и ES th Sur Hart ROSS E O TER ane м-р INN Co Er A к | vee hd gy ey 1 ar Ber) с > Ae E ML SAAC AD byt ana, a Pua, À vn is Lit SUN SE APR DT, a f er Соч К EN ? ns AT, VS his ne у | И eS ee A aches EN il el ñ Ps) a ПИТ у БА AT ie TARA far ) VER A Lath aca, | FEUX Щи: SIAR AE AUS UE ES MANIERE TES 4 у frat À hi. A FERRER EL Taye po TSO ES OI. dpt sach, ed PA oy Wi иг, ON Come) tm) hey rin Ms we lhe у j : | , : | [ANT OMe 19 Ô FA wy Dam er ve | CARO N 5 f f “a ná ; A AAA aD landes UI MAA ола! FE de | | } Loe ta E Ga Em Ya TA he OG HE a ALERT: RM 4 hi wu Wish: Eh} | ' у ТА me | sto LEUR “et IN рт `` ИИА, ICO ia Le oy 8) 4 LE р a aie : sty D AE rt Cr AA, AMA re y A a ‹ A ‘ | t à Г. Ar ee À AT PPT Ara re. Dé VA x т Fr, A EU: amara) ‘au L 2 K h ut e Lh TM À « ? a . À AQU у . De DE LPS SEC aout if 4 we iis ‘al ae ey” У | у TR И ee en CORRE ES м à ar Ca Bi a La | sey CAE > tu Sa | is у svn Pues adi eee ee ithe RE RC ER ден saan yd HU ART PQ le EU) té be 9 | Fe ANT ae: sh da | hae ки ЛА MALACOLOGIA, 1967, 6(1-2): 155-174 THE ANATOMY AND RELATIONSHIPS OF A SOUTH AFRICAN FERRISSIA (BASOMMATOPHORA: ANCYLIDAE) D. 5. Brownl British Medical Research Council Institute for Parasitology, Durban, Republic of South Africa ABSTRACT Ferrissia, introduced by Walker (1903) to accommodate Ancylus rivularis Say of North America, has subsequently been regarded as a worldwide genus of freshwater limpets with fine radial sculpture on the apex of the shell. On anatomical evidence, the New and Old World species were classified in the subgenera Ferrissia s.s. and F. (Pettancylus) respectively by Hubendick (1964). The genus Gundlachia Pfeiffer, originally characterised by the septate shell, was restricted by Hubendick to comprise ancylids of Central and South America; African species previously assigned to that genus are, apparently, septate forms of Ferrissia. The range of Ferrissia in Africa extends from the Mediterranean to the South African Cape. Published information about the anatomy of African species is restricted to the radula, jaw and external characters. The present account deals with certain features of the internal and external anatomy of non-septate specimens of F. burnupi (Walker, 1912) collected in Natal province, Republic of South Africa. One object of the study was to obtain information for com- parison with other Old World Ferrissia species, in particular Е. tenuis (Bourguignat) which has been reported to transmit human schistosomiasis in India. Features of Ferrissia burnupi indicating a close relationship with the Petiancylus group of species are: the poor pigmentation, the lack of fusion between the plates composing the jaw, the large uterine gland complex, and, above all, the structure of the male copulatory organ with its small penis and vestigial penis sheath into which opens a long flagellum. Although the dif- ferences between К. burnupi and Г. tenuis in the radular teeth, seminal vesicle and flagellum may be of specific rank, F. buynupi may be worthy of considera- tion as a possible intermediate host of Schistosoma haematobium in Natal. Anatomically Ferrissia burnupi resembles the non-septate Italian form of Watsonula wautieri Mirolli, according to information given by Mirolli (1960) and Hubendick (1964), but the copulatory organ differs from that described by Wautier et al. (1966) for French specimens identified as Gundlachia wautieri. The left pleural ganglion of Ferrissia burnupi is distinguishable from, though intimately connected with, the left parietal ganglion. Fusion between the corresponding ganglia of the right side has been carried even further. The lateral ganglia in the visceral loop are named left and right pleuro-parietal ganglia and the asymmetrically placed ganglion lying between them the ab- dominal ganglion, a terminology which expresses the composition and function of these 3 ganglia in Ferrissia more accurately than various previous termi- nologies, and may be applicable to some other ancylids. There are separate connectives between the pedal ganglion and the cerebral and pleuro-parietal ganglia on each side. lpresent address: c/o Medical Research Council, 20 Park Crescent, London W. 1, England. (155) 156 D. 5. BROWN INTRODUCTION Ferrissia was introduced by Walker (1903) as a section of Ancylus for the North American species A.rivularis Say. It has subsequently been regarded as a genus comprising freshwater limpets with fine radial sculpture on the apex of the shell from many parts of the world. Hubendick (1964) defined 2 subgenera with different male copulatory organs and gave the distribution of Ferrissias.s. as North and Middle America and the West Indies; the species of Africa, South and East Asia, Australia and Oceania were placed, together with Watsonula wautieri Mirolli of Italy in F. (Pettancylus), with Ancylus tasmanicus Tenison-Woods as type species. At least some species of each subgenus are capable of forming septate shells, in which an upper anda lower portion are partially separated. The presence of a shell septum closing the posterior part of the aperture has been regarded by many authors as a characteristic of Gundlachia Pfeiffer, 1849, and for this reason several African ancylids have been classified in that genus. In the case of the South African species equeefensis (Walker) and G. burnupi Walker, Walker (1923, 1926) believed that their generic position was confirmed by the structure of the radular teeth. However, the central radular tooth of G. equeefensis, redescribed by H. Watson (in Connolly, 1939), and that of G. burnupi, illustrated by H. B. Baker (in Walker, 1926), have, like the Ferrissia species of South Africa, sym- metrical cusps instead of the asymmetri- cal cusps found in Gundlachia species of South and Central America. North American septate ancylids placed in Gundlachia by some authors were trans- ferred to Ferrissia by Basch (1959a, 1963), who concluded that South American species belonging to the former genus are distinguished by the asymmetrical cusps on the centralradular teeth. Like- wise, Hubendick (1964) restricted Gund- lachia to comprise Neotropical species with asymmetrical cusps on the central tooth; these species are further differ - entiated by the structure of the male copulatory organ. Although no de- scription of the copulatory organ of a septate African ancylid has been published, it appears unlikely that Gund- lachia occurs in the southern part of the continent, and the present author follows Connolly (1939) by includingin Ferrissia the South African forms that were formerly placed in Gundlachia. The range of Ferrissia in Africa ex- tends from Egypt to the South African Cape, although, in spite of considerable collecting activity, no species have been recorded from the Congo or Angola (Pilsbry & Bequaert, 1927; Wright, 1963). Twelve species have been de- scribed from South Africa, Rhodesia, and Mozambique (Walker, 1912, 1923, 1926; Connolly, 1925), and the genus appears to be particularly diverse and abundant in southern Africa. Of the other 2 ancylid genera occurring in Africa, Ancylus reaches the southern limits of its range in Ethiopia. The species of Burnupia Walker, 1912, known from Africa only, have punctate shell apices. Little information has been published about the anatomy of African Ferrissia species. The radulae of 2 South African forms have been described (Walker, 1923; Watson, in Connolly, 1939), and the radula and jaw of F. junodi Connolly, from Mozambique (Connolly, 1925). H.B. Baker (in Walker, 1926) gave an account of some external features and also the radula of Gundlachia burnupi (transfer - red to Ferrissia and renamed F. clifdeni by Connolly, 1939). Brown (1965) des- cribed the radulae of aphallic Ethiopian specimens belonging to 2 species. The present account deals with certain fea- tures of the external and internal ana- tomy of non-septate specimens of Fer- vissia burnupi (Walker, 1912) that were collected from a single population in Natal province, Republic of South Africa. Future revision may show that some or all of the other 5 species of Ferrissia recorded from Natal (Connolly, 1939) ANATOMY OF FERRISSIA 157 are synonymous with Е. burnupi, which is the senior species described from that region. One reason for undertaking the study was to obtain information for comparison with other Old World Fer- rissia species, in particular F. tenuis (Bourguignat) of India. Ferrissia tenuis has been reported to serve as an intermediate host of Schisto- soma haematobium in a focus 250 km south of Bombay (Gadgil & Shah, 1955; Gadgil, 1963); it is consequently de- sirable to compareF. tenuis with African species of Ferrissia since these ancylids are widely distributed in the regions of Africa where human urinary schisto- somiasis is endemic. In South Africa, Ferrissia occurs abundantly in a variety of habitats including permanent lakes, temporary pools, and small stony streams. Porter (1938) examined 45 “wild” specimens for trematode cer- cariae and reported a “furcocercous monostome” and a “holostome” from F. burnupt. Cercariae of African schistosomes have not been reported from African Ferrissia, although, as far as the author is aware, the susceptibility of the snails has not been tested experi- mentally. MATERIAL AND METHODS The specimens of Ferrissia burnupi used in this study were collected by the author from Amahlongwa River, 5 miles northeast of Umzinto, Natal province, Republic of South Africa; at bridge on road between Umzinto and Umkomaas, within Umahalangwa Mission Reserve (South Africa 1: 250,000 topo-cadastral map, sheet 3030 Port Shepstone). Co- ordinates: south 30° 15’, east 30° 43’. 17 July 1964. Collector's number 359. National Snail Collection (Institute for Zoological Research, Potchefstroom University, South Africa) catalogue No. 68.48.66. The limpets were collected from dead leaves in residual pools on sandbanks when the river was at a low level. The majority of shells were thickly encrusted with a black deposit, which was re- movable with oxalic acid solution. External features were observed in snails that were active or had been narcotised in a 5% solution of Nembutal. Dissections were made of 30 specimens fixed in 5% formol saline at 60° C, and preserved in 80% ethanol. Two speci- mens were fixed in Bouin’s Fluid, serially sectioned and stained with Mallory’s Triple stain for histological study. OBSERVATIONS Shell (Figs. 1-4) The largest shells were 4.0 mm long. No septate specimens were found. The lateral margins are straight or slightly convex. There are numerous fine radial ribs on the apex, and coarser ribs on the anterior slope, which extend to the Shell margin in some small specimens. Irregularly placed circular ridges are conspicuous on some parts of the shells. In their height, the present shells resemble Ancylus (Ferrissia) burnupi Walker, 1912, and A. (F.) equeefensis Walker, 1912, both described from Equeefa River, which lies approximately 10 miles southwest of the Amahlongwa River at Nkwifa. The apex, in its shape and position relative to the posterior margin of the shell, is most similar to that of F. equeefensis. Walker (1923) stressed the importance of differences between the radulae of these species, but he appears to have illustrated worn teeth of F. equeefensis. In view of the variation in shell form that may be ob- served in series of Ferrissia from a single locality, and of the fact that F. burnupi and РЕ. equeefensis have a common type locality, the former name, which has page priority, is employed for the present material. Animal (Figs. 5, 8, 9) In an active animal both the anterior and posterior ends of the foot are bluntly rounded. No grooves or macroscopic glandular openings were observed onthe 158 D. S. BROWN mn Ni ul | NA Wie x 44 PQ | AU 9: Я gi 5 LM OS KI’ PA IN 1-0 mm 0:5 mm ne D FIGS. 1-7. Figs. 1-4. Ferrissia burnupi. Dorsal and lateral views of shells. Fig. 5. Ex- ternal features of left side of animal removed from shell. Fig. 6. Alimentary canal. Fig. 7. Stomach viewed from right side. $: female genital aperture; с’: male genital aperture. ANATOMY OF FERRISSIA 159 List of abbreviations (except nervous system) AG albumen gland AN anus BM buccal mass BW cut edge of body wall С carrefour CA caecum CI cilia CR crop DG digestive gland EG egg HYG) eye FL flagellum FP floor of pulmonary cavity FT foot GI gizzard GU glandular uterus HD hermaphrodite duct IN intestine KI posterior loop of kidney tubule KI’ anterior loop of kidney tubule LM left anterior shell muscle MA membrane of André ME mantle edge MM median muscle MU muscular part of uterus NG nidamental gland МВ non-glandular ridge of uterus OD oviduct ODG opening of digestive gland OE oesophagus ОР operculate suture of egg capsule foot. A small post-tentacular lappet (PL, Fig. 5) is attached to the postero-lateral base of each tentacle. Each eye (EY) is situated within the median anterior part of the tentacle base. The female genital aperture lies beneath the anterior end of the pseudobranch. When the male copu- latory organ is present there is a male aperture behind the left tentacle close to the post-tentacular lappet. The term pseudobranch may be applied to the whole of the flap that projects from the body between the left anterior shell muscle (LM) andthe posterior shell muscle (PM). The posterior part of the pseudobranch (PS, Figs. 5, 9), consisting of a single lobe with 2 longitudinal folds (“auriform lobe” of Mirolli, 1960), OS osphradium OV ovotestis 19 penis PA pulmonary aperture PC pulmonary cavity PE pericardium PI pigmented band РК proximal sac of kidney PL post-tentacular lappet PM posterior shell muscle PR preputium PS pseudobranch PSH penis sheath PY pyloric stomach QM quaternary membrane RCG right cerebral ganglion RE rectum RO position of renal opening on ventral sur- face of mantle SA sarcobelum SG _ salivary glands SP spermatheca TS thin strip in dorsal wall of uterus TT terminal tail of egg capsule UR ureter UT uterus VA vagina VD vas deferens VE ventricle VH external surface of visceral hump probably performs a respiratory func- tion. The rectum (RE) passes forwards through the anterior part of the pseudo- branch, which may be called the anal lobe, to the anus (AN) on the dorso- lateral surface of this lobe. In living animals a capacious hypo- peplar cavity (a cavity protected by the mantle but lying outside the pulmonary aperture; Harry, 1964) is formed by the extension of the peripheral region of the mantle to the edge of the shell. The pulmonary aperture (PA, Figs. 5, 10) is situated behind the left anterior shell muscle (LM), and leads to a pulmonary cavity that is bounded dorsally and pos- teriorly by the pericardium (PE) andthe anterior loop of the kidney tubule (KI’), 160 D. 5. BROWN FIGS. 8 & 9. Dorsal view of Ferrissia burnupi removed from shell. Fig. 8. The organs seen in situ as if the mantle were transparent. Fig. 9. Organs in situ after the mantle, kidney, peri- cardium, and digestive gland have been removed. Median wall of pulmonary cavity indicated by heavy broken line. ANATOMY OF FERRISSIA 161 RIGHT SIDE 025mm LERT SIDE FIG. 10. Ferrissia burnupi. Transverse section at vaginal opening (viewed towards the pos- terior end of the animal). and medianly by the anterior part of the digestive gland (DG, Figs. 8, 10). The osphradium (OS, Figs. 5, 8) may be observed through the dorsal surface of the mantle, slightly anterior tothe lateral edge of the left anterior shell muscle; it is an invagination from the ventral surface of the mantle that is lined with ciliated epithelium. There is little dark pigment inthe foot, the tentacles, or the superficial tissues of the cephalic region, and pigment is most densely concentrated in the labial palps. Internally there are numerous flecks of dark pigment in the tissue sheathing the buccal mass and in the lining of the body cavity. Inthe majority of specimens pigmentation of the mantle is confined to a short median band onthe dorsal surface between the anterior shell muscles (PI, Fig. 8), but in one animal there was a wide band anterior to these muscles. Many shining bodies resembling oil droplets were observed in the tissue of the mantle edge in living animals, and also present in some individuals were small opaque white granules. These granules do not appear to be comparable to relatively large white glands in the mantle edge of Burnupia caffra (Krauss), which discharge a milky secretion when the animal is prised from the substratum (unpublished observations). There are many cilia on the external surfaces of the body, particularly on the pseudobranch. Fine particles were ob- served to be drawn under the anterior edge of the shell of stationary animals and expelled from beneath the posterior 162 D. 5. BROWN shell margin. Organs of the visceral hump (Figs. 5, 8, 9) The organs lying within the visceral hump may be seen through the overlying tissue without dissection (Fig. 8). Most of the space on the right side of the visceral hump is occupied by the digestive gland (DG), beneath which lies the albumen gland (AG). The oesophagus is deflected to the left side, so that the stomach (PY) lies in the left side of the body. The apex of the visceral hump is occupied by the ovotestis (OV), which is surrounded by the posterior part of the digestive gland. The dorsal attachment of the posterior shell muscle (PM, Fig. 8) lies to the left of the median line. The 2 anterior shell muscles (LM, RM) are widely separated and between them the connection between the mantle and the body wall of the cephalic region is relatively thick. A small median muscle (MM), which isin- serted in the cephalic body wall at the base of the connection with the mantle, passes between the anterior shell muscles into the cavity of the visceral hump; its dorsal end is attached to epi- thelium at the posterior end of the median band of pigment on the mantle (PI). The kidney tubule (KI, KI’, Figs. 5, 8) extends posteriorly from the anterior end of the proximal sac (PK), forms a pos- terior loop (KI) near the posterior shell muscle and then runs forward to near the left anterior shell muscle. Here, it makes a double anterior loop (КГ), and continues as ureter (UR, Fig. 8) posteriorly once more tothe renal opening (RO), whichlies on the ventral side of the mantle lateral to the posterior shell muscle. A large part of the kidney tubule is contained within the eave-like projection of the mantle that overlies the pulmonary aperture and the pseudobranch (Fig. 10). The pos- terior part of the pericardium (PE, Fig. 8) extends towards the proximal sac of the kidney (PK), to which it is probably connected by a reno-pericardial duct as observed by Hubendick (1958, 1960b, 1964) in other species of Pettancylus; in transverse sections of F. burnupz long cilia were observed at the presumable position of this duct. The cephalo-pedal cavity The cavities of the visceral hump and the cephalo-pedal region are partially separated by horizontal membranes divided into anterior and posterior parts (MA, Fig. 9), which extends between the floor of the pulmonary cavity and the right side of the body, where itis attached to the inner surface of the right shell muscle and to the body wall behind that muscle. A similar membrane in Ferris - sia tarda was named the membrane of André by Hoff (1940). The median muscle (MM) passes through a gap between this membrane and the junction between the mantle and the body wall of the cephalic region. Beneath the membrane of André lie the distal genital organs, oesophagus, and the anterior lobes of the albumen gland. The right side of the pedal cavity is largely occupied by the albumen gland (AG), and the left side by the uterus (GU, Fig. 10) and nidamental gland (NG). The alimentary canal (Figs. 6-9, 11, 12) The jaw (Fig. 12) consists of a single row of numerous’ overlapping, but unfused, plates with fine teeth on their median edges; in an animal of 3.2 mm shell length there are approximately 25 plates in each side of the jaw. The radula sac lies beneath the vis- ceral loop of the central nervous system and extends no further posteriorly than the pedal ganglia. In 5 radulae examined the number of teeth in a transverse row is 27 or 29 (Fig. 11). The cusps of the central tooth are extremely small, but it appears that 2 symmetrical median cusps are present and perhaps also a pair of lateral cusps. There are 3 main cusps on the inner lateral teeth. The crown of each tooth is progressively elongated towards the lateral margin of the radula, addition- al small cusps appear on the lateral part of the crown, and an interstitial cusp arises between the ectocone and the mesocone. This interstitial cusp is best ANATOMY OF FERRISSIA 163 ото 12m A > 0:25 mm FIGS. 11-14. Ferrissia burnupi. Fig. 11. Radula (half of a complete transverse row of teeth). Fig. 12. Jaw (oblique lateral view). Fig. 13. Reproductive organs in situ. Fig. 14. Diagram- matic longitudinal section of male copulatory organ reconstructed from serial sections (details of the internal folds of the preputium ommitted). 164 D. 5. BROWN developed in teeth 9-12. Marginal teeth numbers 11 and 12 bear about 10 cusps of which the largest correspond to the 3 major cusps of the inner lateral teeth. The salivary glands (SG, Fig. 6) are attached to the dorsal surface of the buccal mass close to the origin of the oesophagus and are joined above the oesophagus. They do not pass throughthe central nerve ring. The oesophagus lies dorsal to the buccal commissure, and after passing between the cerebral and the visceral commissures, turns dor- sally and gradually widens. The crop (CR) leads to the muscular gizzard (GI), which is succeeded by the relatively thin- walled pyloric region (PY) of the stomach (Figs. 6, 7). Circular and longitudinal muscle fibres are visible externally on the wall of the gizzard, which contains pieces of grit measuring up to 0.25 x 0.20 mm in an animal of 3.5 mm shell length. A caecum (CA) opens into the pyloric stomach at the attachment of the intestine. The main anterior and pos- terior ducts from the digestive gland unite just before their entry (ODG) near the junction of the caecum. The intestine (IN, Figs. 6, 9) forms a loop on the left side of the visceral hump followed by a loop on the extreme right; the rectum (RE) follows the median side of the auri- form lobe of the pseudobranch, and passes through the anal lobe to the anus situated on its dorso-lateral surface (Figs. 5, 9). Reproductive system (Figs. 9, 10, 13-15) The ovotestis consists of 3-5 lobes connected to a common atrium, eachlobe being formed of a group of 5-7 acini (OV, Figs. 9, 13). The proximal half of the hermaphrodite duct (HD) is translucent and relatively wide, but possesses no definite seminal vesicle; the duct be- comes very narrow before entering a swelling of the carrefour (C, Fig. 13). The albumen gland (AG, Fig. 13) is translucent, bluntly lobed, and has a hard texture in animals preserved in alcohol. Most of the volume of the albumen glands that were serially sectioned is made up of secretion, cell nuclei being confined to a superficial layer (Fig. 10). A narrow albumen duct enters a swelling of the carrefour that is separate from the one at the entrance of the hermaphro- dite duct. A very short oviduct (OD, Fig. 13) leads from the carrefour to the uterus (UT), which is sharply bent at its distal end, so that the slender vagina (VA) passes in a postero-lateral direction to the body wall. A narrow strip (TS, Fig. 15) along the dorsal surface of the uterus is thin-walled and broken open in many preserved specimens. Two distinct glandular regions are present in the uterine complex: (1) except for the dorsal strip and near the vagina, most of the uterus wall is thick- ened with glandular tissue (GU, Figs. 10, 15), and (2) the posterior part of the unterine complex is formed by the nida- mental gland (NG), cells of which also occur in the wall of the distal part of the oviduct. These regions correspond to the “glandular region of the uterus” of Mirolli (1960) and the “nidamental gland” of Hoff (1940) and of Mirolli. There is a wide connection between the lumina of the uterus and the nidamental gland (Fig. 15, sections d, e) and they are separate only in the posterior part of the uterine complex (Fig. 15, sections f-h). Muscular tissue is developed at the distal end of the uterus (MU, Fig. 15) near its junction with the vagina; there is a ridge of non-glandular tissue (NR, sections b, c) on the floor of the anterior part of the uterus. Cilia (CI, Fig. 15, section a) were present on the internal antero-lateral surface of the uterus and also in the lumen of the oviduct. The spermathecal duct attaches to the proximal end of the vagina, near to the uterus (Fig. 13). The duct is slightly longer than the club-shaped spermatheca (SP). A male copulatory organ was present in 3 specimens (including one that was serially sectioned) out of 30 that were examined. The penis sheath (PSH, Fig. ANATOMY OF FERRISSIA 165 FIGS. 15-20. Ferrissia burnupi. Fig. 15. Dorsal view of uterus and transverse sections at positions a-h. Fig. 16. Dorsal view of egg capsule. Fig. 17. Lateral view of hatched egg capsule. FIGS. 18-20. Nervous system of Ferrissia burnupi. Fig. 18. Dorsal view of central ganglia. Fig. 19. Left lateral view of central ganglia. Fig. 20. The peripheral nerves of the cerebral, pleuro-parietal, and abdominal ganglia seen in dissection. A, abdominal ganglion; B,B’, buccal ganglia; C,C’, cerebral ganglia; P,P’, pedal ganglia; PA,PA’, parietal regions of pleuro-parietal ganglia; PL, pleural region of left pleuro-parietal ganglion; 1,1’, n. gastri- cus; 2,2’, nervi pharyngeales; 3,3’, n. tentacularis; 4,4’, n. frontolabialis superior; 5,5’, n. labialis medius; 6, n. pallialis sinister; 7, n. pallialis dexter anterior; 8, n. pallialis dexter posterior; 9, п. analis; 10, п. genitalis; 11,11’, п. cervicalis superior; 12,12’, n. cervicalis inferior; 13,13’, n. pedalis superior; 14,14’, n. pedalis medius; 15,15’, n. pedalis inferior. 166 D. S. BROWN 14) is reduced to a vestige; it cannot be distinguished from the preputium ex- ternally (PR), and appears to be perma- nently invaginated within the preputium. A large flagellum (FL, Figs. 13, 14) is attached to the penis Sheath close to the vas deferens (VD); the lumen of the flagellum opens into the lumen of the penis sheath at the base of the penis (P). The penis is small and encircled by a ridge, the sarcobelum (SA), which pro- jects internally from the junction between the penis sheath and the preputium. The vas deferens (VD, Fig. 13) emerges from the body wall laterally to the attachment of the preputium, and follows the length of the preputium before joining the proximal end of the copula- tory organ. Neither the proximal part of the vas deferens nor the prostate gland were found in the 3 euphallic specimens examined, nor in any aphallic snails. Egg capsules collected from dead leaves, or laid on glass surfaces in the laboratory, were approximately circular in outline and contained a single egg each (EG, Figs. 16, 17). The capsule opens by a suture (OP) at its edge (“operculate suture” of Bondeson, 1950) and appears to possess a terminal tail (TT) and quaternary membrane (QM) as described by that author for Ancylus fluviatilis Miiller. Nervous system (Figs. 18-20) Paired cerebral, buccal and pedal ganglia are present (CC’, BB’ and PP’, Fig. 18). A small pleural ganglion (PL) may be distinguished beneath the left parietal ganglion (PA, Fig. 19). As the dorsal part of this pleural ganglion is intimately connected with the parietal ganglion, and fusion between the pleural and parietal ganglia has been carried even further on the right side, it is convenient to refer to a pleuro-parietal ganglion on each Side of the animal. A single abdominal ganglion (A, Fig. 18) is situated asymmetrically on the left side of the visceral loop between the pleuro-parietal ganglia. The ganglia and their peripheral nerves will be described successively. The terminology of the nerves is that used by Lever et al. (1965) in their description of the nervous system of Australorbis (=Biomphalaria) glabratus. The median surfaces of the buccal ganglia (B, B’) are connected by a long, slender commissure which passes under the oesophagus. In Biomphalaria glabrata а п. veceptacularis radulae originates from each point of attachment of the commissure to a buccal ganglion. The presence of these nerves in Ferrissia burnupi was not definitely established, but is suggested by strands of pigmented tissue. From the dorso- anterior surface of each ganglion origi- nates a n. gastricus (1, 1’) giving one branch to the oesophagus and another to the buccal mass. At least 2 nervi pharyngeales (2, 2’) originate from the ventro-lateral surface of each buccal ganglion, near to the attachment of the cerebro-buccal connective. The left cerebral ganglion (C) is slightly larger thanthe right (C’). Medio- dorsal bodies are present near to the attachment of the cerebral commissure. There are swellings at the bases of the peripheral nerves, but large lateral lobes of the kind described by Lever (1957) and Wautier et al. (1961) were not observed. Three nerves pass anteriorly from each cerebral ganglion: a n. tentacularis (3, 3’) originates from the dorso-lateral surface of each ganglion and innervates the post-tentacular lappet; an. fronto- labialis superior (4, 4’) has a more ventral origin and runs to the anterior edge of the dorsal buccal lip; the most ventral in origin and stoutest of all, the n. labialis medius (5, 5’) goes to the ventral part of the lateral buccal lip. Despite careful dissection in the vicinity of the eye, an. opticus could not be found. It is thought that the eye may be in- nervated by a fine branch from the n. tentacularis (3) or from the n. fronto- labialis superior (4), which passes close to the eye. The cerebro-buccal con- nectives are attached to the median ANATOMY OF FERRISSIA 167 surfaces of the cerebral ganglia (Figs. 18, 19). A cerebro-pedal connective originates from the antero-ventral parts of each cerebral ganglion, and a con- nective joins the posterior part of each cerebral ganglion to a pleuro-parietal ganglion. The position of the left pleural ganglion (PL, Fig. 19) is indicated by a swelling, better developed in some individuals than in others, beneath the left parietal ganglion (PA). In whole mounts at a magnification of x 160, the cerebro- pedal and pleuro-parietal to pedal con- nectives were readily distinguishable, and also fibres passing from the pleural region of the left pleural-parietal ganglion to the cerebral ganglion. It was not possible to determine whether the latter fibres originated within the pleural region, in which case they would con- stitute a pleuro-cerebral connective, or passed directly from the parietal to the cerebral ganglion. A relatively small swelling beneath the right parietal ganglion appears to represent the pleural region of the right pleuro-parietal ganglion. No nerves originate from the pleural regions of either pleuro-parietal ganglion. The dorsal (parietal) part of the left pleuro-parietal ganglion is conspicu- ously bigger than the right one (PA, Fig. 18). It bears a single, stout n. pallialis sinister (6) whichpasses dorso-laterally through a cleft in the left anterior shell muscle to the lower surface of the mantle where one of its branches innervates the osphradium (OS, Fig. 20). From the right pleuro-parietal ganglion originate 2 nerves, a п. pallialis dexter anterior (7) and а п. pallialis dexter posterior (8), which both run towards the body wall and mantle of the right side. The former was traced as far as the dorsal edge of the median surface of the right shell muscle, andthe latter was observed to pass through this muscle in a dorso- lateral direction. The abdominal ganglion (A) is attached to the pleuro-parietal ganglia by a short connective on the left side, anda long one on the right side. It bears 2 nerves on its posterior surface, of which the thickest appears to correspond to the n. analis (9) of Biomphalaria glabrata, pro- ceeding posteriorly in the body cavity beneath the viscera to the inner surface of the ventral wall of the pseudobranch, and giving at least one branch to the viscera (Fig. 20). The other nerve extends to the vagina and may corres- pond to the n. genitalis (10) of В. glabrata. Neither the n. intestinalis nor the n. cutaneus pallialis of B. glabrata were observed in Ferrissia burnupi. The pedal ganglia (P, Р’) lie close together and are connected by 2 fine commissures (Fig. 18) Two nerves originate from the dorso-lateral surface of each pedal ganglion, а п. cervicalis superior (11, Fig. 19) passing antero- laterally to the body wall, and а п. cervicalis inferior (12) extending laterally to the body wall (Fig. 19). No nerve corresponding to the n. columellaris of Biomphalaria glabrata was observed. The ventral region of each pedal ganglion bears 4 nerves which innervate the musculature of the foot: an anterior n. pedalis superior (13), a lateral n. pedalis medius (14), and 2 posterior nervi pedales inferiori (15). A statocyst is situated on the lateral median surface of each pedal ganglion. The aorta lies beneath the left pleuro- parietal ganglion and runs across the median surface of the left cerebral ganglion to a haemocoel lying anterior to the pedal ganglia ( “preganglionic sinus” of Boer & Lever, 1959). In Ferrissia burnupi, the wall of this haemocoel is darkly pigmented and over- lies the anterior nerves originating from the pedal ganglia. DISCUSSION Various terminologies have been used by different authors to describe the ganglia of the visceral loopinthe central nervous systems of Ferrissia and other Ancylidae. According tothe basic theory of the gastropod nervous system 168 TABLE 1. to F. (Pettancylus). Feature Radula Jaw Salivary glands Pseudobranch Digestive gland Ovotestis Seminal vesicle Nidamental gland Uterus Spermathecal duct Ferrissia rivularis| Ferrissia austra- Hoff (1940), Hubendick (1964) North America 21-1-21 (maxi- mum), central bi- cuspid (Hoff); 19-1-19, central with 2 major and 2 minor cusps (Hubendick) Dorsal scales partially fused (Hubendick) Separate (Hoff); joined (Hubendick) Unfolded or slighty folded (Hoff); folded (Hubendick) Single opening into stomach (Hoff) and (Hubendick) 5-7 lobes (Hoff) 5 - many lobes (Hubendick) Diverticulum from hermaphrodite duct (Hoff) Attached to uterus by narrow duct (Hoff) Small, with little glandular de- velopment (Hoff) Slightly longer than spermatheca (Hoff) D. S. BROWN Иса Hubendick (1960b, 1964) Australia 13-1-13, central with 2 major and 2 minor cusps Not fused Joined Slightly folded Single opening into stomach 2 lobes Thin walled, bladder-like diverticulum No information Large uterine gland complex Slightly longer than spermatheca | than spermatheca Ferrissia burnupi South Africa 14-1-14 (maxi- mum), central probably with 4 minute cusps Not fused Joined Folded Single opening into stomach 3-5 lobes Slight dilatation of hermaphro- dite duct Intimately associated with uterus Large, with glandular walls Slightly longer Some anatomical features of Ferrissia (Ferrissia) rivularis and 3 forms belonging Ferrissia tenuis Hubendick (1958, 1964) India 17-1-17, central with 2 major and 2 minor cusps Not fused Joined Folded (1964) Single opening into stomach (1964) 4-5 lobes (1964) Similar to F. australica (1964) No information No information No information ANATOMY OF FERRISSIA 169 by Basch (1963) (Pelseneer, 1906), the parietal ganglion is peculiar to the Euthyneura and may be regarded as having arisen, to provide an origin for pallial nerves, as a result of fusion between the infra-intestinal and abdominal ganglia. The visceral loop of a dextral basommatophoran contains 3 ganglia, which are from left to right, named withregard to homology with other Gastropoda: parietal, abdominal,2 and supra-intestinal. This terminology has been simplified by later authors (e.g., Fretter & Graham, 1962; Morton, 1964; Lever etal., 1965) who refer tothe lateral ganglia as left and right parietals. In the Ancylidae and other higher Euthyneura the cerebral, pleural, and parietal ganglia on each side become closely associated and fusions may take place. In the unfused condition, the cerebral and pleural ganglia of one side are connected to each other, and each is connected separately to the pedal ganglion of the same side; the parietal 2The name abdominal is preferable to visceral because this is a composite ganglion inner- vating the body wall as well as the viscera. TABLE 1. (continued) Ferrissia vivularis| Ferrissia austva- | Fervissia burnupi | Ferrissia tenuis НВ Hoff (1940), Иса Hubendick Hubendick (1958, Hubendick (1964) (1960 b, 1964) 1964) North America Australia South Africa India Copulatory Flagellum at- Long cylindrical Long cylindrical Moderately long organ tached to proximal] flagellum opening | flagellum opening | cylindrical fla- end of preputium into vestigial into vestigial gellum opening (Hoff); short penis sheath, at penis sheath, at into vestigial pear-shaped base of short base of short penis sheath, at flagellum open- penis penis base of short ing into long in- penis vaginated penis sheath, moder- ately long penis (Hubendick) Aphallic Reported for Many individuals F. (F.) fragilis ganglion, being situated in the visceral loop, has no direct connection with the pedal ganglion. The pleural ganglion may be fused, partly or completely, with either the cerebral or the parietal ganglia, and the connectives to the pedal ganglion provide evidence of which process has occurred. The presence of 2 separate connectives to the pedal ganglion indicates that the pleural has fused with the parietal ganglion, since if the pleural had fused withthe cerebral ganglion it is likely that their combined connectives would be indistinguishable. To designate the composite pleuro- parietal ganglia as pleural ganglia (Hubendick, 1964, in the case of the American Ferrissia tarda) is un- satisfactory, because the pleural ganglia are regarded as bearing peripheral nerves only in the more primitive forms of Euthyneura. The condition of the post-cerebral ganglia in Ferrissia burnupi is similar to that described by Pelseneer (1901) for Gundlachia sp. from New Zealand, which, according to Hubendick (1964) belongs to F. (Pettancylus). Pelseneer recognised 3 ganglia inthe visceralloop, 170 D. 5. BROWN namely, a left pleural + parietal, an abdominal, and a supra-intestinal + right pleural. The lateral ganglia thus appear to be homologous with those termed pleuro-parietal in F. burnupi. Hoff (1940) used the terms left anterior visceral, left posterior visceral, and right visceral for the 3 post-cerebral ganglia of Ferrissia tarda. His de- scription and figure showing 2 con- nectives with the pedal ganglion suggest that his left anterior visceral and right visceral ganglia represent fused pleural and parietal ganglia. Two parietal ganglia and one abdominal ganglion were illustrated by Lever (1957) in Ferrissia sp. of North American origin (identified as F. shimekii (Pilsbry) by Lever, un- published), but no information was given about the pedal connectives. That form shows marked differences to F. burnupi in the position of attachment of the cerebro-pedal connectives to the cere- bral ganglia, in the peripheral nerves arising from the cerebral ganglia, and in the presence on these ganglia of large lateral lobes. Three post-cerebral ganglia similar in size and arrangement to those of Ferrissia burnupi were illustrated by Mirolli (1960) for Watsonula wautieri from Italy. These ganglia were named left and right parietal and visceral by Wautier et al. (1961) in French material, later identified (Wautier, 1964) as Gund- lachia wautieri; no information- about pedal connectives was given. Although the nature of the lateral ganglia in the visceral loop can be interpreted with confidence in no more than a few species of Ferrissia, it appears likely, from the general Similarity of these ganglia in all the forms that have been studied, that partial fusion of the pleural and parietal ganglia is commoninthis genus. Pleuro-parietal ganglia lying on either side of an abdominal ganglion have been observed in Ancylus tapirulus by Hubendick (1960a), in Burnupia caffra by Brown (unpublished observations), and in the Acroloxidae by Hubendick (1962). An- cylus tapirulus resembles Ferrissia burnupi in the presence of a small left pleural ganglion, which is distinct from the pedal ganglion but is closely associ- ated with the cerebral and parietal ganglia. While the formation of pleuro- parietal ganglia appears to occur fre- quently in freshwater limpets, the single connective to the right pedal ganglion of Laevapex fuscus illustrated by Basch (1959b) suggests that fusion of the cerebral and pleural ganglia can occur. The anatomical characteristics of representatives of the 2 subgenera of Ferrissia may be compared in Table 1. Information for F. rivularis, the type species of the genus, is derived from Hubendick (1964) and from the account by Hoff (1940) for F. tarda, which is Synonymous with F. rivularis according to Basch (1963). Anatomical differences described between the various North American species of Ferrissia s.s. are only slight and lie mainly in the size of the flageilum and the pseudobranch (Basch, 1963). Included in Table 1 are F. burnupi, F. tenuis, and F. australica (Tate), which, according to Hubendick (1964), is closely related to the type species of Pettancylus. Ferrissia burnupi resembles РЕ. australica, but differs from F. rivularis, in respect of its radular formula (small number of teeth), jaw (Separate scales), large uterine gland complex, and copu- latory organ (long flagellum attached to vestigial penis sheath containing short penis). These resemblances particularly in the copulatory organ, are reasons for classifying F. burnupi in the subgenus Pettancylus as defined by Hubendick (1964). In addition, F. burnupi is poorly pigmented, which is the normal condition in that group according to Hubendick. The relations between Ferrissia burnupi and Е. tenuis are not easily determined in the absence of any analysis of variation in taxonomic characters at the specific level within the subgenus Pettancylus. The cusps on the radular teeth of F. burnupi are less pointed, and there are fewer teeth in a transverse ANATOMY OF FERRISSIA 171 row than in F. tenuis. No distinct seminal vesicle was observed in F. burnupi; the flagellum attached to the copulatory organ is somewhat larger inthat species. At present, these differences may be regarded as of specific rank, although it is possible that differences in organs associated with the reproductive system are related to different phases of repro- ductive activity. In view of the close relationship between the South African F. burnupi and the Indian F. tenuis, and the presence in Natal of a human Indian population that might be susceptible to the Indian strain of Schistosoma haema- tobium if it were introduced, F. burnupi is worthy of consideration as a potential intermediate host. Ferrissia burnupi should be compared with Watsonula wautieri Mirolli of Italy, since Hubendick (1964) regarded the latter as a member of Pettancylus and suggested that it had been introduced into Europe in recent time. Africa is an obvious source of such an intro- duction. According to the information given by Mirolli (1960) and Hubendick (1964) for aphallic, non-septate speci- mens of W. wautieri, it appears that F. burnupi has fewer ovotestis lobes and a Shorter spermathecal duct, while in other respects the anatomies of the 2 species are similar. However, from the de- scription of the copulatory organ by Wautier et al. (1966) for 3 euphallic specimens from France, identified as Gundlachia wautieri, that species re- sembles Ferrissia s.s. rather than F. (Pettancylus) in respect of the well de- veloped penis sheathand moderately long penis (0.2 - 0.3 mm). It can only be concluded that the relations between the Italian and French limpets, and their affinities with other ancylids deserve further study. ACKNOWLEDGEMENTS My thanks are due to Dr. R. Elsdon- Dew for accommodation in the Insti- tute for Parasitology, Durban, South Africa, Mr. C. H. J. Schutte for pre- pex fuscus, paring serial sections, Dr. G. Mandahl- Barth and Dr. B. Hubendick for cri- ticising the manuscript of this paper in the course of preparation, Mrs. J. Brown for typing the manuscript, and to Mrs. A.Gismann for her careful editing. LITERATURE CITED BASCH, P. F., 1959a, Status of the genus Gundlachia (Pulmonata: Ancyl- idae). Occ. Pap., Mus. Zool., Univ. Michigan, 602: 1-9. 1959b, The anatomy of Laeva- a freshwater limpet (Gastropoda, Pulmonata). Misc. Publ., Mus. Zool., Univ. Michigan, 108: 5-56. 1963, A review of the recent freshwater limpet snails of North America (Mollusca: Pulmonata). Bull. Mus. comp. Zool., Harvard Univ., 129: 399-461. BOER, H. H. & LEVER, J., 1959, On the anatomy of the circulatory system in Ferrissia shimekii (Ancylidae, Pul- monata); especially onthe blood supply of the central nervous system. Koninkl. Nederl. Akad. Wetensch., Amsterdam, C 62: 76-83. BONDESEN, P., 1950, A comparative morphological-biological analysis of the egg capsules of freshwater pul- monate gastropods. Natura Jutlandica, Aarhus, 3: 1-208. BROWN, D.S., 1965, Freshwater gastro- pod Mollusca from Ethiopia. Bull. Brit. Mus. (nat. Hist.) Zool., 12 (2): 39-94. CONNOLLY, M., 1925, The non-marine Mollusca of Portuguese East Africa. Trans. roy. Soc. South Africa, 12: 105- 220. 1939, A monographic survey of the South African non-marine Mol- lusca. Апп. 5. African Mus., 33: 1-660. FRETTER, V. & GRAHAM, A., 1962, British Prosobranch Molluscs. Ray Society, London, xvi + 755p. GADGIL, В. K., 1963, Human schisto- somiasis in India. Indian J.med.Res., 51: 244-251. 172 D. S. BROWN GADGIL, R. K. & SHAH, S. N., 1955, Human schistosomiasis in India. Part 2. Infection of snails with S. haema- tobium. Ibid., 43: 695-701. HARRY, H. W., 1964, The anatomy of Chilina fluctuosa Gray reexamined, with prolegomena on the phylogeny of the higher limnic Basommatophora. Malacologia, 1 (3): 355-385. HOFF, C. C., 1940, Anatomy of the ancylid snail Ferrissia tarda (Say). Trans. Amer. microsc. Soc., 59: 224- 242. HUBENDICK, B., 1958, On the family Ancylidae with special reference to Ferrissia tenuis (Bourguignat), the suspected intermediate host of Schistosoma haematobium in India. Proc. VI Intern. Cong. Trop. Med. and Malaria, 2: 17-21. 1960a, The Ancylidae of Lake Ochrid and their bearing on intra- lacustrine speciation. Proc. zool. Soc. Lond., 133(4): 497-529. 1960b, A note on “Pettancylus” australicus (Tate). J. malac. Soc. Australia, 4: 32-38. 1962, Studies on Acroloxus (Moll. Basomm.). Medd. Göteborgs Mus. Zool. Avd., 133: 1-68. 1964, Studies on Ancylidae. The subgroups. K. Vet. O. Vitterh. Samh. Handl. F. 6. Ser. B., 9(6): 1-72. LEVER, J., 1957, Some remarks on neurosecretory phenomena in Ferris- sia sp. (Gastropoda Pulmonata). Proc. K. Nederl. Akad. Wiss. Amsterdam, Ser. C, 60(4): 510-552. LEVER, J, DE VRIES, C. M. & JAGER, J. C., 1965, On the anatomy of the central nervous system and the lo- cation of neurosecretory cells in Aus- tralorbis glabratus. Malacologia, 2(2): 219-230. MIROLLI, M., 1960, Morfologia, bi- ologia, e posizione sistematica di Watsonula wautieri, n.g., n.s. (Basom- matophora, Ancylidae). Mem. Inst. Ital. Idrobiol., 12: 121-162. MORTON, J. E., 1964, Molluscs. 3rd ed. Hutchinson, London, 232 p. PELSENEER, P., 1901, Etudes sur les Gastéropodes Pulmonés. Mem. Acad. roy. Belg. Cl.’ Sci, 11-4054: Ё 1906, Mollusca. A Treatise on Zoology; Part V/AEdMENMERATY Lankester. Black, London, 355 p. PILSBRY, H. A. € BEQUAERT, J., 1927, The aquatic mollusks of the Belgian Congo with a geographical and eco- logical account of Congo malacology. Bull. Amer. Mus. nat. Hist., 53: 69- 602. PORTER, A., 1938, The larval Tre- matoda found in certain South African Mollusca with special reference to Schistosomiasis (Bilharziasis). Publ. S. Afr. Inst. med. Res., 52: 1-492. WALKER, B., 1903, Notes on eastern American Ancyli. Nautilus, 17(2): 13- 18713) 29=31. 1912, A revision of the Ancyli of South Africa. Nautilus, 25: 139. 1923, (published 1924), The Ancylidae of South Africa. Printed for the author, 82p. 1926, Notes on South African Ancylidae. I. Occ. Pap., Mus. Zool., Univ. Michigan, 175, 6 p. WAUTIER, J., 1964, Watsonula ou Gundlachia? Bull. mens. Soc. Linn. Lyon, 33: 201-203. WAUTIER, J., PAVANS DE СЕССАТТУ, M., RICHARDOT, M., BUISSON, B. & HERNANDEZ, M. L., 1961, Note sur les complexes neuro-endocriniens de Gundlachia sp. (Mollusque Ancyl- idae). Ibid., 30(4): 79-87. WAUTIER, J., HERNANDEZ, M.-L. € RICHARDOT, M., 1966, Anatomie, histologie et cycle vital de Gundlachia wautieri (Mirolli) (Mollusque Basom- matophore). Ann. Sc. nat. Zool., 12th series, 8 (4): 495-566. WRIGHT, C. A., 1963, The freshwater gastropod Mollusca of Angola. Bull. Brit. Mus. (nat. Hist.) Zool., 10: 447- 528. ANATOMY OF FERRISSIA RESUMEN ANATOMIA Y RELACIONES DE UNA FERRISSIA SUDAFRICANA (BASOMMATOPHORA: ANCYLIDAE) D. 5. Brown El género Ferrissia, introducido por Walker (1903) para acomodar Ancylus rivularis Say de Norte América, ha sido subsecuentemente reconocido como un género de dis- tribución mundial, caracterizado por la fina escultura radial del ápice. Por evidencia anatómica, las especies del Nuevo y Viejo Mundo fueron clasificadas respectivamente en los subgéneros Ferrissia s.s. y F. (Pettancylus) por Hubendick (1964). El género Gundlachia Pfeiffer, originalmente caracterizado por el septum de la conchilla, fue restricto por Hubendick para ancylidos de Central y Sud América; especies africanas previamente asignadas a este género son, aparentemente, formas septadas de Ferrissia. La distribución de Ferrissia en Africa se extiende desde el Mediterráneo al Cabo de Buena Esperanza. La información publicada sobre la anatomía de las especies africanas, tratan sólo la rádula, mandíbula y caracteres externos. El presente trabajo trata ciertos aspectos de la anatomía interna y externa de los individuos no septados de F. burnupi (Walker, 1912), colectados en la Provincia de Natal, República de Sud Africa. Un objeto del estudio era obtener información para comparar esta con otras especies de Ferrissia del Viejo Mundo, en particular F. tenuis (Bourguignat), que ha sido indicada como trasmisora de esquistosomiasis humana en India. Aspectos de Ferrissia burnupi que indican estrecha relación con el grupo Pettan- cylus, son: la escasa pigmentación, la falta de fusión entre las placas que componen la mandíbula, el largo complejo glandular uterino y, sobre todo, la estructura del órgano copulador masculino con su pene pequeño y vestigios de una vaina penial en la cual se abre un largo flagelo. Aunque las diferencias entre F. burnupi y F. tenuis en la rádula, vesícula seminal y flagelo pueden ser de rango especifico, F. burnupi merece consideración como un posible huespedo intermediario de Schistosoma haemato- bium en Natal. Anatomicamente Ferrissia burnupi se asemeja a las formas italianas no septadas de Watsonula wautieri Mirolli, de acuerdo a Mirolli (1960) y Hubendick (1964), pero el órgano copulatorio difiere de aquel descripto por Wautier y otros (1966) para ejemplares de Francia identificados como Gundlachia wautieri. El ganglio pleural izquierdo de F. burnupi no se puede distinguir del parietal izquierdo, aunque está intimamente conectado con éste. La fusión entre los ganglios correspondientes del lado derecho es más avanzada. Los ganglios laterales en la torsión visceral son llamados pleuro-parietales izquierdo y derecho y el asimetri- camente colocado entre ellos, ganglio abdominal, una terminología que expresa la composición y función de esos tres ganglios en Ferrissia más seguramente que en previa terminología, y puede aplicarse a otros ancylidos. Hay connectivos separados entre los ganglios pedal, cerebral y pleuro-parietal en cada lado. ABCTPAKT АНАТОМИЯ И РОДСТВЕННЫЕ ВЗАИМООТНОШЕНИЯ ЮЖНО- АФРИКАНСКИХ FERRISSIA (ВАЗОММАТОРНОВА: ANCYLIDAE) Д. С. БРОУН Род Ferrissia, установленный Уолкером (Walker, 1903) для Ancylus rivularis Say из Северной Америки, рассматривался как всесветно- распространенный род пресноводных улиток, макушка раковины у которых обладает тонкой радиальной скульптурой. По своему анатомическому строению виды Ferrissia из Старого и Нового 173 174 D. 5. BROWN Света оыли отнесены Хубендиком (Hubendick, 1964) к подродам Ferrissia 5.3. и РЕ. (Pettancylus), соответственно. Род Gundlachia Pfeiffer, первоначально характеризовавшийся раковиной с септами, был съужен Хубендиком до объема анциллид Центральной и Южной Америки; африканские виды, ранее относимые к этому роду, являются, видимо видами рода Ferrissia, обладающими септами. В Африке Ferrissia распространены от Средиземного моря до южной ее оконечности. Имеющиеся опубликованные данные по анатомии африканских видов сводятся к строению радулы, челюстей и к наружным признакам. Настоящая работа касается внутренней и внешней анатомии без-септовых форм Ferrissia burnupi (Walker, 1912), собранных в провинции Наталь, Южно- Африканская Республика. Целью настоящего исследования было получение данных для сравнения этих Форм с другими видами Ferrissia Старого Света, особенно с РЕ. tenuis (Bourguinat), который стал известен, как промежуточный хозяин возбудителя человеческого шистозомиазиса, В Индии. Признаками, указывающими Ha близкае родство Ferrissia burnupi с видами группы Pettancylus, являются следующие: слабая пигментация, отсутствие слияния между пластинками челюстей, крупный маточный железистый комплекс и, кроме того, етруктура мужского копулятивного органа с его маленьким пенисом и рудиментарной его оболочкой, куда входит длинный Ффлагеллум. Хотя различия между F. burnupi и Е. tenuis в строении зубцов радулы семенного пузырька и Флагеллума могут иметь видовое значение, Е. burnupi заслуживает внимания, как возможный промежуточный хозяин Schistosoma haematobium в Натале. Анатомически Ferrissia burnupi похожа Ha бессептовую итальянскую Форму Watsonula wautieri Mirolli как это следует из работ Миролли (Mirolli, 1960) и Хубендика (Hubendick,1964), Ho её копулятивный орган отличается от описанного Вотье и др. (Wautier et al, 1966) для Французских экземпляров, определенных как Gundlachia wautieri. Левый плевральный ганглий y Ferrissia burnupi отличим от левого париетального ганглия, хотя и тесно C ним связан. Слияние между соответствующими ганглиями правой стороны происходит даже еще дальше. Боковое ганглии в висцеральной петле названы левым и правым плевро-париетальными ганглиями, а ассимметрически-расположенный ганглий, лежащцй между ними - абдоминальным ганглием; эта терминология более точно выражает состав и Функцию этих трех ганглиев y Ferrissia, чем предыдущие названия и может быть применена и к некоторым другим анциллидам. Имеются отдельные коннективы между ножным ганглием, церебральным и плевро-париетальными ганглиями с каждой стороны. MALACOLOGIA, 1967, 6(1-2): 175-188 CHROMOSOME NUMBERS IN RELATION TO OTHER MORPHOLOGICAL CHARACTERS OF SOME SOUTHERN AFRICAN BULINUS (ВАЗОММАТОРНОВА: PLANORBIDAE)! D. $. Brown2, С. H. J. Schutte’, J. В. Burch4 and В. Natarajan® ABSTRACT Chromosomes were counted in preparations of ovotestis tissue from Bulinus (Bulinus) collected at 87 localities in southern Africa. A basic haploid set of 18 chromosomes was present in all samples. In 3 samples of the group of B. natalensis, 1-3 extra chromosomes were observed, with different numbers of chromosomes occurring in different meiotic cells of the same individuals in one sample. Two samples of laboratory specimens of B. truncatus from Egypt and Iran were, as expected, tetraploid: n=36. Previous work has suggested that a haploid chromosome number of n=18 is characteristic of the tropicus species group and n=36 (or higher multiples) is characteristic of the truncatus species group. Although all the southern African specimens studied possessed a basic chromosome number of 18, some popu- lations had smoothly curved mesocones of the 1st lateral teeth of the radula associated with the tropicus group, while others had the angular-sided meso- cones of the truncatus group and in some cases included aphallic specimens, also associated with that group. Thus, chromosome numbers do not always show a correlation with these other characters in respect of the 2 established species groups. Therefore the natalensis species group is introduced to con- tain forms which are morphologically similar to the truncatus group but possess only 18 chromosomes. To this end the samples studied were divided according to the shapes of the mesocones (which showed some correlation with shell shape) into the tropicus or natalensis groups, or were placed in an “inter- mediate’ category. The importance of the separation of species groups of Bulinus lies in the fact that some of the intermediate hosts of Schistosoma spp. with terminal-spined ova are included in the truncatus group, and may possibly exist in the nata- lensis group, while members of the tropicus group apparently do not transmit human bilharziasis. lContribution No. 18, Intermediate Hosts of Schistosomiasis Program, Institute of Malacology, Ann Arbor, Michigan, U. S. A. 2British Medical Research Council, 20, Park Crescent, London W. 1. Present address: Insti- tute for Zoological Research, University of Potchefstroom, South Africa. 3Bilharzia Research Unit, South African Council for Scientific and Industrial Research, Nels- pruit, Transvaal, Republic of South Africa. Contribution published by permission of the South African Council for Scientific and Industrial Research. Museum and Department of Zoology, University of Michigan, Ann Arbor, U. S. A. Supported by a research grant (АТ 07279) and Research Career Program Award (5-K3-AI-19, 451) from the National Institute of Allergy and Infectious Diseases, U. S. Public Health Service. 5Museum of Zoology, University of Michigan, Ann Arbor, U. 5. A. Supported by a research grant (GB-787) from the National Science Foundation, Washington D. C., U. S. A. (175) 176 BROWN, SCHUTTE, BURCH AND NATARAJAN The tropicus group was found to occur over the greater part of South Africa including western Cape Province, whereas the natalensis group is apparently confined to the warmer moderate and lower altitudes of the northern and east- ern parts of the area. INTRODUCTION This paper deals with southern African material referable to the tropicus and truncatus species groups (Mandahl- Barth, 1957) of the planorbid genus Bulinus, which constitute the subgenus Bulinus s.s. according to the usage of Walter (1962), Burch (1964) and Natarajan, Burch & Gismann (1965). Few of the taxa that have been established within these species groups can be un- equivocally defined, and the limits ofthe groups themselves are not entirely clear. Since some members of the truncatus group serve as intermediate hosts of schistosomes with terminal-spined ova and members of the tropicus group apparently do not transmit human bil- harziasis, an improvement in our know- ledge and classification of the bulinine snails may contribute to a better under- standing of their parasites. Species of the tropicus group occur in Africa south of the Sahara and are un- known in countries bordering the Mediterranean Sea. The northwestern limit of the range of this group appears to lie in West Cameroon (Wright, 1965), and it should be noted that a considerable part of the range formerly attributed to it in northwest Africa (Mandahl-Barth, 1957) was based upon records of Bulinus guernei (Dautzenberg), which is now re- garded as belonging to the truncatus group by Wright (1965) and Mandahl- Barth (1965). In eastern Africa the tropicus group extends from Kenya to the coast of Cape Province (Mandahl-Barth, 1957; Azevedo et al., 1961; van Eeden, Brown € Oberholzer, 1965). The trun- catus group, as formerly understood, occurs primarily in many Mediterranean and Middle Eastern countries, and also in central and southern Africa where the southernmost recorded localities are in South West Africa (В. natalensis; Mandahl-Barth, 1965), Transvaal (B. depressus; van Eeden, 1964; Schutte, 1966), and Natal (B. depressus; van Eeden et al., 1965). Burch (1964) and Natarajan et al. (1965) have reported on 35 samples of Bulinus from central Africa, several countries of the Mediterranean area, from Asia minor, and Western Aden Protectorate, and found that the species of the tropicus and the truncatus groups investigated possessed haploid sets of 18 and 36 chromosomes respectively. Burch (1964) commented on the combi- nation in the truncatus group of a haploid set of 36 or higher multiples of 18, with the capacity to transmit Schisto- soma haematobium, and suggested that chromosome numbers might help in identifying potential intermediate hosts. The present paper describes the results of an investigation of chromo- some numbers in relation to several taxonomic characters previously em- ployed, i.e., the shell, radula (shape of the mesocone of the first lateral tooth) and copulatory organ (commonly absent in the truncatus group). The purpose of our work was to investigate the distri- bution in South Africa of Bulinus popu- lations apparently belonging to the trun- catus group, and to determine whether or not a correlation existed between chromosome numbers and the morpho- logical characters mentioned above. Identification of our material is based upon the radula, which has been given importance in definitions of the truncatus and tropicus groups (Mandahl-Barth, 1957). MATERIAL AND METHODS Specimens intended for cytological examination, collected between June 25, SOUTHERN AFRICAN BULINUS 177 Bechuanaland South Africa Cd Rhodesia iy 3 5 Y © с A \ AAA AH Grahamstown®, xe 200k. B. tropicus e B. natalensis A ‘intermediate X FIG. 1. Localities in southern Africa from which samples of Bulinus were examined and as- signed to the species groups of В. tropicus, В. natalensis or to an ‘intermediate’ category, on radular characteristics. 1964 and March 28, 1965, were killed and preserved in Newcomer’s (1953) fluid after the apices of the shells had been cracked to allow rapid penetration of the fixative. Chromosomes were ob- served in ovotestis tissue prepared by the acetic-orcein squash technique (La Cour, 1941). Observations were made with Nikon microscopes using 100X (n.a. 1.25) oil immersion objectives and 10, 20 and 30X oculars. The damaged shell and the animal of each specimen were retained and labelled so that they could be correlated with the appropriate ovotestis preparation. Each animal was dissected to determine whether the copulatory organ was present or absent, and an unstained preparation was made of each radula in the manner described by Schutte (1965); further observations on shells, radulae and copu- latory organs were made from specimens preserved in alcohol. Specimens were obtained from a total of 87 localities situated in the Republic of South Africa, Swaziland or Mozam- bique (Fig. 1 and Table 1); records containing full details of these localities have been deposited in the Experimental Taxonomy Section of the Zoology De- 178 TABLE 1. Chromosome numbers and other morphological features of various populations of South African Bulinus Chromosome Copulatory organ : number Locality* Radula type ee Ka studied | aphallic 1. Mhlangana R., Durban (N) 18-21 - intermediate 0 2. Mhlatuzane R., Durban (N) 18 - intermediate 0 3. Umsunduzi R., Valley of Thousand Hills (N) 18 - tropicus 0 4. Pietermaritzburg (N) 18 36 natalensis Y 5. Mooi River at town (N) 18 - tropicus 0 6. Newcastle (N) 18 - tropicus 0 7. Bisana (CP) 18 - tropicus 0 8. Tongaat (N) 18 - intermediate 0 9. Wewe В. , Tongaat (N) 18 36 tropicus 0 10. Shakaskraal (N) 18 36 tropicus 0 11. Ncanaweni R. , Stanger (N) 18 - tropicus 0 12. Stanger (N) 18 - intermediate 0 13. Big Bend, Swaziland 18 - intermediate 0 14. Big Bend, Swaziland 18 36 tropicus 0 15. Pongola Settlement (T) 18 - tropicus 0 16. Sterkspruit, Lydenberg (T) 18 = tropicus 0 17. Kokstad (CP) 18 - tropicus 0 18. Ixopo R., at town (N) 18 - tropicus 0 19. Kokstad (CP) 18 - tropicus 0 20. Cedarville (CP) 18 - tropicus 0 21. Cedarville (CP) 18 - intermediate 0 22. Matatiele (CP) 18 - tropicus 0 23. Matatiele (CP) 18 - tropicus 0 24. Swartberg (CP) 18 - tropicus 0 25. Swartberg (CP) 18 - tropicus 0 26. Swartberg (CP) 18 - tropicus 0 27. 23km N. Swartberg (CP) 18 - tropicus 0 28. 40 km N. Swartberg (N) 18 - tropicus 0 29. Mtubatuba (N) 18 - natalensis 0 30. Ncemane В. , Hluhluwe (N) 18 36 intermediate 0 31. Msunduzi R. , Hluhluwe (N) 18 - intermediate 0 32. Msunduzi R. , Hluhluwe (N) 18 - intermediate 0 33. Mzinene R. , Hluhluwe (N) 19 - intermediate 0 34, 35. Ladysmith (N) 18 - tropicus 0 36. Bethlehem (OFS) 18 - tropicus 0 37. 46 km W. of Bethlehem (OFS) 18 - tropicus 0 38. 70 km W. of Bethlehem (OFS) 18 36 tropicus 0 39. 96 km W. of Bethlehem (OFS) 18 36 tropicus 0 40. Winburg (OFS) 18 - tropicus 0 41. Christiana (T) 18 - intermediate 0 42. Pietersburg (T) 18 36 intermediate 0 43. Bandoleirkop (T) 18 - tropicus 0 44. Thabina R. , Tzaneen (T) 18 - natalensis 8 45. Nylstroom (T) 18 - natalensis 6 46. Palmeira, Mozambique 18 = intermediate 0 BROWN, SCHUTTE, BURCH AND NATARAJAN *Except for localities 60 and 61, which are from the Middle East, provinces of South Africa are abbreviated as follows: Natal (N), Cape Province (CP), Orange Free State (OFS), Transvaal (T). SOUTHERN AFRICAN BULINUS 179 Table 1 (continued) Chromosome Copulatory organ b Locality* В Radula type Nos. 18 - 47. Winterton (N) tropicus 48. Estcourt (N) 18 36 tropicus 49. Nottingham Road (N) 18 intermediate 50. Roodepoort, Warmbaths (T) 18 natalensis 05 51. Leeupoort, Nylstroom (T) 18 natalensis 0+ 52. Nylstroom (T) 18 intermediate 0 53. Nelspruit aquaria (T) 18 natalensis 0 54. Buffelspruit, Malelane (T) 18 natalensis 0 55. Thankerton Creek, Hector- spruit (T) 18 intermediate 0 56. Ngwetspruit, Komatipoort (T) 18 natalensis 0 57. Border Gate, Swaziland 18 intermediate 0 58. Border Gate, Swaziland 18 intermediate 0 59. Simondsdal, Lake Chrissie (T) 18 tropicus 0 60. Abis, Egypt 36 truncatus 0 61. Iran 36 truncatus 4 62. Merebank, Durban (N) natalensis 2 63. Durban (N) 18 intermediate 0 64. Rietvlei (N) 18 intermediate 0 65. Ladysmith (N) 18 tropicus 0 66. Van Reenen (OFS) 18 intermediate 0 67. Van Reenen (OFS) 18 tropicus 0 68. Harrismith (OFS) 18 intermediate 0 69. Harrismith (OFS) 18 tropicus 0 70. 113 km NW of Harrismith (OFS) 18 tropicus 0 71. Villiers (T) 18 tropicus 0 72. Villiers (T) 18 intermediate 0 73. Villiers (T) tropicus 0 74. Tzaneen (T) natalensis 5 75. Idutywa (CP) tropicus 0 76. Heidelberg (CP) tropicus 0 77. Knysna (CP) tropicus 0 78. Grahamstown (CP) tropicus 0 79. Port Alfred (CP) tropicus 0 80. Alexandria (CP) tropicus 0 81. Grahamstown (CP) intermediate 0 82. Breakfastvlei (CP) intermediate 0 83. Peddie (CP) tropicus 0 84. 46 km NW Piet Retief (T) tropicus 0 85. Lothair (T) tropicus 0 86. Ermelo (T) tropicus 0 87. Ermelo (T) tropicus 0 88. Amersfoort (T) tropicus 0 0 89. Sani Pass (N) tropicus +Aphallic specimens were present in the original samples from which these specimens were descended. 180 BROWN, SCHUTTE, BURCH AND NATARAJAN partment, British Museum (Natural History), and in the Institute for Zoologi- cal Research, University of Potchef- stroom, South Africa. Most of the samples were fixed in the field, but laboratory-bred descendants of snails collected at localities 50, 51, 52, 53 and 59 were used. Observations were also made on laboratory stocks of Bulinus truncatus truncatus originally obtained in Egypt and Iran (60 and 61). OBSERVATIONS Chromosomes (Table 1) A constant number of 18 chromosomes was present in meiotic (haploid) cells of all but 3 of the 87 samples from southern Africa. Between 1 and 16 individuals were examined from each sample. Eighteen pairs of chromosomes were observed whenever it was possible to make counts in diploid cells. In the 3 exceptional samples, although the basic number was also 18, between 1 and 3 extra chromosomes were ob- served in meiotic cells. АП 3 specimens of sample 33 had 19 chromosomes. Six- teen specimens of sample 1 were exam- ined having 18, 19, 20 and 21 chromo- somes inindividual specimens in the pro- portions 5:5:5:1. Different numbers of chromosomes were present in different meiotic cells from each of the 2 speci- mens examined from locality 62; 18, 19 and 20 in one individual, and 18 and 19 in the other. Such supernumerary chromo- somes have already been reported for B. natalensis from Southern Rhodesia (Burch, 1964). Two samples of Bulinus truncatus truncatus from Egypt and Iran (locali- ties 60 and 61) possessed 36 chromo- somes in meiotic cells. This polyploid number has been reported before for В. t. truncatus from these 2 countries by Burch (1964). Radula (Figs. 2 and 3) Each radula was classified according to the most frequent shape of the meso- cones (middle cusps) of the first lateral Fes SON pus 204 FIG. 2. First lateral teeth of radulae of: Bulinus of the tropicus group. 1, locality 88; 2, locality 67; 3, locality 25 (the meso- cones are typical of the tropicus type). Bulinus of the “intermediate” group. 4-6, locality 68; 7-11, locality 1 (the mesocones of 4, 8 and 10 are of the intermediate type; 5 and 7 are of the natalensis type; 6, 9 and 11 are of the tropicus type). teeth into 1 of the 3 following categories: 1. tropicus group: Sides of mesocone smoothly curved (Fig. 2: 1-3, 6; Fig. 3: 4, 19), or slightly angu- lar (Fig. 2: 9, 11). These shapes most closely resemble the “tri- angular” mesocone described by Mandahl-Barth (1957) for the tropicus group. 2. natalensis group: Both sides of mesocone angular, with the sides usually converging to- wards the base (Fig. 2: 5, 7; Fig. 3: 1-3, 5-14, 16-18). This shape resembles the “arrow- head” mesocone described by Mandahl-Barth (1957) for the truncatus group. 3. Intermediate group: One side of the mesocone angular (Fig. 2: 4, 10; Fig. 3: 15). Other meso- cones which, because of the fluting of their edges, could not be classified in either of the 2 preceding groups (Fig. 2: 8; SOUTHERN AFRICAN BULINUS 181 er ey ES aaa rada NS Ba Bs FIG. 3. First lateral teeth of radulae of: Bulinus of the natalensis group. 1, locality 29; 2, locality 4; 3, locality 44; 4, 5, lo- cality 45; 6, locality 53; 7, 8, locality 50; 9, 10, locality 51; 11, locality 54; 12, lo- cality 56; 16, 17, locality 75; 18, 19, lo- cality 62 (most mesocones are typical; 4 and 19 are of the tropicus type). Bulinus truncatus truncatus. 13, locality 60; 14, 15, locality 61 (mesocone 15 is of the intermediate type). see also Schutte, 1965), are placed here. Whenever possible 5 radulae were prepared from each sample of snails. The mesocones on the first lateral teeth in the unworn rows were examined. Individual radulae usually had meso- cones of all the 3 types described above and were classified in the group of whichever type of mesocone was preva- lent. Each sample was then classified according to its most frequent type of radula. Data on radulae of other specimens from localities 16, 50, 51, 53, 55, 59 and 60 of the present series were given by Schutte (1965), who presented a de- tailed analysis of the frequencies of various shapes of mesocone. No de- tailed analysis was carried out in the present investigation, but in material from the same localities the most fre- quent Shapes of mesocone were found, with one exception, to correspond well with those reported by Schutte. Material from the exceptional locality (55: Thankerton Creek, Hectorspruit) was classified by Schutte in the tropicus group but is included in our ‘inter- mediate’ category. This difference may be due to the fact that whereas Schutte determined the prevalent type of meso- cone in each sample of radulae con- sidered as a whole, our identifications are based on separately classified radulae. More than 50% of the samples classi- fied in the natalensis group and approxi- mately 25% of those placed in the tropi- cus group included radulae of the inter- mediate category, and about 80% of the samples placed inthe intermediate group included radulae of either or both the tropicus and natalensis types. However, no natalensis sample possessed any tropicus radulae although mesocones of this type were present in some radulae. Similarly, no tropicus sample contained any natalensis radulae although meso- cones of this type were present in some radulae. Shell (Figs. 4 and 5) The tropicus and natalensis groups of Samples possess in general distinct forms of shell, but an attempt to express accurately the degree of correlation between mesocone shape and shell form cannot be made until the variation of these characters has been studied further. Characteristics commonly pre- sent in tropicus group shells are: the long spire which is conical and sharply pointed, the evenly rounded whorls, and the acute upper angle of the aperture. These features are illustrated in shells 182 BROWN, SCHUTTE, BURCH AND NATARAJAN Y 9 Y SO 0.508 PHASE: AA FIG. 4. Shells of Bulinus classified according to the prevalent shape of mesocone on the first lateral teeth of the radula. tropicus group: e, f, locality 20; g, h, locality 22; i, j, locality 34; k, 1, locality 90 (these shells are of the form commonly found in the tropicus group of samples, and also represent the range of variation in their particular samples). “intermediate” group: a-d, locality 1 (these shells illustrate the wide variation present in some samples possessing radulae of the “intermediate” type). SOUTHERN AFRICAN BULINUS 183 Ss co 2 2 © ones FIG. 5. Shells of Bulinus classified according to the prevalent shape of mesocone on the first lateral teeth of the radula. natalensis group: a, m, locality 75; b,c, locality 62; d, e, f, locality 29; в, h, locality 44; i, j, locality 45; k, locality 50; 1, locality 51; m, п, о, locality 53. representing the extremes of variation whorls of some shells. in 4 samples of the tropicus group (Fig. The majority of the shells in the 4, е-1). A microsculpture of fine spiral natalensis group samples have com- lines is present on the post-embryonic paratively short spires, and in further 184 BROWN, SCHUTTE, BURCH AND NATARAJAN contrast to the tropicus group have shouldered whorls and an obtuse upper angle of the aperture (Fig. 5). A microsculpture of fine spiral lines, better developed than in the tropicus group, is frequently present on the post- embryonic whorls. Some of the samples classified as intermediate on the basis of their radulae possessed shells of the tropicus type, while others had shells of the natalensis type. Still other samples contained a wide variety of shells, including forms resembling both the tropicus and nata- lensis types (Fig. 4, a-d). Copulatory organ (Table 1) Aphallic individuals were present in 5 out of the 12 samples of the natalensis group (localities 4, 44, 45, 62 and 74) and in 2 further cases had been present in the original stock from which the euphallic specimens here examined were descended (localities 50 and 51). Aphallic specimens were also present in the samples of В. truncatus from Iran (locality 61). Some specimens of the tropicus group that were heavily infected with larval trematodes possessed a copu- latory organ much reduced in size, but no aphallic specimens were found. IDENTIFICATION OF MATERIAL Many shells in samples classified in the group in which non-angular meso- cones were prevalent have a pointed, relatively high spire and a smoothly curved columella, in these respects resembling Physa tropica Krauss 1848. It is possible that distinct but related forms are present in South Africa (see Mandahl-Barth, 1957, for synonymy of В. tropicus) and were represented in our samples and therefore our material is referred to the tropicus group without specific identification. Some of the samples classified in our group in which angular-sided mesocones were prevalent have shells with a short spire and a straight or twisted columella, thus resembling Physa natalensis Küster 1841; several of these samples were collected near the type locality of that species in the valley of the Umgeni River in Natal. Other samples in this group have shells with considerably shorter spires, resembling in this respect Bulinus hemprichii depressus Haas 1936, which was included in the synonymy of Bulinus natalensis as a member of the truncatus group by Mandahl-Barth (1965). In the prevalance of angular- sided mesocones and the presence in some populations of aphallic individuals our samples resemble the forms included by Mandahl-Barth (1957) inthe truncatus group; cytologically, however, theSouth African material, with 18 pairs of chromosomes, is diploid, in contrast to the polyploid chromosome numbers possessed by all the northern African or Middle Eastern material of the trun- catus group that has been investigated (Burch, 1964; Natarajan et al., 1965; Burch, 1967). Schutte (1966) suggested that the South African Bulinus apparently belonging to the truncatus group ex- amined by him should be placed in a separate group and it seems appropriate to refer our samples to the species group of Bulinus natalensis. DISCUSSION Snails from the northern Transvaal identified as Bulinus depressus Haas by van Eeden (1964) have characters of the ‘truncatus’ group according to van Eeden (unpublished observations) and Schutte (1966). В. depressus is recorded from the eastern Transvaal and northern Natal by van Eeden et al. (1965). Since shells with short spires in some of our natal- ensis group samples are similar to the form identified as B. depressus from eastern and northern Transvaal (G. Oberholzer, personal communication; Fig. 5, a-j and m), it seems probable that bulinines with some morphological characters of the ‘tvuncatus’ group, but having 18 pairs of chromosomes, are widely distributed in north easternSouth Africa at moderate and low altitudes. SOUTHERN AFRICAN BULINUS 185 A form of “В. tropicus’ collected at Grahamstown, Eastern Cape having angular mesocones onthe lateral radular teeth (Stiglingh, van Eeden & Ryke, 1962), appears to be the southernmost popu- lation known to have that characteristic of the ‘truncatus’ group. The distribution pattern of the B. natalensis group in South Africa suggests that these snails are adapted to warm climatic conditions, in contrast to the tropicus group which is more widely distributed, occurring in the subtropical region and also at relatively high altitudes inthe temperate climatic region. The existence of populations which we have classified as ‘intermediate’ could be regarded as evidence that genetic exchange between Bulinus tropicus and B. natalensis may take place. However, the nature of these populations is obscure and it is possible that a more detailed analysis of morphological characters would in some cases reveal 2 or more distinct forms. Some of these samples could be classified in either one or the other of the tropicus or natalensis groups depending on which taxonomic character was given most importance, i.e. shell and/or shape of the mesocone of the first lateral tooth vs. egg mass proteins and/or chromosome number. For example, specimens from Mhlangana River (locality 2) near Durban have been identified according to their shells and radulae as В. natalensis belonging to the truncatus group (Mandahl-Barth, personal communication), but according to biochemical evidence this population is most closely related to the B. tropi- cus group (Wright & Ross, 1965). Further, it has 18 pairs of chromosomes, which, according to Burch (1964) places it in the B. tropicus group. The apparent lack of correlation between these taxo- nomic characters suggests that further study of the “subgenus Bulinus s.s.” in southern Africa will provide a significant contribution to our understanding of the relationship existing between the charac- ters that are at present applied in the distinction of species groups. A form of Bulinus with n=18 occurring in Ethiopia (Lake Awasa) was listed by Natarajan et al. (1965) as Bulinus “? sericinus” following information supplied with the specimens by C. A. Wright, but considered to belong to the tropicus species group according to the chromosome numbers. The same form was Classified as Bulinus sp. belonging to the truncatus group by Brown (1965) according to the shapes of the radular mesocones and the existence of aphallic individuals. This Ethiopian bulinine snail conforms to our definition of the B. nata- lensis group, which therefore appears to be widely distributed in Africa. Little information has been published on the ability of Bulinus of the natalensis group to transmit species of Schisto- soma. Laboratory-bred descendants of snails (n=18) collected in Lake Awasa, Ethiopia, could not be infected with 5. haematobium from Western Aden Pro- tectorate (Brown, 1964). Attempts to infect South African ‘B. depressus’ (Table 1, 50 and 51) with strains of S. haematobium and S. mattheei from the eastern Transvaal were also un- successful (Schutte, 1966). Positive results have been reported by Pitchford (1965: 109, 114) and Schutte (loc. cit.) in the case of snails obtained from out- door aquaria at the Nelspruit laboratory, that were found to be capable of trans- mitting South African strains of S. haematobium and S. mattheei, and also Iranian strains of S. haematobium and S. bovis. Pitchford and Schutte referred to these snails as Bulinus sp. belonging to the truncatus group and believed them to be descended from specimens col- lected locally in South Africa. The several specimens of this stock from these outdoor aquaria that were examined cytologically possessed 18 pairs of chromosomes (Table 1, 53) and itthere- fore may be possible that naturally occurring populations of the Bulinus natalensis group are capable of trans- mitting schistosomes. One or more extra chromosomes have been reported in meiotic cells of South 186 BROWN, SCHUTTE, BURCH AND NATARAJAN African and Southern Rhodesian speci- mens of the B. natalensis group; these represent approximately 4% of all the populations of African Bulinus s.s. that have been examined (Burch, 1964; Natarajan et al., 1965; present paper). The nature and origin of these extra chromosomes are, at present, unknown. Although extra chromosomes occur in a few populations, all “Bulinus s.s.” in southern Africa have been found to be basically diploid, i.e., to have 18 pairs of chromosomes. Polyploidy has not been found to occur in any specimens southof Tanzania. This homogeneity contrasts with the remarkable variety of haploid chromosome numbers, i.e., 18, 36, 54 and 72, known in Ethiopian populations (Burch, 1967). ACKNOWLEDGEMENTS Accommodation provided by Professor R. Elsdon-Dew for D. S. Brown at the Institute for Parasitology, Durban, South Africa is gratefully acknowledged. We are indebted to Dr. G. Mandahl-Barth (Danish Bilharziasis Laboratory) and Mr. G. Oberholzer (Institute for Zoo- logical Research, University of Potchef- stroom) for their comments on the identification of some of our material, and to Mrs. A. Gismann for her valuable criticisms of this paper in the course of its preparation. LITERATURE CITED AZEVEDO, J. FRAGA de, MEDEIROS, L. do C. M. de, FARO, M. M., da COSTA, XAVIER, M. de LOURDES, GANDARA, A. F. & MORAIS, T. de, 1961, Freshwater mollusks of the Portuguese overseas Provinces, III. Mollusks of Mozambique. Estudos, Ensaios e Documentos Ultramar, Lis- bon, 88: 1-394. BROWN, D. S., 1964, The distribution of intermediate hosts of Schistosoma in Ethiopia. Ethiopian med. J., 2(4): 250-259. 1965, Freshwater gastropod Mollusca from Ethiopia. Bull. Brit. Mus. (Nat. Hist.), Zool., 12(2): 37-94. BURCH, J. B., 1964, Cytological studies of Planorbidae (Gastropoda: Pulmon- ata). I. The African subgenus Bulinus 3.5. Malacologia, 1(3): 387-400. 1967, Chromosomes of inter- mediate hosts of human bilharziasis. Malacologia, 5(2): 127-135. LA COUR, L., 1941, Acetic-orcein; A new stain-fixative for chromosomes. Stain Techn., 16: 169-174. MANDAHL-BARTH, G., 1957, Inter- mediate hosts of Schistosoma. African Biomphalaria and Bulinus: II. Bull. Wild Hlth Org., 17: 1-65. 1960, Intermediate hosts of Schistosoma. Some recent information ation. /bid., 22: 565-573. 1965, The species of the genus _ Bulinus, intermediate hosts of Schisto- soma. Ibid., 33(1): 33-44. NATARAJAN, R., BURCH, J. B. € GISMANN, A., 1965, Cytological studies of Planorbidae (Gastropoda: Pulmonata). I. Some African Planor- binae, Planorbininae and Bulininae. Malacologia, 2(2): 239-251. NEWCOMER, E. H., 1953, A new cyto- logical and histological fixing fluid. Science, 118(3058): 161. PITCHFORD, R. J., 1965, Differences in the egg morphology, and certain biological characteristics of some African and Middle Eastern schisto- somes, Genus Schistosoma, with terminal-spined eggs. Bull. Wld Hlth Org., 32: 105-120. SCHUTTE, C. H. J., 1965, Notes on the radular mesocone as a criterion for distinguishing between the truncatus and tropicus groups of the genus Bulinus (Mollusca, Basommatophora): Ann. Mag. nat. Hist., 8: 409-419. 1966, Observations on two South African bulinid species of the truncatus group (Gastropoda, Planor- bidae). Ann. trop. Med. Parasit., 60: 106-113. STIGLINGH, I., VAN EEDEN, J. A. & RYKE, P. A., 1962, Contributions to the morphology of Bulinus tropicus SOUTHERN AFRICAN BULINUS 187 (Gastropoda: Basommatophora: Plan- orbidae). Malacologia, 1(1): 73-114. VAN EEDEN, J. A., 1964, Die voorkoms en verspreiding van Bilharziatussen- gashere in die nordelike munisipale gebeid van Johannesburg en verder noordwarts tot by die Haartebees- portdam. Tydskr. Natuurw., 4: 52. VAN EEDEN, J. A., BROWN, D. S. € OBERHOLZER, G., 1965, The distri- bution of freshwater molluscs of medical and veterinary importance in south-eastern Africa. Ann. trop. Med. Parasit., 59(4): 413-424. WALTER, H. J., 1962, Punctation ofthe embryonic shell of Bulininae (Planor- bidae) and some other Basommato- phora and its possible taxonomic- phylogenetic implications. Mala- cologia, 1(1): 115-137. WRIGHT, C. A., 1965, The freshwater gastropod moliuscs of West Cameroon. Bull. Brit. Mus. (Nat. Hist.), Zool., 13(3): 73-98. WRIGHT, C. A. & ROSS, G. C., 1965, Electrophoretic studies of some planorbid egg proteins. Bull. Wld Hlth Org., 32(5): 709-712. RESUMEN EL NUMERO CROMOSOMATICO EN RELACION A OTROS CARACTERES MORFOLOGICOS DE ALGUNOS BULINUS DEL AFRICA MERIDIONAL (BASOMMATOPHORA: PLANORBIIDAE) D. S. Brown, C. H. J. Schutte, J. B. Burch and R. Natarajan Se contaron los cromosomas en preparaciones de tejidos del ovotestis de Bulinus s.S., colectados en 87 localidades del Africa meridional. Un juego basico de 18 cromosomas estaba presente en todas las muestras. En 3 muestras del grupo de B. natalensis se observaron de 1 a 3 cromosomas extras, con números diferentes en distintas células meióticas de la mismos individuos de una muestra. En dos muestras de ejemplares de laboratorio de B. truncatus del Egipto e Irán, el número, como se esperaba, fué tetraploideo: n=36. Trabajos previos han sugerido que el número de cromosomas haploideos n=18, es caracteristico del grupo de especies de В. tropicus y n=36 (o mayores múltiples) del grupo truncatus. Aunque todos los ejemplares estudiados del sur de Africa, poseian un numero basico de 18, algunas poblaciones tenian los mesoconos del primer diente lateral de la radula, suavemente curvos, asociados con el grupo tropicus, mientras que en otros los mesoconos tenian costados angulosos como en el grupo de truncatus, y en algunos casos inclufan ejemplares afalicos también asociados a ese grupo. De tal manera, el número cromosomático no siempre muestra una correlación con esos u otros caracteres de los dos grupos establecidos. Por consiguiente, in- troducimos aqui, un grupo de especies de B. natalensis, para incluir las formas que son morfologicamente similares al grupo de truncatus, pero que $010 poseen 18 cromosomas. Con esta finalidad las muestras estudiadas fueron divididas de acuerdo a la forma de los mesoconos (que mostraron correlación con la forma de la concha) dentro de los grupos tropicus o natalensis, o fueron colocados como una categoria intermedia. La importancia de la separación de Bulinus en grupos de especies, se destaca por el hecho de que algunos de los huespedes de Schistosoma spp., con ova terminando en espinas, estan incluídos en el grupo de truncatus, y posiblemente existe en el grupo de natalensis, mientras que los miembros del grupo de tropicus aparentemente no trasmiten bilharziasis humana. El grupo de tropicus se encontró en la mayor parte de Sud Africa, incluyendo la Provincia del Cabo, mientras que el grupo de natalensis parece estar confinado a las áreas de menor o moderada, altitud, más calidas, y más al norte y al oeste de aquella región. 188 BROWN, SCHUTTE, BURCH AND NATARAJAN ABCTPAKT ОТНОШЕНИЕ ЧИСЛА ХРОМОСОМ К ДРУГИМ МОРФОЛОГИЧЕСКИМ ПРИЗНАКАМ НЕКОТОРЫХ ЮЖНО-АФРИКАНСКИХ BULINUS (BASOMMATOPHORA: PLANORBIDAE) Д.С. BPOYH, K.X. ШЮТТЕ, ДЖ.Б. БЁРЧ И P. НАТАРАДЖАН При исследовании тканей гермафродитной железы у Bulinus (Bulinus), собранных в 87 местах Южной Африки, у них было также подсчитано число хромосом. Основной гаплоидный набор из 18 хромосом имелся у моллюсков из всех проб. В 3 пробах в группе В. natalensis были обнаружены 1-3 дополнительных хромосом, при различном числе хромосом, встреченных в различных мейотических клетках одних и тех же экземпляров животных из одной и той же пробы. Две пробы лабораторных зкзкмпляров В. truncatus из Египта и Ирана были, как и ожидалось тетраплоидными: п = 36. Из предыдущих работ следовало, что число гаплоидных хромосом равное 18, характерно для видов из группы tropicus, a n=36 (или больше множественных хромосом), свойственно видам из группы truncatus. Хотя у всех изученных южно-африканских видов основное число хромосом было 18, некоторые популяции имели гладко-изогнутые мезоконы первый боковых зубцов радулы, что присуще группе видов tropicus, в то время, как другие имели угловатые сбоку мезоконы, характерные для группы truncatus, и включали в некоторых случаях афалические экземпляры, также связанные с этой группой. Таким образом, число хромосом не всегда коррелирует с другими признаками, как это было установлено для 2, указанных выше групп видов. Поэтому группа видов natalensis представляется, как имеющая Формы, морфологически сходные с видами группы truncatus, но обладающая 18 хромосомами. Исходя из этого, изученные пробы моллюсков были разделены (в соответствии с формой мезоконов, которые имеют некоторую корреляцию с формой раковины) на Tpynnsltropicus или natalensis или были отнесены к "промежуточной" категории. Важность разделения видовых групп Bulinus заключается в TOM, что некоторые моллюски - промежуточные хозяева Schistosoma spp. (с заостренными на конце яйцами) включались в группу truncatus и, возможно могут оказаться среди Форм из группы natalensis, в то время, как моллюски из группы tropicus, видимо не являются переносчиками человеческого билхарциазиса. Было найдено, что виды группы tropicus встречаются на большей части Южной Африки, включая западную часть Капской провинции, в то время как группа natalensis, видимо связана с более теплыми, умеренными и низкими широтами северной и восточной частей ареала. MALACOLOGIA, 1967, 6(1-2): 189-198 DISTRIBUTION OF CYTOLOGICALLY DIFFERENT POPULATIONS OF THE GENUS BULINUS (BASOMMATOPHORA: PLANORBIDAE) IN ETHIOPIA D. $. Brown! and J. В. Burch2 ABSTRACT The genus Bulinus has, like the rest of the Planorbidae, a basic chromosomal complement of 18 pairs of chromosomes. However, in the B. truncatus species group there are tetraploid species with 36 pairs and some populations, ap- parently belonging to that group, have 54 or 72 pairs of chromosomes. The species groups of B. tropicus and B. natalensis have 18 pairs of chromosomes and are referred to in the present paper, together with the B. truncatus group, as the subgenus “Bulinus 5.3.” The diploid (n=18) members of the complex occur predominantly in southern Africa, and are known as far north as Lake Tana in Ethiopia and as far west as Senegal. The tetraploid (n=36) species occur predominantly in northern Africa and the Mediterranean and Middle Eastern regions, and extend southwards to Mauritania, Senegal, Ghana and Tanzania. The hexaploid (n=54) has been recorded only from Ethiopia and the octoploid (n=72)from Ethiopia and Western Aden Protectorate. Apparently only the tetraploid, and perhaps the octoploid, serve as intermediate hosts to Schistosoma haematobium under natural conditions. The present paper combines the results of recent cytological observations on 22 samples of “Bulinus s.s.” collected by the authors in Ethiopia with earlier data. From a total of 28 localities, cytologically different forms have been ob- tained with the following frequencies: n=18 (10), n=36 (8), n=54 (4) and n=72 (6). Diploid and tetraploid populations are widely distributed, while the hexaploid and octoploid forms have been found only on the highland plateau northwest of the Rift Valley. Ethiopian “Bulinus s.s.” have been classified previously in the B. truncatus species group. In different populations formerly identified as B. truncatus sericinus we have observed diploid, tetraploid and octoploid numbers. It there- fore seems likely that study of morphology in relation to chromosome number will lead to a revised classification of these snails. Both diploid and tetraploid “Bulinus s.s.” occur in lakes Awasa and Zwai; the 2 forms have distinctive shells and each is apparently restricted to a characteristic habitat. Cytologically different “Bulinus s.s.” have otherwise been recorded from the same waterbodies only in Tanzania. The fact that different forms have not been found living together in any Ethiopian locality may be due to competition between the forms, based on adaptation to particular eco- logical conditions. The tetraploid form of “Bulinus s.s.” appears to be most relevant in relation to the transmission of Schistosoma haematobium in Ethiopia, and its apparent rarity may serve, in conjunction with other epidemiological factors, as a barrier to the establishment of foci of urinary bilharziasis. lscientific Staff, British Medical Research Council, 20 Park Crescent, London W. 1. Support in Ethiopia was provided by the Haile Sellassie I University, Addis Ababa. 2Museum and Department of Zoology, University of Michigan, Ann Arbor, Michigan, U. S. A. Supported by research grant (AI 07279) and a research career program award (5-K3-AI-19, 451) from the National Institute of Allergy and Infectious Diseases, U. S. Public Health Ser- vice. (189) 190 BROWN AND BURCH It is possible that future study will lead to the detection of some cytological form, resistant to S. haematobium, and successful enough in competition with susceptible snails to serve as a means for their biological control. INTRODUCTION The planorbid genus Bulinus inhabits freshwaters in the African continent and the Mediterranean region, extending eastwards in southwestern Asia to Iran. These snails are the subject of intensive study because some forms serve as intermediate hosts to Schistosoma Species causing human vesical and bovine bilharziasis. The genus Bulinus is unusual among Basommatophora in that it includes forms having widely different chromosome numbers, which appear to constitue a polyploid series. The basic haploid chromosome number in the genus Bulinus is 18, as in other Planorbidae (Burch, 1965). However, in the B. truncatus (Audouin) group there are tetraploid species with 36 pairs and some populations, apparently belonging to that group, have 54 and 72 pairs (Burch, 1964, 1967). The B. truncatus group was defined by Mandahl-Barth (1958) with particular reference to the arrow- head shaped mesocones on the lateral radular teeth. The polyploid B. trun- catus group appears to belong to a com- plex of closely related forms that in- cludes 2 diploid (n=18, 2n=36) species groups, i.e., the В. tropicus (Krauss) group that has mesocones of triangular shape (Mandahl-Barth, 1958), and the B. natalensis (Ktister) group, which has arrowhead-shaped radular mesocones resembling those of B. truncatus (Brown, Schutte, Burch & Natarajan, 1967). The morphological distinctions between these 3 species groups are not entirely clear and it seems advisable to consider them together, from the cytological point of view. All snails of this complex from populations in Africa known or suspected to serve as intermediate hosts of Schisto- soma haematobium under natural con- ditions, that have been examined, have been found to be tetraploid (n=36; 2n=72). The species groups mentioned above may be referred to briefly as the sub- genus “Bulinus s.s.”. Although nomen- clatural objections may be raised tothis usage, it provides a means of dis- tinguishing the complex at present under consideration from the other members of the genus Bulinus, contained in the species groups of B. (=Physopsis) afri- canus (Krauss) and B. forskalii (Ehren- berg) (Mandahl-Barth, 1958). Those 2 groups are, so far as is known, diploid, but do contain species that serve as intermediate hosts to Schistosoma haematobium. Observations onthe chromosome num- bers. of. “Bulinus в.в”. ¡have been published by Burch (1964, 1967), Natara- jan, Burch & Gismann (1965) and Brown et al. (1967). A total of approximately 130 samples have been examined from Africa, the Mediterranean region and southwestern Asia, including 8 from 7 Ethiopian localities (Burch, 1964, 1967). From these observations, the diploid and various polyploid forms appear to have different distribution patterns (Fig. 1), although our knowledge of these is far from complete. However, differences between the 2 most intensively sampled countries, South Africa and Ethiopia, are striking: from South Africa only the diploid number (n=18) with occasionally 1-3 extra chromosomes has been re- ported, but in Ethiopia the diploid and all of the 3 known polyploid numbers are present. The present paper gives the results of cytological observations on 22 samples of Bulinus collected in Ethiopia by the authors, separately or together, in 1965. Observations on the ecology of snails in the Ethiopian Rift Valley lakes were made by one of us (Brown) in that year, and also in 1962, with the help of Dr. M. V. Prosser (Haile Sellassie I University). ETHIOPIAN BULINUS 191 FIG. 1. Haploid chromosome numbers of “Bulinus s.s.”, comprising the species groups of truncatus Audouin, tropicus Krauss and natalensis Küster. Data are from about 130 samples and cover Corsica, Sardinia, Iran, Iraq, Egypt, Sudan, Mauritania, Senegal, Gambia, Ghana, Ethiopia, Western Aden Protectorate, Kenya, Tanzania, Zambia, Rhodesia, Mozambique and South Africa, from Burch (1964, 1967, and unpublished in the case of Corsica, Mauritania and Gambia); Natarajan, Burch & Gismann (1965); and Brown, Schutte, Burch & Natarajan (1967). Distribution in Ethiopia (indicated by broken line) is shown in Fig. 2. 192 BROWN AND BURCH TABLE 1. Chromosome numbers of “Bulinus s.s.” from 22 Ethiopian localities Haploid Collection Locality chromosome number number 65/8 15 km west of Webbe Shibeli bridge on road be- tween Shashamanna and Dodolo, Arussi 18 province 65/14* Lake at Langhei village, Harar province 18 65/28 Western shore of Lake Zwai, Arussi province - 18 65/44 43 km northeast of Dangila on Bahar Dar road, Gojjam province 18 65/61* Northeastern shore of Lake Ashangi, Tigre province - 18 65/62* Southern shore of Lake Haik, Wallo province 2 18 65/65* Lake Bishoftu (Biete Mengest), Shoa province 18 65/77 30 km towards Shashamanna from Soddu/Aba Minch road junction, Sidamo province 18 65/29 Northeastern shore of Lake Awasa at entrance of Black River, Sidamo province : 36 65/46 Bahar Dar town, Gojjam province - 36 65/50 50 km northeast of Bahar Dar town on Gondar road, Begemedir province o 36 65/53 37 km south of Gondar on Gorgora road, Begemedir province х 36 65/58 1 km west of Aduwa road junction on Aksum road, Tigre province . 36 65/66* Northwestern shore of Lake Zwai, Arussi province 36 65/22 15 km west of Debra Birhan on Jihur road, Shoa province 54 65/40 21 km southeast of Engiabaia (Injibara) on Debra Markos road, Gojjam province À 54 65/85 21 km north of Shano, Shoa province z 54 65/54* 38 km north of Gondar on Asmara road, Begemedir province 3 72 65/84 14 km north of Shano, Shoa province à 72 65/86 1 km north of Debra Birhan, Shoa province A 72 65/87 3 km north of Debra Birhan, Shoa province : Ta 65/89 4 km south of Shano, Shoa province Е na * The asterisks designate populations previously classified in B. truncatus sericinus by Brown (1965). * ETHIOPIAN BULINUS 193 The new observations are discussed in relation to information previously published for Ethiopia and other parts of Africa. We are indebted to Mrs. A. Gismann for her valuable criticism of this paper. METHODS Material for cytological examination was fixed in Newcomer’s (1953) fluid immediately after collection. Cyto- logical techniques employed were those used in our previous cytological studies, i.e., acetic-orcein squash preparations (La Cour, 1941) observed with Nikon microscopes using oil immersion ob- jectives (n.a. 1.25) and 10X and 30X oculars. Chromosomes were counted in up to 5 individuals from each locality. Material for morphological study is preserved in alcohol and has been de- posited in the Experimental Taxonomy Section of the Zoology Department, British Museum (Natural History). OBSERVATIONS Details for samples of Bulinus, col- lected between the end of July and the middle of September, 1965, are listed in Table 1. Their distribution is shown in Fig. 2. DISCUSSION Because of the presence of arrowhead- shaped mesocones on the lateral radular teeth and the frequent absence of the male copulatory organ, Ethiopian bulinines with various shell forms have been classified in the Bulinus truncatus species group, as В. truncatus sericinus (Jickeli) and Bulinus sp., by Brown (1965) and Mandahl-Barth (1965). Burch(1967) found topotypical material of В. t. schackoi (Jickeli) (-sericinus, according to Brown, 1965; Mandahl-Barth, 1965) to be tetraploid (n=36), whichis the usual condition in the B. truncatus group. Our cytological studies have shown some of the populations previously classified in B. t. sericinus to be diploid, while others are polyploid (Table 1), and it thus seems likely that a study of morphology in relation to chromosome number will lead to a revised classification of these snails. The Ethiopian diploid populations that have been examined have arrow- head shaped mesocones on the lateral radular teeth, and are therefore com- parable to the B. natalensis group of forms occurring in southern Africa (Brown et al., 1967). In the total of 28 Ethiopian localities from which “Bulinus s.s.” have been examined cytologically, different forms have been obtained with the following frequencies: n=18(10), n=36(8), n=54(4) and n=72(6). The total of localities is made up of 1 reported by Burch (1964, Lake Awasa), 6 given by Burch (1967, excluding a further sample from the same part of Lake Awasa), and 21 reported in the present paper (excluding Lake Bishoftu, from which material was examined by Burch, 1967). The dupli- cate samples had, for each lake, the same numbers of chromosomes. Both diploid and tetraploid snails are widely distributed (Fig. 2), although diploid populations have not been found north of Lake Ashangi or Lake Tana. The hexa- ploid (n=54) and the octoploid (n=72) are comparatively restricted in occurrence, but have extensive ranges on the high- land plateau northwest of the Rift Valley (Fig. 2). В. “sericinus” with 72 pairs of chromosomes has also been reported from Tarbak in the highland of Western Aden Protectorate (Natarajan et al., 1965). In each of the lakes Awasa and Zwai, situated in the southern Ethiopian Rift Valley (Fig. 2), occur 2 cytologically different forms of “Bulinus s.s.”: the diploid Bulinus sp. previously reported from Lake Awasa by Brown (1965), anda tetraploid form. These 2 bulinines have distinctive shells and each is apparently restricted to a characteristic habitat. Collections were made on the eastern shore of Lake Awasa and the western shore of Lake Zwai: on each shore 2 194 BROWN AND BURCH ей € LOY 1 < === \ & sani ~ TIGRE SS 16 ER = ^ ^ - / o) er SANT iso SR Gondar. ое ` > = > ИИ АА 7 L. Ashangi II => ! :@ их ” I SA AX Nie at ” 4 ype Y ) ses Nie “a / L FETE >. ‘ = у / ( DC N iz / a a ® sa \ / \ PA No. \ / a м \ Е «>> Bahar 250 ne Е ae Haik / > => 1 ле 4 © / | ! Son \—/ Y En = bledo: AN Ce WALLO = r SS 4 ® Ter DR , a u а 1 de OJJAM sale \ a / e E ER \ © Debra Marcos De nee 2 \ 1 ER Sea Se, Debra Birhan \ ( = == CN j ' N 7 OS A \ A : à \ ` SES CES 1 ПА НОА , [ e | sh te MARNE Ca ea RATE Ama NE ve See A \Shano oe AO \Harar S tS = 3 AL «Y : Addis Ababa >). i Der en x Sa + ~ e SEE ES И | DI HARAR > US hats EN Г. Bishoftu ‘5 s ’ LS > Er ! Dr . Gé 2 1102001 “ E MZ e--- ce ` SS n= DESSEN SEN = er L. Zwal >> ee ae a Ve ASA FT LRT es y= АВЕ, \ SEE SSeS MNT e+Shashamanna , ae \ o © N O en = x = Y \ E , L. Awasa MERE \ ‘ie р On i „iD “fp SIDAMO ı ı, O!l.n=T18 \ AY ef 4; e -- / \ ` \y / Ss No \ C2 Ne \ A “8 Gee Sr o n-36 \ DEREN VA { J 1.0 \ Sell Ne IN n = 54 OS no 1 LR A n = 72 ии \ Ny = = = = m + = S 2000 m < Ре ` + \ — > + Я \ / N =" IS 200 km L. Rudolf S—— 7 ии = = / = $ КЕМУА IS / © FIG. 2. Distribution in Ethiopia of diploid (п=18) and various polyploid populations of “Bulinus s.s.” (from Burch, 1967 and present paper). Each symbol represents 1 locality, except for the octoploid near Debra Birhan where 2 localities are represented by 1 symbol. ETHIOPIAN BULINUS 195 types of littoral habitat were in- vestigated. 1) On gently shelving bottoms exposed to wave action, beds of aquatic grass extend from the edge of the water into a depth of about 2 m, i.e., to as much as 100 m offshore in Lake Zwai. The substratum is composed largely of sand and fine gravel. The molluscan fauna comprises few species: Melanoides tuberculatus (Miiller), Lymnaea natalen- sis Krauss (found in Lake Zwai only), Gyraulus bicarinatus Mandahl-Barth (Brown, 1967), and Bulinus sp. with 18 pairs of chromosomes. Bulinus sp. was very abundant in both lakes. 2) In situations protected from wave action there are Swamps with luxuriant vegetation, including Papyrus and Nymphaea spp., which in some places forms a floating bog. At the times of the observations the water was brown in colour and clear, in contrast to the cloudy water of the open lakes. The substratum is composed largely of dark mud with a high organic content. The molluscan fauna is comparatively rich, comprising: Biomphalaria sudanica (Martens), Segmentorbis sp., Lentorbis sp., Anisus coretus (de Blainville) (Brown, 1967), A. natalensis (Krauss) (found in Lake Zwai only), Lymnaea natalensis, Sphaerium sp. (found in Lake Awasa only), and Bulinus sp. with 36 pairs of chromosomes. Bulinus Sp. was rare in both lakes. In view of the demonstrated differences between the habitats of the cytologically different bulinines in lakes Awasa and Zwai, and the fact that no more than 1 cytological form has been obtained from any locality in Ethiopia, it is believed that different forms might well be adapted to particular ecological conditions. All the known localities for the hexaploid and octoploid forms in Ethiopia are situated at an altitude of more than 2000 m, and those reported in the present paper are small, cold-water streams like those described by Burch (1967). However, the numbers of individuals in our samples are not large, and it ispossible that cytological examination of more specimens might reveal that different forms occur together in Ethiopia, as found by Burch (1967) in several small bodies of water in Tanzania. Even so, the homogeneity we have found in each sample suggests that other cytological forms, if present in these localities, are rare. A form might have a low density in some localities because it had recently originated or arrived there, but, with regard to the diploid and tetraploid forms, which are widely distributed with overlapping ranges, it seems that the apparent dominance of one or other form in any habitat may depend on whichever is better adapted to the particular eco- logical conditions, and perhaps be the outcome of competition between the forms. Because of the apparently close relationship between Bulinus “truncatus sericinus” and В. truncatus truncatus of Egypt, these snails have been regarded as an actual or potential intermediate host of Schistosoma haematobium in Ethiopia (Ayad, 1956; Brown, 1964). There are, however, few records of human infection with S. haematobium from areas where В. “truncatus seri- cinus” has been recorded, i.e., the high- land plateaux and the southern part of the Rift Valley. The B. (“Physopsis”) africanus species group, whichis of great epidemiological importance in some parts of Africa, is comparatively rare in Ethiopia. Endemic foci of urinary bilharziasis have been reported reliably only from the lower Awash River valley (Russell, 1958; Aklilu Lemma, 1965), where the intermediate host is probably B. (“Physopsis”) abyssinicus (Martens) (Brown, 1967), as is also the case in the lower Webbe Shibeli river valley of Somalia (Ayad, 1956). With regard to the apparent rarity of 5. haematobium in Ethiopia, Ayad suggested, among other epidemiological considerations, that B. “truncatus sevicinus” in Eritrea might not be susceptible to infection; Brown (1964) pointed out that the great varia- bility in shell form of that taxon in Ethiopia, considered as a whole, in- dicated the existence of local races that might also differ in their physiology and capacity to act as intermediate hosts. 196 BROWN AND BURCH Support for those early speculations is provided by our finding that B. “trun- catus sericinus” comprises several cytologically different forms. Of these, the tetraploid (n=36) may be capable of serving as an intermediate host to S. haematobium, in view of the fact that the members of the B. truncatus group known to do so in northern Africa and southwestern Asia are tetraploid (Burch, 1964). Ethiopian octoploid (n=72) popu- lations may also be susceptible to infection, as evidence obtained by Wright (personal communication, in Burch, 1964) implicates an octoploid form of “Bulinus s.s.” in the trans- mission of S. haematobium in Western Aden Protectorate. There is no evi- dence that a naturally occurring diploid (n=18) population of the species groups here discussed serves as an intermediate host in Ethiopia or elsewhere. The tetraploid appears to deserve most con- sideration in relation to the epidemiology of urinary bilharziasis in Ethiopia, since the known distribution of the higher poly- ploids is confined to a high altitude zone where the climate is, it may reasonably be assumed, generally too cool to permit full development of larval S. haemato- bium. At lower altitudes, where the climate may be suitable for that parasite, a greater number of diploid than tetra- ploid populations has been found. It therefore appears that the low frequency of tetraploid populations may serve, in conjunction with other factors such as a low density of human populations, as a barrier to the establishment of foci of urinary bilharziasis. A detailed study of the susceptibility to schistosome infection of cytologically different “Bulinus s.s.”, from Ethiopia and elsewhere, and of their ecology, seems worthwhile, as the detection of forms resistant toinfection and success- ful in competition with susceptible snails might be of potential practical value. The introduction of a non-susceptible form into natural habitats or newly created water bodies might serve asa means for the biological control of sus- ceptible snails. This possibility seems most likely of realization using diploid forms, which are apparently resistant to infection and live in a variety of habitats and climatic zones. LITERATURE CITED AYAD, N., 1956, Bilharziasis survey in British Somaliland, Eritrea, Ethio- pia, Somalia, Sudan and Yemen. Bull. Wild Hlth Org., 14: 1-117. BROWN, D. S., 1964, The distribution of intermediate hosts of Schistosoma in Ethiopia. Ethiopian med. J., 2(4): 250-259. 1965, Freshwater gastropod Mollusca from Ethiopia. Bull. Brit. Mus. (nat. Hist.) Zool., 12(2): 37-94. 1967, Records of Planorbidae new for Ethiopia. Arch. Moll., in press. BROWN, D.:08S. sSCHUT TE) CHIEF da, BURCH, J. B. & NATARAJAN, R., 1967, Chromosome numbers in re- lation to other morphological charac- ters of some southern African Bulinus (Basommatophora: Planorbidae). Malacologia, 6(1): 175-188 BURCH, J. B., 1964, Cytological studies of Planorbidae (Gastropoda: Basom- matophora). I. The African subgenus Bulinus s.s. Malacologia, 1(3): 387- 400. 1965, Chromosome numbers and systematics in Euthyneuran snails. Proc. First Europ. malac. Congr., 1962: 215-241 [publ. by Conch. Soc. Great Britain & Ireland and Malac. Soc. London]. 1967, Chromosomes of inter- mediate hosts of human bilharziasis. Malacologia, 5(2): 127-135. LA COUR, L., 1941, Acetic-orcein. A new stain-fixative for chromosomes. Stain Techn., 16: 169-174. LEMMA, AKLILU, 1965, Report on bil- harziasis survey in the lower Awash Valley, February 4-12, 1965. Unpub. doc., Haile Sellassie I University, Addis Ababa. MANDAHL-BARTH, G., 1958, Inter- ETHIOPIAN BULINUS 197 mediate hosts of Schistosoma. African binae, Planorbininae, and Bulininae. Biomphalaria and Bulinus. Wld Hlth Malacologia, 2(2): 239-251. Org. Monograph ser., 37: 1-89. NEWCOMER, E. H., 1953, A new cyto- 1965, The species of the genus logical and histological fixing fluid. Bulinus, intermediate hosts of Schisto- Science, 118(3058): 161. soma. Bull. Wld Hlth Org., 33: 33-44. RUSSELL, H. B. L., 1958, Final Report. NATARAJAN, R., BURCH, J. B. & GIS- Pilot Mobile Health Team, Ethiopia. MANN, A., 1965, Cytological studies Wld Hlth Org. document EM/PHA/62, of Planorbidae (Gastropoda: Basom- Ethiopia 13, 1958. matophora). II. Some African Planor- ABCTPAKT РАСПРОСТАРНЕНИЕ ЦИТОЛОГИЧЕСКИ-РАЗЛИЧНЫХ ПОПУЛЯЦИЙ РОДА BULINUS (ВАЗОММАТОРНОВА: PLANORBIDAE) В ЭФИОПИИ Д. 0 BEOYH Vi ik: “Bs БЕРУ Основное число гаплоидных хромосом у видов рода Bulinus, как и y других Planorbidae равно 18, но в группе видов Bulinus truncatus встречаются Формы с полиплоидным числом (п = 36); возможно, что среди этих Форм встречаются и такие, которые имеют п = 54 ип=72. Эта группа, а также группы видов В. tropicus и В. natalensis являющиеся диплоидными (п = 18) могут рассматриваться, как подрод “Bulinus s.s.”. Диплоидные и различные полиплоидные формы внутри этого комплекса имеют и несколько различные районы распространения. Диплоидные виды встречаются преимущественно в Южной Африке и на север вплоть до Эритреи, а на запад - до Мавритании. Тетраплоидные виды (п = 36) встречаются, главным образом в Северной Африке, в Средиземноморском и Средне-Восточном районах, а также в Мавритании, Сенегале, Гане, и Танзанье. Гексаплоидные (п = 54) были найдены только в Эфиопии, а октоплоидные (п = 72) в районах Эфиопии и Западного Аденского протектората. В природе, повидимому лишь тетраплоидные и, возможно, октоплоидные формы служат промежуточными хозяевами Schistosoma haematobium. В настоящей работе сведены результаты как современных цитологических наблюдений, сделанных на 22 пробах “Bulinus s.S.”, собранных авторами в Эфиопии, так и имеющиеся более ранние данные. В целом, цитологически-различные Формы из 28 мест были получены со следующей частотой встречаемости: п = 18 (10), n=36 (8), n=54 (4) и n=72 (6). Диплоидные u тетраплоидные популяции распространены более широко, в TO время, как гексаплоидные и октоплоидные формы были найдены лишь на высокогорном плато, к северо-западу от Рифтовой долины. Эфиопские формы “ Bulinus s.s.” ранее относили к группе видов В. truncatus. В различных популяциях, ранее определенных как В. truncatus sericinus, мы нашли диплоидные, тетраплоидные и октоплоидные формы по числу хромосом. Нам кажется поэтому, 198 BROWN AND BURCH что изучение морфологии моллюсков в связи с количеством хромосом приведет к пересмотру их классификации. Как диплоидные, так и тетраплоидные “Bulinus в.5.” встре- чаются на озерах Эйвеза и Цвейи; из них 2 Формы имели различимые друг от друга раковины и каждая была, видимо приурочена к определенному местообитанию. Цитологически- различные “Bulinus s.s.” были найдены в одном и том же водоеме только в Танзанье. Тот факт, что различные формы этой группы не были найдены живущими вместе ни в одном месте в Эфиопии может быть происходит из за конкурренции между отдельными формами и основано на адаптации их к определенным экологическим условиям. Тетраплоидная форма “Bulinus s.s.” кажется наиболее подходя- щей, как передатчик Schistosoma haematobium в Эфиопии, однако ее очевидная редкость распространения может здесь служить (вместе с другими эпидемиологическими факторами) барьером для образования очагов уринарного билхарциазиса. MALACOLOGIA, 1967, 6(1-2): 199-222 SOME OPISTHOBRANCHS FROM SAPELO ISLAND, GEORGIA, U.S. A.l Eveline Marcus and Ernst Marcus2 ABSTRACT The paper deals with 2 pelagic and 10 intertidal and subtidal opisthobranchs from the southeastern coast of the U.S.A. Four species are described: Okenia sapelona, Doridella burchi, Tritonia (Candiella) bayeri misa and Armina wattla. This last species differs from the solely American species of Armina by its strong caruncle, and from A. undulata and the other species with strong caruncle by the radula. Okenia sapelona resembles O. mediterranea, hence belongs to the opisthobranchs whose range can be traced from the Tertiary Tethys Sea. The remaining 2 new forms are related to species from the warm water region of the western Atlantic. The known littoral species of the present collection are inhabitants of that region too, with the exception of Doris verrucosa, which occurs also in the eastern Atlantic. The subspecies Pleurobranchaea hedgpethi hamva is suppressed, because in the present material the direction of the flap over the genital apertures was often found to be oblique, i.e. neither dorsal (P. hedgpethi hamva) nor anterior (P. h. hedgpethi). In 1962, Dr. J. B. Burch of the Uni- versity of Michigan initiated a study of the mollusks of the southeastern U.S.A. for the Sapelo Island Research Foun- dation and the Institute of Malacology. This paper is based on the specimens collected by Dr. Burch and his students, and by Mr. Milton 5. Gray, professional collector for the Marine Institute, University of Georgia. The zoological materials described here are part of the collections housed by the Marine Institute, Sapelo Island, Georgia, U.S.A. SYSTEMATICS AND DISTRIBUTION List of species Order Anaspidea, acea Superfamily Aplysi- Aplysiidae, Aplysiinae 1. Aplysia (Varria) morio Ver- rill, 9101 Order Notaspidea, Superfamily Pleuro- branchacea Pleurobranchidae, Pleurobranch- inae 2. Pleurobranchaea hedgpethi Abbott, 1952 Order Doridoidea, Suborder Eudoridoidea Infraorder Cryptobranchiata Dorididae, Doridinae 3. Doris verrucosa Cuvier, 1804, Fig. 1 Infraorder Phanerobranchiata, Superfamily Suctoria Goniodorididae 4. Okenia sapelona, new spe- cies, Figs. 2-6 lContribution No. 143 from the Marine Institute, University of Georgia, Sapelo Island, Georgia, U. S. A., and No. 4 from the Southeast American Mollusks Program, Institute of Malacology. The collection of material on which this publication is based was supported by funds from the Sapelo Island Research Foundation. 2Caixa Postal 6994, Säo Paulo, Brazil. (199) 200 MARCUS AND MARCUS Corambidae 5. Doridella burchi, new spe- cies, Figs. 7-12 Suborder Porostomata Dendrodorididae 6. Dendrodoris krebsii (Mörch, 1863) 7. Doriopsilla pharpa Marcus, 1961 Order Dendronotoidea Tritoniidae 8. Tritonia (Candiella) bayeri misa, new Subspecies, Figs. 13-14 Scyllaeidae 9. Scyllaea pelagica Linne, 1758, Fig. 15 Order Arminoidea Suborder Euarminoidea Arminidae 10. Armina wattla, new species, Figs. 16-20 Order Aeolidioidea Suborder Acleioprocta Fionidae 11. Fiona pinnata (Eschscholtz, 1831) Suborder Cleioprocta Favorinidae, Facalaninae 12. Dondice occidentalis (Engel, 1925) 1. Aplysia (Varria) morio Verrill, 1901 Aplysia morio Verrill, 1901: 25, pl. 3, figs. 5, 5a. ? Aplysia modesta Thiele, 1910: 124, DAS IE MITA Aplysia (Уатта) morio Eales, 1960: 328-332, figs. 28-29. Occurrence: Sapelo Island, 319 02 min. 00 sec. N, 800 02 min. 25sec. W, 42.4 m depth, 1 young specimen, collected on July 9, 1963. Further distribution: Bermuda (origi- nal locality), and from Rhode Island to Florida and Texas. Since a single young, colorless speci- men, 12 mm in length, 8 mm in width, and 6 mm in height cannot be classified with certainty, the following characters were considered as decisive for the classification: soft skin; large and thin parapodia; short, elliptical tail; closed, invisible mantle foramen; rhach- idian tooth with excavated head, long cusp and large basal denticles; details of the denticulation of the lateral teeth; and the caecum lying flat on the surface of the digestive gland. The juvenile specimen of Aplysia mo- desta was included in the list of syno- nyms on the authority of Eales. In Thiele’s short description nothing con- trasts with the characters of A. morio. The locality, Lovango Cay, between St. John and St. Thomas, VirginIslands, was not included in the range of A. morio by Dr. Eales. 2. Pleurobranchaea hedgpethi Abbott, 1952 Pleurobranchaea hedgpethi Abbott, 1952: 1-2, pl. 1, figs. 1-8. Pleurobranchaea hama Marcus, 1957a: 21-27, figs. 40-52. Pleurobranchaea hedgpethi Marcus, 1960b: 253-254, fig. 6. Pleurobranchaea hedgpethi hamva Marcus, 1961b: 141; 1967a (in press), fig. 56. Pleurobranchaea hedgpethi Nijssen- Meyer, 1965: 143-145, figs. 1-2. Occurrences: Sapelo Island, Georgia, between 31° 33 min. 30 sec. to 30055 min. N and 790 37 min. 30 sec. to 800 24 min. W., 30-88 m depth; a total of 24 speci- mens collected in February and October, 1961, May, 1962, and June, July and August, 1963. Further distribution: Point of Cape Hatteras, North Carolina off Savannah Beach, Georgia, 70-95 m depth; Dry Tortugas, Florida, 51 m depth; Gulf of Mexico from Port Aransas (original locality) to the Bay of Campeche; coast of Surinam, 28 m depth; coast of Rio de Janeiro and Sao Paulo to 25° S., Brazil. The preserved specimens are 5-43 mm OPISTHOBRANCHS FROM GEORGIA 201 in length, 3.5-22 mm in width and 3-13 mm in height. Where the pigment is preserved, the caudal spur is black. In some specimens there are residues of a brown dorsal network. The flap above the genital apertures either has adorsal direction or a direction intermediate between dorsal (as in Pleurobranchaea hamva) and anterior (as in P. hedg- pethi). This oblique position, also ob- served in Nijssen-Meyer’s slug from Surinam, leads us to abandon a special designation for specimens witha dorsally directed flap. 3. Doris verrucosa Cuvier, 1804 (Fig. 1) Doris verrucosa Cuvier, 1804: 451, 467, pl. 73, figs. 4-7; Eliot, 1910: 96-98 (incl. var. mol- lis), not the reference (: 147) to Alder and Hancock, 1856, Family lad BALA gilts у. Ihering, 1915: 142; Pruvot-Fol, 1954: 232-233, figs. 86a-e (verrucosa), figs. 87a-h (januarii); Marcus, 1955: 127-131, figs. 102- 108; 1957b: 420, fig. 90 (Sáo Paulo). Staurodoris verrucosa Bergh, 1878: 579-583, pl. 63, figs. 20-23, pl. 64, figs. 2-7; у. Ihering, Catarina); Bergh, 1894: 161-162, pl. 5, figs. 16-18 (western Florida); 1904: 38-39 (South Carolina); Eliot, 1906: 337-339 (incl. var. mol- lis); Bergh, 1907: 46-47, pl. 11, figs. 26-27; Nobre, 1938-1940: 52, pl. 10, fig. 3. Staurodoris januarii Bergh, 1878: 583- 585, pl. 63, fig. 24, pl. 64, figs. 8-12 (Rio de Janeiro); 1880: 37-40, pl. C, figs. 13-25, pl. D, fig. 22 (Rio de Jan- eiro); v. Ihering, 1886: 230 (synonymized with verrucosa). Doridigitata derelicta 1929: 763; Johnson, 1934: 158; Lange de Morretes, 1949: 116. 1886: 230-232 (Sta. O’Donoghue, Occurrences: 1) Sapelo Sound, Geor- gia, November 30, 1961, 3 specimens; 2) Sapelo Sound, 12 m depth, January 26, 1962, 2 specimens. Further distribution: West Atlantic: South Carolina; Manatee Bay, western Florida; Rio de Janeiro, Säo Paulo, Santa Catarina, Brazil. East Atlantic: British Isles, France, Portugal; Medi- terranean; South Africa. Original locality unknown (Bergh, 1878, p 579- 580), possibly Mediterranean. The largest of the present preserved specimens is 35 mm long, 20 mmbroad, and 13 mm high. Its sole is 28x14 mm. The big vesicular warts, up to 1.8 mm in diameter, are separated from one another and do not coalesce. The an- terior pedal border is bilabiate and slightly notched. The color of the live animals was brownish-yellow (locality 1) and yellow- ish-green (locality 2); preserved speci- mens are whitish. In one slug the spicules are preserved and form ridges between the warts. The tentacles are barrel-shaped and grooved. The rhino- phores have 16 perfoliations. The rhino- phoral pits bear one large papilla on either side and smaller ones infront and behind. There are up to 14 gills. The rim of the branchial pit has a variable number of papillae, up to 9 big ones and 9 small ones. The labial cuticle, the radula, and the gut are as previously described. A major distinction between Doris and Archidoris Bergh, 1878, lies in the prostatic ental portion (Fig. 1, pr) ofthe male duct, which is not thickened in Archidoris. The latter genus has a pleurembolic penis, while the penis is the acrembolic type in Doris. The ectal part of the efferent duct (e) of Doris is coiled within a muscular sheath, at the inner end of which inserts a retractor muscle (re). In the present material, and also in a re-examined specimen from Бао Paulo, we found the disposition of the seminal receptacles slightly different from our earlier dia- gram (Marcus, 1955, fig. 106). The insemination duct (id) begins at the 202 MARCUS AND MARCUS FIGS. 1-6. Fig. 1. Doris verrucosa. Diagram of reproductive organs. FIGS. 2-6. Okenia sapelona, n. sp. Fig. 2. Dorsal view of living slug, drawn by Dr. J. B. Burch; stippled parts: pale yellow; black parts: bright yellow. Fig. 3. Side view of preserved slug. Fig. 4. Poly- gonal jaw elements, conical in profile. Fig. 5. Half-row of radula. Fig. 6. Diagram of repro- ductive organs. entrance of the vagina (v) into the sper- Discussion of Doris verrucosa matheca (spa), similar to Bergh’s figure 23 (1880, pl. C) of material from Rio Pruvot-Fol (1954: 234) maintains de Janeiro. Doris ocelligera Bergh, 1881a (p 95-98, OPISTHOBRANCHS FROM GEORGIA 203 KEY TO LETTERING IN FIGURES a ampulla ag albumen gland an anus b brain с caruncle (wattle) e efferent duct fg female gland mass (i.e. albumen and mucus glands) ft foot ftg foot gland g gill ga genital aperture(s) h hermaphrodite duct hn hyponotum id insemination duct io inner oviduct mu mucus gland mt marginal tooth n nidamental duct (outer oviduct) no notum nr notal ridges O ovotestis p penis pr prostate r rhachidian tooth ra radula re retractor muscle rh rhinophore rha rhinophoral appendage тр renal pore sp spermatocyst spa spermatheca it intermediate tooth (or inner lateral tooth) spo spermoviduct } jaw 1 lateral lamellae la lateral appendages lt lateral tooth (teeth) m mouth ma male atrium pl. 4, figs. 11-21) as a separate species. The original material came from Trieste; у. Ihering had a specimen 4 mm long from Naples (1886: 232-233), and Pruvot-Fol had several slugs 10-12 mm long from Banyuls, France. The length of the living animals in the origi- nal diagnosis, 0.5 cm, is a misprint for 5 cm. The descriptions of Bergh, v. Ihering, and Pruvot-Fol are not quite uniform, but hardly justify a specific separation of D. ocelligera. The repro- ductive organs were recorded only by Bergh. Staurodoris pseudoverrucosa v. Iher- ing (1886: 233) from Naples with conical tubercles and 5 bipinnate gills cannot be united with Doris verrucosa. Its repro- ductive organs were not described, so that its generic position remains un- known. t tentacle tu tubercle У vagina va velar appendage ve veil 4. Okenia (Okenia) sapelona, new species (Figs. 2-6) Occurrence: Sapelo Island, Georgia, November, 1963; 2 specimens. Holotype, UMMZ 230616; Paratype UMMZ 230617. The general color is an iridescent pale blue; the rhinophores and some spots on the gills are maroon. The following areas or spots are bright yellow (Fig. 2): dots on the lateral margins and the corners of the veil (ve), a broad fleck between the first ap- pendages, a center stripe excluding the bases of 5 dorso-median pale blue tubercles (tu), 2 pairs of larger roundish spots in front of the gills (g), smaller irregular spots forming a row on each side of the back, and a medianblotchbe- hind the gills. The tips of the lateral appendages (la) and of the gills are pale 204 MARCUS AND MARCUS yellow, and the postbranchial bright yellow area passes into a light yellow one which extends backwards with 2 lines. The measurements of the 2 live speci- mens (Holotype and Paratype re- spectively) were: body length 7.6 and 5.3 mm; width of body 2.2 and 1.9 mm; length of rhinophores 1.7 and 1.4 mm; diameter of rhinophores 0.3 and 0.2 mm; length of the longest gill in both speci- mens 1.1 mm; length of central pig- mented stripe 3.4 and 2.0 mm; Holo- type, UMMZ 230616; Paratype, UMMZ 230617. The shape of the veil is different in the 2 specimens. In the larger speci- men its anterior border is convex (Fig. 2); in the smaller one it is straight and slightly concave in the middle. The tentacles are triangular flaps. The foot corners are rounded, not projecting. The tails of the pre- served specimens end witha longer point than they showed in life; the foot is narrower than the back. The pallial ridge bears 11 soft appendages on each side in the larger slug, 9 in the smaller slug. These appendages have blunt tips when alive; when preserved they are pointed. The hindmost appendages are double. In the smaller slug the 6th right papilla is also double. Two pairs of the appendages stand in front of the rhinophores and 2 behind the bran- chiae. The gills are unipinnate, the larger animal has 9, the smaller ani- mal has 7. Five low, pointed tubercles stand on the center stripe (Fig. 3). In front of the gills there is a pair of low bosses, whose position corresponds to the posterior pair of the large, roundish spots mentioned above. The labial cuticle bears a ring-shaped thickening around the mouth composed of smooth, polygonal elements (Fig. 4), which are conical in side view. Some- times their tips are slightly curved. The radula (Fig. 5) of the larger slug consists of 12 rows of teeth (radular formula: 1.1.0.1.1). The inner (inter- mediate or lateral) tooth is 94и long, the outer (marginal) tooth 28u long. The former bears 20-24 denticles оп Из inner side, of which those near the tip are stronger than those near the base. There is a boss near the root of the cusp similar to Vayssiére’s figure 15 (1919: pl. 4) of Okenia dautzenbergi. The scale- shaped marginal tooth has a single short point. The reproductive organs (Fig. 6) are Similar to those in other species of Okenia. The hermaphrodite duct (h) is thin, the ampulla (a) cylindrical inthe dissected slug. A narrow spermoviduct merges into the female gland mass (fg). From there the male duct goes out and soon becomes prostatic (pr). This glandular portion of the duct bends back on itself and is sharply set off from the narrow efferent duct (e), whichis sheathed by an outer muscle layer. It functions as an acrembolic penis (p), and its terminal section bears cuticular spines. The nidamental duct (n) and the vagina (v) open together. The vagina leads to a spacious spermatheca (spa). The long and curved insemination duct (id) leaves the spermatheca near the entrance of the vagina. These 2 spermathecal ducts correspond to what is called the serial type of the seminal receptacles. The canal of the spermatocyst (sp) arises from the insemination duct. Discussion of Okenia (Okenia) sapelona For the principal literature and the valid species of Okenia Menke, 1830, we refer to our list (Marcus, 1957b: 438). Since then 3 new species, allfrom Japan, have been added to the genus (Baba, 1960: 80-81; Hamatani, 1961: 363-365). One of these Japanese species, O. plana, has recently also been found in San Francisco Bay (Steinberg, 1963: 65, 71). These new Pacific species belong to the subgenus Okenia, characterized by ap- pendages on the central area of the back between rhinophores and gills. When these appendages are poorly developed, they can only be seen in side view as low tubercles (Fig. 3). The 2 species of Okenia described from western OPISTHOBRANCHS FROM GEORGIA 205 Atlantic warm waters belong to the sub- genus Okenia s.s. (Marcus, 1957b: 434- 442), while those reported from the New England area (Johnson, 1934: 156) belong to the subgenus /daliella Bergh, 1881. This latter subgenus does not have appendages in the center of the dorsum. The species from western Atlantic warm waters are Okenia eve- linae and O. impexa. O. evelinae has been recorded from the southern middle Brazilian coast and Miami (Marcus, 1960а: 162). O. impexa has been re- corded from Brazil and Beaufort, North Carolina (Marcus, 1961b: 144). The labial cuticle of Okenia impexa is thin and smooth, and its marginal tooth bears a sharp principal cusp and 2 further basal secondary cusps. The thick labial cuticle of O. evelinae pro- jects with a kind of bilabiate beak on each side of the mouth; this beak is simple, not composed of separate ele- ments. The lateral appendages do not extend backwards beyond the level of the gills. In front of the latter there are 2 median unpaired papillae and 1 lateral pair. The male copulatory organ is pleurembolic, without spines, and ends with a penial papilla. Several of the East Atlantic and Medi- terranean species of Okenia in Pruvot- Fol’s classification (1954: 308-311), O. elegans, O. dautzenbergi, and O. medi- terranea, must be compared with O. sapelona. In Vayssiére’s last publi- cation (1930) O. dautzenbergi is annexed to O. elegans. The marginal tooth of O. elegans elegans and O. elegans daut- zenbergi, although 7-8 times smaller than the lateral tooth, is rather similar to it in shape. O.e. elegans has 17-22 gills, O. e. dautzenbergi 14 gills. The broad foot appears to project beyond the pallial ridge in dorsal view. O. medi- terranea, the species nearest to O. sapelona, has no papillae down the mid- line, but 2 pairs of tubercles in front of the gills, and the surface of the labial elements bears numerous points and tubercles. Pruvot-Fol (1954: 311) said that O. mediterranea and O. amoenula, a South African species, are similar, but according to Macnae (1958: 368-369), the latter species belongs to the subgenus Idaliella. 5. Doridella burchi, new species (Figs. 7-12) Occurrences: Georgia, 1) off Cabretta Island, 4.3 m depth, November 28, 1962; 2a) Blackbeard Creek, between Black- beard and Sapelo Islands, about 1/4 mi. from Raccoon Bluff, tide flat, December 1962; 2b) same locality, May 14, 1964; 3) Sapelo Sound, 16-26 mdepth, April 10, 1963; 4) off Nanny Goat Beach, Sapelo Island, May 17, 1963 (type locality). A total of 182 specimens, all on Alcyoni- dium (Bryozoa, Ctenostomata). Holo- type, UMMZ 230618; Paratypes, UMMZ 230619. Living slugs are up to 8 mm long, but the average size of these specimens was somewhat less than 7 mm. The largest preserved specimen was 7 mm inlength, 5 mm in width and 3 mm in height. The collection also contained many smaller animals, some as little as 1.5 mm in length. The notum ofthis species is transparent, slightly opaque, the foot rather opaque. Through the notum the following organs and structures were seen: the pumping heart in the hind part of the visceral mass, the dark brown digestive gland and white lobes around it, which are the follicles of the ovo- testis, the yellowish-tan outline of the visceral mass, and the outline of the foot. Round brown spots are numerous in the center (Fig. 7), decreasing in number towards the periphery. Many of these spots are about 150. in diameter, but much smaller spots also occur. The ramified pigment cells forming these spots lie in the deepest layer of the notum, and therefore they appear near the surface of the hyponotum in ventral view (Fig. 8). Yellow marks are scat- tered between the brown spots. In pre- served animals the foot, the rhinophores, and the gills are whitish. The notum is almost circular when 206 MARCUS AND MARCUS FIGS. 7-11. Doridella burchi,n. sp. Fig. 7. Dorsal view of living slug. Fig. 8. Ventral view of same. Fig. 9. Side view of same. Figs. 7-9 drawn by Dr. J. B. Burch. Fig. 10. Inter- mediate tooth of radula. Fig. 11. Three rows of marginal (or outer lateral) teeth viewed from different angles. the animal is at rest, but pointed behind in locomotion (Fig. 9). The notal border is not notched behind, but a bundle of muscles that originates between the gills and inserts on the border of the hypo- notum, may produce a temporary emar- gination of the notum. As in other corambids the integument consists of a deciduous cuticle which is known to be periodically shed and renewed (Mac- Farland & O’Donoghue, 1929: 9), an epi- dermis which secretes the cuticle and its pegs, and a thick layer of con- nective tissue without spicules. The pegs in the present species are broader than those in Corambe pacifica MacFarland € O’Donoghue, 1929, (pl. 1, Figs. 3-4). The cells of the connective tissue are scarce. Numerous large glands are sunk into the connective tissue. The foot (ft) is cordate and bilabiate in front (Fig. 8), rounded behind, though sometimes narrower, sometimes broad- er, according to contraction and re- OPISTHOBRANCHS FROM GEORGIA 207 laxation. At rest the foot and the head and its veil (ve) are covered by the notum. In locomotion (Fig. 9) the head is protruded. In the gliding animal the tentacles touch the substrate. Asin most other corambids, the rhinophores bear 2 lamellae on either side. They are similar to those of Corambella baratar- iae (Harry, 1953: fig. 6), whichaccording to Franz (1967, p 75) is a synonym of Doridella obscura Verrill, 1870. In the present species, however, the border of the rhinophoral pit is not scalloped in our specimens of C. burchi. In these the border is a broad collar whose outer edge is smooth, while the inner edge has some slight radial folds (Fig. 7). In the preserved specimens the rhinophores are completely withdrawn. The eyes lie deep under the skin, a little in front of the cerebral ganglia, on either side of the crop. On account of the coa- lescence of the cerebral and pleural ganglia, the central nervous system agrees better withthat of Corambe testu- dinaria (Fischer, 1889 (Hoffmann, 1936: figs. 554 A, B) than with that of C. pacifica (MacFarland & O’Donoghue, 1929: pl. 2, figs. 8, 9). In the groove between the hind end of the notum and the foot, whichrepresents an extremely reduced mantle cavity (Hoffmann, 1934: 330), a pair of gills (g) lies on either side of the midline (Fig. 8). Each pair consists of a smaller anterior, more ventral, and a larger, posterior, dorsal plate, the former with about 4 dorsal and 4 ventral leaves, the latter with about 9 leaves on either face. The anus opens betweenthe gills, not on a papilla; the renal pore lies beside the anus. The muscle fibres connecting the anal region with the hind margin of the hyponotum have already been mentioned. There are 3 branchial glands at the base of each pair of bran- chiae. The genital apertures lie on a lobed papilla immediately behind the right tentacle. A high epithelium covers the genital papilla. The cavity of the mouthis lined with a thick cuticle. The radula consists of about 40 rows. Each half-row con- tains 1 inner lateral or intermediate tooth and 5-6 (in the older portion of the radula 4) outer lateral or marginal teeth. The intermediate tooth (Fig. 10) is 88u long, 47u high, and bears 3-7 denticles on the inner surface of the hook. The first marginal tooth is 34y long. The different aspects of the marginal teeth viewed from different angles are shown by the 3 half-rows of Fig. 11. The male and female follicles of the ovotestis are separate as in Doridella obscura and C. carambola, but in con- trast to Corambe evelinae. The her- maphrodite duct (Fig. 12, h) and the ampulla (a) have the usual characters. The spermoviduct (spo) divides outside the female gland mass (fg). The male branch is glandular (pr) in its entire length, from the bifurcation of the spermoviduct to the root of the conical penis (p). The base of the latter is thick, cushion-like, its tip pointed, its epithelium ciliated. The ciliated vagina (v) runs straight inwards from the female genital aperture and has several con- strictions. It begins with its own outer opening immediately next to the penis and ends in a spacious spermatheca (spa) containing unorientated sperm. The insemination duct (id) leaves the sperma- theca laterally, well away from the en- trance of the vagina. Its coiled course before it enters the oviduct is simplified in the diagram (Fig. 12). The spermato- cyst (sp) which lodges orientated sperms is annexed to the ental portion of the inner oviduct (io). The latter passes into the glandular oviduct whose con- volutions constitute the so-called female gland mass (fg). The outer oviduct or nidamental duct (n) opens separately from the vagina. As mentioned above, the papilla on which the 3 genital aper- tures lie, has a high, folded epithelium. The egg string described by Dr. J. B. Burch (personal communication) corre- sponds to a Single planed coil of clear jelly, similar to that observed in other 208 MARCUS AND MARCUS corambids, e.g., Doridella obscura (Verrill, 1873: 30) and Corambe pacifica (MacFarland & O’Donoghue, 1929: 20). The 3 turns observed in the present species sometimes overlap slightly. They contain 2-3 layers of eggs, each egg in its own capsule. The eggs are spaced from one another by distances equal to their own diameter. Mac- Farland & O’Donoghue calculated 1500 eggs in the egg string of С. pacifica.. The corambids feed upon Bryozoa. The bryozoan on which Doridella burchi was found was identified as Alcyonidium cf. verrilli by Mr. Milton S. Gray of the Marine Institute, University of Georgia. As this erect, branching bryo- zoan species is recognizable by a much firmer consistency than that of A. gela- tinosum and A. hirsutum, both of which grow in a similar form (Osburn, 1932: 444), we presume that this identification is correct, although we are not aware of other records of A. verrilli south of Chesapeake Bay (Osburn, 1944: 14). The species is named for Dr. J. B. Burch, who first recognized it as an undescribed species and who provided drawings of the living animals and many observations which were included in our description. Discussion of Doridella burchi The family Corambidae comprises 11 species known from the European west coast (Netherlands, France) and the Atlantic and Pacific coasts of North and South America. The bibliographic records up to 1929 can be found in Mac- Farland & O’Donoghue (1929: 2-3). Fur- ther species were described by Harry (1953), Marcus (1955, 1958a, 1959), and Lance (1962a). The 2 genera currently distinguished are Corambe Bergh (1869: 359, footnote), and Doridella Verrill (1870, р 405). In Corambe there is a median notch in the posterior border of the notum, in Doridella the border is not notched. The posterior notch of Corambe has been observed in living C. testudinaria, C. pacifica and C. evelinae as a con- stant structure. In preserved C. evelinae and in C. lucea, which is not known alive, the notch may be contracted and reduced to a mere suture. In living Doridella the contraction of the notal border may produce a temporary emar- gination, probably for the better flow of water around the gills, but preserved specimens of Doridella showing a notch have not been described. The first corambid described having an entire notal border is “Doridella” obscura Verrill, 1870, a species which occurs from Vineyard Sound, Mas- sachusetts to New Jersey. (Verrill published a more detailed description in 1873: 307, 400-401, 664, pl. 25, fig. 173a, b). The type species of Corambella, C. depressa Balch, 1899, was first col- lected at Long Island, New York, hence within the range of D. obscura. The gills of C. depressa are plate-like, as probably those of the type species of Corambe, C. sargassicola Bergh (1871: 1295, pl. 11, fig. 24, pl. 12, fig. 1). The same type of gills occurs in Doridella obscura (Franz, 1967, fig. 1A). АП other species with posterior notal notches have plume-like gills. If the gills were used for the separation of the genera, Doridella would become a synonym of Corambe. Then the species with plume-like gills would have to be united under a new name. One species with an entire notal border, Doridella steinbergae Lance, 1962a(: 35), would come under this new genus. Thiele (1931: 430) mentioned the twisted inner lateral tooth of Coram- bella depressa Balch (1899: pl. 1, fig. 15) in the diagnosis of the genus, but we do not accept it as a specificfeature. Evidently this tooth had been deformed by too strong a pressure of the cover glass. The reticulate pattern of the notum, the single pair of lamellae sur- rounding the central cone of the rhino- phore, and the anal opening on a papilla are perhaps specific characters of C. depressa; the genital aperture on the left side is an anomaly or a misin- terpretation. OPISTHOBRANCHS FROM GEORGIA 209 Dovidella steinbergae differs widely from all other species of Doridella by the plume-like branchiae and the smooth rhinophores. Doridella obscura Verrill, 1870, D. baratariae (Harry, 1953), D. carambola Marcus, 1955, and D. burchi are related to one another; Franz (1967: 74) even united D. obscura and D. baratariae. The minute penial papilla of D. carambola and its 2 pairs of branchial glands sepa- rate this species from the others. The shape of the marginal teethcan hardly be used for specific separation, because their aspect varies when viewed from different angles. Both D. carambola and D. baratariae are smaller species than D. burchi. A preserved specimen of D. baratariae from Biscayne Bay, Florida, is 3 mm long and 2.25-2.4 mm wide. Its penis is 0.3 mm in length and does not have the dilated base (Marcus, 1960b: fig. 13) of D. burchi (Fig. 12). The vagina of D. carambola is narrow and the mar- gins of the rhinophoral pits are scalloped. The discrepancy in the length of the lateral and first marginal tooth is less in D. baratariae (Harry, 1953: fig. 5) than in D. burchi. The entire notal margin of Doridella batava (Kerbert, 1886; van Benthem Jutting, 1922: 400-401, figs. 5-6a, b) led Engel (1936: 106-107) to the correct generic allocation, but this species is still placed in the genus Corambe by some authors (e.g., Swennen, 1961: 205). Doridella obscura and D. batava are probably different species, because the former is pointed behind, while the latter is broadly rounded. But the data availa- ble are hardly sufficient for separating the remaining species of Doridella from D. batava. However, the very weak denticulation of the intermediate tooth and the broad irregularly shaped marginal teeth without cusps seem to be peculiar characters of D. batava. 6. Dendrodoris krebsii (Mörch, 1863) Rhacodoris Krebsii Mörch, 1863: 34. Doriopsis Krebsii Bergh, 1875a: 87- Si pli hes. 8=23: Doriopsis Krebsii var. pallida Bergh, 1879: 44-49. Doriopsis atropos Bergh, 1879: 49-64. Dendvodoris atropos Marcus, 1957b: 443-447, figs. 146-154; 1962, fig. 19. Dendrodoris krebsii Marcus, 1963: 35 (D. atropos synonymized). Dendrodoris atropos Collier & Farmer, 1964: 389-391, figs. 2 G-H and pl. 5. Dendrodoris krebsii Marcus, 1967a; 1967b: figs. 62, 63 (in press). Occurrence: Dredged of Sapelo Island, Georgia, September 12, 1963; 1 specimen. Further distribution: Atlantic Ocean: Florida; Bahamas; Virgin Islands (origi- nal localities); Antilles; Curacao; coast of southern middle Brazil. Pacific coast oí Lower California; Gulf of Cali- fornia; mainland of Mexico. At present, Sapelo Island and Puerto Peñasco, Sonora, Mexico, are the northernmost localities for D. krebsiz, whose range testifies to the central American marine continuity admitted even for the Lower Pliocene (Ekman, 1953: 37). The preserved slug is 30 mm long, 15 mm broad and 14 mm high. Its notum and foot have frilled borders. The rhinophore has about 20 leaves; there are 6 tripinnate gills. The surface of the notum and the borders of the rhinophoral and branchial pockets are smooth. Collier & Farmer found the base of the penis thickened in their east Pacific material. We also found this part to be thicker in specimens from the Gulf of California, collected by Dr. Peter E. Pickens, than in our Atlantic specimens from the Lesser Antilles and Brazil. This character could not be examined in the present specimen, because part of the hook-bearing section of the penial papilla is protruded, and only completely re- tracted penes are comparable. 7. Doriopsilla pharpa Marcus, 1961 Doriopsilla pharpa Marcus, 1961b: 146, figs. 19-21. 1) Sapelo Occurrences: Georgia, 210 FIGS. 12-15. Fig. 12. Doridella burchi, 13-14. Tritonia bayeri misa, n. ssp. teeth of radula. Fig. 15. Scyllaea pelagica. Sound, November 30, 1961, 1 specimen; 2) Wallburg Creek, 18-23 m depth, February 20, 1962, 4 specimens; 3) Sapelo Island, 1-2 miles east of sea buoy, dredged from 18-23 m depth, 2 specimens. Further distribution: Beaufort, North п. Sp.. Fig. 13. Preserved slug, MARCUS AND MARCUS D )) Ny D ) ) 51) 2129) ‘ey ) SU 14 Е м tt, FIGS. dorsal view. Fig. 14. Inner Diagram of reproductive organs. Diagram of reproductive organs. Carolina. The biggest of the preserved speci- mens is 18 mm long, 10 mm wide and 4 mm high. The foot is 16 mm long, 7 mm broad. The living slugs were yellow, but in the preserved specimens this ground color had disappeared. How- OPISTHOBRANCHS FROM GEORGIA 211 ever, in preserved Specimens the dark brown chromatophores of the connective tissue, mentioned inthe first description, were visible. The present material agrees with that from Beaufort, М. C., except that the spermatheca is larger and the prostatic section of the efferent duct is considerably wider. These characters are functional and not of systematic value. The dark specks and the 12 rhinophoral perfoliations distinguish D. pharpa from the 2 other species of Doriopsilla found in American Atlantic warm waters, D. leia Marcus (1961b: 144) and D. areolata Bergh, 1880. The latter 2 species lack the specks, and have 8 (D. leia) and 20- 25 (D. areolata) perfoliations. More- over, D. leia is soft and smooth, D. pharpa firm and slightly bossed on the notum. The pedal commissure of D. leia is distinct, and in D. pharpa the pedal ganglia are contiguous. The notal bosses of D. areolata are more distinctly set off than those of D. pharpa, and the hindmost muscular section of the oesophagus (Marcus, 1962: fig. 18, zi)is only half as long. When the white epi- dermal net of D. areolata is present, this species is easily identified. But sometimes this net is absent, making identification more difficult, as was the case in the single specimen of D. areo- lata reported from the West Atlantic Ocean (Marcus, 1962: 472). In Dendrodoris and Doriopsilla, which suck their food, the anterior gut is so much modified that the term “oral tube” and “buccal bulb” are inadequate and should be replaced by “oral vestibule” and “pharynx” respectively. In both genera the oral vestibule iS more or less dilatable, and the anterior portion of the tubular pharynx often protrudes into the posterior part of the vestibule. In Dendrodoris, a bilobed posterior oral gland (ptyaline gland, Bergh), or a pair of glands, opens into the vestibule. In Doriopsilla such glands are absent. Where ptyaline glands have been errone- ously described for species of Doriop- silla, they are really the ductless lymphatic blood glands. The buccal ganglia lie far from the nerve ring in Dendrodoris, whose cerebro-buccal con- nectives are long, while they are apposed to the pedal ganglia in Doriopsilla. Other morphological characters fre- quently mentioned in the literature are subject to variations and therefore cannot be used as diagnostic taxonomic characters. They are: soft consistency of the body in Dendrodoris and a stiff, leathery one in Doriopsilla; a nearly smooth (Dendrodoris) or strongly warty notum (Doriopsilla); spicules scarce (Dendrodoris) and abundant (Doriop- silla). The pharyngeal or salivary glands cannot be used taxonomically in view of Pruvot-Fol’s (1952: 414) and our (Marcus, 1962: 474) negative search for them in east and west Atlantic material of the type species of Doriopstlla. 8. Tritonia (Candiella) bayeri misa, new subspecies (Figs. 13-14) Occurrences: OffSapelo Island, Geor- gia, 1) 31° 33 min. 30 sec. N, 799 37 min. 30 sec W, 77 m depth, August 4, 1962, 1 specimen; 2) 31° 26 min. 32sec. N, 79 42 min. 13 sec. W, 89-77 mdepth, August 4, 1963, 2 specimens (type locality). Holotype, UMMZ 230620; Paratype, UMMZ 230621. The preserved animals are 2.4, 3.0, and 3.5mm in length, the last oneis 2 mm broad without the short gills. The backis smooth. The foot is nearly as broad as the body, round and bilabiate in front, tapering behind. The cephalic veilis entire, not bilobed. There are 4 digitiform velar appendages (Fig. 13, va) between the grooved tenta- cles (+). The smooth rim of the rhinophore sheath bears a single process (rha). The 9 branchial tufts (g) on either side alternate in length, the longer gills dichotomize. The genital aperture (ga) lies under the 3rd right tuft, the anus (an) between the 4th and 5th, behind the middle of the body. The length of the jaws (j) is more than half that of the body. The masticatory 212 MARCUS AND MARCUS border is set with several rows of conical teeth. The radula (Fig. 14) com- prises 26 rows with 10 teeth per half- row (radular formula: 9.1.1.1.9). The median cusp of the tricuspidate rhachi- dian tooth (r) is a little larger than the lateral’ cusps. The intermediate tooth (it) has an inner concavity and a row of outer denticles. The outermost of these is stronger thanthe others. The following teeth (lt) are hook-shaped, but the 1st of them (lt 1), sometimes also the 2nd, (lt 2), bears a small denticle between cusp and base. In spite of the small size of the slugs, their reproductive organs contained mature sperms. The genital system agrees with that of T. (C.) bayeri bayeri Marcus, 1967a, found in the area of Miami. The longish form of the ampulla (curved, sausage-shaped) differs from the globular one in T. (C.) b. bayeri. The name of this subspecies is the latinized form of the French Mise, an abbreviation of Marquise, wife of a Marquis. Discussion of Tritonia (Candiella) bayeri misa The new form has 4 velar appendages against 2 in the larger T. (C.) b. bayeri, which is preserved 7-11 mm in length. Since, in the tritoniids, the number of these appendages 15 known to in- crease with growth, the larger number in the smaller form is a distinctive character. Minor peculiarities of T. (С.) b. misa are the strong 1st denticle of the intermediate tooth (it) and the occasional occurrence of denticles onthe 2nd lateral tooth (lt 2). The (longish, not globular) shape of the ampulla is a functional character withno systematic importance. 9. Scyllaea pelagica Linné, 1758 (Fig. 15) Alder & Hancock, 1848: family 2, pl.5 (anatomy); 1855: pl. 46, suppl. fig. 27 (radula); Bergh, 1875b: 319-342 (including the varieties marginata, ghomfodensis, sin- ensis, ovientalis), pls. 40, 42-43, 44, figs. 1-18, pl. 45, figs. 16-18; Odhner, 1936: 1097 (color after Verrill, 1878), 1098 (synopsis of species of Scyl- laea), figs. 7, 30, 31a; Baba, 1949: 89, 168-169, figs. 112-113, pl. 36, fig. 130 (colored); Pruvot-Fol, 1954: 367-368, figs. 143a- J; Marcus, 1963: 36-37, figs. 65-66; Abe, 1964: 87, pl. 29, fig. 101 (color- ed). Occurrence: Off Georgia coast in Gulf- weed drift, 319 04 min. N, 80° 28 min. W, 4 specimens. Further distribution: Pelagicinwarm and warm-temperate waters, clinging to floating seaweed and feeding on hydroids. Occasionally farther north (Marcus, 1961b: 148). Living slugs reach 60 mm in length (Barnard, 1927: 210) when extended; the largest preserved specimen at hand is 30 mm long, 16 mm high including the lobes, and 8 mm broad. As 2 figures (Odhner, 1936; Baba, 1949) of the reproductive organs of S. pelagica are published in papers not easily accessible, and the reproduction of the 3rd figure (Pruvot-Fol, 1954) is mediocre, we give a new diagrammatic drawing (Fig. 15) of this peculiar system. The hermaphrodite glands (о) are globular. The specimen we dissected had 3 of these glands, but up to 6 have been recorded (Baba, 1949). The her- maphrodite ducts (h) unite, so that a Single duct enters the tubular, coiled ampulla (a). The short spermoviduct (Spo) goes into the female gland mass (albumen gland, ag), in which the male and female ducts separate. The male duct is glandular, prostatic (pr) in its inner, and muscular in its outer course (e). The outermost part, the ejaculatory duct, winds through a muscular, unarmed, conical penis (p) lodged in a narrow male atrium. The wide vagina (v) leads to the spermatheca (spa) whichis small, though bigger than in Odhner’s figure (1936: OPISTHOBRANCHS FROM GEORGIA 213 fig. 30). It contains debris, probably remains of sperm and male Secretion, so that it is not functionless (loc. cit.: 1068). The chambered spermatocyst (sp) is a small organ, apposed to the albumen gland (ag). It is connected with the gland mass by a short insemination duct (id) near the entrance of the sperm- oviduct (spo). Some folds of the glandular oviduct project as a spiral over the surface of the mucus gland (mu). The genital apertures lie between the right rhinophore and the first dorsal lobe on the side of the body. 10. Armina wattla, new species (Figs. 16-20) Occurrences: Sapelo Island, Georgia, 11-19.5 miles from sea buoy, 16.5-19 m depth, January 31, February 13 and March 13, 1961; a total of 6 specimens. Holotype, UMMZ 230622; Paratype, UMMZ 230623. The largest of the animals was 24 mm long, 15 mm broad and 8 mm high. The sole measured 20 mm in length, 9 mm in width. The smallest slug measured 15 mm. The preserved slugs were whitish with black pigment inthe furrows between the notal ridges, on the base of the rhinophores, in the folds of the caruncle, in the middle of the veil, on the sides of the foot, and on the sole. Preserved, the crests of the notal ridges are white, but they may havebeen yellow alive, as some vestiges indicate. There are about 36 notal ridges (nr) in the biggest animal, which run para- llel to the mid-line. Broad and narrow ridges generally alternate on the sides, while in the middle the narrow ridge is often absent. On the anterior border of the notum there begin 20-24 broad and narrow ridges, the rest originate farther behind. In front (Fig. 16) the notum is frilled by the ridges and notched in the middle; it is pointed behind. The pores of the marginal glands, Bergh’s cnidopores, are numerous, but in most Specimens they are difficult to see. There are about 28 branchial leaves (Fig. 17, g), the innermost of which lie in an open pocket over the viscera; 3-4 lateral lamellae (1) arise from the branchiae. Farther behind there are 18-22 or, when all primordia are counted, up to 29 lamellae. They all run obliquely outwards, the posterior ones in a more pronounced way. The semilunar veil (Fig. 16, ve) is as broad as the foot, its corners are bent upwards. Its upper border is weakly undulate. The dark middle of the veil contrasts with the colorless margins. Nuchal papillae are not developed, but a strong caruncle (c) arises from the upper or posterior border of the veil. The caruncle is folded transversely, is concave behind, where it borders the rhinophoral pits, and ends with a point on either side of the rhinophores (rh). These have 12 longitudinal leaves which are confluent on the tip and further divided downwards. The anterior border of the foot (ft) is bilabiate and notched; its corners are slightly prominent and rolled upwards. The pedal gland (ftg, Fig. 17) is marked by a furrow 5 mm long. The shape of the buccal lip (m) varies, being either trapezoid or elliptic. The genital aperture (ga) lies under the gills, the anus (an) behind the middle or in the 2nd third of the body. The renal pore (rp) is between the anal and genital Openings, about equally distant from both. The yellow jaws are small, about 3x 1.5 mm; the masticatory process is undulate after treatment with KOH. The cutting edge has 3-4 rows of denticles in front. These denticles look like corn (maize) on the cob. The rows are more numerous in the rear; on the free pro- cess there are about 10 rows of pointed denticles measuring up to 40y in length. The radula (Fig. 18) comprises 40 rows with about 46 lateral teeth (lt) per half- row. The rhachidian toothis 140u broad, 90u high. The median cusp is flanked by 5-7 denticles, 1-2 of which sit on the central cusp. The intermediate tooth (it) 214 MARCUS AND MARCUS FIGS. 16-20. Aymina wattla, n. sp. Fig. 16. Anterior end of animal, frontal view. Fig. 17. View from the right side. organs. Fig. 20. Two everted penes. has a broad base and about 7 outer denticles. Furthermore there are 1-2 big inner points which lie farther behind than the outer denticles. They are difficult to see, because they are over- lapped by the cusp. The 3 first lateral teeth may bear up to 4, exceptionally 5, Fig. 18. Inner teeth of radula. Fig. 19. Diagram of reproductive denticles. There are, however, many half-rows without any denticles. The lateral teeth increase in size towards the middle of the half-row and then de- crease outwards. The hermaphrodite duct (Fig. 19, h) is rather short, the ampulla (a) globular. OPISTHOBRANCHS FROM GEORGIA 215 The short spermoviduct (spo) bifurcates into the inner oviduct (io) and the sperm duct. The latter begins as a long, winding efferent duct (e) followed by a prostatic portion (pr). The ectal, 3rd part is con- voluted and thin; it reachesthe muscular penis (p). The penis is lodged in the male atrium (ma); in 2 specimens it was protruded. The shape ofthe everted male organs (Fig. 20), conical in one animal, cylindrical in the other, shows that it cannot be used as a Specific character. The inner oviduct (io) passes into the inner portion of the glandular oviduct, the albumen gland. This organ is simplified in the diagram (Fig. 19, ag); it is tubular as in the species examined previously (Marcus, 1960a, fig. 67; 1961a, figs. 148, 154). The mucus gland (mu) is wide and richly folded. The long vagina (v) leads from the external aperture to an ample, spherical seminal receptacle, the spermatocyst (sp). From there the sperm descend again, enter the nidamental duct (n) and passinwards to the inner oviduct, where the eggs are fertilized (Marcus, 1960a, fig. 67, f). The broad bicuspid caruncle (wattle) suggested the name of this species. Discussion of Armina wattla The only previously known Armina of the Atlantic coasts of the Americas is А. mülleri (у. Ihering, 1886: 223-230, pl. 9, fig. 1) from Santa Catarina, Sdo Paulo, and north of Rio de Janeiro (Marcus, 1960a: 170; 1967a). Evidently Nijssen-Meyer’s specimen from Sur- inam also belongstothat species; differ- ences she mentioned (1965: 149) can be considered as intraspecific variations. For the intermediate tooth Nijssen- Meyer (: 148) indicates: “...at least 6 denticles on the inner side of the cusp”. As her figure 4 shows, this is a lapsus for “outer side”. A. mülleri has 2 small but recognizable caruncles and a median boss between them. Therefore we can not unite it with A. semperi (Bergh, 1866: 37-42, pl. 3), whose caruncle is so minute (:39) that it is almost invisible on the cited figure 1. Pruvot-Fol (1933), the only one of the later authors who dealt with A. semperi and mentioned the caruncle, also called it “presque nulle”. The original locality of A. semperi lies on the southwestern coast of Mindanao; it has been further re- ported from Japan, the Arabian Sea (Gulf of Oman), the Gulf of Aden, and the northern Red Sea. Armina mülleri differs from A.wattla by the shape of the caruncle, which in the former, consists of 2 swellings with- out folds that are separated by a median boss. The rhachidian tooth of A. mülleri has a width ranging from 0.2 mm ina preserved slug 31 mm long, to 0.25 mm in preserved animals 39 and 16 mm long, against 0.14 mm in a 24 mm specimen of A. wattla. In the latter species, the lateral denticles of the rhachidian tooth are a little more numerous. In Nijssen-Meyer’s and in our material of A. mülleri the reduction of the denticu- lation of the lateral teeth is less pro- nounced than in A. wattla (see Fig. 18). However, v. Ihering’s description of A. mülleri does not show this difference. A species of the Pacific South American coast, Armina cuvieri (d’Orbigny, 1837: 198, pl. 17, figs. 1-3) from Valparaiso is practically unknown; its pyriform male copulatory organ cannot be evaluated, because the above description of A. wattla as well as the literature (у. Ihering, 1886: 225; Marcus, 1960а: 173; 1961a: 44) show that the shape of the penis is variable, at least in preserved animals, probably due to contraction. In an earlier exposition (Marcus, 1961a: 44) we discussed the 4 following species of Armina from the west coast of North America and indicated their bibliography: A. californica (Cooper, 1862), A. vancouverensis (Bergh, 1876), A. columbiana O’Donoghue, 1924, and A. digueti Pruvot-Fol, 1955. Lance (1962b: 51-54) has since described Armina con- volvula from the northern part of the Gulf of California, and has recognized the isolated position of that species. In 216 MARCUS AND MARCUS fact, A. convolvula belongs to Histiomena Môrch, 1859, known from the Pacific coast of Nicaragua (Marcus, 1966: 189). While Bergh (1881b: 172), v. Ihering (1886: 226), Eliot (1905: 238), and Pruvot- Fol (1933: 114) did not admit a broad intra-specific variability of the radula in Armina, Steinberg (1963: 65) does. She unites the 4 species of Armina of the North American Pacific coast from Panama to Vancouver Island under the oldest name. Her opinion will probably be accepted, though a comparison of the reproductive organs of several speci- mens is still desirable. For our pur- poses, i.e. the distinction of A. wattla from the warm temperate western Atlantic, which has so many faunal re- lationships with the eastern Pacific, it will be sufficient to note the weak caruncles of the 3 first Pacific species. As for A. digueti Pruvot-Fol, 1955, whose caruncle is not described, it differs from A. wattlaby its coarse white ridges, among the broad interspaces of which there course thin black ridges. Comparing the further species of Ar- mina with A. wattla, we found a strongly developed caruncle withtransverse folds in A. {igrina Rafinesque, Bergh’s Pleuro- phyllidia undulata Meckel, 1823 (1866: 18-19, pl. 1). This species is recorded from the western, central (Sargasso Sea) and eastern warm and warm temperate Atlantic Ocean (for range see Marcus, 1966: 191). The radula of A. tigrina differs widely from that of A. wattla, especially by the high and narrow rhachi- dian tooth with 15-30 lateral denticles on either side of the median cusp. Another species that should be compared with A. wattla is A. natalensis (Bergh, 1866: 34; Barnard, 1927: 213) from the coast of Natal. It has a similar strong car- uncle, but its rhachidian tooth (Bergh, 1866: pl. 6 B, fig. 7) is very broad, and the number of lateral lamellae is much higher than in A. wattla. Lateral teeth nearly without denticles, asin A. natalensis, occur also in several other species (Bergh, 1907: 102-103), but all of these have small caruncles, or (Baba, 1949: 162, A. major) longitudinal ridges on the veil. 11. Fiona pinnata (Eschscholtz, 1831) Marcus, 1961a: 50-51, figs. 173-179, references, distribution, description; Bayer, 1963: 460-465, figs. 5-7, be- havior, feeding, growth and reproduction. Occurrences: Off the Georgia Coast in Gulf Stream Drift, 319 01 min. М, 790 52 min. W, May 1, 1962; numerous specimens together witha pre-adult male of the pycnogonid, Anoplodactylus brasiliensis Hedgpeth, 1948 (: 222, 224). Further distribution: Pelagic and gregarious in warm and temperate seas, original locality: Sitka, Alaska, on a piece of wood washed ashore. Dr. Wolfram Noodt and Rudolf Róttger col- lected this species and its egg masses on floating feathers with barnacles about 250 km off Peru in October, 1965 while on board the research ship “Anton Bruun”. The largest of the preserved speci- mens at hand is 17 mm long, which corresponds to the maximum length known of living animals, 25 mm. 12. Dondice occidentalis (Engel, 1925) Caloria occidentalis Engel, 1925: 73- 76, figs. 7-15; Dondice occidentalis Marcus, 1958b: 62-65, figs. 97 (:54), 98-104 (: 63); 1960a: 186-187, figs. 87-90; 1963: 48; Edmunds, 1964: 27-28. Occurrences: Georgia, 1) Sapelo Sound, 16-26 m depth, April 16, 1963, 2 specimens; 2) 17 1/2-15 1/2 mi. 102° from Sea buoy, 15 m depth, 11 speci- mens. Further distribution: Beaufort, North Carolina; Miami, Florida; Jama- ica; St. Martin, Bonaire, Lesser Antilles; Guanta, Venezuela; Sao Paulo, Brazil. The preserved specimens reach20 mm in length. The jaws are covered witha OPISTHOBRANCHS FROM GEORGIA 217 black epithelium. The radula has 17 teeth (radular formula: 0.1.0), whose median cusp is flanked by 4-7 denticles. As in Edmunds’ material, the gonopores lie immediately behind the archof cerata from the anterior liver, at the front of the interhepatic space. ZOOGEOGRAPHIC REMARKS Two species of the present collection, Scyllaea pelagica and Fiona pinnata, are widely distributed pelagic species and occur in all seas of middle and low latitudes. Seven are littoral species peculiar to the warm western Atlantic region, which extends from Cape Hatteras to southern Brazil, probably to northern Santa Catarina. The new form Tritonia bayeri misa belongs to this West Indian group, because it is related to J. b. bayeri from the Miami area, Florida. Doridella burchi is near the widely distributed D. obscura, whose range extends, according to Franz (1967: 75), from Massachusetts to Texas. Only 1 species of the present West Indian ele- ment, Dendrodoris krebsii, isalsoknown from the tropical west coast of North America. Relicts of the Tethys Sea, which existed up to Middle Tertiary times, are Doris verrucosa and Okenia sapelona. The former is known from South Carolina to Santa Catarina, and from the British Isles to South Africa. The latter is related to a species, O. mediterranea, known from Naples and the French Mediterranean coast. One species of the present collection cannot be allotted to any of these groups: Armina wattla. It differs by its caruncle from the solely American species of the genus, and is separated from the re- maining species of Armina by its radula. ACKNOWLEDGEMENTS Grateful acknowledgement is made to Dr. J. B. Burch, Museum of Zoology, University of Michigan and the Marine Institute, University of Georgia for making available for study the speci- mens described inthis paper. In addition, Dr. Burch is to be thanked for providing drawings and detailed annotations of living Okenia and Doridella used in the present descriptions. А note of gratitude also goes to Mr. Milton S. Gray of the Marine Institute for collecting and pre- Serving the majority of the specimens discussed in this work. LITERATURE CITED ABBOTT, R. T., 1952, Twonew opistho- branch mollusks from the Gulf of Me- xico belonging to the genera Pleuro- branchaea and Polycera. Florida State Univ. Stud., 7: 1-7, pls. 1-2. ABE, T., 1964, Opisthobranchia of Toyama Bay and adjacent waters. Hokuryu-Kan, Tokyo. ix + 99 pp., 36 pls. ALDER, J. & HANCOCK, A., 1845-1855, A monograph of the British nudi- branchiate Mollusca, with figures of all the species. The Ray Society, London. Parts 1-7: 438 p, 84 pls. BABA, K., 1949, Opisthobranchia of Sagami Bay. Iwanami Shoten, Tokyo. 194 + 13 p, 50 pls. 1960, The genera Okenia, Goniodoridella and Goniodoris from Japan. Publ. Seto Marine Biol. Laborat., 8(1): 79-83, pls. 7-8. BALCH, F. N., 1899, List of marine Mollusca of Coldspring Harbor, Long Island, with descriptions of one new genus and two new species of nudi- branchs. Proc. Boston Soc. nat. Hist., 29(7): 133-162, pl. 1. BARNARD, K. H., 1927, South African nudibranch Mollusca, with descrip- tions of new species; and a note on some specimens from Tristan d’Acunha. Ann. South African Mus., 25(art. 6): 171-215, pls. 19-20. BAYER, F. M., 1963, Observations on pelagic mollusks associated with the siphonophores Velella and Physalia. Вас mar: Sei: ¿Gulf “€: Carib;, (13 (3): 454-466. BENTHEM JUTTING, У. 5. S. van, 1922, Zoet- en Brakwatermollusken (p 391- 218 MARCUS AND MARCUS 410, figs. 1-15). In: Redeke, H. C., Flora en Fauna der Zuiderzee, viii + 460 р, 2 pls. Te Helder. BERGH, R., 1866, Bidrag til en mono- graphi af pleurophyllidierne. Natur- hist. Tidskr., ser. 3, 4(1-2): 1-80, pls. 1-4; 207-380, pls. 5-9. 1869, Bidrag til en monografi af phyllidierne. Naturhist. Tidskr., ser. 3, 5: 358-542, pls. 14-15. 1871, Beiträge zur Kenntniss der Mollusken des Sargassomeeres. Verh. zool. bot. Ges. Wien, 21: 1272- 1308, pls. 11-13. 1875a, Neue Nacktschnecken der Südsee, Ш. J. Mus. Godeffroy, 8: 53-100, pls. 7-11. 1875b, Malacologische Unter- suchungen. Heft 8. т: Semper, C., . Reisen im Archipel der Philippinen. Wissenschftl. Resultate, 2. Theil, 1: 315-343, pls. 40-45. 1878, Idem, Ней 13. Ibid. 2: 547-601, pls. 62-65. 1879, Die Doriopsen des atlantischen Meeres. Jahrb. malak. Ges., 6: 42-65, no plates. 1880, Malacologische Unter- suchungen, 4. Jn: Semper, C., Reisen im Archipel der Philippinen. Wissen- schaftliche Resultate, 2. Theil, Suppl. zu 2, 1: 1-78, pls. A-F. 1881a, Idem, 4. Ibid., Suppl. zu 2, 2: 79-128, pls. G-L. 1881b, Beiträge zur Kenntniss der japanischen Nudibranchien, I. Verh. zool. bot. Ges. Wien, 30: 155- 200, pls. 1-5. 1894, Die Opisthobranchien. Part 13 of the reports on the dredging operations by the “Albatross”. Bull. Mus. comp. Zool. Harv., 25(10): 125- 233, pls. 1-12. 1904, Malacologische Unter- suchungen, 6, 1. Abt. Nudibranchiata. In: Semper, C., Reisen im Archipel der Philippinen. Wissenschaftliche Resultate, 9: 1-56, pls. 1-4. 1907, The Opisthobranchiata of South Africa. Trans. South African phil. Soc., 17(1): 1-144, pls. 1-14. COLLIER, C. L. & FARMER, W. M., 1964, Additions to the nudibranch fauna of the east Pacific and the Gulf of California. Trans. San Diego Soc. nat. Hist., 13(19): 377- 396, pls. 1-6. COOPER, J. G., 1862, On some new genera and species of California Mollusca. Proc. Calif. Acad. nat. Sci., 2: 202-207. CUVIER, G., 1804, Mémoire sur le genre Doris. Ann. Mus. Hist. nat., 4: 447- 473, pls. 73-74. EALES, N. B., 1960, Revision of the world species of Aplysia. Bull. Brit. Mus. (Nat. Hist.), 5(10): 267-404; frontispiece. EDMUNDS, M., 1964, Eolid Mollusca from Jamaica with descriptions oftwo new genera and three new species. Bull. mar. Sci. Gulf & Carib., 14(1): 1-32. EKMAN, S., 1953, Zoogeography of the sea. Sidgwick & Jackson, London. xiv + 417 p. ELIOT, С.М. E., 1905, Nudibranchs from the Indo-Pacific: I. J. Conchol., 11: 237-256, pl. 5. 1906, Notes on some British ” nudibranchs. J. mar. Biol. Ass. U.K., n.! ser.;. 7(3):0333-382 pme; 1910, A monograph of the British nudibranchiate Mollusca. The Ray Society, London, Pt. 8 (Supple- mentary), 198 p., 8 pls. ENGEL, H., 1925, Westindische opistho- branchiate Mollusken. I. Bijdr. Dierk., 24: 33-80. 1936, The Netherlands nudi- branchiate Mollusca. Zool. Mededeel., 19: 103-116. FRANZ, D. R., 1967, On the taxonomy and biology of the dorid nudibranch Doridella obscura. Nautilus, 80(3): 73- 19. HAMATANI, I., 1961, Preliminary account of a new species of Okenia from Osaka Bay, Japan. Publ. Seto Marine Biol. Laborat., 9(2): 363-365. HARRY, H. W., 1953, Corambella bara- tariae, a new species of nudibranch from the coast of Louisiana. Occasion. Pap. Marine Laborat. Louisiana State Univ., no. 8: 1-9, figs. 1-7. HEDGPETH, J. W., 1948, The Pycno- gonida of the western North Atlantic and the Caribbean. Proc. U.S. Natl. OPISTHOBRANCHS FROM GEORGIA 219 Mus., 97(3216): 157-342, 53 figs., 3 charts. HOFFMANN, H., 1932-40, Opisthobran- chia. Bronn’s Klassen..., 3: Mollusca: Abt. Il: Gastropoda. Akadem. Ver- lagsges. Leipzig. Buch 3. Teil I: XI 21247 p, 1 phe Tels 1:90 p. IHERING, H. v., 1886, Zur Kenntniss der Nudibranchien der brasilianischen Küstern. Jb., malak. Ges., 13: 223- 240, pl. 9. 1915, der brasilianischen Ktisten. Bl. malak. Ges., 47: 133-143. JOHNSON, C. W., 1934, List of marine Mollusca of the Atlantic coast from Labrador to Texas. Proc. BostonSoc. nat. Hist., 40(1): 1-204. KERBERT, C., 1886, Over het geslacht Corambe Bergh. Tijdschr. Nederl. Dierk. Vereen., ser. 2, deel 1 (1885- 87), afl. 2(1886): 137-138. LANCE, J. R., 1962a, A new Siiliger and а new Corambella from the north- western Pacific. Veliger, 5(1): 33-38, pl, б. Die Opisthobranchien Nachr. - 1962b, A new species of Armina from the Gulf of California. Veliger, 5(1): 51-54. LANGE de MORRETES, F., 1949, Ensaio de catalogo dos moluscos do Brasil. Arq. Mus. Paranaense, 7: 5-216. MacFARLAND, F. M. & O’DONOGHUE, C. H., 1929, A new species of Cor- ambe from the Pacific coast of North America. Proc. Calif. Acad. Sci., 18(1): 1-27, pls. 1-3. MACNAE, W., 1958, The families Poly- ceridae and Goniodorididae in Southern Africa. Trans. R. Soc. South Africa, 35(4): 341-372, pls. 17-18. MARCUS, Er. (or Ev. € Er.), 1955, Opisthobranchia from Brazil. Bol. Fac. Filos. Ciénc. Univ. Sáo Paulo, Zoologia, 20: 89-262, pls. 1-30. 1957a, Sea-hares and side- gilled slugs from Brazil. Bol. Inst. Oceanogr. Sao Paulo, 6(1955): 3-49, pls. 1-8. 1957b, On opisthobranchia from Brazil (2). J. Linn. Soc. (Zool.), 43(292): 390-486. thobranchia. lusks from California. North Carolina. the Lesser Antilles. the Gulf of California. 1958a, Notes on Opistho- branchia. Bol. Inst. Oceanogr. Sdo Paulo, 7(1956): 31-78, pls. 1-8. 1958b, On western Atlantic opisthobranchiate gastropods. Amer. Mus. Novit., 1906: 1-82. 1959, Lamellariacea und Opis- Lunds Univ. Aarsskr., N. F., Avd. 2, 55(9): 1-135. 1960a, Opisthobranchs from American Atlantic warmwaters. Bull. mar. Sci. Gulf & Carib., 10(2): 129- 203. 1960b, Some opisthobranchs from the northwestern Gulf of Mexico. Publ. Inst. mar. Sci. Univ. Texas, 6(1960): 251-264. 1961a, Opisthobranch mol- Veliger, 3 (Suppl. 1): 1-85, pls. 1-10. 1961b, Opisthobranchia from J. Elisha Mitchell sci. Soc., 77(2): 141-151. 1962, Opisthobranchs from Florida and the Virgin Islands. Bull. mar. Sci. Gulf & Carib., 12(3): 450- 488. 1963, Opisthobranchs from Stud. Fauna Curacao, 19(79): 1-76. 1966, Opisthobranchs from tropical West Africa. Stud. trop. Oceanogr., Miami, 4(1): 152-208. 1967a, Tropical American opisthobranchs. (In press, #64.) 1967b, Opisthobranchs from (In press, ibid.) MORCH, O. A. L., 1863, Contribution a la faune malacologique des Antilles Danoises. J. Conchyl., 11: 21-43. NIJSSEN-MEYER, J., 1965, Notes ona few opisthobranch Mollusca from Suri- nam (Guianas). Zool. Mededel., 40(19): 143-150. NOBRE, A., 1938-40, Moluscos marin- hos e das aguas salobras. Fauna malacologica de Portugal. Comp. ed. Minho Barcelos, Pôrto, xxxii + 808 p, 106 pls. ODHNER, N. H., 1936, Nudibranchia Dendronotacea, a revision of the sys- 220 MARCUS AND MARCUS tem. Mem. Mus. Roy. Hist. Nat. Belg., ser. 2, fasc. 3(Mélanges Pelseneer): 1057-1128, 1 pl. O’DONOGHUE, C. H., 1929, Report on the Opisthobranchiata. Res. Cam- bridge Exped. Suez Canal 1924. Trans. zool. Soc. Lond., 22, pt. 6(17): 713-841. ORBIGNY, A. d’, 1835-46, Voyage dans l’Amérique Méridionale, 5, 3: Mol- lusques, xliii + 758 p, Atlas, 85 pls. OSBURN, R. C., 1932, Bryozoa from Chesapeake Bay. Ohio J. Sci., 32(5): 441-446, pl. 1. 1944, A survey of the Bryozoa of Chesapeake Bay. Dept. Res. Educat. State of Maryland. Board of Natur. Resources, Publ. 63: 1-59, pls. i-5. PRUVOT-FOL, A., 1933, Opistho- branchiata. Mission Robert Ph. Dollfus en Egypte. Mém. Inst. Egypte, 21: 89-159, pls. 1-4. 1952, Compléments à la con- naissance anatomique de Doriopsilla areolata Bergh. Bull. Soc. zool. Fr., 77(5-6): 411-414. 1954, branches. Mollusques opistho- Faune Fr., 58. Paul Lechevalier, Paris. 460 p, 1 pl. 1955, Les Arminiadae. Bull. Mus. Hist. nat. Paris, ser. 2, 27(6): 462-468. STEINBERG, J. E., 1963, Notes on the opisthobranchs of the west coast of branchiata. In: North America, ill, IV. Veliger, 6(2): 63-73. SWENNEN, C., 1961, Data on distribution reproduction and ecology of the nudi- branchiate molluscs occurring in the Netherlands. Netherl. J. Sea Re- search, 1(1-2): 191-240. THIELE, J., 1910, Molluskenfauna West- indiens. Zool. Jb. Syst. Suppl. 11(2): 109-132, pl. 9. 1931, Handbuch der sys- tematischen Weichtierkunde, 1. Gustav Fischer, Jena. vi + 778 p. VAYSSIERE, A., 1919, Recherches zoologiques et anatomiques sur les mollusques opistobranches du Golfe de Marseille. 2me suppl. Ann. Mus. Hist. nat. Marseille, Zool., 17: 53- 92, pls. 4-6. 1922-34, Gasteropoda Opistho- Joubin, L. (éd.), Faune et Flore de la Méditerranée. Paris, Invertebrata, 2 (pages and plates not numbered). VERRILL, A. E., 1873, Report upon the invertebrate animals of Vineyard Sound and the adjacent waters. Rep. U. S. Comm. Fish Fisheries, for 1871-1872, suppl. 18: 295-778, 38 pls. 1901, Additions to the fauna of the Bermudas from the Yale Expedition 1901. Trans. Connecticut Acad. Arts Sci., 11(1,: art. 2): 19-624 PS: ADDENDUM The disposition of the holotype and paratype specimens of the opistho- branchiate mollusks described in Marcus & Burch (1965, Marine euthyneuran Gastropoda from Eniwetok Atoll, western Pacific, MALACOLOGIA, 3(2): 235-262) is as follows [catalogue numbers are those of the Museum of Zoology, University of Michigan, Ann Arbor, Michigan, U. $. A.] Haminoea musetta Marcus € Burch, Holotype, UMMZ 230624 Haminoea musetta Marcus € Burch, Paratypes, UMMZ 230625 Haminoea linda Marcus € Burch, Holotype, UMMZ 230626 Haminoea linda Marcus € Burch, Paratypes, UMMZ 230627 Chromodoris briqua Marcus € Burch, Holotype, UMMZ 230629 Herviella mietta Marcus & Burch, Holotype, UMMZ 230630 Herviella mietta Marcus & Burch, Paratype, UMMZ 230631 Onchidella evelinae Marcus & Burch, Holotype, UMMZ 230632 Onchidella evelinae Marcus & Burch, Paratype, UMMZ 230633 OPISTHOBRANCHS FROM GEORGIA RESUMEN ALGUNOS OPISTOBRANQUIOS DE LA ISLA SAPELO, GEORGIA, ESTADOS UNIDOS E. Marcus and E. Marcus Este trabajo trata sobre 2 opistobranquios pelagicos, y otros 10 de las zonas entre mareas y Sub-mareas, de la costa sudeste de Estados Unidos. Se describen cuatro especies: Okenia sapelona, Doridella burchi, Tritonia (Candiella) bayeri misa, у Armina wattla. La última especie difiere de la única conocida Armina americana рог su fuerte carúncula, y de undulata y otras especies por su rádula. Okenia sapelona se asemaja a O. mediterranea, y por lo tanto pertenece a los opistobranquios cuya existencia puede trazarse desde el Mar Tethys del Terciario. Las dos restantes son de la región de aguas cálidas del Atlántico occidental. Las especies litorales de la presente colección habitan también aquellas aguas, con excepción de Doris verrucosa que aparece tambien en el Atlántico oriental. La subespecie Pleurobranchaea hedgpethi hamva se elimina, porque en el presente material, la dirección del plegado sobre los orificios genitales es con frecuencia oblícua, no dorsal (P. h. hamva) ni tampoco anterior (P. h. hedgpeth:). ZUSAMMENFASSUNG UBER EINIGE OPISTHOBRANCHIER VON DER SAPELOINSEL, GEORGIA E. and E. Marcus Die Arbeit behandelt 2 pelagische Opisthobranchier und 10 aus der Gezeitenzone und dem Sublitoral von der Sapelo Insel, Georgia. Neu sind: Okenia sapelona, Dori- della burchi, Tritonia (Candiella) bayeri misa und Armina wattla. Die letzte unter- scheidet sich durch die starke Karunkel von den rein amerikanischen Armina-Arten sowie durch die Radula von A. undulata und den anderen Arten mit starker Karunkel. O. sapelona ähnelt der O. mediterranea, gehört also zu den Opisthobranchiern, deren Verbreitung auf das tertidre Tethysmeer zurückgeführt werden kann. Die 2 Übrigen neuen Formen sind mit Arten der westatlantischen Warmwasserregion verwandt. Gleichfalls Bewohner dieser Region sind die bekannten litoralen Arten der vorliegenden Sammlung, mit Ausnahme von Doris verrucosa, die auch im Ostatlantik vorkommt. Die Unterart Pleurobranchaea hedgpethi hamva wird aufgegeben, weilim vorliegenden Material der Fortsatz über den Geschlechtsöffnungen oft schräg gerichtet ist, d.h. weder nach oben (P. h. hamva), noch nach vorn (P. h. hedgpethi). RESUMO SOBRE ALGUNS OPISTOBRANQUIOS DE ILHA DE SAPELO, GEORGIA E. e E. Marcus O trabalho trata de dois opistobrânquios pelagicos edez litorais, da zona das marés e abaixo desta, da ilha de Sapelo, Georgia. Formas novas зао: Okenia sapelona, Dori- della burchi, Tritonia (Candiella) bayeri misa, e Armina wattla. A Ultima difere das 221 222 MARCUS AND MARCUS espécies puramente americanas de Armina pela carúncula forte e pela rádula de А. undulata e das outras espécies com carúncula forte. O. sapelona assemelha-se a O. mediterranea, рог isso pertence aos opistobranquios cuja distribuiçäo pode ser reconduzida ao mar Terciário da Tethys. Asduas novas formas restantes sáo aparen- tadas com espécies da regio quente do Atlántico ocidental. Também as espécies ja conhecidas da presente coleç4o sdo habitantes desta regido, com excecäo de Doris verrucosa que ocorre no Atlantico ocidental e oriental. A subespécie Pleurobranchaea hedgpethi hamva foi suprimida, porque a direç4o do lóbulo em cima das aberturas genitais 6 frequentemente obliqua, nem para cima (P. h. hamva), nem para diante (P. h. hedgpethi). АБСТРАКТ О НЕКОТОРЫХ OPISTHOBRANCHIA ИЗ РАЙОНА о. САПЕЛО (ДЖОРДЖИЯ, С. I. A.) OMAP RYIC UT 3. WMAP YC В работе рассматриваются 2 пелагических и 10 литоральных и сублиторальных видов Opisthobranchia из вод, омывающих BO- сточное побережье С. Ш. А. Описываются 4 вида: Okenta sapelona, Doridella burchi, Ттйота (Candiella) bayeri misa и Аттта wattla. Последний вид отличается от американского вида рода Armina сильно развитым карункулом, а от A. undulata и других видов, имеющих большой карункул - радулой. Okenia sapeloma похожа на О. mediterranea, следовательно относится к Opistho- branchia, распространение которых прослеживается, начиная с третичного моря Тетис. Остальные 2 новых Формы родственны видам из тепловодного района западной Атлантики. Все известные литоральные виды настоящей коллекции - также обитатели этого района, за исключением Doris verrucosa, который встречается также и в восточной Атлантике. Подвид Pleurobranchaea hedgpethi hamva - закрывается, поскольку в настоящем материале кожный вырост над генитальными отверстичми часто был направлен косо Т.е. ни дорзально (как y P. hedgpethi hamva), ни кпереди (как y Р. h. hedgpethi). ar a ви MALACOLOGIA, 1967, 6(1-2): 223-230 REVISION OF THE GENUS HERVIELLA (OPISTHOBRANCHIA: EOLIDACEA) Robert Burn! ABSTRACT Herviella Baba (1949) (Opisthobranchia: Eolidacea) is especially character- ized by a single row of cerata in the right liver, a penial stylet and a ‘serial’ spermatheca. Muessa Marcus (1965), with the same characteristics, is a syno- nym... Eight, species are known from the western Pacific Ocean including the new species H. burchi described in this paper. Two subgenera are distinguished; Herviella s.s., with the central radular cusp longer than the lateral denticles, contains the species H. yatsui (Baba) type species, H. affinis Baba, H. burchi sp. nov., H. evelinae (Marcus), H. claror Burn and H. exigua (Risbec); Mar- ciella subgen. nov. , with the lateral denticles nearly or as long as the central cusp, contains the species H. mietta Marcus (type species) and H. albida Baba. Noumeaella Risbec (1937) with similar genital characters and an arch of ce- rata in the right liver appears to be closely related to Herviella. The 2 genera form a distinct group within the family Favorinidae, and withsome reservations are placed together in a new subfamily, Herviellinae. Cleioproct eolids with a single row of cerata in the anterior or right liver are few in number and their known dis- tribution is restricted to the western Pacific Ocean. Eight species appear to be united by this taxonomically important anatomical characteristic. The current revision of both the generic and specific units involved is derived from the literature, examination of pre- served specimens of H. burchi and H. mietta, together with field notes on living Specimens of these 2 species, and a study of living specimens of H. claror. The writer is indebted to Dr. J. B. Burch, Museum of Zoology, University of Michigan, Ann Arbor, Michigan, U. S. A., for the opportunity to examine some of the Herviella material described by Dr. Ernst Marcus and himself (1965), as well as to study his field notes made at the time of collection. This research was undertaken while the writer was a recipient of a grant from the Sci- ence and Industry Endowment Fund, Commonwealth Scientific and Industrial Research Organization, Melbourne, Australia. THE GENUS HERVIELLA The primary generic unit involved in this revision is Herviella Baba (1949: 107, 180), the type of which, H. yatsui (Baba, 1930), is now known anatomically (Baba, 1966b). Based upon the type Species, it appears that the following characteristics are diagnostic for the genus: a single row of cerata in the right liver; anterior of foot expanded and rounded; rhinophores simple; jaws high anteriorly and narrow posteriorly, masticatory edge with a single row of denticles; penial stylet present; and female ducts with the spermatheca ‘serial’ (i.e., it is formed by a swelling of the oviduct or vagina). The recently constituted genus Muessa Marcus (1965: 282), type M. evelinae Marcus (1965: 283) described from a single very small preserved specimen, lonorary Associate in Conchology, National Museum of Victoria, Melbourne, Australia. (223) 224 R. BURN is similar to Herviella, except that the rhinophores and tentacles are annulate in the preserved Holotype, and the jaws are stated tobe oblongin shape. Annulate rhinophores and tentacles occur in a preserved specimen of dH. mietta examined for this revision, but field notes and published descriptions indicate that these appendages are smooth in life. Therefore, it is presumed that living Muessa have smooth rhinophores and tentacles. The oblong shape of the jaws depends upon how they are orientated for observation. In Muessa (Marcus, 1965: fig. 38) the jaws are Shown with the masticatory border in the horizontal position. If the figure is turned so that the upper anterior mar- gin of the jaw is in the vertical position, then when compared with the figure of the jaw of H. yatsui (Baba, 1966: pl. 1, fig. 4), it can be seenthat the differences are not objective. Consequently, Iregard Muessa to be identical with and a junior synonym of Herviella. As a result of this synonymy, there are 7 species that definitely can be assigned to the genus Herviella. An eighth species, Aeolidia exigua Risbec (1928: 245), in which the position of the anus is not known, is tentatively ascribed to Herviella (Burn, 1963: 18; Marcus & Burch, 1965: 251). These 8 species are sharply divided inthe shape of the radular teeth. In H. yatsui, H. affinis, H. burchi, H. evelinae, H. claror and H. exigua the median cusp is longer than the 3 to 5 denticles on each Side. Herviella mietta and H. albida have the outermost of the 4 to 9 denticles oneach Side nearly or as long as the median cusp and with the intermediary denticles shorter. These 2 species can be sepa- rated by this characteristic into a sub- genus Marciella subgen. nov., with H. mietta Marcus (1965) designated as the type species. The jaw of Herviella burchi is inter- FIG. 1. Specific differences of the species of Herviella. fine stippling, opaque white; heavy stippling, black; mediate in shape between those of H. yatsui, H. evelinae and Н. claror, and therefore it does not seem justified to distinguish the latter 2 species even subgenerically on jaw shapes. Marcus € Burch’s Herviella claror (1965: 251) differs from H. claror Burn (1963: 18) in colour pattern and the shape of the jaw and radular teeth. Here it is described as H. burchi sp. nov. The following characterizations of the species of Herviella are drawn from the literature unless otherwise stated. By means of line drawings, Fig. 1 shows specific differences as they occur in the colour pattern of the anterior portion of the body and the cerata (columns 1 & 2), the jaws and masticatory border (column 3), and the radular teeth (column 4). Subgenus Herviella s.s. H. yatsui (Baba, 1930). The lateral denticles of the radular teeth are shorter than the central cusp and they generally decrease in height toward the lateral margins. H. yatsui (Baba, 1930: 121; 1937: 328; 1949: 107, 180; 1966b: 1; Abe, 1964: 70). Japan; Pacific and Japan Sea coasts, common. Body yellowish white; black U-shaped pigmentation pattern on the head continuing on to the tenta- cles; black specks occur onthe median part of the back; rhinophores with a black band at their mid-length; cerata with an upper and lower cluster or ring of black spots and an opaque white band at the tip. Jaws high anteriorly, taper- ing sharply behind, concave dorsally, with 15-20 denticles. Radula with 18-25 teeth; central cusp short; 4-5 denti- eles on each side. Penial stylet with 3-6 spines along the concave side; spermatheca spherical. Like Herviella mietta and H. albida, the denticles of the masticatory borders are conical. The spines onthe penial stylet of H. yatsuz are unique Type species: Colours are indicated as follows: oblique hatching, red, orange or yellow. REVISION OF HERVIELLA Color pattern of Color pattern and Jaw shape (a) and anterior body shape of cerrata masticatory border (b) SO Los H. yatsui after Baba 1966b H. affinis after Baba 1966b H. burchi after Marcus and Burch 1965 (ceras drawn from field notes) H. evelinae after Marcus 1965 (Anterior body reconstructed) H. claror after Burn 1963 H. exigua after Risbec 1928 H. mietta after Marcus and Burch 1965 (Jaw drawn from own observations) H. albida after Baba 1966a 225 226 among the Eolidacea. H. affinis Baba (1960: 303; 1966b: 4; Abe, 1964: 71). Japan; Pacific and Japan Sea coasts, not common. Body yellowish-white, without U-shaped pigmentation pattern on the head, but instead with black specks covering the head, back, sides and lower half of the cerata; rhinophores with a black band at their mid-length; cerata with an orange ring below their tips. Jaws high anteriorly, tapering behind, dorsally concave; with 10-12 large oblique denticles. Radula with 13-14 teeth, central cusp long and wide, 3-4 denticles on each side. Penial stylet long and curved. Oblique or raking denticles occur also in Herviella burchi, H. evelinae and H. claror. All 4 species have black speckling on the body and 3-4 denticles on each side of the wide central cusp of the radula. Shape of the jaw and colour patterning, par- ticularly on the cerata, separate these 4 species. H. burchi sp. nov. (4. claror Marcus & Burch, 1965: 251, fig. 28-30; non H. clavor Burn, 1963: 18). Marshall Islands; Eniwetok Island, three speci- mens. Body white, without U-shaped pigment pattern on the head, with black specks and white spots on the back and side of the body (clear trans- verse areas occur between the cerata groups on opposite sides of the body); the lower part of the rhinophores and tentacles speckled with black pigment; cerata with an orange ring at the distal one-third and an opaque white band above and below this. Jaws high an- teriorly, tapering slightly behind, slightly convex dorsally; with 6 large inclined denticles. Radula with 11 teeth, central cusp broad and blunt, 3-4 lateral denticles on each side. Penial organ not known. The Holotype is a preserved speci- men 4.5 mm long in the collections of the Museum of Zoology, University of Michigan, (cat. no. 230634). A total of 3 specimens were collected by Dr. R. BURN William H. Heard, April 2-12, 1960. Only the Holotype was available for study, the other 2 specimens having been preserved for cytological studies. The new species differs from Her- viella claror Burn (and other species of the genus) in colour pattern, jaw Shape and the blunt central radular tooth. The ovoid shape of the jaws is especially distinctive. The species is named for Dr. J. B. Burch, who allowed me to examine the type specimen and his field notes, sketches and photographs made at Eniwetok Atoll. H. evelinae (Marcus, 1965: 283). Caro- line Islands; Ifaluk Island, one speci- men. Body yellowish (?) in life, with black specks on the back, head, tenta- cles and rhinophores. The jaws narrowed behind, with 8 large oblique denticles having rough edges. Radula with 14 teeth, central cusp broad with curved sides and 3-4 lateral denticles on each side. Penis with along stylet; the spermatheca exists only as a small dilation. The broad radular teeth, rough- edged masticatory denticles and the posteriorly narrowed jaws (referred to as ‘oblong’ by Marcus, 1965: 282) are the diagnostic characteristics of H. evelinae. A. clavov Burn (1963: 18). Australia; northern New South Wales (Woody Head), 2 specimens. Body white with black speckles on the back, head, tentacles and rhinophores; cerata with an orange band below the tip and black speckling on the anterior side. Jaws slightly narrowed behind; with 6 large oblique denticles having smooth edges. Radula with 13 teeth, central cusp long with straight sides and having 3 denticles on each side. Penis with a curved stylet. Herviella claror and H. evelinae have similarly shaped jaws, but those of H. claror are broader posteriorly and the masticatory denticles have smooth edges. The long taper of the central radular tooth is unlike that of REVISION OF HERVIELLA 227 any other Herviella. An orange band on the cerata occurs also in Я. affinis, and H. burchi, and a red one occurs on the cerata of H. exigua. H. exigua (Risbec, 1928: 245; 1953: 134). New Caledonia; Kouaoua Bay, 3 speci- mens. Body yellowish with minute black speckles grouped together to form greyish areas, tentacles and rhinophores with a black band in their middle portion, cerata with a redband below their tips. Jaws witha single row of strong denticles. Radula with about 12 teeth, the central cusp long and tapering and having 3 denticles of equal length on each side. The penial stylet is long and curved. The position of the anus is not known for Herviella exigua, therefore its placement in the genus Herviella is somewhat doubtful. The red band on the cerata and even height of the lateral denticles are the distinctive characteristics for the species. Subgenus Marciella subgen. nov. Type species: NH. mietta Marcus € Burch (1965). The marginal lateral denticles of the radular teeth are nearly or as long as the central cusp; intermediary lateral denticles are shorter than the marginal denticles. The subgenus is named for Dr. Ernst Marcus of Brazil who has added im- mensely to the knowledge ofthe Eolid- acea and other Opisthobranchia. H. mietta Marcus & Burch (1965: 251). Marshall Islands; Eniwetok Island, not uncommon. Body white below, black above, the pigmentation ex- tending on to the cerata (except at the tips), head, tentacles (dorso-median line) and rhinophores (middle third black, with a short pigmentation line below). Jaws high anteriorly, narrow behind and deeply concave dorsally; with at least 30 sharp denticles, the largest below (personal observation). Radula with 18 teeth, the central cusp short and pointed, with 8-9 thin denti- cles on each side, the outermost denti- cle as long as the centralcusp. Penial stylet curved. The jaws figured for H. mietia in Fig. 1 are from the Holotype (Univer- sity of Michigan, Museum of Zoology, cat. no. 230630); they measured 0.9 mm in both length and height. The Holotype is a specimen with heavy black pigment which, in the original description, is called the second colour type. The first colour type (Uni- versity of Michigan, Museum of Zoology, cat. no. 230631) has morpho- logically defective jaws by being deeply incised (Marcus & Burch, 1965: fig. 34; confirmed by personal obser- vation). Radular and other charac- teristics are in accord despite the fact that this first colour type has a light and transparent body, white granules and sometimes black pigment on the back, the head with a black pattern, the rhinophores with a black band and the cerata clear with yellow digestive glands. This is the most distinctive species of the genus. Theimportant diagnostic characters are carrot-shaped cerata that are predominently blackin colour, black pigment on the body, and 8-9 denticles on each side of the small central cusp. The jaws are unlike those of any other Herviella (see Fig. 1, column 3); the anterior margin is more erect and evenly convex, the masticatory border is deeper and bears more (about 30) denticles, the dorsal margin is deeply concave, and the posterior end is narrow and squarely truncate. Both H. mietta and H. albida have a somewhat similar pattern of black pigment on the head, otherwise they are quite different. It is doubtful that the next species, Herviella albida, should be classified in the subgenus Marciella with H. mietta. Except for the long marginal denticles of the radular teeth, Н. albida belongs to the subgenus Herviellas.s., as indicated by the fusiform cerata, the configuration of the anterior edge of the jaws and the long ‘legs’ of the radular tooth. Raising Marciella to a 228 R. BURN full genus may be justified when more knowledge is available, particularly concerning the reproductive organs. For the present, however, it is main- tained as a subgenus, solely to include the 2 species withlong marginal denti- cles. H. albida Baba (1966a: 361). Japan; Inland Sea (Seto, Kii), one specimen. Body yellowish-white with scattered white spots on the head and back, black pigment present in a U-shape pattern on the head, a pigment line on the tentacles, a pigment band on the rhinophores and in lines laterally between the groups of cerata; cerata with 2 bands of opaque white in their upper halves. Jaws high anteriorly, narrow behind, deeply concave dorsal- ly; with 16-18 pointed denticles. Radula with 20 teeth, the central cusp with a long taper, and with 3-4 denti- cles on each side and nearly as long as the central cusp. The penial stylet is short and curved. The other 2 Japanese species, Herviella yatsui and H. affinis, closely resemble H. albida in body shape, general body colouration and shape of the jaws. Herviella albida is separated from these 2 species by the long mar- ginal lateral denticles of the radular teeth. Fewer lateral denticles (3-4) and much less black pigment dis- tinguish H. albida from H. mietta. DISCUSSION Herviella shows an unusual character in the structure of the female repro- ductive ducts. The spermatheca is a dilation of the vagina with 2 separate openings, one to the vagina proper and the other to the oviduct and gland mass. Thus it may be termed ‘serial’ fol- lowing the terminology of similar parts in the doridacean opisthobranchs. In the species of the Eolidacea, the sper- matheca is a blind sac with a nar- rower stalk attached at the inner end of the vagina (Cleioprocta) or nearer the outer end (Acleioprocta), or it may have a separate external opening near that of the vagina (Some, but not all, Pleuroprocta). From the literature, there appears to be only 2 cleioproct species comparable to Herviella: Palisa papillata Edmunds (1964: 12, fig. 10A; = Moridilla kris- tenseni Marcus & Marcus, 1963: 44) and Noumeaella rehderi Marcus (1965: 282, fig. 35). In P. papillata the sper- matheca is a dilated section of the vagina; in N. rehderi it is lobated. A similar ‘serial’ spermatheca is reported in a number of species of the dendronotacean genus Doto (Marcus, 1957; 1961: 36-41, figs. 129, 134, 138, 140, 146; Marcus, E. € E., 1960: 166, Не. 51) Smbiech: like Herviella, grow to little more than 10 mm in length and have very slender bodies. Therefore, it would seem that smallness of size may have led to this parallel development within related groups. Palisa has 5 or 6 rows of cerata in the right liver and, therefore, in the present eolid classification, belongs to the family Facelinidae. Like Noumea- ella, Palisa has rhinophores that are thickly papillate on their rear edges, but unlike Noumeaella, the penis is unarmed. The present classification places both Herviella and Noumeaella among the genera of the family Favorinidae (characterized by the right liver in an arch or single row), subfamily Favorin- inae (characterized by cerata in a single series). In Noumeaella the right liver forms an arch, the rhinophores are papillate as mentioned above andthe foot corners are tentaculiform. The simi- larity of the genital organs, both with penial stylet and serial spermatheca, suggest that these 2 favorinids shouldbe grouped together, perhaps even so far as to warrent a subfamily of their own, Herviellinae. LITERATURE CITED ABE, T., 1964, Opisthobranchia of To- yama Bay and adjacent waters. 99 + REVISION OF HERVIELLA 229 9 p, 36 pls. [Not seen] BABA, K., 1930, Studies on Japanese nudibranchs. (3). A. Phyllidiidae. B. Aeolididae. Venus, 2(3): 117-125, pl. 4. Tokyo (Hokuryu-Kan). 1937, Opisthobranchia of Japan (II). J. Dept. Agr. Kyushu Imp. Univ., 5(7): 289-344, pls. 1, 2. 1949, Opisthobranchia of Sagami Bay. 194 + 7 p, 50pls. Tokyo (Iwanami Shoten). 1960, The genus Herviella and a new species, H. affinis, from Japan. Publ. Seto Mar. Biol. Lab., 8(2): 303- 305. 1966a, Record of Herviella albida n. sp. from Seto, Kii, Japan. Ibid., 13(5): 361-363, pl. 15. 1966b, The anatomy of Her- viella yatsui (Baba 1930) and H. affinis Baba, 1960. Zbid., 14(1): 1-6, pl. 1-2. BURN, R., 1963, Descriptions of Aus- tralian Eolidacea. 1. The genera Catriona and Herviella. J. Malac. Soc. Australia, 7: 12-20. EDMUNDS, M., 1964, Eolid Mollusca from Jamaica, with descriptions of two new genera and three new species. Bull. Mar. Sci. Gulf and Caribbean, 14(1): 1-32. MARCUS, E., 1961, Opisthobranch mol- lusks from California. Veliger, 3 (Suppl. 1): 1-85, pl. 1-10. 1965, Some Opisthobranchia from Micronesia. Malacologia, 3(2): 262-286. MARCUS, E. € E., 1960, Opisthobranchs from the American Atlantic warm waters. Bull. Mar. Sci. Gulf and Caribbean, 10(2): 129-203. 1963, Opisthobranchs from the Lesser Antilles. Stud. Fauna Curacao, 19(79): 1-76. MARCUS, E. € BURCH, J. B., 1965, Marine euthyneuran Gastropoda from Eniwetok Atoll, western Pacific. Malacologia, 3(2): 235-262. RISBEC, J., 1928, Contribution a l’étude des Nudibranches Néo-Calédoniens. Faune Colon. Franc., 2(1): 1-238, pl. 1- 12, A-D. 1937, Note préliminaire au sujet de nudibranches Néo-Calé- doniens. Bull. Mus. Hist. nat., Paris, (2), 9(2): 159-164. [Not seen]. 1953, Mollusques Nudi- branches de la Nouvelle-Calédonie. Faune Un. Franc., 15: 1-189. RESUMEN REVISION DEL GENERO HERVIELLA (OPISTOBRANCHIA: EOLIDACEA) R. Burn Herviella Baba 1949 (Opistobranchia: Eolidacea) se caracteriza especialmente por una hilera de “cerata” en el higado derecho, un estilete penial y una espermateca “serial”. Muessa Marcus 1965, con las mismas caracteristicas es un sinónimo. Ocho especies son conocidas del Pacifico occidental, incluyendo la nueva H. burchi aqui descripta. Se distinguen dos subgéneros; Herviella s.s. con la cúspide del diente raquideo más larga que los dentículos laterales, contiene las especies H. yatsui (Baba) tipo, H. affinis Baba, H. burchisp. п., H. evelinae (Marcus), H. claror Burn y H. exigua (Risbec); Marciella subgénero nuevo, con los dentículos laterales casi tan largos como la cúspide central, contiene la especie H. mietta Marcus (tipo), y Н. albida Baba. Noumaeaella Risbec 1937 con caracteristicas genitales similares y un arco de “cerata” en el hígado derecho, parece estar muy relacionada a Herviella. Los dos géneros forman un grupo distinto dentro de la familia Favorinidae, y con algunas reservas se juntan en una nueva subfamilia, Herviellinae. 230 R. BURN ABCTPAKT РЕВИЗИЯ РОДА HERVIELLA (OPISTHOBRANCHIA: EOLIDACEA) P. BEPH Род Herviella Baba (1949) (Opisthobranchia: Eolidacea) характеризуется следующими особенностями: один ряд папилл (cerata) в правой печени, пениальный стилет и "сериальная" сперматека (семеприемник). Род Muessa (Marcus, 1965), имеющий те же признаки, является синонимом. Из западной части Тихого океана известно 8 видов рода Herviella, включая новый вид Н. burchi, описанный в настоящей работе. Различаются 2 подрода: Herviella s.s., у которого центральный зубец радулярной пластинки длиннее латеральных зубчиков. Сюда относятся - Н. yatsui (Baba) тип; Н. affinis Baba, Н. Битс sp. nov., Н. evelinae (Marcus), Н. claror Burn, dH. exigua (Risbec); Marciella subgen. nov., с латеральными зубчиками почти или такой же длины, как и центральный зубец; OH включает 2 вида: Н. mietta Marcus (типовой вид) u Н. ааа Baba. Noumeaella Risbec (1937) co сходным строением гениталий и с дугообразным расположением папилл в правой печени, видимо представляет собою род, близко-родственный Herviella. Эти 2 рода образуют хорошо-очерченную группу внутри семейства Favorinidae и, с некоторыми оговорками, выделены в новое подсемейство Herviellinae. MALACOLOGIA, 1967, 6(1-2): 231-241 ERVILIA CONCENTRICA AND MESODESMA CONCENTRICA: CLARIFICATION OF SYNONYMYI J. D. Davis Department of Zoology Smith College Northampton, Massachusetts, U. $. A. ABSTRACT A small pelecypod, Mesodesma concentrica Holmes (1860), was described from fossil material at Simmons Bluff, Yonges Island, South Carolina, U.S. A. A similar mollusk, Ervilia concentrica Gould (1862), was described from dredgings on the North Carolina coast. Comparison of hinge structure, pallial markings and external characteristics of lectotypes shows that these 2 forms are identical and synonymous. Furthermore, comparison of features possessed by Holmes” type material and specimens of M. arctatum Conrad shows no basis for including the synonymized species in the genus Mesodesma. Thus, a newly designated lectotype is described for E. concentrica along with a corrected taxonomic citation. Nomenclatural Synonymy Ervilia (Turton) 1822, Conchylia Dithyra Insularum Britannicarum: The Bi- valve Shells of the British Islands. p 55-56, pl. 19, fig. 4. Ervilia concentrica (Holmes) Plate II, Figs. 3-6 Mesodesma concentrica Holmes 1860, Post-Pliocene Fossils of South Carolina, p 44, pl. 6, fig. 10. Simmons Bluff, Yonges Island, South Carolina. Ervilia concentrica Gould 1862, Proc. Boston Soc. nat. History, 8: 281-282. [No figure.] Coast of North Carolina. Ervilia concentrica Gould 1862, Otia Conchologia, р 239. [No figure.] Coast of North Carolina. GENERIC CONSIDERATIONS The genus Ervilia was established by Turton (1822) to accommodate a lentil- shaped shell previously described as Mya nitens by Montagu (1808). Thus the type for the genus is M. nitens (by monotypy). The genus has a world- wide distribution in tropical and tem- perate waters; fossil forms are known beginning with the Tertiary. All species of Ervilia, fossil and recent, have certain characteristics in common. The valves are small, rarely exceeding 10 mm in length, and they are somewhat compressed and have con- centric striations (Fig. 1). Radiating striae are present on most species, but these usually are reduced mid-laterally. Occasionally, the radiating striae are restricted to the posterior portion ofthe Shell. Variability of these striae are probably determined in part by erosion. The valves are fragile and often trans- lucent. The umbo is usually slightly closer to the anterior end, although still close enough to the mid-point to give the shell lPublication No. 255 of the Department of Zoology, Smith College, Northampton, Massachusetts, US А: (231) 232 POSTERIOR MARGIN == ( RADIATING STRIAE (often absent) VENTRAL MARGIN EXTERIOR — RIGHT VALVE GROOVE TO ACCOMMODATE RIDGE ON LEFT VALVE GREATER CARDINAL CHONDROPHORE PIT >. E DORSAL MARGIN >) S ANTERIOR MARGIN “| == POSTERIOR MARGIN PALLIAL eae Wwe POSTERIOR ADDUCTOR и 7 PALLIAL LINE PROJECTION VENTRAL MARGIN INTERIOR — RIGHT VALVE BAER г ANTERIOR MARGIN POSTERIOR MARGIN CONCENTRIC RIDGES VENTRAL MARGIN EXTERIOR —LEFT VALVE CHONDROPHORE PIT moy PIT FOR GREATER CARDINAL Pa > == LESSER CARDINAL № We POSTERIOR MARGIN A ey —— ANTERIOR MARGIN PALLIAL SINUS DORSAL MARGIN (with ridge fitting into groove of right valve) POSTERIOR ADDUCTOR SCAR > ANTERIOR ADDUCTOR SCAR PALLIAL LINE PROJECTION VENTRAL MARGIN INTERIOR—LEFT VALVE FIG. 1. General shell morphology of Ervilia concentrica (Holmes). ERVILIA CONCENTRICA AND MESODESMA CONCENTRICA 233 я = da a — — re FIG. 2. Comparison of Ervilia concentrica and Mesodesma arctatum. Left column: Е. con- centrica; the upper 2 valves are the new lectotype selected from the Holmes’ material (AMNH 11291); the bottom valve is the paratype selected by Whitfield & Hovey (1901) as a representative specimen of the same collection. Right column; M. arctatum (MCZ 214394) from Nauset Beach, Orleans, Cape Cod, Massachusetts. Scale in millimeters. an oval shape. Some species are ros- lower edge of the pallial sinus merges trate posteriorly, which tends to accentuate the displacement of the umbo toward the anterior end. The hinge region is relatively simple. The right valve has a prominent cardinal tooth just anterior to a large chrondro- phore pit. Posterior to this pit is a smaller depression for the lesser cardi- nal tooth of the left valve. Theleft valve has a pit anteriorly for the greater cardi- nal tooth of the right valve. Adjacent and posterior to this pit is the chondrophore pit followed by the lesser cardinal tooth. There are essentially no lateral teeth, and the ligament is much reduced. The inner surface of the valves dis- plays a distinctive pattern. The pallial sinus is deep, extending nearly to be- neath the umbo. Posteriorly, where the with the pallial line, the fused lines bend downward and outward toward the margin of the valve. In summary, Ervilia is characterized by concentric ridges on small, com- pressed, oval valves; radiating striae that are often restricted to the posterior region; a very large cardinal tooth in the right valve; the lack of lateral teeth; a deep pallial sinus; and by adownward- projecting combination of the pallial line and the ventral sinus margin. This last feature is perhaps the most distinctive diagnostic feature for generic identifi- cation of both fossil and recent speci- mens. Characteristics of the genus Meso- desma in the western North Atlantic have been reviewed elsewhere (Davis, 234 J. D. DAVIS 1964, 1965), but because that genusis al- so involved in this paper a few mor- phological characters of both fossil and recent forms will be mentioned. Mem- bers of the genus Mesodesma are much larger than those of Ervilia (see Fig. 2). The 2 species of Mesodesma foundinthe western North Atlantic, M. deauratum (Turton) 1822, and M. arctatum (Conrad) 1831, are commonly about 35 to 40 mm long and occasionally reach 50 mm in length. The shells are acutely truncate posteriorly (the most distinctive diag- nostic feature) and quite thick and only moderately compressed. Serrated later- al teeth are present and the cardinal teeth are somewhat reduced, but a large chondrophore is present. Features ofthe hinge area are shownin Pl. 1, Figs. 1 € 2. DISCUSSION Holmes (1860) described a new fossil pelecypod taken from material at Simmons Bluff, Yonges Island, South Carolina, and named it Mesodesma con- centrica. The original Holmes de- scription follows: “Small shell, very inequilateral, con- centrically and finely ribbed. Anal mar- gin compressed. Posterior extremity of sheil prolonged, narrowed, wedge- shaped. “This shell closely resembles Meso- desma arctata Gould; the anterior ex- tremity is not truncated, but regularly rounded. The concentric striae are quite characteristic. Found in sand of sea beaches. ” Two years later, Gould (1862) de- scribed a small bivalve which he named Ervilia concentrica. His description follows: “Ervilia concentrica. T. minuta, ob- longo-ovata, pellucida, nitida, (seniori- bus, incrassatis, margaritaccis) con- fertim sed profecto concentrice arata: umbonibus paullo postmedianis; extre- mitate antico acutiori quam extremitate postico. Long. 6+; alt. 4; lat. 3 milim. “Dredged off the coast of North Caro- lina. Coast Survey. “This little shell, which seems to be abundant along the whole southern coast. is quite different from anything before described. ” Dall € Simpson (1902) later gave the following, more informative description under the same name: “Shell small, scarcely inflated. Pos- terior end narrower. Surface finely con- centrically ridged. Having delicate radial riblets most conspicuous on the anterior end. “Hinge — right valve with single tri- angular tooth in front of the small tri- angular resilium and a feeble one behind it. Left valve with a double cardinal. Pallial sinus faint, deep. Color whitish or pink. Length 5, height 3.5, diam. 2 mm. ” Later workers have frequently questioned the validity of these 2 species, and it is the purpose of this paper to examine this taxonomic problem and show that Mesodesma concentrica and Ervilia concentrica are, indeed, both synonyms and homonyms and that the species involved shouldbe excluded from the genus Mesodesma. As noted above, when Holmes de- scribed the species Mesodesma con- centrica in 1860, he concluded his re- marks by saying, “Found in sand of sea beaches”. Yet his description is in- cluded as a part of his survey of the Post-Pliocene Mollusca of South Caro- lina. A search for the shells usedin the original description led to a lot of Specimens in the paleontological col- lections of the American Museum of Natural History, New York City. The information accompanying the material reads as follows: “No. 11291 Type Am. Mus. nat. Hist. Holmes Mesodesma concentrica Hol. Р.Р. Foss. Post-Pliocene Fossils of S. C., p. 44, Pl. 6, Fig. 10. Up. Miocene, Simmons’ 5. С.” Whitfield & Hovey (1901) indicated that they considered this lot to be the “type” of Mesodesma concentrica. When I first studied this material a single ERVILIA CONCENTRICA AND MESODESMA CONCENTRICA Comparison of hinge structure in Mesodesma arctatum and Ervilia concentrica. Fig. 1. Right valve of M. arctatum. Fig. 2. Left valve of M. arctatum. Fig. 3. Right valve of E. concentrica. Fig. 4. Left valve of E. concentrica. Scale as in- dicated. 236 J. D. DAVIS right valve was glued to a small diamond- shaped piece of green cardboard. Dr. Norman Newell, Curator of Fossil In- vertebrates at the American Museum of Natural History, informed me that the cardboard was probably fastened to the valve by Whitfield and Hovey. Newell further informed me that Whitfield and Hovey placed a question mark after the listing of this material in the American Museum catalogue. Itis Newell’s opinion that this notation may indicate Whitfield and Hovey were not sure that this was the type material. It is my opinion, however, that they were probably questioning the validity of the genus and not the type. It is known that much, if not all, of the Holmes’ material did eventually come to reside in the col- lections of the American Museum of Natural History. The presence of the word “type” and Holmes’ name in the upper right corner of the label strongly suggests that either this is the original type material or it at least came from the describer’s collection and repre- sents a series of paratypes. In either case, it represents a starting point for discussion of the species, M. con- centrica. After careful study of the material I disagree with Holmes’ contention that the species “... closely resembles Meso- desma arctata Gould”. In fact, there is little to suggest any relationship be- tween the 2 forms. Most obvious is the difference in size, readily apparent in Fig. 2. The valves in the Holmes lot of M. concentrica do not exceed 10 mm in length. On the other hand, the original type of M. arctatum, designated by Con- rad — not Gould, has apparently been lost, but the lectotype designated by Davis (1964) has dimensions of length 25.2 mm, height 18.0 mm. Comparison of other features confirms further this lack of similarity. Study of the hinge areas, as shown in Pl. 1, reveal little similarity between the 2 forms. The order and arrangement of teeth and pits are entirely dissimilar. In addition, shells of Mesodesma arc- tatum are much thicker and different in form. For example, M. arctatum, as indicated previously, has a trun- cate posterior margin. By contrast, it is the posterior margin of valves in the Holmes lot that is drawn out — the exact opposite of the situation in M. arctatum. Comparison of pallial lines and sinuses complete the picture; there appear to be relatively few similarities between these 2 bivalves. Therefore, the species so named M. concentrica by Holmes must be reconsidered as it cannot be accepted as a representative of the genus Mesodesma. As indicated previously, Gould de- scribed Ervilia concentrica 2 years later. Although his description was not overly informative, the name persisted and the Dall & Simpson description (1902) strengthened the identity of the bivalve involved. Gould’s type material was dredged up off the coast of North Caro- lina in the Coast Survey. According to Johnson (1964), after Gould’s death the original Gould material was sold to the PLATE 2. Ervilia concentrica (Holmes) 1860. FIGS. 1-2. Paratype selected as representative by Whitfield & Hovey (1901) from AMNH 11291 (AMNH 11291/1:3); length 7.2 mm, height 4.9 mm. Fig. 1, exterior — right valve. Fig. 2, interior — right valve. FIGS. 3-6. Lectotype designated from AMNH 11291 (AMNH 11291/1:1, AMNH 11291/1:2); length 6.6 mm, height 4.3mm. Fig. 3, exterior — right valve. Fig. 4, exterior — left valve. Fig. 5, interior — right valve. Fig. 6, interior — left valve. FIGS. 7-8. Lectotype designated by Johnson (1964) from Gould Type Collection, MCZ 169092; length 6.4 mm, height 4.1 mm. terior — right valve. Fig. 7, exterior — right valve. Fig. 8, in- ERVILIA CONCENTRICA AND MESODESMA CONCENTRICA 237 238 J. D. DAVIS New York State Museum. In 1959, that portion of the collection described as “The Gould Type Collection” was placed on permanent loan to the Museum of Comparative Zoology, Cambridge, Massachusetts. From this material Johnson designated the lectotype of E. concentrica as the lot numbered MCZ 169092. He also designated 3 paratypes, Museum of Comparative Zoology (MCZ) cat. no. 169093, and 1 paratype, U. S. National Museum (USNM) cat. no. 611263, all from the original lot. Examination of the lectotype desig- nated by Johnson reveals that Ervilia concentrica possesses all of the generic features discussed previously (the radiate striae are reduced but still visible). The most significant aspect is encountered when the Gould lectotype of Е. concentrica is compared with the Holmes material from the American Museum of Natural History. I have done this with great care, and it was readily apparent that all specimens are identical. All features are the same — shape, dimensional proportions, hinge structure, pallial lines and pallial sinuses. These features can be com- pared in Plate 2. Thus, without question, Mesodesma concentrica Holmes and Ervilia con- centrica Gould are synonyms (and homonyms also — an unusual com- bination). Holmes discovered the species first but assigned it to the wrong genus. Gould encountered the shell 2 years later during marine dredging and assigned it to the correct genus, Ervilia, and at the same time gave it the same species name which Holmes had givento his so-called species of Mesodesma. Therefore, Holmes named the species but Gould put it in the correct genus. On the basis of Holmes’ earlier de- scription and the assumption that the material inlot AMNH (American Museum of Natural History) cat. no. 11291 has direct lineage, at least, fromthe Holmes Collection, I have designated 2 matching right and left valves from this lot as the lectotype for Ervilia concentrica. The additional valves (including the single right valve to which Whitfield & Hovey (1901) attached the green cardboard marker) have been designated para- types. The valves of the lectotype have been cataloged under AMNH 11291/1: 1 (right valve) and AMNH 11291/1:2 (left valve). The single right valve designated by Whitfield and Hovey is identified as a paratype AMNH 11291/ 1:3; The remaining paratypes are cataloged under the original number. The lectotype was found to be the only pair of matching left and right valves in the lot and was selected to provide representation of the entire shell. The paired matching valves are 6.6 mm long and 4.3 mm high at the umbo. The valves are opaque white with many concentric ridges of equal height. Radiating striae are absent. The valves are moderately compressed and un- equilateral (therefore being unevenly oval in shape). The posterior end is fairly rostrate, placing the umbo an- teriorly to ihe midpoint of the shell. The beaks are turned inward and some- what posteriorly. Internally, the hinge area of the right valve possesses a large cardinal tooth slanting obliquely downward and for- ward anteriorlytoafairly large chondro- phore. There are no lateral teeth. The pallial sinus is deep, extending nearly beneath the umbo. Posteriorly, the pallial line bends up to join the ventral margin of the sinus. These 2fused lines then curve downward and project out toward the ventral margin (see Pl. 2, Figs. 5 and 6). The hinge area of the left valve is somewhat different. The dorsal valve margin produces a projection anterior to the pit accommodating the large cardinal tooth of the opposite valve. Posterior to this pit is the chondro- phore with no intervening ridge. Fur- ther posterior another extension of the margin protrudes like a very small cardinal tooth. There are no lateral teeth, but the dorsal margin of the left valve has a laterally-projecting ridge ERVILIA CONCENTRICA AND MESODESMA CONCENTRICA which is accommodated in a corres- ponding groove in the right valve. A brief description of the single valve previously singled out by Whit- field & Hovey (1901) is provided for purposes of identification and com- parison. It is a right valve and is 7.2 mm long and 4.9 mm high at the umbo. It is opaque buff-white with many concentric ridges all of about equal height. Radiating striae are not visible. Like the lectotype, the valve is moderately compressed and unequi- lateral with an uneven oval shape. Other features, external and internal, are essentially identical to the features de- scribed for the right valve of the lecto- type. Some crystallized glue is attached to part of the exterior where the green marker was fastened by Whitfield and Hovey. ACKNOWLEDGEMENTS Appreciation is expressed to the following for loan of specimens for study: Dr. Norman D. Newell, American Museum of Natural History; Drs. Horace G. Richards and R. Tucker Abbott, Academy of Natural Sciences of Philadelphia; Dr. William J. Clench, Museum of Comparative Zoology; and Drs. Joseph Rosewater and Druid Wilson, United States National Museum. I also wish to thank Mr. Richard E. Pettit of Ocean Beach, South Carolina, for aid in locating Holmes’ material, and Drs. H. G. Richards, William J. Clench and N. D. Newell for reading the manuscript and making helpful suggestions. LITERATURE CITED CONRAD, T. A., 1831, Descriptions of 239 15 new species of recent and three of fossil shells, chiefly from the coast of the United States. J. Acad. nat. Sci. Philadelphia, 6: 256-258. DALL, W. H. & SIMPSON, C. T., 1902, The Mollusca of Porto Rico. Bull. U. S. fish. Comm. for 1900, 20: 351- 524. DAVIS, J. D., 1964, Lectotype desig- nation for Mesodesma arctatum. Nautilus, 78: 3-6. 1965, Mesodesma deauratum: Synonymy, holotype, and type locality. Nautilus, 78: 96-100. GOULD, A. A., 1862, Descriptions of new genera and species of shells. Proc. Boston Soc. nat. Hist., 8: 280- 284. 1862, Otia Conchologia: De- scriptions of shells and mollusks from 1839 to 1862. Boston, Gould and Lin- coln, 256 pp. HOLMES, F. S., 1860, Post-Pliocene fossils of South Carolina. Charleston, South Carolina, Russel and Jones, 122 pp , 28 pls. JOHNSON, R. E., 1964, The recent Mollusca of A. A. Gould. Bull. U. 5. Nat. Mus., 239: 1-182. MONTAGU, G., 1808, Supplement to Testacea Britannica with additional plates. London, England. 183 pp, pls. 17-30. TURTON, W., 1822, Conchylia Dithyra Insularum Britannicarum: The bi- valve shells of the British Islands. |Ervilia р 55-56, Mactra deaurata р CONE a В.Р. & HOVEY YE. ©. 1901, Catalogue of types and figured specimens in the Geological Depart- ment of the American Museum of Natural History. Bull. Amer. Mus. nat. Hist., 11: 464-466. 240 J. D. DAVIS RESUMEN ERVILIA CONCENTRICA Y MESODESMA CONCENTRICA CLARIFICACION DE SINONIMIA J. D. Davis Un pequeño pelecipodo, Mesodesma concentrica Holmes (1860), fué descripto sobre materiales fósiles de Simmons Bluff, Isla Yongues, Carolina del Sur, Estados Unidos. Otro molusco similar, Ervilia concentrica Gould, se describió en 1862, proveniente de rastreos en la costa de Carolina del Norte. La comparación de la estructura de las charnelas, marcas paleales, y caracteristicas externas de los lectotipos, demuestra que esas formas son idénticas y sinónimas. También se demuestra que, comparando los caracteres que poseen los materiales tipo de Holmes, con ejemplares de M. arctatum, no hay base para incluir la especie sinonimizada en el genero Mesodesma. Así, un nuevo lectotipo designado se describe para E. concentrica, junto con la correc- ción de la cita taxonómica. Sinonimia Nomenclatural Ervilia (Turton) 1822, Conchylia Dithyra Insularum Britannicarum: The Bivalve Shells of the British Islands. p 55-56, pl. 19, fig. 4. Ervilia concentrica (Holmes) Plate II, figs. 3-6. Mesodesma concentrica Holmes, 1860, Post-Pliocene Fossils of South Carolina. p. 44, pl. 6, fig. 10. Simmons Bluff, Yonges Island, South Carolina. Ervilia concentrica Gould, 1862, Proc. Boston Soc. nat. Hist., 8: 281-282. [No figure.] Coast of North Carolina. Ervilia concentrica Gould, 1862, Otia Conchologica, p. 239. [No figure.] Coast of North Carolina. ABCTPAKT ERVILIA CONCENTRICA Y MESODESMA CONCENTRICA (ПО ПОВОДУ ИХ СИНОНИМИИ) ДЖ. Д. ДЭВИС Мелкий двустворчатый моллюск Mesodesma concentrica Holmes (1860) был описан из отложений Симмонс Влафф, о. Йонгс, Южная Каролины, США. Сходная Форма Ervilia concen‘rica Gould (1862) была описана из драгажных сборов у берегов Северной Каролины. Сравнение строения замка и синуса, а также наружных признаков лектотипа показывает, что эти 2 формы идентичны и являются синонимами. Кроме того, сравнение признаков экземпляров из типовой коллекции Холмса и экземпляров М. arctatum Conrad не дает оснований включать эти виды-синонимы в род Mesodesma. Таким образом, вновь установленный лектотип описан для E. concentrica и приводится исправленные ссылки на систематические указания. ERVILIA CONCENTRICA AND MESODESMA CONCENTRICA 241 Номенклатурная синонимия Ervilia (Turton) 1822, Conchylia Dithyra Insularum Britannicarum: The bivalve Shells of the British Islands. p 55-66, pl. 19, fig. 4. Ervilia concentrica (Holmes) Plate II, figs. 3-6. Mesodesma concentrica Holmes 1860, Post-Pliocene Fossils of South Carolina. p 44, pl. 6, fig. 10. Simmons Bluff, Yonges Island, South Carolina. Ervilia concentrica Gould 1862, Proc. Boston Soc. nat. Hist., 8: 281-282. [No figure.] Coast of North Carolina. Ervilia concentrica Gould 1862, Otia Conchologica, p 239. [No figure.] Coast of North Carolina. l Le en h IA te a ри RI N UN \ вика nn Er y р EN a i hi: в. AA 1 ou ' gist Ba AE | Re > oy Mal à Le rent ds Pe Te ! | | A A IAN | AA rd a! races, PE, Del ARMÉE MTS wink diy И: i и И т Г NAS A di HET anil da Nu am 4 wi | Be lid nl ‘Autor и и: № ть DIRES. nh lg. ¡AS A Py au x le Sr owes ot Resid ow” am Le ta ve, ur Le TL mi ru)! VOL. 6 NO. 3 MUS. COMP. ZOOL. JUNE 1968 LIBRARY JUL 15 1968 HARVARD UNIVERSITY MALACOLOGIA International Journal of Malacology Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift MALACOLOGIA ANNE GISMANN, General Editor 19, Road 12 Maadi, Egypt The University of Michigan The University of Michigan О. А. В. Ann Arbor, Mich. 48104, U.S.A. Ann Arbor, Mich. 48104, U.S.A. EDITORIAL BOARD CONSEJO EDITORIAL SCHRIFTLEITUNGSRAT CONSEIL DE REDACTION РЕДАКЦИОННАЯ КОЛЛЕГИЯ Р. О. AGOCSY М. A. HOLME W. L. PARAENSE Magyar Nemzeti Muzeum Marine Biological Assoc. U.K. Centro Nacional de Pesquisas Baross U. 13 The Laboratory, Citadel Hill Malacolôgicas, C. P. 2113 Budapest, VIII., Hungary Plymouth, Devon, England Belo Horizonte, Brazil H. B. BAKER B. HUBENDICK J. J. PARODIZ 11 Chelten Road Naturhistoriska Museet Carnegie Museum Havertown Göteborg 11 Pittsburg, Penn. 15213 Pennsylvania 19038, U.S.A. Sweden U.S.A. С. В. BOETTGER Technische Universitat Braunschweig Braunschweig, Germany A. H. CLARKE, JR. National Museum of Canada Ottawa, Ontario Canada C. J. DUNCAN Department of Zoology University of Durham South Rd., Durham, England Z. A. FILATOVA Institute of Oceanology U.S.S.R. Academy of Sciences Moscow, U.S.S.R. E. FISCHER-PIETTE Mus. Nat. d’Hist. Natur. 55, rue de Buffon Paris V®, France A. FRANC Faculté des Sciences 55, rue de Buffon Paris V®, France P. GALTSOFF P. O. Box 167 Woods Hole, Mass. 02543 U.S.A. T. HABE National Science Museum Ueno Park, Daito-ku Tokyo, Japan A. D. HARRISON Department of Biology University of Waterloo Waterloo, Ontario, Canada K. HATAI Inst. Geology & Paleontology Tohoku University» Sendai, Japan J. M. HUBER, Managing Editor Museum of Zoology С. Р. KANAKOFF Los Angeles County Museum 900 Exposition Boulevard Los Angeles, Calif. 90007, U.S.A. A. M. KEEN Department of Geology Stanford University Stanford, Calif. 94305, U.S.A. М.А. KLAPPENBACH Museo Nacional Historia Natural Casilla de Correo 399 Montevido, Uruguay Y. KONDO Bernice P. Bishop Museum Honolulu, Hawaii 96819 U.S.A. H. LEMCHE Universitetets Zool. Museum Universitetsparken 15 Copenhagen @, Denmark AKLILU LEMMA Faculty of Science Haile Sellassie I University Addis Ababa, Ethiopia N. MACAROVICI Laboratoire de Géologie Université “Al. I. Cuza” Iasi, Romania D. F. McMICHAEL Australian Conservation Found. Macquarie University, Eastwood М. 5. W. 2122, Australia J. Е. MORTON Department of Zoology The University of Auckland Auckland, New Zealand W. K. OCKELMANN Marine Biological Laboratory Grgnnehave, Helsinggr Denmark J. B. BURCH, Associate Editor Museum of Zoology A. W. B. POWELL Auckland Institute and Museum Auckland, New Zealand R. D. PURCHON Chelsea College of Science and Technology London, S. W. 3, England $. G. SEGERSTRALE Institute of Marine Research Biological Lab., Bulevardi 9 A Helsinki 12, Finland R. V. SESHAIYA Marine Biological Station Porto Novo, Madras State India F. STARMUHLNER Zool. Inst. der Universität Wien Wien 1, Luegerring 1 Austria J. STUARDO Instituto Central de Biologia Universidad de Concepcion Cas. 301, Concepcion, Chile W.S.S. VAN BENTHEM JUTTING Noordweg 10 Domburg The Netherlands J. A. VAN EEDEN Inst. for Zoological Research Potchefstroom Univ. for C. H. E. Potchefstroom, South Africa C. M. YONGE Department of Zoology The University Glasgow, Scotland A. ZILCH Senckenberg-Anlage 25 6 Frankfurt am Main 1 Germany гурмана A MALACOLOGIA, 1968, 6(3): 243-251 ЭТАПЫ РАЗВИТИЯ НЕОГЕНОВЫХ НАЗЕМНЫХ МОЛЛЮСКОВ ПРЕДКАВКАЗЬЯ A. А. Стеклов Геологический институт Академии Наук СССР Мосвова, CCP ABCTPAKT Изучение неогеновых наземных моллюсков Предкавказья показало их большое разнообразие и широкое распространение в осадках среднего и верхнего миоцена и верхнего плиоцена. По своей зоогеографической структуре эти моллюски принадлежат к 4 разным группам: 1)группе тропических психро- и термофилов, вымерших на Кавказе, в Европе и северной Азии, 2) группе восточного Средиземноморья, 3) европейской группе и 4) группе видов, тождественных с современными северокавказскими. На протяжении неогена роль тропической группы угасала и увеличивался удельный вес средиземноморских и европейских видов, а общий облик фауны приближался к современному. Анализ распространения разных групп моллюсков позволяет в общих чертах восстановить историю изменения климатических условий Кавказа за неогеновое время. Отечественная палеонтология располагает весьма скудными, можно сказать ничтожными, сведениями о фауне наземных моллюсков неогенового времени. В континентальных отложениях неогена, широко распространенных на территории Сибири и Средней Азии, остатки раковин этой группы моллюсков по-видимому встречаются сравнительно редко. Иначе обстоит дело в Крымско- Кавказской провинции, где ископаемые остатки наземных моллюсков встреча- ются в изобилии и имеют хорошую сохранность. Указания на их присутствие в Предкавказье и Закавказье, а также в Крыму, на Украине и в Молдавии можно найти в работах еще прошлого столетия. До сих пор, однако, эта интересная группа моллюсков не подвергалась у нас специальному изучению, а те редкие отрывочные описания, которые появились в нашей научной литературе за последние 70-80 лет (Эйхвальд, 1850; Синцов, 1875, 1877, 1837; Анлрусов, 1902: АлизадЕе, 1936, 1954: Волкова, 1939, 1953; Коробков и Смирнов, 1959), включая и наиболее полную (с описанием 16 видов) работу В. В. Богачева по Куринской низменности (Богачев, 1935), не дают об этой фауне сколько-нибудь ясного представления. Естественно поэтому, что исследователям, трактующим вопросы истории малакофауны территории Советского Союза, приходится опираться лишь на данные палеонтологов Западной Европы, где ископаемые наземные моллюски собираются и изучаются уже более 100 лет. Попыткой в какой-то степени восполнить этот пробел является проводимая мной в последние годы работа по исследованию остатков наземных моллюсков из континентальных неогеновых отложений Предкавказья (Стеклов, 1959, 1961, 1962a, 6, 1964). Изучение ископаемых наземных брюхоногих (243) 244 A. A. СТЕКЛОВ Предкавказья показало большое разнообразие их по систематическому составу. В ископаемом состоянии в неогене найдены представители почти всех семейств, составляющих ныне богатейшую малакофауну Кавказской провинции. Из известных на Кавказе 20 семейств в неогене не найдены лишь представители Oleacinidae, Endodontidae, Bradybaenidae, Vitrinidae и Trigonochla- mididae, то есть 5 семейств. Что касается двух последних, то можно предполагать их присутствие в собранной коллекции. Кроме того, в ней присутствуют виды Strobilopsidae и Subulinidae - семейств, вымерших. ныне не только на Кавказе, но и во всей Европе. Наибольшим распространением в неогене пользовались Pomatiasidae, Ellobiidae, Valloniidae, Limacidae, Pupillidae, Enidae, Clausiliidae и Helicidae. Первые 3 семейства представлены каждое 1 родом, как и в современной Фауне. Остатки слизней, по понятным причинам не сохраняющиеся в ископаемом состоянии во всем разнообразии видов, пока систематически не обработаны. Зато Pupillidae, Enidae, Clausiliidae и Helicidae характеризуются значительным разнообразием, охватывающим в общей сложности до 30 родов и более 70 видов, что составляет примерно 60% всего видового состава фауны. Остатки представителей Aciculidae, Succineidae, Cochlicopidae, Ferussaciidae, Zonitidae и Parmacellidae встречаются редко. Брюхоно среднего миоцена собраны в единственном местонахождении у станицы Костромской в Майкопо-Пабинском районе. Встреченный здесь комплекс отличается большим своеобразием и насчитывает 21 вид, 12 родов, 7 семейств: Cochlicopa sp., Gastrocopta (Albinula) cf. acuminata Klein, С. (Sinalbinula) fissidens Sandberger, G. (Sinalbinula) nouletiana Dupuy, G. (Sinalbinula) farcimen Sand- berger, Vertigo (Vertigo) cf. ovatula Sandberger, У. (Vertilla) angulifera O. Boettger, Truncatellina sp., Pupilla triplicatoidea Steklov, P. signataeformis Steklov, Microstele wenzi Fischer, М. caucasica Steklov, M. buryaki Steklov, Pupilorcula karaganica Steklov, Vallonia sandbergeri Deshayes, V. subcyclophovella Gottschick, Chondrula (Mastus) forcarti Steklov, Caecilioides sp., Opeas minutum Klein, Zootecus insularis caucasicus Steklov, Caucasotachea kubanica Steklov. Остатки верхнемиоценовых ‘улиток собраны во многих местонахождениях Дагестана, Чечено-Ингушетии, Северной Осетии, Кабардино-Балкарии, а также к востоку от Ставрополя, в Майкопо-Лабинском районе и у станицы Верхне- Баканской в районе между Анапой и Крымском. Кроме того, интересные находки были сделаны в меотических отложениях на Керченском полуострове и в нижнем сармате южной Украины (в последнем местонахождении материал был собран доктором Л. С. Белокрысом). Среди очень разнообразных верхнемиоценовых улиток чаще встречаются и особенно характерны Caspicyclo- tus praesieversi Steklov, Pomatias rivulare Eichwald, Carychium plicatum Steklov, Gastrocopta (Vertigopsis) magna Steklov, G. (Albinula) acuminata Klein, G. (Albinula) ukvainica Steklov, С. (Sinalbinula) nouletiana Dupuy, С. (Sinalbinula) fissidens Sand- berger, Vertigo (Vertigo) ovatula Sandberger, V. (Vertigo) antivertigo callosa Reuss, Pupilla mutabilis Steklov, Microstele wenzi Fisher, Vallonia ex gr. lepida Reuss, У. subcyclophorella Gottschick, Strobilops (Strobilops) ukrainica Steklov, 5. (Strobilops) costata Clessin, S. (Eostrobilops) caucasica Steklov, Zebrina gumsiana Steklov, Chondrula (Mastus) caucasica strigata Steklov, Retowskia matyokini Steklov, Euxin- ophaedusa volkovae Likharev, Serrulina nazranica Likharev, Quadriplicata farsica Likharev et Steklov, Hawaiia antiqua Riedel, Nesovitrea petronella L. Pfeiffer, Monacha (?) externa Steklov, Caucasotachea andrussovi Steklov, C. fortangense Steklov. Кроме названных, реже встречаются еще другие виды Acicula, Carychium, Vertigo, Negulus, Truncatellina, Pupilla, Microstele, Chondrus, Jaminia, Chondrula (Mastus), А. А. СТЕКЛОВ 245 Euxinophaedusa, Laeviphaedusa, Pontophaedusa, Oxychilus, Daudebardia, “Limax>” Heli- cella, Tropidomphalus, Helicodonta, Caracollina, Helix. Плиоценовые (акчагыльские и апшеронские) брюхоногие собраны также из многих местонахождений, главным образом Дагестана и районов Сунженского хребта, хотя в меньшем количестве они встречаются и западнее (к северу от Минеральных Вод, на Кубани ниже Армавира и в других местах). В плиоценовых отложениях особенно часто встречаются Chondrula (Chondrula) microtvaga psedachica Steklov, С. (Chondrula) microtraga sunzhica Steklov, С. (Chondrula) tchetchenica Steklov, Euxina aff. somchetica UL. Pfeiffer, Helicella sunzhica Steklov, H. libidinosa Steklov, H. crenimargo L. Pfeiffer, Monacha(?) praeorientalis Steklov, Tropidomphalus psedachica Steklov, а также, более редко - Pomatias rivulare Eichwald, Gastrocopta (Albinula) zamankulense Steklov, G. (Sinalbinula) calumniosa Steklov, Vertigo (Vertigo) antivertigo antivertigo Droparnaud, У. (Vertilla) angustior Jeffreys, Trun- catellina cylindrica Ferussac, T. dentata Steklov, Vallonia aff. pulchella Muller, Jaminia (Bollingeria) pupoides Krynicki, Retowskia schlaeflii pliocenica Steklov, Quadriplicata intermedia Likharev, Caucasotachea (?) maslovae Steklov, Helix cf. buchi L. Pfeiffer VOR Пытаясь обобщить имеющийся материал, можно наметить в составе неогеновой малакофауны Предкавказья несколько разнородных групп, различающихся в зависимости от зоогеографических связей и древности входящих в них видов. 1. Группа древних видов, вымерших в настоящее время не только на Кавказе, но и в Европе и северной Азии, ближайшие которым - преимущественно психро- и термофилы, обитатели тропических и субтропических лесов, - сохранились в юго-восточной, южной и изредка, центральной частях Азии, в Америке, Африке, Австралии (Gastrocopta, Micro- stele, Pupilla mutabilis, Р. belokrysi, Negulus, Strobilops, Hawatia, Zootecus, Opeas). К этой xe группе приходится причислить и такие, не имеющие прямых аналогов в современной фауне Формы, как Euxinophaedusa. 2. Группа видов, вымерших в настоящее время на Северном Кавказе, ближайшие которым распространены преимущественно в области восточного Средиземноморья (Турция, Иран, Закавказье, Греция, реже - Балканы и север Африки). Эта группа охватывает преимущественно обитателей засушливых и жарких областей (аридных субтропиков) - Chondrula (кроме С. tchetchenica и С. caucasica strigata), Jaminia ledereri, Imparietula, Мопасйа, - и в меньшей степени субтропических мезо- и психрофилов, таких как Caspi- cyclotus, Retowskia, Pontophaedusa, Laeviphaedusa, Quadriplicata и др. 3. Группа видов, ближайшие которым ныне распространены преимущественно в Европе, а также в большинстве (а некоторые почти исключительно) и на Кавказе (Acicula, Zebrina, Parmacella, Нейсейа, Caucasotachea и Ос Es также такие виды, как Vertigo angulifera, Pupilla triplicatoidea, Chondrula tchetchenica, С. caucasica strigata). С экологической точки зрения эта группа, охватывает как виды умеренно-теплых степей (Chondrula, Helicella), Teer обитателей смешанных лесов (Vertigo, Caucasotachea) и горных областей (Pupilla triplicatoidea, Daudebardia). 4. Группа видов, тождественных с обитающими и ныне на Северном Кавказе (Pomatias vivulare, Vertigo antivertigo antivertigo, V. angustior, Truncatellina cylindrica, Jaminia pupoides, Euxina somchetica, Nesovitrea petronella, Helicella crenimargo и др.) M таких, ближайшие которым пользуются широким рас- пространением в Палеарктике, или Голарктике (Carychium suevicum, Pupilla submuscorum, Vallonia subcyclophorella, У. aff. pulchella и др.). 246 А. А. СТЕКЛОВ Хотя намеченное подразделение является ‘условным и даже в какой-то степени искусственным, так как оно, не отражая в полной мере зоогеографической природы ассоциаций моллюсков разных моментов геологического времени, существенным своим критерием имеет степень близости форм - обитающим ныне на Северном Кавказе, - оно тем не менее дает возможность отчетливо продемонстрировать процсс перестройки кавказской малакофауны на протяжении неогена.. Группа широкораспространенных и тождественных северокавказским видов не имеет представителей в среднемиоценовой малакофауне. Начиная же с верхнего миоцена и особенно в плиоцене она представлена уже целым рядом видов. В верхнем плиоцене вполне отчетлива близость ископаемых Форм рецентным вплоть до полной идентификации некоторых видов (Vertigo antivertigo antivertigo, У. pusilla, Truncatellina cylindrica, Jaminia pupoides, Zebrina hohenackeri, Euxina tschetschenica, Helicella crenimargo и Ip. ). Группа видов восточного Средиземноморья представлена в среднем миоцене двумя видами - Pupilla signataeformis и Chondrula ротсайь а в верхнем миоцене - семью: Caspicyclotus praesieversi, Jaminia ledereri, Retowskia matyokini, Pontophaedusa praefuniculum, Laeviphaedusa miocaenica, Serrulina sieversi, Pagodulina. Из верхнеплиоценовых видов к этой группе принадлежат Retowskia schlaeflii pliocenica, Chondrula likharevi, С. exgr. microtraga, Imparietula, Zebrina hohenackeri. Группа древних тропических видов составляет ядро среднемиоценовой малакофауны Предкавказья (Microstele wenzi, М. buryaki, М. caucasica, Gastrocopta fissidens, С. nouletiana, С. farcimen, Opeas minutum, Zootecus insularis caucasicus) и широко представлена в верхнем миоцене (Gastrocopta magna, С. acuminata, С. ukrainica, Pupilla mutabilis, P. belokrysi, Microstele, Strobilops caucasica, 5. costata, S. ukrainica, Euxinophaedusa volkovae, Е. steklovi, Negulus, Hawaiia antiqua и др.) В плиоцене роль этой группы сходит Ha нет. К ней можно отнести всего лишь два вида Gastrocopta. Группу древних тропических видов составляют в основном Формы, относящиеся по систематическому положению к родам, полностью вымершим в Европе и северной Азии, или имеющим в Фауне этих областей единичных представителей обычно E разорванным ареалом распространения, подчеркивающим их реликтовый характер. В меньшей степени сюда входят вымершие виды родов, пользующихся более широким распространением. Так, род Microstele, в настоящее время известен Ha Цейлоне, в Индии и восточной Африке, Negulus, - только в Африке, Opeas - в тропической зоне Азии, Африки и Америки, Zootecus - в Индии, Северной Африке и на островах Зеленого Мыса. Семейство Strobilopsidae, богато представленное в позднем палеогене и неогене Европы, а также в миоцене Предкавказья и Южной Украины, в настоящее время распространено в юго-восточной Азии и обеих Америках. Гастрокопты из подродов Albinula и Vertigopsis сохранились только в Америке. Сравнивая относительный объем и структуру каждой из выделен-ных групп в разные моменты неогенового времени, мы отчетливо видим, как на протяжении неогена с одной стороны угасала роль экзотических тропических элементов в кавказской малакофауне и возрастала роль группы форм, связанных со Средиземноморьем, и как, с другой стороны, общий облик фауны приближался к современному. Если в среднемиоценовой фауне тропические виды составляли более 60% всего ее состава, а в верхнемиоценовой - половину, то в верхнем плиоцене удельный вес этой группы не превышает А. A. СТЕКЛОВ 247 5- 6%. Наоборот, удельный вес средиземноморских видов (в целом) возрастает с 33% в среднем миоцене до 60% в верхнем плиоцене. Bee средиземноморские виды среднего миоцена входят B группу, связанную с восточным Средиземноморьем. В верхнем миоцене влияние последней группы уменьшается, и соответствующие виды составляют около половины всех средиземноморских, в верхнем же плиоцене - еще меньше. Факт распространения в миоцене на большой площади Европы и Кавказа полностью вымерших ныне на этой территории (или сохранившихся в виде единичных и редких реликтов) родов и даже семейств, представленных близкими и тождественными видами, позволяет сделать вывод об общности неогеновой малакофауны всей этой области. Можно допускать, что экзотическая группа в миоценовой Фауне Кавказа является дериватом некогда единой тропического типа фауны, распространенной в пределах палеогеновой, или еще более древней суши на месте Евразии. Последние ee потомки выжили и сохранились главным образом в юго-восточной области Азиатского материка, где, очевидно, удержались наиболее благоприятные для этого ландшафтные, в первую очередь климатические условия. Эти соображения согласуются с представлением о мало диференцированном тропическом климате Евразии в начале миоцена. Субтропическими чертами, вероятно, отличался климат Европы и Кавказа и в течение всего миоцена. В это время уже происходил процесс вытеснения тропических элементов фауны группой нового, средиземноморского типа. С первыми средиземноморс- кими элементами мы сталкиваемся в среднем миоцене. При этом обращает на себя внимание, что среднемиоценовые представители группы ближе всего стоят к рецентным видам, обитающим в области жаркого и засушливого климата в основном в странах Ближнего Востока и Закавказья. Так, Pupilla signataeformis близка P. signataMouss., обитающей в Закавказье, Средней Азии и Иране (а также - на севере Китая). Современные родичи Chondrula (Mastus) forcarti обитают в Греции и Турции. Тем самым, находит подтверждение идея некоторых исследователей о возникновении очагов ксеротермизации в центре Евразии ещё в глубокой древности (Давиташвили, 1956). В верхнем миоцене на Северном Кавказе еще обитает ряд Форм, близких видам, сохранившимся ныне только в Закавказье: Caspicyclotus praesieversi, Retowskia matyokini, Jaminia ledereri, Pontophaedusa praefuniculum, Serrulina sieversi. Верхнемиоценовый климат Северного Кавказа был по-видимому, влажным, субтропическим. Об этом свидетельствует не только разнообразие и относительное богатство азиатской группы в составе Фауны верхнего миоцена, не только присутствие форм, близких закавказским видам, но и расцвет мезо- и психрофильных групп средиземноморского типа - обилие Pomatias, крупных уплощенных гелицид, Clausiliidae, и др. Попутно можно отметить интересный Факт бедности кавказской верхнесарматской и меотической фауны настоящими Helix, тогда как в Крыму и на южной Украине Нейх является преобладающим элементом фауны этого времени. В этом факте можно видеть свидетельство достаточно отчетливой климатической диференциации, возникшей к концу миоцена. Территория северного побережья Сарматского моря имела по-видимому климат гораздо более умеренный, чем кавказский остров, а в дальнейшем полуостров. Последнее обстоятельство обычно не учитывается палеонтологами, изучающими морскую сарматскую фауну, а его следовало бы принимать в расчет. Кроме того намечается диференциация ландшафтной обстановки в области самого Предкавказья, достаточно отчетливо 248 A. A. СТЕКЛОВ проявляющаяся с конца сарматского времени и особенно в меотическом веке. Видимо, уже в конце миоцена заложились те черты различия климатических условий двух противоположных окончаний Кавказского полуострова, которые наиболее ярко проявили себя в конце понтической эпохи (время образования керченских железных руд и первой половины балаханской серии Азербайджана). Верхнесарматские и меотические ассоциации западной части Предкавказья содержат в преобладающем количестве виды психрофильных групп. Среди верхнесарматских моллюсков района р.Аргудан (между Нальчиком и Орджоникидзе) уже появляется значительная примесь таких мезофильных лесных элементов, как Pomatias и Caucasotachea, и вместе отмечается относительное сокращение роли видов субтропической группировки (Gastrocopta, Strobilops, и др.). Меотический же комплекс р. Фортанги (к востоку OT Орджоникидзе) характеризуется резким преобладанием мезофилов и даже существенной примесью ксерофильных элементов (Mastus, Zebrina, Jaminia, Monacha, Helicella), играющих заметную роль и в других местонахождениях восточного Предкавказья. Таким образом, прогрессирующая климатическая диференциация и возникновение очагов ксеротермизации привели в миоцене к распаду единой тропической фауны палеогена, к вымиранию ее в Европе и на Кавказе и выработке новой Фауны средиземноморского типа, окончательно сформировавшейся на рубеже миоцена и плиоцена. Верхнеплиоценовая Фауна особенно резко отличается по своей структуре от более древних, и приближается по типу к современной кавказской, сохраняя при этом и отчетливые черты своеобразия - присутствие реликтов древней тропической группы и Форм, очень близких рецентным видам Закавказья. Ядро верхнеплиоценовой фауны составляют виды средиземноморской группы, часть которых конхиологически тождественны рецентным. Большинство этих видов, однако, встречается довольно редко. Чаще же и в изобилии встречаются остатки немногих видов Chondrula в. str. и Нейсейа. Это обстоятельство само по себе свидетельствует о широком развитии в верхнем плиоцене в Предкавказье степных ландшафтов. Анализ морфологических адаптаций раковины акчагыльских и апшеронских Chondrula приводит к выводу о высокой ксеротермности предкавказского климата, акчагыла, в значительной степени затем сглаживающейся с началом второй половины верхнеплиоценового времени. В заключение этого краткого обзора можно сказать, что наши знания о неогеновых наземных моллюсках юга CCCP еще очень ограничены. Изложенное показывает, между тем, что изучение этой группы Фауны может дать интересный материал и для биостратиграфии континентальных отложений, и для палеоклиматологии, и для истории развития фауны. ЛИТЕРАТУРА Ализаде, К.А., 1936, Фауна акчагыльских слоев Нафталана. Тр. Азерб. нефт. Heavies MC Cle ин-та, вым. O43 1-36. 1954, Акчагыльский ярус Азербайджана. Изд. Ак. наук АзССР, Баку: 1-344. Андрусов, Н.И., 1902, Материалы к познанию Прикаспийского неогена. Дкчагыльские пласты. Tp. Геол. Ком., т. ХУ, №4: 1-153. А. А. СТЕКЛОВ 249 Богачев, B.B., 1935, Пресноводные и наземные моллюски из верхнетретичных отложений бассейна реки Куры. Тр. Азерб. фил. Ак. наук СССР, геол. сер. вым. ХШ: 1-96, Давиташвили, Л.Ш., 1956, К изучению экогенеза травянистых мезофильных и ксерофильных фитоценозов. Сообщения Ак. наук Груз. CCCP, т. ХУП, NE a7 Эберзин, А.Г., 1960, О находках Parmacella в CCCP. Палеонтол. журнал NES ad. Эйхвальд, E., 1850, Палеонтология России. Новый период. СПб: 1-284. Коробков, И.А., Смирнов, Л.Н., 1959, О нахождении Parmacella (Mollusca, Pulmonata) в верхнетретичных и четвертичных отложениях Бадхыза, и Карабиля. Палеонтол. журнал №: 100-104. Лихарев, И.М., 1962, Клаузилииды (Clausiliidae) . Сер. "Фауна CCCP. Mommecrus, т pls Bb dios ЭТУ Лихарев, M.M., Раммельмейер, E.C., 1952, Наземные моллюски фауны CCCP. ASS Еее AVE (COCCI М. ее ТА Лихарев, И.М., Стеклов, A.A., 1965, Новые виды Clausiliidae (Mollusca, Pulmonata) из миоценовых отложений Предкавказья. Палеонтол. журнал №2: 128-133. Матекин, П.В., 1950, Фауна наземных моллюсков Нижнего Поволжья и ее значение для представления об истории современных лесов района. зоологи. журнал, т. Hada №3 MOS 205. Синцов, И.Ф., 1875, Отчет о геологических исследованиях в Бессарабии в 1873 году. Зап. Новороссийского Общ. Естествоисп., т. Ш, вып. 1: 31-46. 1877, Описание новых и малоисследованных Форм раковин из третичных образований Новороссии. Зап. Новороссийского Общ. HeTCCPBOMCHIs 4 т. У, KBB snp 23. 1897, Описание некоторых видов неогеновых окаменелостей, найденных в Бессарабии и в Херсонской губернии. Зап. Новорос- сийского 00m. HeTreerBonem.),) то ЖИ, (Ban. 2: 39-88. Стеклов, А.А., 1959, О фауне наземных гастропод неогеновых отложений Восточного Предкавказья. Вестн. Моск. ун-та, сер. биол., почв., ор ООО о 4 WLS РОС 1961, Первая в СССР находка ископаемых Strobilopsidae (Mollusca, Pulmonata). Палеонтол. журнал №: 50-54. 1962а,Неогеновые виды кавказскаго рода Retowskia (Mollusca, Pulmonata). Палеонтол. журнал №2: 71-75. 19620, Роль наземных брюхоногих моллюсков в стратиграфии неоге- новых континентальных отложений Северного Кавказа. В сб. "Геология Центрального и Западного Кавказа": 141-157. 1964, Стратиграфическая роль ископаемых наземных гастропод на примере рода Chondrula. Изв. Высш. учебн. заведений, "Геология и разведка", №: 22-38. Волкова, Н.С., 1939, К стратиграфии верхнетретичных отложений Ставрополья. Тр. по геол. и полезн. ископ. Сев. Кавказа, вып., AAA 1953, Фауна нижней части верхнего сармата окрестностей г. Армавира. Тр. Всесоюзн. геологич. ин-та, "Палеонтология и стратиграфия". р. 52-84. 250 А. А. СТЕКЛОВ Pilsbry, H., 1916-1935, Manual of Conchology. Ser. 2, у. xxiv-xxviii, Acad. Nat. Sci. Phila- delphia. Riedel, A., 1963a, Ein rezenter Hawaiia-Fund aus Afghanistan und ein fossiler aus dem Kauk- asus (Gastropoda, Zonitidae). Ann. Zool. , 21(5): 34-41. 1963b, Fossile Zonitidae (Gastropoda) aus dem Kaukasus. Ann. Zool. , 21(15): 273- 287. Rosen, О., 1914, Katalog der schalentragenden Mollusken des Kaukasus. Изв. Кавказского музея, т. 4, вым, А=3 1-12. Sandberger, F., 1870-1875, Die Land- und Susswasserconchylien der Vorwelt. Wiesbaden, р 1-1000. Simionescu, J.& Barbu, J., 1940, La faune sarmatienne de Roumanie. Mem. Inst. Geol. al Rom., 3: 1-194. Wenz, W., 1915, Die fossilen Arten der Gattung Strobilops Pilsbry und ihre Beziehungen zu den lebenden. N. Jahrb. Min. Geol. und Pal. , 2(2): 63-88. 1921, Uber die zoogeographischen Beziehungen der Land- und Susswassermollusken des europäischen Tertiars. Centralbl. Min. , Geol. Pal. р. 687-694; 713-721. 1923, Gastropoda extramarina tertiaria. т: Diener “Fossilium Catalogus”, pts. 17, 18, 20-23: 3-1862. Berlin. 1938-1944, Gastropoda. Handbuch der Palaozoologie, 6(2-7): 1-720; 1507-1639. Ber- lin. 1942, Die Mollusken des Pliozáns der rumánischen Erdôl-Gebiete als Leitversteine- rungen ftir die Aufschlussarbeiten. Senckenbergiana, 24: 1-293. Wenz, W.& Zilch, A., 1959-1960, Gastropoda. Handbuch der Paläozoologie. 6(2): 1-834. Born- träger, Berlin. Westerlund, С.А., 1884-1890, Fauna der in der paläarktischen Region lebenden Binnenconchy- lien. i-vii. Lund. ABSTRACT DEVELOPMENT STAGES OF NEOGENE TERRESTRIAL MOLLUSKS OF CISCAUCASIA Dr. A. A. Steklov Geological Institute Academy of Sciences Moscow, U.S.S.R. A study of the Neogene terrestrial Mollusca of Ciscaucasia has revealed the variability and wide distribution of these mollusks in sediments of Middle and Upper Miocene and Upper Pliocene age. These mollusks belong to 4 different groups in regard to their zoogeographical structure: 1) a group of tropical psychro- and thermophylls now extinct in the Caucasus, Europe and North Asia, 2)a group of the eastern Mediterranean, 3) a European group, and 4) a group of species identical to the recent North-Caucasian ones. Throughout the Neogene period the role of the tropical group decreased and the relative im- portance of the Mediterranean and European species increased, approaching the composition of the recent fauna. An analysis of the distribution of various mol- lusk groups enables us to reproduce in general the history of the climatic changes of the Caucasus during the Neogene time. А. A. СТЕКЛОВ RESUMEN PERIODOS DE DESARROLLO DE LOS MOLUSCOS NEOGENICOS DE CISCAUCASIA A. A. Steklov El estudio de los moluscos del epígrafe reveló su amplia distribución y variabilidad en sedimentos del Mioceno Medio-Superior y Plioceno Superior. Estos moluscos pertenecen a 4 grupos diferenciados en su estructura zoogeográfica: 1) psycro- y termo-filo tropical al presente extintos en el Caucaso, Europa y norte de Asia. 2) grupo del Mediterraneo oriental. 3) grupo europeo, y 4) un grupo de especies idénticas a las del Reciente Nor-Caucasico. A través del periodo Neogeno. el rol del grupo tropical decreció, aumentando la importancia relativa de los mediterraneos y europeos, acercandose así a la composición actual. Un análisis de la distribución de los diferentes grupos nos capacita para reconstruir, engeneral, la historia de los cambios climaticos del Caucaso durante el Neogeno. 251 icon : Ple $ ant Mo il ide Na) у re te у A SE are р ум ¡AS Pie, PEN oh a wy iat de uh y sl Je vat a L Ar "4 Ил oa hf, et ht ar MAR e na ‘basin wha posi doré sh st une МР naan un MLD MN 1 NP Ta er ss fats be RR NR RAR Mu ET he AH FAN TN Я АА a a Г 5 У if, (il D RUE Me mue: han Ep h api им QUE ЗИ ое оу AN us ya QUE APR IS y NE № | 0 y Lo Ныне qu qu AN ES WU | ме los ОКН, И A u 414 ve, HR i Dr “lu A he | LES Bun" à Be A А NT DRA ER Ha! HUE Co $ cone ME MUNIE AU NE N u j u er h И hy) et fee) me LEA DE, uae PET а SL, Lt éd ii de + | déc Tavs te le! | К ¡ada ee бы el: ВИ Hd imp nina ОИ A am a Pere ALO a cé | Е 4 VENT" hp Аи, ‚Ц Ария arenes AT Ab 7 ua dane) |. dl ou leo tintos a al À lo déni vw Kr FT ig рай Me | 2 De IDEE а [CE EUR as os als eis i AT dig dr. peat? y би: Ми ves i a} “rs a von LAS mar Ont ent, TR Dar Ne MALACOLOGIA, 1968, 6(3): 253-265 PLANORBULA CAMPESTRIS (GASTROPODA: PLANORBIDAE) FROM THE CUDAHY FAUNA (KANSAN) OF MEADE COUNTY, KANSAS, WITH NOTES ON THE STATUS OF THE SUBGENERIC CATEGORIES OF PLANORBULA Barry B. Miller Department of Geology Kent State University 1 Kent, Ohio 44240, U.S. A. ABSTRACT Over 200 fossil shells of Planorbula campestris (Dawson) have been collected from 2 localities of the Cudahy fauna (Kansan), Meade County, Kansas. Study of shell sculpture and internal lamellae in these materials and in all available Recent lots of P. campestris and P. armigera in collections of the U.S. National Museum and Museum of Zoology, University of Michigan, revealed several characteristics which have not been previously reported. These include: (1) the presence of multiple sets of denticles in both Planor- bula campestris (as many as 4) and P. armigera (as many as 2); (2) a maximum number of denticles per set, which consistently never exceeded 5 in P. cam- pestris and 6 in P. armigera; and (8) 2 distinctive types of sculpture, con- sisting in P. campestris of strongly developed, continuous, incised spiral striae which interrupt the slightly raised, evenly spaced, incremental growth lines. In P. armigera surface ornamentation consists of less regularly spaced growth lines which are crossed by unevenly spaced, weakly developed, and discon- tinuous spiral striae and fine lirae. The shell characteristics here recognized to separate these 2 species, placed by Baker (1945) in the subgenus Planorbula s.s., are at least as important as the shape of the lower palatal lamella used by him to distinguish the 2 sub- generic catagories of Planorbula, Planorbula s.s. and Haldemanina. It is sug- gested that Baker’s subgeneric catagories of Planorbula are probably not valid. INTRODUCTION During the summer of 1958, field parties from the University of Michigan Museum of Paleontology made col- lections from 2 localities of the Cudahy fauna in Meade County, Kansas. A total of 41 species of mollusks (Table 1) were recovered from the matrix col- lected from the Cudahy Ash Mine, SE 1/4 sec. 2, T. 31 S., В 28 W2(Meade Co., K. U. Loc. 10), and Sunbrite Ash Mine, NE 1/4 sec. 26, T. 32S.,R. 28 W2 (Meade Co., K. U. Loc. 17), Meade County, Kansas. Included among these mollusks were over 200 shells of Planor- bula campestris. The present report is based ona study of this fossil material, and of Recent lots of Planorbula campestris and P. armigera examined in the collections of the United States National Museum (USNM) and the University of Michigan Museum of Zoology (UMMZ). Its pur- lContribution Number 5, Department of Geology, Kent State University. 2Interpretation of the U. S. Geological Survey Topographical maps here quoted is explained in Malacologia, 1966, 4(1): 27-28; the position of Meade County is shown. 16а: 193. (253) 254 B. B. MILLER TABLE 1. Mollusks collected duringthe summer of 1958 from the Sunbrite and Cudahy localities of the Cudahy fauna, Meade County, Kansas cu Species UMMZ No. of UMMZ No. of Cat. No. specimens * Cat. No. specimens* Pisidium casertanum 216747 (7/2) - P. obtusale 218322 (8/2) - Carychium exiguum 216753 (121) 216705 (220) Stagnicola caperata 216751 (14) 216727 (7) S. exilis 218320 (55) = Fossaria dalli 216752 (16) 216722 (38) Omalodiscus pattersoni 216731 (79) 216725 (9) Gyraulus circumstriatus 216755 (34) 216702 (21) G. deflectus 216754 (13) 216700 (175) G. parvus 216756 (54) 216703 (64) Helisoma trivolvis 216732 (105) - Planorbula campestris 216730 (200) 216723 (3) P. cf. P. armigera - 216724 (3) Promenetus exacuous kansasensis 218321 (170) - Р. umbilicatellus 216745 (5) 216701 (16) Ferrissia parallela 218324 (25) - Physa skinneri 216737 (250) - Physa sp. (immature) 216760 (9) - Aplexa hypnorum 216757 (12) 216721 (36) Cionella lubrica - 216718 (400) Strobilops labyrinthica 216735 (28) 216729 (350) Gastrocopta armifera 216738 (25) - G. holzingeri 216759 (1) 216711 (9) G. tappaniana 216739 (23) 216726 (70) Pupoides albilabris 216742 (4) - Pupilla тизсотит 216741 (39) 216720 (75) Vertigo elatior 218325 (7) 216713 (60) V. milium 216750 (53) 216712 (100) V. ovata 216749 (45) 216716 (16) Vallonia cyclophorella 216746 (29) = V. gracilicosta 216743 (28) 216709 (175) V. pulchella 216744 (6) 216710 (30) Oxyloma sp. 218323 (14) 216728 (9) Discus cronkhitei 216734 (24) 216715 (30) Deroceras aenigma 216740 (525) 216719 (350) Euconulus fulvus 216736 (20) 216714 (85) Punctum minutissimum - 216706 (45) Nesovitrea electrina 216733 (25) 216708 (45) Hawatia minuscula 216758 (20) 216704 (33) Zonitoides arboreus 216748 (9) 216707 (23) Stenotrema leai - PANS AUT (150) * Numbers in excess of 100 have been estimated volumetrically. PLANORBULA CAMPESTRIS 255 pose is to: (1) record the first un- equivocal fossil occurrences of P. cam- pestris and to list the mollusks with which they were found associated; (2) present the results of obServations made on the shell characters of P. campes- tris; and (3) bring together information on the ecology and geographic dis- tribution of P. campestris. METHODS AND MATERIALS All of the Recent lots of P. campes- 1715 examined in the U. 5. National Museum collections (20 lots, 68 speci- mens) were studied through transmitted light to determine the number of lamellar sets present in each shell. A technique suggested by Walter (1962) was followed to make the shells more translucent. The shells were soaked several minutes in a full-strength solution of sodium hypochlorite which usually removed sufficient amounts of dirt, organic mat- erial and periostracum to permit viewing internal shell structures through the outer wall. Seventeen lots (73 speci- mens) of Planorbula armigera from the U. S. National Museum collections were Similarly examined. Two individuals, one in lot USNM 8970 and USNM 511386, had 2 lamellar sets. Twenty-nine fossil shells of Planor- bula campestris from the Sunbrite Ash Mine locality (UMMZ 216730) were selected for study of their internal lamellae. These shells were permitted to stand over-night in a commercially prepared oil (refractive index 1.60) used in the determination of the index of refraction in minerals and sold under the trade name “Shillaber’s Certified Index of Refraction Liquids”. All but 2 Shells were made sufficiently trans- lucent by this treatment to permit view- ing of the internal lamellae without destroying the shell. During examination each shell was viewed alternately from the right and left sides 3, with the light passing through the shell at right angles to the plane of coiling. In these 2 po- sitions the location of the large trans- verse basal and smaller upper palatal lamellae could be readily established when these lamellae occurred several whorls back from the aperture. Shells in which more than one set of lamellae were visible within the last whorl con- tained a parietal lamella, as well as a basal, lower and upper palatal lamel- la within each Set. In the earlier whorls it was not possible to see all the lamellae in a set. In these instances it was assumed that when an upper and basal palatal lamellae were observed in the same region of the shell, a complete set of lamellae was probably present at this position. The results of these obser- vations are summarized in Tables 2 and 3b SYSTEMATIC DISCUSSION Class Gastropoda Subclass Euthyneura Order Basommatophora Superfamily Ancyloidea Family Planorbidae Genus Planorbula Haldeman, 1840 Planorbula campestris (Dawson) 1875 Plate I, Figs. 1-4, 6-8 Segmentina armigera var. campestris Dawson, 1875, p 349-350. Segmentina (Planorbula) christyi Dall, 1905 bp 99 Pie mias 10,71% Planorbula (Planorbula) campestris (Dawson), Baker, 1945, p 176, pl. 118, Fig. 8; Pl. 119, Figs. 13-15. Original Description “Segmentina armigera var. cam- SRight and left side refer to orientation of the shell with respect to the living animal, wherein the outer lip of the aperture is po- sitioned anteriorly. 256 TABLE 2. USNM catalog number Parietal 1 633931 2 180300 382128 252447 и 519925 471342 470842 и 471327 471253 11 и п 471251 471317 (2) (4) (3) В. В. MILLER Measurements of 68 Recent Planorbula campestris in the United States National Museum collections. The development of lamellae in the set situated nearest the aperture is indicated by: A=absent; P=present; W= weakly developed. The number of sets of lamellae in excess of 1 are indicated by the number in parenthesis following the catalog number. Apertural lamellae Palatal No. of Height | Diameter Parietal wie ah mis A A A A 6-3/8 3,2 8.9 A A A A N 6-1/4 3.2 8.6 P P P P P 6-1/2 Bal 8.5 P P P P P 6-1/3 2.9 8.0 P P P P pP 5-1/23 2.8 6.7 P P P P P 63 3.0 8.5 А А А А А 73 4. 0 10.9 А А А А А 6-3/4 3.9 1187 A A A A A 5-3/4 853 8.1 А А А А А 6-1/23 3.4 9. 0 А А А А А 6-5/8 37 9.7 A A A A A 6-1/2 ST 10.2 A A A A A 5-3/43 23 5.5 A A A A A 5-7/8 3.0 8.6 A A A A A 6-1/2 3.4 8.9 A A A A A 5-1/2 3.0 8.3 A A A A A 6-5/8 3.1 8.5 A W A A A 6 832 8.3 A A A A A 5-3/43 3.5 8.7 А А А А А 5-3/4 3.2 MAT A A A A A 6-1/4 3.3 8.3 A A A A A 6-1/4 3.8 9.2 A A A A A 6-1/4 3.3 8.4 A W A A A 6 3.0 7.9 A A A A A 5-7/8 2.9 7.9 A A A A A 6 3.0 7:7 A A A A A 6 Bot 8.2 A A A A A 4-3/43 278 4.6 A A A A A 63 32 73 A A A A A 6-1/83 3.0 8.2 A A A A A 6-3/8 3.3 9.6 A P A A A 5-3/43 2.8 7.7 A A A A A 6-1/2 IE 7.8 A A A W W 6-3/8 al 9.1 A A A A A 6-1/4 3.4 9. 0 A A A A A 7-1/2 4.2 121 A A A A A 6-3/8 3.4 9.6 A A A A A 6-1/4 3.0 9.0 P P P P P 5-1/4 2.0 5.6 P P Р Р Р 5-7/8 2.4 6.1 А А А А А 6-1/2 3.5 9. 4 А А А А А 5-3/43 2.9 7.0 PLANORBULA CAMPESTRIS 257 Table 2 (continued) Apertural lamellae m Diameter catalo : Palatal tubercle 470483 A A A A A 3.6 9.8 601439 A A A A A 3.0 8.8 " A A A A A 3.15 9.3 " (2) A P P P P 3.6 9. 6 601437 A A A A A 2:9 8.2 " A A A A A 1.8 De 2 471252 A A A A A 5.2 12. 0 " A A A A A 72013) 6.4 " A 12 P 12 W 3.8 82 " A A A A A Sent! 9.8 rt ale A P P Р Р 3.1 9.7 n 5 A P P P P BD 9. 4 " A W A A A 3.2 Tork и А Р МУ 12) 12 2519 6.7 " A W A W W 3.0 072 и А А А А А 2015 He n (2)6 A A P P P 2.9 6.7 " A A A A A 2.6 6.5 601438 A A A A A Se U 10.0 " A P P 12 P Deo 6.1 "n (2) A 12 12 12 1% 2.3 6.7 и А Р Р iP P O 6.2 и А Р Р 12 P 2. 4 102 " A P P 12) P 2.4 6.3 № (2) A 12) P 12 P Be 5.4 " A P P P 12 2.4 6.4 1 Holotype of Planorbula christyi 2 Paratypes of P. christyi 3 Shell broken 4 Lamellae located 7/8 whorl behind aperture pestris, Pointe du Chéne. Dufferin. Traders’ Road. 500 mile Lake. This is a large fine variety characteristic of the prairie region, which I have dis- tinguished by the above varietal name. The normal form, with the usual number of whorls (4) is abundant in the Lake of the Woods, and surrounding wooded region. Specimens seldom at all ex- ceed 6.5 mm. The variety campestris occurs abundantly in some pools and coulees of the Red River Valley, and prairie region westward. They are 5 Lamellae located 1-1/8 whorls behind aper- ture 6 Second set of lamellae weakly developed 7 Aperture deflected downward much larger, with more whorls, and only in young specimens show the teeth. Colour generally wax yellow or pale brown. Diameter of largest specimens from 10.5 mm to 12.5 mm, whorls often six, specimens to 7.5 mm often, but not invariably show teeth; above this size no teeth were recognized” (Dawson, 1875: 349-350). Emended Description Shell medium, ultradextral. Whorls rounded, about 6 1/2, slowly increasing 258 FIGS. 1-3. FIG. 4. FIG. 5. FIG. 6. FIGS. 7-8. B. B. MILLER PLATE I Planorbula campestris (Dawson) right (=umbilical), left (=spire-pit) and apertural views. Cudahy fauna (Kansan), Meade County, UMMZ 216730a. X 10. Planorbula campestris (Dawson), view showing apertural lamellae. Same locality. UMMZ 216730b. X 10. Planorbula armigera (Say), view of portion of right side of shell showing details of sculpture. Recent, Hudson, Summit County, Ohio. X 40. Planorbula campestris (Dawson), enlarged view of right side of shell showing details of sculpture. Cudahy fauna (Kansan), Meade County, Kansas. UMMZ 216730c. X 40. Planorbula campestris (Dawson), peripheral and left (spire-pit) view of shell showing distribution of lamellae. Two transverse basal palatal lamellae can be seen as white bars in left view. Re- cent, Belleville, Ontario. UMMZ 185968a. X 8. PLANORBULA CAMPESTRIS 259 260 B. B. MILLER TABLE 3. Measurements of 29 fossil shells of Planorbula campestris (UMMZ 216730) from the Sunbrite Ash Mine locality No. of Height | Diameter N whorls mm mm eee sets 4-1/8 1.7 3.5 3* 4 1.5 4,5 11 5-1/8 2.5 6.7 1* 5 7.25 5.5 1* 4-1/2 2.0 Ake lf none* 4-1/4 2.5 5.5 2* 5-1/2 DAT 6.75 none* 4-3/4 2.0 5. 0 1*1 4 2.5 5.5 2* 5-1/2 2.5 6.5 none* 3-1/2 1.75 3.25 none* 5-1/8 2.3 5.8 none* 5-1/2 2.25 5.8 none* 3-1/2 1.25 2.75 none* 4 1.6 4.0 2*2 5 1.75 A 2 4-1/4 il 4.6 none* 4-7/8 2.75 о попе* 5-5/8 2.8 15 none* 5 Zaid 6.5 none* 4-1/4 1.1 3.25 2* 4-1/4 1.5 3.75 1* 4-1/4 1.0 3.5 none* 4-3/4 1.0 4. 0 43 4-1/4 113 315 none* 5-1/8 1.5 4.5 2*4 4-1/2 1.6 4.0 11 4-1/2 1.25 3.0 29 6 3.0 8.5 22 * Shell broken 1 One set in last whorl 2 Shell too opaque to permit examination by transmitted light. 3 Three sets in last whorl 4 First set 1/4whorl behind aperture; 2nd set 1-1/4 whorls behind aperture 5 Two set in last whorl. in diameter, coiled in same plane, over- lapping, the last whorl not deflected to left near aperture; right umbilical side flat to slightly depressed below the general plane; periphery rounded; left side rounded to subcarinate around edge of spire-pit. Spire-pit rounded, shallow, about 1/2 of total shell diameter, exhibiting all of the earlier whorls. Sutures weakly impressed. Apertural lip thickened within. Sculpture con- sisting of fine, close, evenly spaced, incremental growth lines which are in- cised by spiral striae. Apertural den- ticles, when all are present, are 5 in number and usually 1/5 of a whorl back from aperture. There are 2 parietal denticles: a large sigmoidal parietal lamella, ascending at about an angle of 70° to axis of shell; below the parietal there may be a small tubercle. There are never more than 3 palatal lamellae: a large slightly transverse basal palatal; a large lower palatal on the periphery of the outer wall, which descends at an angle of about 25° from the plane of coiling; anda short, thick upper palatal lamella which descends from the plane of coiling at an angle of about 45°. As many as 4 sets of lamellae may be present. Measurements See Tables 2 and 3. Geologic Range Taylor (1966) has referred some poorly preserved specimens of Planor- bula, collected from a number of Plio- cene and Pleistocene age deposits in Wyoming and Idaho, to P. campestris. The oldest of these occurrences is from the Middle Pliocene Teewinot Formation. The oldest unequivocal fossil oc- currences of this species are from the Cudahy and Sunbrite Ash Mine localities of the Cudahy fauna (Middle Pleistocene, Crooked Creek Formation) (Taylor, 1965; Miller, this report). Distribution Southeastern Ontario, west to British Columbia, north to the Great Slave Lake, south in the Great Plains to east central South Dakota. 261 PLANORBULA CAMPESTRIS *(uosMmeq) 51445242 DINQAOUDIF JO 10191912951 °Т “OLA e 92194141550 119593 Y 9249147990 |ISSO Y 262 B. B. MILLER The accompanying map (Fig. 1) is based on published records and materials examined in the United States National Museum (USNM) and University of Michi- gan Museum of Zoology (UMMZ) col- lections. The following is a list of peripheral localities: Ontario: Belleville (UMMZ 185968). South Dakota: Duell County, slough, Coteau Hills (UMMZ 90359). North Dakota: Ramsey County, road pond west of Garske (UMMZ 30094). Wyoming: Grand Teton National Park, Moran, small pond south of dam (UMMZ 184678). Montana: Flathead County, Kalis- pell (USNM 519925). British Columbia: Nulki Lake, small marsh near Vanderhoof (USNM 601437; 601438). Mackenzie District: Fort Smith, MacKenzie River (USNM 180300); MacKenzie River, 30 miles above Fort Providence (Whittaker, 1924). Ecology Planorbula campestris is typically a species of temporary ponds that occur in relatively open grassland areas. It is often associated in these situations with Stagnicola caperata, S. palustris, Aplexa hypnorum, Promenetus exacuous, P. umbilicatellus and Planorbula armi- gera (Mozley, 1938). It has been collected in Grand Teton National Park from small ephemeral ponds that are subject to considerable fluctuations in water level (Beetle, 1965). The ponds at these localities are situated in open meadowland at an elevation of about 6700 feet and apparently represent habitats that are quite similar to those from which P. campestris has been col- lected in the northern Plains (Mozley, 1938). Remarks Planorbula campestris is distinct from the 3 other species of Planorbula s.s., P. armigera, “P. crassilabris” and“P. jenksii”, considered valid by Baker (1945: 176) in the: (1) nature of its shell sculpture; (2) possession of never more than 3 palatal lamellae; and (3) absence of a flexure in the last whorl near the aperture. In P. campestris, spiral striae incise and interrupt the slightly raised incremental growth lines, producing an evenly distributed reticu- lated pattern that covers the entire shell exterior (Pl. I, Fig. 6). Sculpture in the other species of Planorbula s.s. is usually weakly developed and unevenly distributed, and consists of discontinuous spiral striae and fine lirae (Pl. I, Fig. 5). The lots of Planorbula campestris and Р. aymigera examined exhibited a great deal of variation in the number of sets of lamellae present in individual shells, the location of lamellar sets and in the degree of development of the individual denticles in each set. The number of lamellar sets in P. armigera was found to range from 0-2, апа, in P. campes- tris, from 0-4. To the writer’s know- ledge, this is the first report of multiple sets of lamellae in Planorbula. In some individuals of P. campestris only one set of lamellae occurs and may be located as much as 1 - 1/8 whorls behind the aperture. Althoughthe minimum number of denticles per set in both P. armigera and P. campestris is not constant, the maximum number of denticles per set never exceeded 6 (2 parietal and 4 palatal) in P.armigera and 5 (2 parietal and 3 palatal) in P. campestris (Fig. 2). DISCUSSION AND CONCLUSIONS The shells of Planorbula campestris that have been recovered from 2 fossil localities of the Cudahy fauna of Meade County, Kansas represent the first re- ported fossil record of this species in the southern Great Plains and extends its known range to the Kansan. The present study of the internal lamellae and shell ornamentation froma representative series of this fossil material and of all the available lots PLANORBULA CAMPESTRIS 263 FIG. 2. A, B, Planorbula armigera (Say). A. Apertural view of lamellae. B. Lamellae as they appear from outside of shell. C, D, P. campestris (Dawson). C. Aper- tural view of lamellae. D. Lamellae as they appear on outside of shell. The lamellae are indicated by numbers: (1) Parietal tubercle, (2) Parietal, (3) Supra- palatal (absent in P. campestris), (4) Upper Palatal, (5) Lower Palatal, (6) Basal Palatal (modified from Winslow, 1921). of Recent Planorbula campestris and P. armigera in the collections of the U. S. National Museum and University of Michigan revealed shell characteristics not previously reported, and indicates that the description of the shell charac- teristics attributed to the subgenus Planorbula s.s. by F. C. Baker (1945) in his monograph on the Planorbidae, is in need of emendation and consequently his subgeneric classification in need of revision. Baker proposed 2 subgeneric cate- gories, Planorbula s.s. and Haldemanina, a separation based on the difference in the lower palatal lamella. The latter subgenus, containing the 1 species P. wheatleyi, was differentiated from all the other species of the genus by the complexity of this lamella which “... is about twice as long as in armigera ... the upper part is bent upward almost at right angles to the transverse lower part, so that this end is on a line with the upper palatal lamella, the whole lamella being bent like an Australian boomerang” (Baker, 1945: 177). As regards Planorbula s.s., which contains P. campestris, P. armigera, “P. cras- silabris”, “P. jenksii”, as well as several doubtful fossil species, Baker states in his description of that group (1945: 173- 174) that there are 6 lamellae “.. a large parietal lamella ... a small tuber- cular lamella ... and 4 palatal lamellae .... Only one set of lamellae occurs in each shell ... the old set appearing to be absorbed before the new oneisformed as the shell increases through growth.” From my observations it appearsthat, while the number of denticles was maxi- mally 6 in P. armigera, it consistently never exceeded 5 in P. campestris. Moreover, the old sets of lamellae are not regularly all absorbed, but multiple sets occur in some individuals of both species, with a maximum of 4 sets ob- served in P. campestris and 2 in P. armigera. These 2 species are further distinguished by 2 types of shell sculp- ture, consisting, in P. campestris, of continuous incised spiral striae that in- terrupt the slightly raised, evenly spaced incremental growth lines. In contrast, the incremental growth lines in P. armi- gera are less regular, and are crossed by irregularly spaced, weakly developed, discontinuous spiral striae andfine lirae. The shell sculpture in“P. crassilabris”, “P. jenksii” and P. wheatleyi is similar to that of P. armigera. In my opinion the difference between the Haldemanina and Planorbula groups is no greater than the differences which distinguish P. campestris fromthe other species within the Planorbulas.s. group. To treat the size and shape of the lower palatal armature in P. wheatleyi as a sub- generic character would justify the creation of a 3rd subgenus to accom- modate the shell characters found in P. campestris. Such a proliferation of subgeneric categories in a genus that probably contains no more than 3 valid living species, P. campestris, P. armi- gera and P. wheatleyi, seems an un- 264 B. B. MILLER necessary complication. In these 3 species the shell characters in question probably represent no more than good specific differences. ACKNOWLEDGEMENTS Appreciation is expressed to Dr. Harald A. Rehder, U.S. National Museum and Dr. Henry van der Schalie, Museum of Zoology, University of Michigan, for the use of their facilities. Dr. Claude W. Hibbard, Museum of Paleontology, University of Michigan, provided the fossil material from the Cudahy fauna. LITERATURE CITED BAKER, F. C., 1945, The molluscan family Planorbidae [collation, revision and additions by H. J. van CLEAVE]. Univ. Illinois Press, Urbana, р 1-530. BEETLE, D. E., 1965, Molluscan fauna of some small ponds in Grand Teton National Park. Nautilus, 78(4): 125- 130. DALL, W. H., 1905, Land and fresh water mollusks of Alaska and ad- joining regions. Smithsonian Institu- tion. Harriman Alaska Series Vol. XII, р 1-171, Pl. 1-2. DAWSON, G. M., 1875, Land and fresh water Mollusca, collected during the summers of 1873-1874, in the vicinity of the forty-ninth parallel -- Lake of the Woods to the Rocky Mountains. In: British N. American Boundary Comm., Report on the geology and resources of the region in the vicinity of the forty-ninth parallel (etc.), App. E., p 347-350, illustr., map, Montreal. MOZLEY, A., 1938, The fresh water Mollusca of Sub-Arctic Canada. Canad. J. Res., 6(D): 93-138. TAYLOR, D. W., 1965, The study of Pleistocene nonmarine mollusks in North America. т: WRIGHT, H. E., Jr., & FREY, D.G., (Ed.), The Quater- nary of the United States. Princeton Univ. Press, Princeton, p 597-611. 1966, Summary of North American Blancan nonmarine mol- lusks. Malacologia, 4(1): 1-172. WALTER, H. J., 1962, Punctation of the embryonic shell of Bulininae (Planor- bidae) and some other Basommato- phora and its possible taxonomic- phylogenetic implications. Mala- cologia, 1(1): 115-137. WHITTAKER, E. J., 1924, Freshwater Mollusca from MacKenzie River Basin, Canada. Nautilus, 38: 8-12. WINSLOW, M. L., 1921, Mollusca of North Dakota. Occ. Paps. Mus. Zool., Univ. Michigan, 98: 1-18. RESUMEN PLANORBULA CAMPESTRIS (GASTROPODA: PLANORBIIDAE) DE LA FAUNA CUDAHY (EDADE KANSANA), DISTRITO MEADE, KANSAS, CON NOTAS SOBRE EL STATUS DE LAS CATEGORIAS SUBGENERICAS DE PLANORBULA B. B. Miller Se colectaron mas de 200 ejemplares fosiles de Planorbula campestris (Dawson), en el condado de Meade, Kansas, en dos localidades de la Fauna Cudahy de edad Kansana. El estudio de la escultura de las conchillas y los denticulos internos, en estos asi como en todos los lotes Recientes disponibles de P. campestris y P. armi- geva, en las colecciones del U. S. Nat. Museum y del Mus. de Zool. de la Universidad de Michigan, revelaron las siguientes caracteriaticas que no se habian informado previamente: (1) Presencia de denticulos en series múltiples en P. campestris (hasta 4) y P. armigera (hasta 2); (2) Numero maximo de dentículos por serie que nunca excede PLANORBULA CAMPESTRIS de 5 en P. campestris o 6 en P. armigera; (3) dos tipos distintos de escultura con- sistentes en P. campestris de continuas estrias espirales de fuerte desarrollo, que se interrumpen en las líneas un poco elevadas, de crecimiento y regularmente espaciadas. En P.armigera las líneas estan espaciadas con menos regularidad, pero cruzadas por estrias y fina liración discontinua débiles e irregularmente espaciadas. Tales caracteristicas, aqui reconocidas, para separar las especies ubicadas por Baker (1945) en el subgen. Planorbula s.s., son tan importantes por lo menos como la forma de las lamelas palatales inferiores que dicho autor usó para dis- tinguir las dos categorias subgenéricas, Planorbula s.s. y Haldemanina. Se sugiere que las categorias subgenéricas de Baker para Planorbula, problamente no son validas. ABCTPAKT PLANORBULA CAMPESTRIS (GASTROPODA: PLANORBIDAE) ФАУНЫ КЬЮДЕХИ (КАНЗАН) ИЗ РАЙОНА МИД KAYHTU, КАНЗАС И ЗАМЕТКИ О ПОЛОЖЕНИИ ПОДРОДОВЫХ КАТЕГОРИЙ PLANORBULA Б. Б. МИЛЛЕР Более 200 ископаемых раковин Planorbula campestris (Dawson) были собраны в двух местах распространения фауны Кьюдехи (Канзан), Мид Каунти, Канзас. Исследование скульптуры раковины и внутренних пластинок в этих и во всех имеющихся современных пробах Р. campestris и Р. armigera из коллекций Национального Музея С. Ш. A. и Зоологического Музея Мичиганского Универ- ситета показало наличие нескольких ранее неизвестных признаков. Сюда относятся: 1) наличие нескольких наборов зубчиков у Planorbula campestris (до 4) и УР. armigera (до 2); 2) максимальное количество зубчиков в каждом наборе никогда не превышает 5 у P. campestrisu 6 y P. armigera 3) имеется 2 ясных типа скульптуры, состоящей у Р. campestris из сильно- развитых непрерывных, вдавленных спиральных линий, прерывающихся слабоприаоднятыми, расположенными на равном расстоянии линиями наростания. У Р. armigeva скульптура поверхности состоит из менее правильных линий наростания, перескающихся с неравномерно-расположенными, слаборазвитыми и непрерывными спиральными линиями и тонкими бороздками. Приведенные характеристики для разделения этих 2 видов (отнесенных Бэкером в 1945 г. к п/роду Planorbula s.s.) яв- ляются не менее важными, чем Форма нижней полатольной пластинки, которую он использовал для разделения двух под- родовых категорий рода Planorbula: Planorbula s.s. и Haldemanina. Автор предполагает, что подродовые категории Бэкера для подрода Planorbula, возможно являются недействительными. 265 f AT A ar Qi PATA Le AN aa) aly | С ие A bide Г aM eae (в y И AA Sur ANA ИХ A y Ут А ee 5 AN 2 у № ik | Le ON A niga! НИИ PT vat de + Pas sy ‹ AY bry + EAS yt APE a РГА NES om y My pinay wid quel ae De fas зави, Clea) Вы АН Mae TARA ER A NP AA OR ie mites M or ae: ЗН PREY ESS A AN OY AN ET “alla nu de DINE MPN ED IN rabo), Reel à, At in % ee ny VAL QU ia? ie PANNE LEE nn a un, RAT ) + Ai KR REN eur a ach En wie? dl CHEN ya KANN skh Pe A AO A | Le 6 y ¿dh 1 { ф À f $3 DAR vy x ud | A i À Hdi yt bia w 4 al 4 ПА } Я | | TE VE LPO ANT 1 o rm i ie i ere CP Rat ee | Wie Ed т I ANS АИ И Я Аа Dre wanS и ий derbi U | | ‘ 12 4) Te A Nite Aes NI CT bie ‘4 2 VA ds A wa pi ME A AR “fi y à A eng а Rk aa TN LENS) y m fee am ly ee CENAR AAA Hd Y К TEA АЙ, y RTE ЕЮ Eo LAS | A an " K PET a RUA WHI | ees bs me tees be | Ин yet PATA A Sa (AS Kan | UE km al Ma и Vi ет MT ET 99 сво FER! dormi rl Bo ны ии ! Mio MEN IR ara ere Ax te | иг . MASSAGE y + M un Eee LAIT TR | RER pt: whe re # { ити e Meh iR A т в fi PO y | У и ai L ne | и iy i de à}, AI к (a. it de вы ME o TON ] en Я 0 Hi eli pos ovis MALACOLOGIA, 1968, 6(3): 267-320 ‘ PHYSIOLOGICAL AND ECOLOGICAL ASPECTS OF PREY SELECTION BY THE MARINE GASTROPOD UROSALPINX CINEREA (PROSOBRANCHIA: MURICIDAE)1,2 Langley Wood Department of Environmental Physiology Virginia Institute of Marine Science Gloucester Point, Virginia 23062, U.S. A. ABSTRACT Behavioral, physiological and ecological relationships between the molluscan predator Urosalpinx cinerea (Say) and its intertidal prey on the east coast of the United States have been examined in field and laboratory studies. In studies of relative attack frequencies in nature and olfactometer responses in the laboratory, barnacles, Balanus spp. , are shown to be significantly more attractive to U. cinerea than either of its other major intertidal prey, oysters and mussels. Field reports of this preference are based upon direct obser- vations of 4,416 snails in 11 intertidal habitats in the continuous east coast range of U. cinerea from Massachusetts to northern Florida. This statistical preference for barnacles is not genetically fixed; indeed, field studies indicate that ecological factors can account for prey selection in inter- tidal habitats. One of these is intertidal conzonation of prey and predator, another is relative abundance of a given prey species within the intertidal zone. The role of external metabolites in bringing the predator to its prey could not be elucidated from my field observations. Experimental evidence is presented to introduce the concept of ingestive con- ditioning, in which the predator’s tendency to respond to effluents from a given prey species is increased after it has ingested living tissues from that species. Ingestive conditioning was partially reversed in juvenile U. cinerea by returning them to their original diets. Circumstantial support for the operation of this process in nature was derived from experiments with snails from single- and multiple-prey habitats. Statistical tendencies of U. cinerea to select barnacles in preference to oysters was partly confirmed in experiments with young and juvenile stages, which were most easily conditioned to barnacles, but not always in the case of adults, whose diverse natural diets were difficult to reverse. Evolutionary aspects of this predator-prey relationship are discussed with particular reference to the adaptive value to the predator of ingestive con- ditioning. Restriction to a single prey species would have disoperative effects, so it is to the predator’s advantage to be capable of feeding upon more than one species. However, different attack techniques are utilized for efficient penetration of various prey species, and these are apparently acquired by in- dividual U. cinerea. By concentrating upon a single species, U. cinerea proba- bly increases its attack efficiency. The mechanism described here as ingestive conditioning provides such a concentrating influence without the irreversibility of genetically fixed prey specificity. lAdapted from portions of a thesis submitted to Cornell University in partial fulfillment of the requirements for the degree of Doctor of Philosophy. 2Contribution Number 229 from the Virginia Institute of Marine Science. (267) L. WOOD 268 CONTENTS Pages I. INTRODUCTION 1. Systematics and morphology ..... 2.2... 269 2. Habitat and distribution .... 269 3. Feeding mechanisms...... 269 4. Responses to changes in the environment . . . . . .. 270 5. Prey preference.......... 271 106 IT. ТУ. 6. Objectives of present study... 273 PREY SELECTION UNDER NATURAL CONDITIONS 1. Introductiontmmidia 2 ee 273 2. Selection of field stations ....274 3. Observational methods . . ... 274 4, Maintenance of live collections оное ARE. 276 5. Summary of observations .... 276 6. Discussion and con- GLUSIONS IN. IT ork epee ees 282 EXPERIMENTAL METHODS 1. Generalomethodi 0 un ; 283 2. Olfactometers: 2... 4 ete 284 3. Identification of individual predator sensi EEE 285 4, Predator maintenance... ...285 5. Possible sources of experimental error... ..286 ORIENTATION IN COMPLEX CURRENTS 1: Introductiont abrio 42.000291 2. Materials and methods ...... 291 3. Results and discussion..... 292 THE EFFECTS OF PREVIOUS EXPERIENCE 1. Introduetion LE ана 294 2. Comparison of natural and laboratory prey sélectionnant dns mr 294 3. Effects of controlled, single- species diets upon olfactory behavior ......295 4. Responses of conditioned snails to odor of one Species atatime...... 301 5. Effects of long-term ex- posure to prey odors. . ..303 6. Effects of short-term ex- posure in olfactometer ... 305 7. Summary of ingestive con- ditioning experiments . . . 308 VI. GENERAL DISCUSSION 1. Revision of preference concept... Sr 309 2. Mechanism of ingestive сопок. 11 311 3. Adaptive aspects of ingestive conditioning’ AAA 313 ACKNOWLEDGEMENTS. ....... 315 LITERATURE "CIFED ARR 316 I. INTRODUCTION The marine gastropod Urosalpinx cinerea (Say) is a copredator, with man, of the Virginia oyster, and as such has attracted considerable attentionfrom the shellfish industry and from those biologists whose research is industry oriented. My initial investigation of the feeding habits of this snail began in such a context, but it was not long before several fundamental problems assumed a central importance The original question “Does this predator have a food preference?” is important to the oyster industry; but the more basic questions which grew from this root have carried the investigation into the fields of ecology, physiology, and biochemistry. The present paper, describing part of this work, will cover field studies of prey selection, and introduce an hy- pothesis to explain results of both field and laboratory investigations. The other part, “Qualitative and quantitative nature of attractance,” will appear separately. Much general information about the species has been compiled by Carriker (1955) for North American populations and by Hancock (1959) for those of the channel coast of England. The following brief review of aspects of the biology of U. cinerea relevant to this report draws heavily from these 2 sources, except where otherwise indicated. PREY SELECTION BY UROSALPINX 269 1. Systematics and Morphology Urosalpinx cinerea is a member of the Family Muricidae, Order Neogastro- poda, and Subclass Prosobranchia (Thiele, 1931). The common name of the family is “The Rock Shells,” as these are gastro- pods adapted to, and commonly found in, habitats which offer a firm substrate. U. cinerea has a strong, muscular foot that can adhere firmly to a smooth rock or shell surface, but can be retracted fully into the rather heavy shell, and there be protected by a horny operculum. Adults range from about 17 to an in- frequent maximum of 58 mm (total spire height, measured from apex to the tip of the siphonal canal). Typically, dis- crete populations may vary considerably in size, color, and shell shape. Vari- ations in size are extreme, and Baker (1951) recognized this variation by taxo- nomic separation. She gave the name Urosalpinx cinerea var. follyensis to specimens collected from a shell midden on the Atlantic shore of the Delaware- Maryland-Virginia Peninsula at Folly Creek, Accomac, Virginia. Carriker (1961) granted subspecific rank to that population, calling it Urosalpinx cinerea follyensis, and referred also to U. c. cinerea, which presumably includes all other Atlantic coast populations. But for purposes of the present work, I will regard U. cinerea as one species without further taxonomic separation. Specific geographic origins of experimental ani- mals will be given. 2. Habitat and Distribution Urosalpinx cinerea is apparently native to the middle Atlantic coast of the United States, but has been inadvert- ently carried to other parts of the world by man. Its present North American range is from eastern Canada to the northern part of Florida’s Atlantic coast; scattered introduced populations have also been reported on the Pacific coast of North America, from northern Cali- fornia to southern Canada. It is dis- tributed vertically from the middle inter - tidal zone down to unknown depths, though it probably does not extend beyond the continental shelf. Within its range, it is found in rock or shell intertidal habitats characterized by intermediate to full oceanic salinities, a wide annual temperature range, good water circulation, and an abundance of sessile epifauna. Under optimal con- ditions of food and hydrography, popu- lations of this predator (juveniles and adults) can achieve seasonal densities as high as 1000/m2 (pers. obs.), though this is atypical. On the other hand, there are potential habitats which seem to be ideal but which have not yet been popu- lated by this or any other predatory gastropod. A possible explanation for this distribution is based on the lack of a pelagic larva, though Carriker (1957) has suggested alternative means of transport for the species. 3. Feeding Mechanisms Two distinct capabilities are requisite for successful predation. First, the predator must be able to detect, locate, and move toward its prey. Second, it must possess apparatus with which it can attack and ingest the prey or portions thereof. It has been amply demonstrated in both field and laboratory studies (Haskin, 1940; Blake, 1960; reviewed by Kohn, 1961) that Urosalpinx cinerea detects the presence of, and migrates to, its prey on the basis of chemical stimuli. Other sensory stimuli, such as light, sound, or the sensation of touch, are either not used at all or (as in the case of touch) are probably important only after the predator has located and arrived at the prey. Such a situation can hardly be regarded as extraordinary, since chemo- reception is the most primitive and ubiquitous sense in invertebrates and is instrumental in initiating and main- taining numerous inter - and intraspecific relationships (Davenport, 1950 et seq.; Bullock, 1953; Hodgson, 1955; Kohn, 1961). 270 L. WOOD An animal which preys upon shelled organisms must be able to penetrate the shell. Such an adaptation has been de- scribed in the case of Urosalpinx cine- rea by Carriker (1943; Carriker, Scott & Martin, 1963), and for other proso- branchs by Fretter & Graham (1962). Shell penetration by these gastropods is accomplished in various ways, de- pending upon the species of predator and prey involved, and upon the individual experience of the predator. In U. cine- yea, the most commonly occurring method of attack is that of shell per- foration. According to Carriker et al. (1963), perforation is accomplished by 2 principal techniques, using different sets of organs. The first of these in- volves the radula and its associated musculature, ensheathed ina protrusible proboscis. This apparatus rasps shell and soft tissue and conveys these tissues into the esophagus. The other makes use of an accessory boring organ, located in the foot, which weakens the shell at the drilling site for removal by the rasping action of the radula. The hole produced by the alternating actions of radula and accessory boring organ is slightly conical and just large enough so that the proboscis can be in- serted through it to the interior of the prey. There, apparently aided by rapid autolysis of some of the firmer tissues such as the adductor muscle (Carriker, 1955), the radula rasps away bits of dead tissue and carries them to the esophagus. As suggested above, not all types of prey must be drilled to be successfully attacked. Such behavioral modifications as are employed in attacks upon barnacles, mussels, andother forms will be mentioned later. 4, Responses to Changes inthe Environ- ment Carriker (1955) has reviewed much of the available information concerning the orientation of Urosalpinx cinerea to various cues from the environment, such as light, gravity, and current. Briefly, it can be said that at summer tempera- tures the species is negatively photo- tactic to bright light but apparently positive to dim light, geonegative and almost uniformly rheopositive. Experimental evidence has been pro- vided by Carriker (1954) which confirms many previous observations in nature to the effect that, at least in the northern part of its Atlantic Coast range, U. cinerea moves downward into the sub- strate with the onset of winter and then upward into its feeding areas in the intertidal zone as water temperatures begin to increase in the spring. The actual temperatures which initiate these movements vary with latitude (author’s unpubl. data). It is noteworthy also that females move upward more actively and in greater numbers at the respective threshold temperatures than do the males. Carriker (1955) and Hancock (1959) have both reported an upward and inshore movement of spawning fe- males. Two functions of this tendency are apparent. First, by depositing their egg capsules upon elevated topographic features, they reduce danger to their young from suffocation. Second, young Snails will be more apt to have access to food of the proper size and kind if they hatch in the richly encrusted middle intertidal zone as opposed to the lower intertidal or subtidal. The role of temperature in regulating reproductive activity in this species was discussed by both Carriker (1955) and Hancock (1959), who gave threshold tem- peratures for copulation and oviposition in various populations of U. cinerea. Minimal temperatures reported for cop- ulation are uniformly lower than minima reported for ovipositing. This disparity permits the suggestion that fertilization may occur in wintering-over places, during spring when water temperatures are increasing. Afterward, females begin crawling upward, preceding males, until their movement is arrested either by increasing intertidal exposure or by their arrival in anarea heavily populated with suitable prey. PREY SELECTION BY UROSALPINX 271 U. cinerea is only limitedly eury- haline. Carriker (1955) reported the work of several investigators who at- tempted to establish the tolerance limits of the species: it can survive hyper- oceanic conditions of salinity, but is generally not present in waters diluted by more than half. Recently I found (unpubl. data) that variations in the gastropod’s nutritive state and thermal history significantly influence its ability to survive salinities below 11 0/00, and confirmed Carriker’s (1955) assertion that the variables of “temperature and time” strongly influence salinity toler- ances. 5. Prey preference Several investigators have reported observations of the prey preferences of Urosalpinx cinerea. Carriker (1955) summarized the literature up to that time, and reported that this predator fed upon “... its own kind, slipper limpets, edible and ribbed mussels, soft and hard clams, scallops, oysters, small crabs, the carrion of fish, and on such lower invertebrates as encrusting bryo- zoans .. On the whole its diet appears to consist principally of small oysters, edible mussels, and barnacles when these are available... The effect of the relative abundance and accessibility of food species on the selection of prey is poorly understood, but it may be conjectured that these factors also influence the diet...” (р 48).. Hancock (1959), reporting on popu- ations of U. cinerea in the Thames estuary, listed as prey species the bi- valves Ostrea edulis (L.) and its spat, Mytilus edulis L., Cardium spp., Paphia spp., the American slipper limpet Cre- pidula fornicata (L.), and the barnacles Balanus spp. and Elminius modestus Darwin. He stated further that “ ... Urosalpinx feeds preferentially on oyster spat. Although barnacles are eaten, tingles [U. cinerea] have been observed drilling young oysters, mussels, and even Crepidula which were covered by live barnacles.” (p 21). Hancock cited Orton’s (1929) counsel that the “in- fluence of environment and habit should be considered” when food preferences are examined. Of interest is a comparison of food preferences of other boring gastropods. Butler (1953) cited experiments in which muricid snails (Thais haemastoma flori- dana Conrad) attacked almost all of the other sessile animals before attacking large oysters. Under laboratory con- ditions the order of preference found by Butler was mussels, spat (presumably Crassostrea virginica Gmelin), barna- cles, clams, hydroids, andfinally mature oysters, but he did not describe his ex- perimental methods in detail. Chew & Eisler (1958) and Chew (1960) investigated the food preferences of another muricid, the Japanese oyster drill “Ocenebra” (=Tritonalia) japonica (Dunker), on the Washington coast. Due mostly to deficient experimental technique, they were unable to report distinct and significant results. Hancock (1959) reported that “Ocene- bra” and Nucella, especially the latter, seem to prefer mussels to oyster spat and brood oysters. Kohn (1959), in an experimental study of the food preferences of Conus spp. in Hawaii, was able to demonstrate at least a partial correlation between the natural food (based on analysis of stomach contents) and preferences ex- hibited in a choice chamber. The crown conch, Melongena corona (Gmelin), was reported by Hathaway & Woodburn (1961) to show a preference for live oysters over shelled oyster meats, but was known to feed upon various species both living and as carrion. In most of the reports of field obser- vations of Urosalpinx cited above, little or no quantitative information is avail- able, nor generally are study techniques carefully described. Such factors as relative access and abundance related to statements con- cerning prey preferences have not been studied. There has been recently anincreasing 272 L. WOOD interest in the physiological and ecological factors underlying prey preference. These can be separated for convenience into the following cate- gories: 1. The rate of growth and metabolic rate of the prey. 2. Genetic predispositions for one prey or another on the part of the predator. 3. Factors inherent in the experience of the predator (habit, “olfactory conditioning”, etc.). 4. Teleological factors, e.g., aprefer- ence based upon some benefit which would accrue to the predator as а result of attack upon a particular prey. The first experimental demonstration of a relationship between metabolism and attractiveness was provided by Haskin (1940, 1950), who used a simple olfacto- meter to show that Urosalpinx cinerea oriented preferentially to water flowing over young, as opposed to old, oysters. The metabolism concept was extended in work done by Blake (1960), who found a positive correlation between increased oxygen consumption of prey and its attractiveness to U. cinerea. It should be pointed out that his respiratory com- parisons were made among groups of the Same prey species or, in a few cases, different but closely related pelecypod species, so that the effects of metabolic or quantitative differences might be ex- pected to have been given increased prominence, on the grounds that quali- tative, interspecific differences were minimal. Genetic determination has not been considered very seriously as a factor in the prey preferences of U. cinerea, principally because it is difficult to imagine а rigid species-specific response by a predator whose prey is so diverse, but Blake (1958) mentioned the genetic factor in connection with the preference phenomenon. In any case, genetic transmission of receptor types Specifically sensitive to the odor of only one prey species remains a theoretical possibility until it is ruled out by convincing evidence. While it has been demonstrated in several animal phyla, principally in in- sects, that the experience of individual organisms can modify their preferences for host or prey species (Thorpe & Jones, 1937), no experimental study of gastropods in this regard has been at- tempted. Nonetheless, several writers have claimed that “habit” or “ex- perience” was responsible for prey preferences in U. cinerea. Galtsoff (pers. comm.), in commenting upon the preference for barnacles by Woods Hole, Massachusetts, populations, stated that they attack barnacles out of habit. Orton’s (1929) comment on the influence of habit has already been cited. He mentioned further that Ocenebra’s feeding habits were the product of its habitat, i.e., the food organism(s) to which it had become accustomed. Cole (1942) suggested that his preference re- sults with Urosalpinx might be explained in the light of the native food of his experimental animals. Implied ascription of prey preferences to teleological factors was made by Galtsoff, Prytherch & Engle (1937), who reported that because barnacles were attacked through the opercular aperture, they were more vulnerable than prey whose calcareous valves hadtobe bored. Engle (1942) stated that U. cinerea showed different growth rates when fed upon various prey. In order of de- creasing growth rates, the prey were Mya, Ostrea (=Crassostrea), Balanus and Mytilus. Engle said that laboratory results were partially supported by field observations in which he found that large numbers of young oysters and Balanus, but very few Mytilus, were killed by U. cinerea. Hanks (1957), in a laboratory study of feeding rates of U. cinerea at various temperatures, found that preda- tors ate more oyster spat than young mussels at all tested temperatures; but stated that this could have as easily been due to the spatial arrangment of prey in laboratory trays as to a preference. PREY SELECTION BY UROSALPINX 273 6. Objectives of Present Study Prey preference is here defined as the tendency of a predator to select a Single prey type out of a number of available types, in a statistically signifi- cant proportion of given opportunities. A clear demonstration of such a prey preference would give rise to several questions of fundamental physiological and ecological significance. First, indi- cation of a consistent preference would suggest that the predator is capable of selecting one species from an array of other species, or can discriminate. Second, should such a preference come about as the result of the individual predator’s experience, and furthermore persist in that individual, it would indi- cate that a type of learning had occurred. Third, since recognition of prey can be based on nothing but chemical stimu- lation, it would follow that these stimuli differ from one prey species to another (in kind, amount, or a combination of both) and furthermore that each species would produce, consistently, an odor characteristic of that species alone. A characteristic, identifying stimulus must be persistent if its use as a discrimi- natory cue by the predator is to have adaptive value. A consistent preference need not be genetic in origin, and is not so defined. The definition of preference given above can be satisfied by a selective tendency growing out of the experience of the individual predator. But the important problems of discrimination, learning, and the nature of stimulus properties, all depend upon demonstration of a statistical preference, regardless of its causes. A failure to demonstrate significant preference would imply the inverse of statements made above. First, it could mean that the predator lacks sensory or integrative apparatus which permits dis- crimination. Or it could mean that while discriminatory apparatus is present, each discrimination is made de novo, regardless of the predator’s previous experience. Or it could mean that all prey organisms produce the same generalized stimulus, and that no one effluent compound identifies a specific prey. Another alternative is thatin each choice situation, the predator’s selection is based either on quantitative com- parison of attractant concentrations or Simply upon the presence of “the” at- tractant in the odor of one prey (and its absence in all others). Hence the first objective of the present research is to demonstrate a statistically Significant preference. Should it be demonstrated, regardless of cause, the second objective is to describe in as much detail as possible those physio- logical and ecological conditions under which the preference is exhibited. Third, if the first 2 objectives can be achieved, it should be possible to contribute a better understanding of the ways in which ectocrines (Lucas, 1955) act as carriers of information to bind together the several trophic components of marine habitats, and to make inferences about the ways in which such a predator- prey relationship has evolved. Two general approaches were fol- lowed: field and laboratory. In the first, an attempt was made to carry out a careful census of attacks on prey in nature and to relate this information to other factors in a variety of habitat types. In the second, predators were given an opportunity to make a selection of prey on the basis of effluent dis- crimination alone, with other factors eliminated (where possible) from the experimental design. П. PREY SELECTION UNDER NATURAL CONDITIONS 1. Introduction While it is true that Urosalpinx cine- vea is found in both intertidal and sub- tidal habitats, I elected to study it ex- clusively in the former for rather com- pelling reasons. The first of these is the relative ease of access to intertidal zones during periods of low water. While 274 L. WOOD underwater breathing apparatus was available throughout the period of study, long experience with its use had taught me that underwater observations and records could not be as detailed as those made on land. Second, I was interested in the effects of zonation upon prey selection, and this consideration elimi- nated the subtidal area altogether. Previous students of preferential feed- ing habits of Urosalpinx have generally reported summary statements without quantitative support, and no systematic study of prey preference in the field is known. To provide acceptable evidence, such a study should include the following categories of information: 1. Relative density of prey species populations versus frequency of attacks per species. 2. Distribution of prey species within predator’s intertidal habitat. 3. Distribution of predators in various parts of the intertidal zone as an incidental result of the combined effects of other factors, such as a. Seasonal temperature regimes b. exposure: immersion ratios (of time) c. reproductive terns. Although such complete information was not obtained from each habitat studied, this outline provided the logical basis for the list given below under “Observational Methods.” 2. Selection of Field Stations behavior pat- It would have been highly desirable to obtain observations from as great a diversity of habitats as possible (one source of diversity being wide latitudinal separation), but it was also necessary for some habitats to be examined as intensively as possible with a view toward detecting possible seasonal or annual changes. The following com- promise was adopted: several stations were visited only occasionally, andafew received concentrated attention, as will be shown in the observations. The stations were selected to cover Urosalpinx cinerea’s East coast range as shown in Fig. 1. Nobska, West Haven, Ocean City, Shark Shoal, Fort Clinch, and Nassau Sound were studied more intensively than the others. A few had such sparse populations of U. cinerea that further study of them was pointless. Others were for various reasons visited only once or twice. The Gloucester Point habitat was examined periodically, but since no deviations from a single- species diet were ever observed, specific examinations are not reported. 3. Observational Methods Field trips were planned so that stations could be visited on ebbing tides. When this was impossible, stations were examined at other stages of tide, with or without diving apparatus, depending upon water depth and temperature. Notes were taken at the station, by either the writer or an assistant, and when possible these were checked against the observations of other investigators. With some exceptions, the following records were made when visiting a station: 1. Date, time of day, and state of tide. 2. Bucket sample of water was secured for salinity determination and sur- face temperature. Air tempera- ture was taken in the shade witha dry, hand-held stem thermometer. 3. Color photograph, including a meter stick or string grid for size and density estimates, of area from which predators were collected. 4. Sketch, with metric measurements, of habitat, with notations of faunal zonation. 5. Estimate of abundance of possible prey species in each intertidal zone, 6. Collection of samples of prey species for later identification. 7. Live collection of all predators in given area of the habitat, withdiag- nosis of feeding disposition at time of removal from substrate. As- cription of feeding was made only if a close examination (with a hand lens if necessary) showed clear evi- PREY SELECTION BY UROSALPINX 00448 91... гии > 2 РА 22224 о YN 700000 < m 10100000 1110117, 141090 1909 RARA Х = LA LA LA = 11101100 o с 11101100 с > ._ VL AAAAALA x Le) - a о LA ALL ZA т o ~ o 22222222 N n aS 000001008 A on о => ИИ Х о Fe © = с PIIIIIIII т m o Bun 141040444 2 y e > о = ..1.14414 = © = n o o 1110411444 = O 3 о a © = 1111144101 oO o o о ® 9 11114111114 o © о = o = aS = 9717110101111) и о 5 ._ <= r — = 1040044111411 зе > > Е 3 3 1111014111 1040 O o - pi 5 == о о 1111111101411 — ra = a = 1110014411111) 144144 oO o oO E, an = n 111144411141 11111144 c = 111111111141 eq 11114111440 [re = o ГИГ VA 111111111141 n $3 fa о = ARARARARA AUR VERARARARA RARA ._ 3 ГИГ ИГУ - -. sr COOLLOIELELL 7% ИИ о о o wo ИЛИЯ 1011114411014 = u © <= n rm НИИ: ssss tts ии > = 7770171111077 ф 77111112 17, 711111111 a о “Y 111111114111 111111141 ПРИ LAT, u [=) ГИ DE ии 111111111111 111 11111111111111111411 447 г 111111114111 ПИЩИ ИКИ AAA LL LL LL ART ALL ALL ETAT ПГ ИИ ПИРИ 44440 CAL ll ll ddl dll ddl tt tt I 147777 Y 11111 4 ИИ ГИ ПИЩИ AL d'à ПИ ИИ И AMAR LARG ERA ERE ИИ УЧИ eee ИИ ИИ НЯ НЯ CII Id ddl ddl ddl dd ИИ 1111 dd ddl dl ИИ ПИРИ ИИ ПИЯ ИИ ии ИИ ИИ | ИИ ИИ ИИ ИИ dt tt dd td ИИ ПИ 1LI1111111111111111111111111111111144 ll 111111111111 111111111111041414417 O ИИ ИИ ИИ ИИ ИЕ О НЯ И 7 IL TI I ddl dd ll tdi dd lit tiittittt De RMS SSSR SR RIT RIRE RER LIT EI IT [LIIII1111111111111111111111111111 Œ ПИ, ПИ 0m ПИРИ ИИ НР — ПИРИ ПИ ИИ À О Я НН - | ПИ ИИ Пи ии ПИ ИИ ИИ = ПИРИ ПИ ии CAL LL II1111111111111111111111111111104111111 А LCL LL LLC LOL LOL CL OL LOL CL LOC OT LOT LOL LOT LL TTT ITS LI TITI DISS DOP. PÉCÉCL CCC CO LOL LL LOL LI OO LCOLOLL CCC LL LOL COL LU L OL LL LL LOT TOI CRIAR TODD DI DRG DRGDDGDDG И АОИ, = ОРГИИ LOL LOL OL OL LNA LOL LOT CO LTT LOL РАЙ SR AIS PS SAS A LI POS SLR AAA ESAS OR PCR DEN AA I DDR AE PIRE PR PEL) — COCO CCC OLE ECOL CMOUEYVLEOTGLEOIVTOLUSIOTGSLATFIOULUOAIDEODELUIEOPPIDLELEL EDI OD ADS DET oO DOE ORL ELE OOOO АСОИ ИО рр @ TOO PO OLR LOGE L DLLD LLC LCL CA SEL UC ДРЕЛИ = ОИСИ АИ РИА ЛИРИЙ Y SO OPS SERES GOL GOGLDOCAUPELAEVALIGIGELCELCELLUTELCOVELLOLPLLDLPEAIIPLDLDYESIDOPILOED oO, СЛОИ po 1110" Ti DP DO CC LOL GC CGE СИИ РАЛЛИ 44444 ИО ИРЛИ АЙ ® /// Y DIO LL GLE NN ES SLIDE SARA LOL LLC OL OL L CCD DADA ADA RANAS ARRASANDO CAVA PES LIVE COS EEC OLE OL TO OUL LOL ROLL LLC LL LOT TLC LOL TLCIT DLL OL LL LOST TT TT II IP IIP II OT 2227272777 7 GAO LCL LODEL LOL OLPC LL ИИ ГРААЛЯ O 011111110111 4 GRO OCR COOL COL CT LOLL OCTO LCL COL CAL L TO LOT LPC UT OIOIITIO1 СЯ En //I/I PP ИИ 4444 CALE ASD SAE EAA TES ASC SPOOL AOA IID OOD FOO IE SE OL LOE ИА BEE GOOLE LILIA LADLE AAO LIAO IL CE LOL LCL TL OT LOL TTL LL LOC L CT CT LOL TT 111 1010907 ИИ ии CUBR ОИ ИЕ ОИ АОИ "1111111111111 141 РЕВ ИРА IATA AIR DEE ADDEOL DIDS 144 И ОО В РР CCAA уг и the Atlantic Coast of the U. 5. А. , from Massachusetts 1ons on ield stat 1. Locations of 14 f to Northern Florida. FIG. irginian ions north and south of Cape Hatteras correspond to the V , respectively. The reg gions faunal re пап and Carol 10r tional behav and of “hibernation.” 9. Observations of local currents, iposi 8. Records of ov dence of drilling, or, in the case of predators feeding upon barna- cles, that the proboscis was pro- truded through opercular plates and other i features, 1c topograph 276 PREY SELECTION BY UROSALPINX relevant characteristics of the en- vironment. 4, Maintenance of Live Collections Neither predators nor prey were pre- served in the field, but were always carried back to the laboratory alive in polyethylene bags with a small amount of seawater. A large polyethylene bottle of filtered seawater was carried on board the field truck so that seawater changes could be made regularly. When summer air temperatures were high enough to make such treatment necessary, the plastic bags were placed in a large refrigerator box where a temperature of 10-15° C could be main- tained. Major field trips covering all or part of the east coast range of U. cinerea were made in June and October 1959; July, September, and October 1960; April and October 1961; May 1963; and July 1964. The longest duration of these trips, from the time of the first col- lection to the time of arrival at the laboratory, was 8 days. Shorter local trips were made on numerous occasions from each of the marine laboratories at which work was being done. These field trips, short and long, were made under a variety of weather conditions, and at no time did mortality in transit exceed 3% of the total collection. In the summary that follows, detailed station descriptions have been omitted. They may be found in Wood (1965b). 5. Summary of Observations All direct observations of attacks upon the 3 major prey types (barnacles, mus- sels and oysters) of Urosalpinx cinerea are presented in Table 1. Thebarnacles were represented by Balanus spp. and Chthamalus, the mussels largely by Mytilus edulis, but also by the mytilid Brachidontes exustus and the oysters by Crassostrea virginica. While it is dif- ficult to compare findings between sta- tions (due mostly to wide variations in conditions of substrate, climate, and prey availability), it seems to me that direct observations of attacks upon specific prey т situ are quite reliable and there- fore present an accurate picture of prey selection for any one station. The sum of 2,451 represents all attacks directly observed during 4 years’ field work at various stations between Woods Hole, Massachusetts, and Nassau Sound, Florida. It should be pointed out that if quantitative censuses had been taken during all visits to 2 of these stations (Nobska and Ocean City), this figure would have been increased perhaps by as much as 1,000, but the total relative frequencies of attacks upon barnacles and mussels, respectively, would be changed very little. The reason is that at Nobska, no attacks on mussels were observed, while at Ocean City, U. cine- yea attacked little else. Thus the sum- mary ratios (58% barnacles, 16% mus- sels, and 26% oysters) probably reflect with reasonable accuracy observed prey selection tendencies in selected habitats during the period of study. It will be noted that the selective tendency toward barnacles decreased towards the south, and a corresponding increase was observed in frequency of attacks upon pelecypods. This regional difference is compared to the obvious climatic differences which exist be- tween northern and southern stations. Differences in prey selection patterns were analyzed by the chi-square con- tingency test of Snedecor (1956) and were significant (P < 0.005). The dividing point is Cape Hatteras, following the zoogeography of Hutchins (1947) and Ekman (1953). While Tables 2a, b show what seems to be a correlation between temperature and prey selection, these 2 factors are assuredly not directly connected. Rather, to explain this apparent corre- lation, it is necessary to examine very carefully 2 other factors, namely, differences in intertidal zonation and in relative abundance of prey species within a given intertidal zone, both of which may themselves be influenced by temperature. L. WOOD 277 TABLE 1. Summary of field observations on prey selection of Urosalpinx cinerea on the East Coast of the U. S. A. (1959-1964) ee 7 Attacks observed on Not Station Barnacles Mussels Oysters Feeding Total (N to S) Dares N N N % N % N % Nobska 9.VII.60 | 110 100 0 0 al 87 197 19 V0 187 100 0 0 - - 94 281 West Haven 7. VII. 60 180 (5 60 25 — - 94 334 25. EX. 60 49 64 27 36 - - 1415 191 Aile Xe Oil i 3 28 97 - - 106 135 20 VES: | 178... 100 0 0 = = ne2 178 Lewes ils IDS 60 91 100 0 0 - - 83 174 Ocean City By 285 BY) ne 0 ne 1003 - - ne ne 11. VI.60 ne 0 ne 100 - - ne ne 2e 2:60 ne 0 ne 100 - - ne ne 21e Vile Gs (“a few”) (“many”) - - ne ne TXT. 64. 3 9 30 91 - - 600 633 6. XII. 64 0 0 0 0 - - 18 18 Gloucester 29. VII. 62 67 100 - - - - 42 109 Shark Shoal LANCE 8 14 0 0 49 86 ne 57 18. VII. 60 15 59 20 15 33 26 32 160 Wrightsville SS Ds 59 93 74 0 0 33 26 42 168 160 1 11 2/7 29 64 70 37 129 Fort Sumter DA Nebo 9 50 7 39 2 Lil 2 20 Bears Bluff 8. X.60 31 23 0 0 104 77 67 202 Fort Clinch 5. 01261 т 62 18 15 29 23 174 298 DOM IV. 61 1 5 6 28 14 67 150 171 26. V6 21 20 54 51 30 29 12 117 Pls WES 55 49 12 11 45 40 ne Le Nassau 2. VII. 60 37 25) 23 14 104 63 81 245 4. 111.61 5 29 2 12 10 59 32 49 23 Vie 0! 2 14 6 43 6 43 45 59 25 Wag OS 140 43 83 25 104 32 52 379 Total observations: 1,421 58 | 403 16 627 26 1,965 4,416 Total attacks observed: 2,451 N = Number 1(-) = Prey species not present in significant numbers 2ne” = Not counted 3Not counted, but no exceptions seen. 278 L. WOOD TABLE 2a. Prey selection of Urosalpinx cinerea in 2 faunal provinces, East Coast, U. S. A. Faunal Observed attacks PRES Barnacles X2 Pelecypods All (north to N Virginian 866 85.6 0. 005 145 14. 3 1, 011 Carolinian 855 38.5 0. 005 885 61.5 1, 440 ALL 1,421 58.0 1,030 42.0 2,451 P= probability TABLE 2b. Temperature regimes in °C in 2 faunal provinces, East Coast, U.S.A. Faunal province (north to south Annual Regimes Virginian Carolinian Summer Regimes? 1 Five-year means, 1954-58, in 0C, converted from values published in Fahrenheit. 2 Including June, July, and August. Source: U. S. Coast and Geodetic Survey Publ. 31-1 (First Ed.): “Surface water temperature and salinity, Atlantic Coast, North and South America.” U. 5. Govern- ment Printing Office, Washington, D. C., 1960. a. Intertidal Zonation Figure 2 summarizes the observed vertical intertidal distributions of the predator andits prey in selected habitats. Since total intertidal amplitudes vary amongst stations by as much asa meter, it was necessary to express the vertical heights of these zones as a ratio to base 100 of the distance above the mean low water (MLW), as estimated in the field. In this way, the relative juxtaposition of the faunal zones may be compared amongst stations. Such a comparison suggests zonation trends which, despite noted exceptions, seem clear. First, there is a general decrease from north to south in distinctness of But in the more southerly habitats ex- amined, there is extensive overlapping and mixing of barnacle and molluscan prey species. Second, as one moves from north to south there is a general tendency for the barnacle-mussel overlap zone to move upward in its relative position in the intertidal, reaching its highest ob- served point at the Wrightsville station (lower part of Fig. 2). Southof Wrights- ville, the relative vertical positions of barnacles and pelecypods characteristic of the Virginian province, withbarnacles above and pelecypods below, undergoes a gradual inversion southward, cul- minating in the nearly complete position reversal at the Nassau Sound station. zonation, In the more northerly 5 Third, vertical distribution of the stations, for example, the barnacle- predator, Urosalpinx cinerea, changes mussel overlap area is of small with latitude. In the present context, amplitude and is quite definite, at least on the outermost, exposed rock surfaces. the most important effects of this shift are two: as shown in the upper part of PREY SELECTION BY UROSALPINX 279 Height above mean low water (mlw) of faunal zones (ст) 70 189 125 110 100 KEY [9] Urosalpinx 80 height above MLW Mussels* 60 Balanus Bs] Chthamalus Gb N © © N 40 E Crassostrea © Q © 20 © © © Колокол ооо оо © © © © © © © © © O) 2 MLW Zonation as percent of total West Haven Lewes Ocean City Height above mean low water (miw) of faunal zones (cm) = 116 200 170 UTA 160 7 100 = = = © = = 5 = REO = N = = \ = fen S = = \ ic No 3 60 a = = me Ее \ a + \ © SENO NEN + Е ENTER SENO Ев Ф REN == 0 о N= lo N= 5 SE Se a NE © = = N ge : Se _ № SI 20 N= Ÿ Be = © N © Ñ MLW = = Wrightsville Fort Sumter Bears Bluff Fort Clinch Nassau Sound FIG. 2. Relative intertidal zonation patterns of Urosalpinx cinerea and its major prey in the northern Virginian province (upper diagram) and the southern Carolinian province (lower dia- gram). Stations are listed from north to south. For easy comparison of distribution patterns, varying vertical distances between high and low water (70-200 cm) have been converted to a common base (100). *Mussel means Mytilus edulis in the Virginian province and Brachidontes exustus in the Caro- linian province. 280 L. WOOD Fig. 2, the relative vertical positions of the barnacle-mussel overlap and that of Urosalpinx differ markedly between the Nobska and Ocean City stations. At Nobska, mussels are few and located in the lowermost intertidal; at Ocean City, mussels are abundant and extend nearly to the middle intertidal. Hence in the northern habitat Urosalpinx is conzonal primarily with the barnacle Balanus balanoides, while at Ocean City in the south it coexists primarily with the mussel Mytilus edulis. Furthermore, at all observed habitats as far south as Fort Sumter in the Carolinian province, there was evidence in favor of the seasonal vertical migration of the predator mentioned on p 270, though at Fort Sumter the evidence was incon- clusive and derived from a single station visit. South of Fort Sumter, it was apparent that no Seasonal migration oc- curred. This radical variation in season- al behavior patterns between northern and southern Urosalpinx populations is almost certainly due to climatic differ- ences, shown in Table 2b. Variations on the part of the prey species may also be due to climate, andthis thought brings up a rather startling omission in the marine biogeographic literature, for very little work seems to have been done concerning the influence of temper- ature upon vertical intertidal zonation. Hutchins (1947) discusses in detail the geographic distribution of Balanus bala- noides and Mytilus edulis with respect to temperature regimes, but does not mention latitudinal differences in temperature regime as a factor in sub- zone variations between cold-temperate and warm-temperate habitats. Doty (1957) notes that temperatures are im- portant in establishing features of vertical zonation, but does not specify the effects of latitudinal differences. Moore (1958) examines the relationship between sea and air temperatures ina single intertidal habitat, but does not examine their influence upon intertidal zonation, nor do Wells & Gray (1960), in a study of subtidal oyster populations immediately north and south of Cape Hatteras, although they cite the tempera- ture gradient which exists there. While the causes of such latitudinal differences in vertical zonation are only of tangential interest in the present work, the observed fact that the differ- ences exist is central to it. As shown in Fig. 2, there is a progressive rise in the Balanus-Mytilus overlap until, finally, at Ocean City, the major part of the Balanus population is above the bulk of the predator population, andthe preda- tor is essentially conzonal with M. edulis. In nearly every habitat studied in the present work, furthermore, it was possible to relate the modal distribution of the predator with the modal distri- bution of its statistically “preferred” prey. Only the exceptions to the general trend need to be noted here. At Shark Shoal, in the summer of 1959, an apparently atypical set of the barnacle Chthamalus fragilis coated the upper surfaces of the rocks forming the seaward part of the jetty. (While this set is here termed “atypical,” it should be noted that a similar set was observed by T. A. & A. Stephenson [1952] at Shark Shoal in 1947.) Thus while C. fragilis was numerically the dominant potential prey organism inthe upper part of the jetty inhabited by the predator, it was apparently not attacked. In several years of observing the feeding habits of this predator, I have never seen an attack upon Chthamalus under natural conditions, andthis same observation has been reported (Barnes, pers. comm., 1959; and Crisp, pers. comm., 1959) in the case of the British Urosalpinx cinerea. Indeed, Chthamalus is usually found so far above the normal intertidal range of the predator that it would seem remarkable if it were attacked. In any case, the fact remains that during the summer of 1959 predators were rarely found on upper surfaces of the rocks with the abundant C. fragilis. Hence they were feeding upon numerically secondary Crassostrea virginica on the sides and lower surfaces of boulders. In the following summer of 1960, the Chthamalus fragilis population at Shark Shoal was overwhelmed by a dense set of Balanus amphitrite. With Chthama- lus gone, predators occurred in abundance upon the upper rock surfaces, PREY SELECTION BY UROSALPINX 281 and were feeding largely upon the domi- nant Balanus. The second noted exception was seen in April 1961 at Fort Clinch. Here, the predator was conzonal with both Balanus spp. and Crassostrea, and to a lesser extent with the mussel Brachidontes ex- ustus. Though barnacles were numeri- cally dominant on the April 1961 visit, most adults seemed to have been recently killed (the suspected cause was industrial pollution of the water by a pulp mill several miles upstream), and only a few Urosalpinx were active. Most survivors appeared to be in distress and were clustered within crevices and between encrusting faunal clumps in a manner reminiscent of this snail’s wintering be- havior in the northern part of its range. Those few which were feeding were in the upper part of the middle intertidal zone, where Crassostrea virginica was dominant. The third exception was observed the following day at Nassau Sound. Here, though the numerous C. virginica were dominant in the upper part of the middle intertidal zone, predators were feeding in equal numbers upon the oysters and on Brachidontes exustus. b. Relative Prey Density In no case observed during the field work reported here were predators feeding preferentially upon a prey form not abundant or common in the habitat. The question of effects of relative prey density upon prey selection was difficult to resolve because of irregular prey distribution and the consequent difficulty of quantitative analysis in most mixed- prey habitats. Theresults of one attempt to secure a rigorous Statistical analysis are shown in Fig. 3, and the techniques for it are given below. The cobbled beach portion of the West Haven habitat in 1960 possessed both major prey species in varying relative abundance. In an attempt to determine the effect of such density variations upon prey selection, I selected 6 quadrats 40 cm on a side and photographed them in color, the field of view coinciding with the quadrat. Three examples were chosen of each of 2 types of relative I80r (On Barnacles) calculated attacks observe dE attacks 160 140 80 60 (On Mussels) Attacks Upon Prey Indicated 40 (On Barnacles) 20 Mussels dominate Barnacles dominate FIG. 3. Correlation of attack frequency (=ob- served attacks) and prey density (=calculated attacks) in non-zoned, mixed prey habitat at West Haven (Connecticut) field station. density: Mytilus -dominant, and Balanus - dominant. As predators from each quadrat were removed, the prey species which they had attacked was noted. In the laboratory, the color slides were placed under a dissecting binocular microscope with transmitted light, and individuals of the 2 prey species were counted. On the basis of these counts, a ratio of expected attacks was calculated, assuming no Selectivity. 282 L. WOOD Figure 3 shows the means of the 3 samples from each of the 2 types of quadrat chosen (Mytilus -dominated, Ba- lanus-dominated). These “calculated attacks” were compared to the means of the actual attacks observed in the 2 types of quadrats, and the differences between them were analyzed by the chi- square method. No significant difference was found. A similar analysis was per- formed with the data obtained from a later visit to West Haven with much the same result. 6. Discussion and Conclusions A superficial inspection of the data contained in Tables 1 and 2 would lead one to conclude that (1) there is a significant total preference for the genus Balanus over all pelecypods by intertidal Urosalpinx cinerea along the east coast of the United States, but that (2) there is a shift from a Balanus preference in the northern latitudes to one for pelecypods in the southern part of the predator’s range. A more careful scrutiny of the data, however, has led me to make thefollow- ing tentative conclusions: 1) In mixed prey habitats, the fre- quency of attacks upon a Specific prey type is a function of the relative density of the prey species present in the habitat. 2) In habitats where prey species are separated from one another asthe result of intertidal zonation, prey selection is an incidental product of the co-existence of the predator with whatever prey Species is dominant in its zone. Note that olfaction and olfactory dis- crimination have not been mentioned in connection with the foregoing field work. The reason for this is simple: no evidence obtained from field studies bears directly upon these concepts. We see upon going into the field that U. cinerea has a certain intertidal range and that this range may place it in company with one or more potential prey species. We see further that where the prey species are mixed, it is possible under certain conditions to predict the relative frequencies of attacks upon those species. However, it is still possible that such co-existence of predator and prey is itself the product of purposeful movement into a specific zone by the predator, and that, further, this move- ment is elicited by the production of chemical attractants by the prey. Hence one cannot know, on the basis of field information alone, whether a certain ratio of barnacie-preferring predators is present in abarnacle zone fortuitously or as the direct result of attraction to the prey. Therefore, conclusions based upon field studies must be qualified until the results of experimental inquiry can be presented. Certain specific questions, based partly on field observations and partly on hypothesis, can be asked: First, can it be demonstrated beyond doubt that Urosalpinx cinerea will make an orientational response tothe effluents from prey species inthe absence of other orientational cues? As much tothe point as anything else is whether, in the con- fused current situations found in nature, such orientations can still be made. Second, is it possible that the predator’s past ingestive experiences т- fluence such orientational responses? This is a question which follows naturally from the observations made at Nobska and Ocean City, where U. cinerea ex- hibits a persistent tendency to continue feeding upon the same prey (different Species in each case), despite op- portunities to change diets. These and other questions will be the subject of the chapters which follow. Ш. EXPERIMENTAL METHODS During the period covered by the field studies described above, experimental investigations of these predator-prey relations were being designed and con- ducted ina series of marine laboratories. For convenience these studies will be referred to in an abbreviated form by letters followed by numbers, the former indicating the place where they were done and the latter indicating their chrono- logical sequence. The place names designated by the abbreviations and other details are explained below: FWS. This series was carried out during the summer of 1956 at atemporary PREY SELECTION BY UROSALPINX 283 РИ ея: Арравети $ Se Male Reservoir | | Overflow ГСС rene ro Cells „2221 Aeration for À Column | (he Air > FIG. 4. Simplified diagram showing passage of seawater through filter and aeration column into temperature-controlled reservoir. From the reservoir, seawater flows by gravity through clear plastic prey cells (at least one of which is always kept empty of prey, as a control) and thence through in-line flow meters to the peripheral compartments of the olfactometer shown in FIG. 5. For a diagram of the more complicated controlled conditions system used in the VIMS Series, UF Sea rt [to Olfactometer] Flow Rate Controls see Wood, 1965a. seawater laboratory set up during my brief period of employment by the Clam and Chesapeake Oyster Investigation, Bureau of Commercial Fisheries, Fish and Wildlife Service, Department of Interior. The laboratory was located on Chincoteague Island, Virginia. IFR. These experiments were car- ried out in the summer and fall, 1959, and terminated in summer, 1960, while I was working in the laboratory of Dr. M. R. Carriker at the Institute of Fisheries Research of the University of North Carolina at Morehead City, North Carolina. OI. This series was conducted during the academic year 1960-61 while I wasa visiting investigator in the Alligator Harbor Marine Laboratory of the Oceanographic Institute of Florida State University, Tallahassee. VIMS. These experiments, beginning in the summer of 1961 and continuing through to the present, have been con- ducted at the Virginia Institute of Marine Science, Gloucester Point, Virginia. The presentation which follows is organized by topic rather than chro- nology, but the sequence of the work can be understood from the experiment numbers. Similarly, the section on methods and materials which follows will describe in a general way the experimental techniques employed throughout the investigation. Departures from these general methods will be specified where necessary. 1. General Method The most important criterion for ex- perimental analysis of prey selection is that the predator be allowed to select prey on the basis of a single variable, while other conditions are kept constant. In a system as complex as the marine environment, it is not always easy to satisfy this criterion, as will be seen in descriptions of my successive attempts to attain an ideal experimental procedure. The procedure used in all olfacto- meter experiments on prey selection 284 L. WOOD consisted in placing prey in plastic or glass cells in such a way that incoming seawater had to flow through and around them (Fig. 4) before reaching the “choice” apparatus (Fig. 5) containing the predators. Thus theoretically preda- tors would orient solely on the basis of chemical stimuli emanating from prey organisms. In most cases, one or more of the “prey” cells were used as “con- trol” cells with seawater only from the same source flowing through them tothe predators. Flow rates through prey cells were adjusted until they were equal; prey populations were placed in the cells an hour or 2 beforehand to permit their adjustment to a new environment; temperature and salinity determinations were made; and, finally, predators were placed inthe olfactometer’s central com- partment and arun was started. Duration of a run varied with temperature (Series OI-3), was set arbitrarily at 30-60 minutes (Series VIMS), or the run was declared ended when an arbitrary pro- portion, usually 3/4,of the predators had made a choice (IFR and Ol). Criteria for “response” (r) were rigid. In the original cross-type olfactometer, snails simply entered a peripheral com- partment. Shortly after the beginning of Series IFR, this model was improved (Fig. 5) by adding a smoothed-curve, inclined ramp before each peripheral compartment; responding predators now had to climb the ramp and passa thresh- old mark which was above the central compartment water level. This meant that, to be counted, individual U. cinerea were required to crawl up a ramp from the central compartment and go up a shallow (1-2 mm) stream of odor-laden water into a peripheral compartment. That so many performed this rather unnatural feat speaksforthe attractive- ness of prey ectocrines. 2. Olfactometers Several olfactometers were employed in this study, and something of their evolution through successive stages of development should be mentioned. An olfactometer, by definition, is designed to permit one or more subject organisms to indicate selection of an odorous substance from a neutral back- ground, or to make a choice of one amongst several odors. The indication itself is usually a matter of orientational movement by a free and intact subject, though of course it is possible, and sometimes more convenient, to assay olfactory stimuli by non-orientational behavior, such as proboscidal extension (Carr, 1967). But in the case of Uro- salpinx cinerea the latter response 1$ not feasible and orientational selection must be employed. From discussions with D. Davenport it appeared that the classical “У” or “Т” maze was not satisfactory on 2 counts. First, depending upon the arrangement of the subject’s chemo- receptors versus its dimensions relative to the maze tube or trough, itis possible that the subject could be stimulated by effluents from only 1 of the 2 arms of the maze. Second, only 1 subject can be tested at a time, so that a great deal of time is spent (especially with slower moving forms, such as snails) securinga statistically useful series of tests. To Davenport’s objecticns can be added a point raised by Putnam (1962), whofound that Aleochara (Coleoptera) tended to repeat inital selections of either right or left arms of a Y-maze, despite the absence of reward or punishment in either arm. Thus non-randomization of successive responses might be based upon cues (either proprioceptive or external) not perceivable by the in- vestigator. At least one way of diminish- ing errors arising from this possibility is to increase the complexity of re- Sponses required by the olfactometer. Davenport’s suggested alternative to the T- or Y-maze was a pie-shaped device consisting of several peripheral compartments surrounding a central drain. Subjects were placed inthe center, the water from prey and control cells was run into peripheral compartments, and subjects were to indicate selection by moving from the center to the chosen peripheral cell. This device was used in trials preliminary to Series IFR, but it was not suitable for slowly moving animals and was rejected. Davenport (pers. comm.) had come to a similar PREY SELECTION BY UROSALPINX 285 decision independently, though Kleere- koper (1961) successfully employed an elegantly automated variation of the Davenport idea in his study of predator- prey relations of the lamprey Petro- myzon marinus and the lake trout, and Sastry & Menzel (1962) used yet another adaptation inaninvestigation of commen- sal relations between the scallop Aequi- pecten and a pinnotherid crab. For a recent thorough discussion, see Davenport (1966). Before Series IFR experiments were finally begun in the fall of 1959, a new olfactometer was designed and con- structed. It was used for all Series IFR experiments, and modified only Slightly for Series OI and VIMS. The design was based upon the follow- ing criteria: 1. There must be a central compart- ment, or starting point, large enough for a sample of at least 25 adult predators. 2. Chance of exposure to water from prey or control cells must be the same for all predators, regardless of their initial placement in the central compartment. 3. Water streams from outer com- partments must be thoroughly mixed in the center compartment, but there must be a point of choice at which the subject may compare a Single “pure” stream with the background mixture. 4. This comparison must result in clearly enumerable selections of prey compartments by subjects. The design which followed from these criteria is shown in Fig. 5. It con- sisted of 4 peripheral compartments, with channels leading into a center one in such a way that a circular current pattern is set up around a central floor drain. Water flowing in from the pe- ripheral compartments is thus entrained in a clockwise whirlpool, and makes several rapid circuits before leaving through the drain. This olfactometer has proved to be a sensitive and accurate device for assaying the attractiveness of prey effluents. 3. Identification of Individual Predators U. cinerea used in experiments in Series IFR, OI, and VIMS were marked for identification by means of a 4 color, 5-digit code. Thus a total of (4x5)2 = 400 identification numbers were available, which exceeded the number of individuals employed in any given series. For convenience, individual predators were further coded as to their sex, place of origin, and selected prey, if any, at time of capture, throughuse of additional color dots placed variously upon their shells. Since it was necessary to dry shells completely prior to application of fast- drying enamel colors, portions of the Shells to be colored were scrubbed, rinsed in distilled water, dried, put through 2 rinses of 95% ethanol, and dried again. A few minutes after color dots were applied, the snails were back in seawater. Apparently no mortality ever resulted directly from this process. 4. Predator Maintenance Predator groups were kept in aquaria supplied with running seawater in all experimental series. Except for studies in which special feeding was required by experimental protocol, predators were not fed while in captivity. Therefore the only food available was that which they could browse from the sides of their containers. Urosalpinx cinerea is a species which can endure long periods of food deprivation even at summer temperatures; during the course of this study, some individuals were known not to have had access to prey for about a year. Upon dissection, some of these were found to be apparently normal except for a loss of “non- essential” body tissue. Individuals of the smaller varieties were not as resistant to starvation as the large snails from the Eastern Shore of Virginia and Maryland, but nonetheless survived periods of gross food depri- vation of 3-4 months. In only 1 experi- ment (OI-3) was mortality a serious problem. It was then primarily re- stricted to a sample from 1 habitat, Nassau, and could not be ascribed to mal- nutrition. In other respects, laboratory popu- lations of U. cinerea were apparently 286 L. WOOD unaffected by captivity. Copulation and oviposition were observed, andthe snails added shell. With few exceptions, food- deprived captive predators displayed rheotactic responses comparable to those of controls which had been fed (see discussion of errors below). Numerous precedents (Haskin, 1950; Carriker, 1955; Blake, 1960) have es- tablished the relative ease with whichU. cinerea can be maintained in captivity. 5. Possible Sources of Experimental Error Since it has been shown (Carriker, 1955) that Urosalpinx cinerea responds to a diversity of external stimuli which could conceivably act as intervening variables in a study of prey selection, considerable care was taken to ensure control of experimental conditions. Measures adopted to this end will be described below. At least 4 physical factors could in- fluence orientation of predators in the olfactometer: temperature, light, gravity, and current. The first was an experimental variable in the Series OI studies, and detailed discussion of its effects will be saved for a subsequent paper. For the present, it'need only be pointed out that ambient seawater temperatures prevailed in Series FWS and IFR, and in the VIMS experiments temperatures were controlled at about 25° C. Water temperatures within the olfactometer did not vary between parts of the apparatus and therefore can be neglected. a. Light It has been shown that Urosalpinx cinerea is sensitive to light. Carriker (1955) summarized reports of other workers in this area (Federighi, 1931; Sizer, unpubl.; Stauber, unpubl.; Cole, 1942): at strong intensities, U. cinerea is negatively phototactic; at weaker intensities, it is positively phototactic; and in “dim light” the phototactic re- sponse is apparently extinguished. Two methods of coping with light were employed. In one, the room was darkened, or a black, opaque hood was placed over the olfactometer; in either case, the responses were made in dark- ness or near darkness. In the second, a broad, diffuse light was placed directly over the olfactometer so that all surfaces were evenly illuminated at an intensity of about 1 foot-candle. The latter method was regarded as the more effective, be- cause in complete darkness, the predators’ activity decreased. In addition, preliminary trials were also run in which a strongly directional light source was placed at one side of the olfactometer to determine maximum effect of such a situation. Regardless of relative orientations of the light vis-a-vis prey compartments, there was no Significant variation in orientation of predators which could not be ascribed to chemotaxis. \ b. Gravity Carriker’s (1955) report indicates that U. cinerea tends to creep upward (geo- FIG. 5. Olfactometer used in all but FWS Series of experiments (shown in FIG. 6). Rounded countours were hand molded in plaster of Paris after which the entire inner surface of the device was coated with inert epoxy paint. Seawater enters square peripheral compartments from corresponding prey cells (or a blank control cell, as shown in FIG. 4) and thence down an inclined ramp where complete mixture takes place in the clockwise gyre of the central compartment before it leaves through central drain. In picture a, India ink is placed in the stream going down the ramp and its progress around the central compartment can be followed in pictures b through d. Passage of materials is so rapid that the olfactometer can clear itself of an ink cloud in about 10 seconds. For this reason, the 4 pictures given here do not represent a sequence from a single dye test, but were selected from several photographed tests in such a way that stages of dye transit through the apparatus are depicted. The response criterion threshold mentioned on р. 284 is at the point of the arrow in picture а. Snails were not counted as responding until after they had crossed this point. PREY SELECTION BY UROSALPINX 287 FIGURE 5. 288 L. WOOD negative) at warmer temperatures and downward when the temperature de- creases beyond the “hibernation” thresh- old. This threshold has been variously reported at between 10 and 15° C. Presumably, at temperatures em- ployed in the present study (15-about 33° C) predators would have been nega- tively geotactic. Their tendency to climb up walls of the olfactometer was early recognized, and standard pro- cedure for dealing with this problem was to remove them from the vertical wall and place them at the bottom of it. Since this procedure was applied to all subjects, since, further, the ramp approaches to all 4 compartments were of nearly identical grade, and since finally the apparatus was carefully leveled prior to commencement of each day’s trials, it was thought that the geo- tactic error was minimal. c. Current It was observed many times in the present study that most predators in- variably turned upcurrent in the ol- factometer whether or not the current was known to be laden with attractant. A few individuals, however, were con- sistently rheonegative (i.e., moved in the same direction as the current); none of these individuals ever made a positive prey selection during an experimental run. While the flow of water into each of the peripheral compartments of the ol- factometer was equalized as nearly as possible by volume/time measurements (usually 250 m1/min into each compart- ment), changes in water pressure and occasional clogging of water lines during a test caused variations in flow rates which were sometimes considerable. Furthermore, even when flow rates into peripheral compartments were identical, there was no effective way of determining differential current ve- locities amongst separate portions of the central compartment. Therefore I decided to test the effects of different flow rates upon prey se- lection, preliminary to Series OI trials. These tests showed that when all cells were empty of prey organisms, the input flow rate difference had to be at least twofold before a significant selective tendency for the higher flow rate com- partment could be demonstrated. On the other hand, the outflow rate from cells containing attractive prey organ- isms could be reduced to nearly zero with little reduction in responses tothose cells. I concluded that a marked current velocity difference could alter selection results, but only when prey were relatively unattractive or when differ- ences in attractiveness of 2 or more prey species were very slight. d. Size and Sex of Predators It has been established (Cole, 1942; Carriker, 1955) that female Urosalpinx cinerea grow more quickly than males and reach a larger size. My ownobser- vations bear this out. It is reasonable to ask whether a prey preference might change as the predator grows’ and matures, or whether selective tendencies might differ between males andfemales. There is every reason to think that food selected by young predators would differ from that of mature individuals, just on the basis of relative size. Newly hatched U. cinerea, with a spire height of about 1.0 mm, could attack a small hydroid or ectoproct, or a recently set barnacle or oyster, with some degree of expected success (pers. obs.). But ona purely mechanical basis it probably could not successfully attack an adult oyster. On the other hand, would there be similar behavioral differences between a predator with a spire height of 15 mm and a 30 mm predator? While presumably the difference in the size of proboscis would reflect those of the predator’s general development, it must be re- membered that in these olfactometer tests, we are dealing not with a gastro- pod’s attack and ingestion armament, but with the capability of its chemo- receptors and central nervous system. Hence in order to say that there would be differences between small and large predators we would have to find changes in receptor systems at some point inthe snail’s development. This in itself con- stitutes an extended investigation, and therefore it was decided to analyze PREY SELECTION BY UROSALPINX existing data for the purpose of de- termining what correlation there was, if any, between prey selection and size or sexin U. cinerea. During Series IFR experiments, the Wrightsville habitat with its distinctly separated Balanus and Crassostrea zones was discovered. Since both prey species were about equally accessible and existed in somewhat equivalent densities, information con- cerning the role of predator size in prey selection could be derived from a study of this population. Therefore every active predator seen was collected, and later, at the laboratory, its spire height was measured and its sex determined. It was found that male-female size differ - ences were greater than size differences between groups feeding upon Balanus spp. and Crassostrea _ virginica. Next, samples of 15 of each sex and original prey group (total of 60 animals), were put through 5 prey selection tests in the olfactometer. The results of these tests coincided withthose derived from nature: minor differences in spire height cannot account for differences in prey selection. On the other hand, inferences drawnfrom the VIMS-37 experiments with young snails indicate that orientational ability may develop after they have hatched. The question of sex-linked differ- ences was solved by an analysis of the data from Series IFR. There were no consistent and significant differences in prey selection which could be attributed to sex. When Series OI experiments were being planned, therefore, it was de- cided that since males and females did not differ significantly in selection tendencies, only males would be used. The advantage of using only males was twofold. It has been reported by Peters (1964), but not confirmed by Gibson (1964), that male Littovina planaxis Phillipi locate females by chemotactic orientation; in any caseafirst advantage was gained by denying male U. cinerea the diversion of females in breeding condition. Second, it had already been noted that when females were ready to oviposit, they would do so regardless of circumstance. Several times during 289 Series IFR, at the end of a particularly long run, it was found that a female had not moved from her original position and had deposited an egg case upon the floor of the central compartment during the run. To summarize: in the experiments of Series FWS, IFR, and OI, only adults were used, while in Series VIMS other stages were employed for specific experiments. As to sex, both were used in Series FWS and VIMS, males and females were run separately in Series IFR, and in Series OI, only males were used, d. Orientation of Predators to one Another At least one gastropod, Nassarius obsoletus, exhibits a type of “schooling” behavior in which aggregations are ap- parently maintained through the medi- ation of chemoreception (Jenner, 1959). It was therefore necessary to determine whether or not any suchbehavior patterns were characteristic of Urosalpinx cine- rea, since if they were, all experiments in which groups of predators were used would be open to extreme doubt. One group of 25 U. cinerea was placed in the central compartment of the olfacto- meter. Prey cells were thoroughly cleaned, and in one of them I placed an estimated 2,0000. cinereafromthe same laboratory stock. There was no signifi- cant response in several repetitions of the test. It was concluded that, at least in the experimental situation adopted in this study, the possibility of a “schooling” tendency based on distant chemo- reception could bé discounted. But distant chemoreception is not the only means whereby one predator can follow another. Gastropods character- istically secrete a continuous sheet of mucus from glands in the propodial groove (Peters, 1964). This mucous layer functions as an adhesive and a lubricant, and while relatively soluble in saline solution, it remains applied to hard substrates for several hours. In the olfactometer, its presence could be detected because of the adhesion of minute water-borne particles of debris. It would have been impractical to attempt erasure of this mucous track witha swab 290 L. WOOD immediately after the passage of each snail from one point to another in the olfactometer, so I decided to determine the effect of the track by means of direct observation: though a succession of predators selected the same peripheral compartments within the space of a few minutes, the tracks of those which followed did not necessarily coincide with those laid down. Nonetheless thorough cleansing of the olfactometer after each test was standard experimental pro- cedure. e. Effects of Food Deprivation Upon Predators Experimentalists sometimes fall into traps of their own making: in the case of some of my early experiments, it was unavoidable. On the one hand, it was necessary to deprive experimental snails of prey on the grounds that to do otherwise was to risk habit formationor olfactory conditioning (Thorpe & Jones, 1937). On the other hand, there was a chance that food deprivation per se might either alter the behavior of the subjects, or impose upon them nutritional stress. Ideally, some means should have been devised for controlling thisfactor. How- ever, it was not possible to schedule selection runs at the same stage of food deprivation for each predator, andit was also impossible to obtain a food which was simulataneously palatable to the snails and free of olfactory stimuli. f. Learning During Experiments When a single subject is used only once, interpretations of results are not hampered by questions of the degree to which the subject was changed by the experiment. For most experiments in the present study, subjects were marked for identification and deliberately re-run several times (up to a maximum of 14 runs in Series OI), precisely for the purpose of determing the extent of in- dividual variations. Most theories of animal learning de- pend in some way upon the idea of re- inforcement, by reward or punishment, of a specific behavioral item. It follows from this that non-reinforced behavior tends not to become a permanent part of the animal’s behavioral patterns. Application of this idea to the present problem would result in the logical con- clusion that because selections were not rewarded, no bias could be placed upon the predator’s later selections. (For a review of the theory of reinforcement, see Brogden, 1951.) At least 2 studies of invertebrate be- havior, however, cast some doubt upon this conclusion. Thorpe & Jones (1937), in a study of host selection in parasitic insects, concluded that exposure to a certain chemical environment at some time during an organism’s life cycle tended to increase responses to those chemical stimuli in laboratory tests. In this and later investigations, Thorpe (1938, 1939, 1956) explored the nature and function of olfactory conditioning as a type of latent learning. Though latent learning has been given only limited attention (homing in lim- pets, Thorpe, 1956) among the molluscs, and no record has been published of olfactory conditioning in this group, it was thought that the possibility could not be ignored. Even without formal reinforcement of a selective response, it was possible that exposure (in a peripheral compartment of the olfacto- meter) to “pure” prey effluent for several hours might lower response thresholds. Since no way of avoiding this situation was found, experiments were designed for the specific purpose of investigating the effects of prolonged immersion in the chemical environ- ments produced by prey organisms. These will be reported separately be- low (VIMS-36c, p 305 and Table 5). The second study which raised the question of learning during experiments is that reported by Putnam (1962) and mentioned briefly above in connection with olfactometer design. Putnam found that 58.9% of his coleopteran subjects, in 20 successive, non-reinforced runs through a Y-maze, chose the same arm PREY SELECTION BY UROSALPINX 291 of the Y at least 15 times. [If their selective behavior had been random, only 4.14% would have done this. There was no group preference for either arm (r for left arm = 717; r for right arm = 743), so the possibility of there being a constant difference between the 2 arms, not detected by the investigator, must be discounted. Analysis of each in- dividual’s successive choices suggested that either the initial choice somehow “programmed” the later ones, or there existed in this sample the kind of genetic “right-handedness” ог “left-handed- ness” reported for other organisms, particularly mammals (Morgan, 1951). Because of the complex design of the olfactometer used in the present investi- gation, the latter eventuality would pre- sent no particular problem. But the possibility of programming is real. Chew & Eisler (1958) and Chew (1960), in studies of prey selection in “Ocenebra” japonica, reported that snails tended to repeat their initial selections regardless of the prey being attacked at time of capture. Chew (1960) failed to subject this finding to more than a cursory analysis and discussion, but did present his original data. To detect possible influence of early choices upon later ones, a consistency analysis was devised and applied to the results of experiment OI-3, which will be presented later (p 305-307). g. Contamination of Control “Blanks” Prey species employed in experiments to be reported here have in each case been the same as, or closely related to, species living outside the laboratories, near or upon the seawater system’s pump intake. Hence it can be assumed that “natural” attractants in varying concentrations would be brought in with incoming seawater, and that this back- ground of “chemical noise” in control compartments could easily diminish the Significance of experimental results. Steps taken in each series to reduce such interference are described briefly be- low. FWS. No preventive measures; sea- water pumped directly from Chinco- teague Bay into experimental apparatus. No controls run simultaneously with prey experiments. IFR. Seawater intake pipes and non- return valves cleaned and flushed regularly. Control water run through absorbent cotton and activated charcoal. OI. Seawater taken from a large reservoir filled at high tide and cleaned as needed. Experimental water passed through aeration column and shell frag- ment filter. VIMS. Seawater system _ cleaned regularly and steamed. Experimental water processed through shell fragment filter, activated charcoal, aeration column, and aeration reservoir, as de- scribed elsewhere by Wood (1965a). IV. ORIENTATION INCOMPLEX CURRENTS 1. Introduction The question of whether Urosalpinx cinerea can orient to its prey in the kinds of confused currents found in nature is of suchfundamental importance that it should be presented first. The experiments in series FWS were con- ducted with a rather primitive olfacto- meter, later discarded for the very defect that made it suitable for the present question: current patterns were so confused that predator responses were usually quite attenuated. 2. Materials and Methods Figure 6 is a diagram of the olfacto- meter used in series FWS. Water from a continuously flowing seawater system was introduced through a “T” into 2 prey compartments. (P, Fig. 6). A different prey was placed in each of these com- partments. Water flowed out of the prey compartments into the predator com- partment (U) by way of 6 acrylic plastic tubes 2.5 cm in diameter arranged to cause water currents to converge at point “X”, which was also the center of the starting area for predators at the 292 L. WOOD A Sea water FIG. 6. Olfactometer used in FWS Series. Compartments P, separated by a partition, hold groups of prey animals. Seawater flows through prey compartments from “tee” fitting and out into predator compartment U through 2.5 cm diameter acrylic plastic tubes. Ar- rows in large compartment U indicate major current patterns as revealed in dye tests. At start of olfactometer test, snails are placed around point X, where current streams from the prey compartments converge. beginning of a test. The entire apparatus was coated with inert black paint and covered with an opaque lid so that tests were conducted in total darkness. Water flow into com- partments “P” was equalized by volume/ time determinations, and the apparatus was leveled before eachtest run. Fifty predators, of about equal size, were used each time, a run lasting 1 hour. Seawater temperatures during experi- ments ranged from 21-290 C. Salinities were determined with a hydrometer calibrated by the U. S. Bureau of Standards, and exhibited a range of 25-28 0/00. Predators were collected from sub- and intertidal seed oyster beds and brought to the laboratory within a few hours, where they were maintained with- out food in running water at ambient temperature and salinity. Prey test samples were young Mytilus edulis from Ocean City and Crassostrea virginica from local seed beds (called “rocks” by Chincoteague oystermen). Prey groups were changed after each experi- ment, and approximately equal amounts of tissue, by whole live weight, were used in each test. In experiments FWS-1,2, 328 preda- tors were used in groups of approximate- ly 50, with no prey added to the olfacto- meter. In FWS-5, 120 predators were tested against a mixture of both prey Species in the 2 prey compartments. In FWS-6, 2,250 predators were tested in groups of 50 with mussels in one compartment and oysters in the other. As each group finished, those individuals which had successfully crawled through the acrylic tubes and into prey com- partments were counted and then kept in a separate container; those which had remained in compartment “U” were Similarly kept apart. In experiments FWS-7,8, predators were randomly se- lected from the groups which had made successful orientations in the previous experiments on the one hand, and from non-responders, on the other. After each test, prey groups were re- moved from compartments “P” and the entire apparatus was scrubbed and flush- ed with seawater. Then fresh prey groups were so placed that they occupied the prey compartment opposite that the species had occupied in the preceeding test. Several minutes’ adjustment time was allowed for the prey to open and begin pumping before a new predator group was placed in compartment “U” around point “X” and a new test run started. 3. Results and Discussion In this experimental series, predators displayed no significant preference for either of the 2 pelecypod species tested. The important question here is whether Urosalpinx cinerea canaccurately locate prey when olfactory trails are confused by large eddies. Hence orientational behavior in an empty olfactometer was compared to that when prey com- partments had mixed prey inthem. Re- PREY SELECTION BY UROSALPINX 293 TABLE 3. Predator responses in complex currents (series FWS) Predators i. x2 Predators Experiment Responding p= Not Responding 7 File N + % N % FWS-1, 2 (no prey) 159 48.5 NS* 169 51.5 FWS-5 (with prey) 91 INS < 0.005 29 24.2 FWS-6 1134 50.4 NS* 1116 49.6 FWS-7 (400 responders from FWS-6) 324 81.0 < 0.005 76 19.0 (300 non-responders from FWS-6) 124 41.3 < 0.005 176 58.7 FWS-8 (448 responders from FWS-7) 339 Dent < 0.005 109 24.3 (252 non-responders from FWS-7) 78 30.9 < 0.005 174 69.1 *NS = not significant P = probability sults are shown at top of Table 3: with no prey, about as many snails moved into prey compartments as stayed out- Side in compartment “U”. With mixed prey, a highly significant majority (75.8%) of the snails moved into a prey compartment. Despite significant results from ex- periment FWS-5, I thought conclusions would be on firmer ground if additional tests were carried out with larger samples of predators. Also, it seemed necessary to ask whether failure of nearly 1/4 of the animals in FWS-5 to make successful responses was due to chance or to a characteristic of in- dividual predators and would, in re- peated tests, manifest itself in their continued failure. Therefore a large sample of predators (2,250) was tested against the same 2 prey species, with combined results shown under FWS-6 in Table 3. Ratios of success and non-success were again about one-half, very close to those in FWS-1 and 2 in which no prey was used. The apparently normal distribution of response ratios was analyzed for signifi- cance by a t-test, which showed that it could have been due to chance. I concluded that under the experi- mental conditions described, and given a large enough sample, only about half the animals could locate prey success- fully. This statement must be con- sidered in the light of current patterns in the FWS olfactometer which were quite complex and, as shown in dye tests, were characterized by numerous large and small eddies (some of the more prominent are indicated by arrows in Fig. 6). The net effect of these sub- sidiary currents was to leave, at the end of the test, aggregations of preda- tors in any or all of 3 primary areas: (1) beneath and around the 6 acrylic tubes (a near miss, presumably), (2) in either or both of the lower right and left corners of the drawing, and (3) against the curved wall of compartment “U,” between the 2 sets of acrylic entry tubes. In the case of the 2 latter loci, aggre- gations could have formed as a result of predators’ having followed subsidiary currents upstream (having missed the tubes on first try) to a point at which olfactory trails became multi-di- rectional, totally confusing, and perhaps inhibitory of all directed movement. In addition to the statistical success with which a large sample of predators could locate prey, there was the question 294 L. WOOD of whether in successive trials the same predators would continually fail (or succeed). In an attempt to determine this point, 700 predators from FWS-6 were retained, of which 400 had made successful responses. As shown in the lower part of Table 3, a majority of those which had been successful in FWS-6 continued to be so in FWS-7, and still so continued in FWS-8. A similar con- sistency was observed in the behavior of those which had failed in the first experiment. In all cases, differences were highly significant. From this experiment it was con- cluded that the characteristic which attenuates statistically the degree to which predators successfully locate prey is persistent, at least throughout the period of time (about a week) covered by the experiment. Otherwise, re-run groups in FWS-7 and 8 would have tended to split their response ratios evenly in successive trials. The nature of the characteristic(s) involved is unknown, but the possibility of learning onthe part of U. cinerea cannot be ignored. Snails which made it into the prey compart- ments were not restrained from pre- dation, and ingestion of prey tissue might have constituted reinforcement. Further discussion of this point will be delayed until a more detailed foundation can be laid. V. THE EFFECTS OF PREVIOUS EXPERIENCE 1. Introduction Two problems will be considered in this chapter. First, it willbe established that Urosalpinx cinerea can in fact exhibit consistent prey preferences under suitable experimental conditions. Second, the effects of individual ex- perience upon the predator’s observed behavior will be examined. Several different approaches to the latter problem have been attempted. (1) Response to a prey species selected in the laboratory was compared with re- sponse to that upon which the predator was feeding at time of capture. (2) Preda- tors in various stages of maturation were maintained upon single-species diets and then given opportunities to make selections in the olfactometer. (3) Chemotactic responses of “naive,” newly-hatched U. cinerea were ex- amined, prior to and after single-species diets. (4) Attempts were made to de- termine whether actual ingestion of prey tissue was necessary to establish a preference, or if it could be done by Simply exposing predators to effluents from a single prey species, either in aquaria or the olfactometer. The general purpose of this chapter, then, is to present and discuss evidence concerning plasticity of the predator’s prey selection behavior. 2. Comparison of Natural and Labo- ratory Prey Selection If there are persistent preferences in Urosalpinx cinerea, it shouldbe possible to elucidate them at least to some degree by comparing natural and laboratory prey responses, providing it is reasonably certain that the predator was feeding upon only one species for some time prior to capture, and providing further that it does not feed upon other prey species following capture. Comparisons of natural and laboratory prey selections in an extended series of experiments (IFR and VIMS) were fruit- less primarily because predators were maintained too long after capture before being tested and secondarily because of the scarcity of nearby natural habitats in which 2 prey species were equally accessible but were also distinctly separated from one another in the inter- tidal zone. Both defects were corrected by discovery of the Wrightsville habitat, only an hour’s drive from Morehead City. a. Materials and Methods Owing to unusual local conditions, the Balanus and Crassostrea zones at Wrightsville were remarkably distinct. About equal numbers of predators were PREY SELECTION BY UROSALPINX 295 oO о Day 3 Days 11-14 indicated 3 3 total possible responses № о to prey 5 Percent of Prey in Natural KEY Day3 Days 11-14 ÈS Oe —n Response /, тии to Balanus Response to Crassostrea Habitat FIG. 7. Experimental prey selection responses of Wrightsville predators from 2 different faunal zones in Series IFR. collected from each of the 2 prey species in their respective zones, returned to the laboratory on the same afternoon, in separate containers, and tested separately as soon as they could be measured and their sex determined (3 days). b. Results It has already been mentioned that many of the preference experiments, done with animals whose prey selections in nature were known, gave negative results (in fact, the snails indicated an almost uniform preference for barnacle effluents, regardless of their original natural prey). But the uniquely clear faunal zonation of the Wrightsville habitat permitted rapid collection and testing of 2 predator groups from known prey, the first experiments terminating on the 3rd day after collection. As shown in Fig. 7, responses by predator groups closely reflected their zonal disposition in the habitat. Further, this corre- spondence persisted through the 14thday after collection, though there was a tendency, especially in the oyster- feeding group, toward attenuation of the of the difference. 3. The Effects of Controlled, Single- species Diets Upon Olfactory Be- havior a. Introduction This section deals with 2 main questions: (1) can “natural” preference of predators be enhanced or intensified by ingestion of only preferred prey, and (2) can “natural” prey preferences be changed by feeding predators a different prey? Should both or either of these questions be answered inthe affirmative, it is reasonable to ask next whether such intensification or reversal may vary with the maturity of the predator. In other words, is the selective behavior of an adult predator more or less plastic than that of a younger one? Finally, a crucial question must be asked: do newly hatched, presumably “naive” Urosalpinx 296 L. WOOD cinerea demonstrate the same kinds of orientational movements to _ prey effluents as adults? Are there “innate” prey preferences? If so, can these be demonstrated by olfactometric technique before the young Snail has had anoppor- tunity to feed upon specific prey? These questions have been recognized as fundamental since the inception ofthe investigation. The present series of ex- periments represents one of many at- tempts to secure such information; but all efforts prior to the VIMS-36, 37 series ended in failure, due partly to high mor- tality rates amongst controlled-diet young predators, and partly to the enor- mous amount of time-consuming labor required to rear young U. cinerea under rigorously controlled dietary con- ditions. Repeatedly, in successive at- tempts to execute a basically simple ex- perimental design, I found that it was possible to rear young only on barnacle diets: mortality was excessive amongst groups fed only newly-set mussels or oysters. It is, of course, not difficult to rear the young on natural substrates, but in this investigation experimental protocol required presentation of only one species of potential prey to the young snails. b. Materials and Methods Adult predators (longer than 15 mm and hatched prior to preceding summer; IFR-3b). Natural rocks from Shark Shoal were placed in running water aquaria together with about 100 Shark Shoal predators. Mixtures of barnacles and oysters were attached tothese rocks in a ratio of about 10:1. After prey attacks had begun, the predators were removed and started on controlled diets of the prey species they had selected from the natural rocks. Control groups consisting of half the predators from each “rock” group were maintained without food. After 9 days, the 4 groups were tested in the olfactometer against effluents from the 2 prey species. Juvenile predators (6-15 mm; hatched in preceding summer; VIMS-36a). Juve- nile predators were collected from West Haven Balanus balanoides populations on 1 July 1963. They were divided into 3 equal groups, one of which was fed nothing but young B. eburneus collected on asbestos plates, one upon young Crassostrea virginica spat cultured on clean shell in the laboratory, and one maintained as a control in an empty aquarium. First tests were carried out on the 6th day of the diet and upon several irregularly spaced days thereafter up to and including the 16th. On that day, juvenile predators were given reversed diets: those which had been fed oysters were given nothing but barnacles, while those which had been eating barnacles were given nothing but oysters. The control group was discontinued. Young predators (1-6 mm; VIMS-37). Egg capsules of Urosalpinx cinerea col- lected on 1 November and 6 December 1964 at Ocean City were kept cooled (6-10° С) until they could be sorted by embryonic’ stage. They were then gradually warmed to 25° C, and allowed to hatch naturally or the capsules were opened under a dissecting microscope. The unfed (except for probable browsing on micro-organisms) young were held for various time periods pending ac- cumulation of sufficient numbers for testing. It was not practicable to main- tain the various age groups separately prior to testing and feeding. The diet prey groups for this experi- ment consisted of young Balanus spp. col- lected on asbestos plates and cleaned under a dissecting microscope of all other visible material; small Crasso- strea virginica spat (3-10 mm) either cultured on clean oyster shell in the laboratory or taken alive from barnacle plates; and young (3-6 mm) Mytilus edulis collected on 1 November 1964 at Ocean City. All prey animals were maintained in running seawater until controlled feeding periods began. Shell surfaces were always inspected for fouling before being placed with young U. cinerea. Young predators were maintained with PREY SELECTION BY UROSALPINX 297 P< 0.005 60 ho] = Le] = 50 ho] £ 4 a о 3 © S = = o 20 o y 2 NE RO 4 = = a о = SS) = Bolonus Not fed Species Ingested For P<0.050 Ba/anus Ba/anus Not fed Crassostrea Nine Days FIG. 8. Experimental prey selection responses of adult predators previously maintained on a single-species diet for 9 days, and of their unfed controls. Series IFR-3b. (P = probability; N.S. = not significant statistically). their respective prey diets in tightly- covered, 4-liter, polyethylene containers through which flowed filtered seawater at the proper temperature. Except for one early group which was kept for a time at 180 С, temperature for all experi- mental containers was about 25° C, the temperature used during olfaction ex- periments. Maintenance temperatures were controlled by mixing cold seawater in a manual valve with heated seawater from an exchange linked tothe building’s oil-burning hot water system. Failure of this system on a few occasions sub- jected the animals to temporary temperature changes. In each case, how- ever, they were kept at the stated temperature for several days thereafter before being tested. As mortality claimed young predators in each of the separate groups, groups were consolidated to save space and to keep experimental numbers as large as possible. Further, individual identifi- cation of predators so small was not possible, and duration of diet for each group can only be estimated. This change of procedure, however, proved im- material since there was no noticeable difference between test responses of young predators in early, as contrasted with late, periods of controlled diet. Throughout these experiments both predators and prey were kept in running seawater at a controlled temperature of about 250 C. During winter, natural planktonic food was augmented by addition of mixed cultures of algae (chiefly diatoms). Both predators and prey showed some growth and suffered little mortality during the experiment. c. Results and Discussion Results of the first feeding experiment, with adults (IFR-3b), are shown in Fig. 8. While the majority of responses of barnacle-feeders were to barnacle effluents, and statistically highly sig- 298 L. WOOD o o Responses ) 80 Reversed (Percent Total 20 Correct Responses to Species Ingested Diet Specles 5 п 9 PU SAS ESO KEY OR ----- +] Fed oysters at time of test o—————O Fed barnacles at time of test 14 16 18 25 32 44 51 Duration of Exposure to Controlled Diet (Days) FIG. 9. Prey selection responses after ingestive conditioning of West Haven juveniles (total height 6-15 mm), all originally feeding on barnacles, in 2 controlled diet sequences (VIMS-36a). The sequences are marked to show order of diets: BF = barnacle-fed, OF = oyster-fed. “Cor- rect” prey is that offered last. Note scale discontinuities in time (horizontal axis). nificant by chi-square test (P < 0.005), responses of the barnacle control group were about even. Results of the oyster- feeder tests were comparable: those which had been allowed to feed upon oysters preferred oyster effluents, though by a less significant margin (P < 0.050), while the oyster control group’s selection was more evenly divided (not signficant). The results of this preliminary experi- ment indicate that selective tendencies of Urosalpinx cinerea are intensified by allowing them to ingest preferred prey. This treatment is designated as ingestive conditioning to distinguish it from a similar process, olfactory con- ditioning, described by Thorpe & Jones (1937) for insect larvae. That ingestive conditioning depends upon the actual in- take of prey tissue rather than upon ex- posure to the odor of the prey (as in the case of the Thorpe & Jones investi- gations) will be shown in a later ex- periment (VIMS-36c, p 303, Table 5). The next inquiry concerned the extent to which ingestive conditioning can re- verse previously demonstrated prefer- ences. The results of experiments with juveniles from West Haven (VIMS-36a, p 296) are shown in Fig. 9. West Haven juveniles, which in nature had fed upon Balanus during spring and early summer of the year in which they were collected, had in pre-diet tests shown a strong preference for the same genus. As already indicated, one group was fed on Crassostrea, another on Balanus, and a third remained unfed. On days 5 and 6 of the diet, tests showed that Crassostrea was becoming highly attractive to the group fed on oysters, PREY SELECTION BY UROSALPINX 299 TABLE 4. Effects of single-species ingestion on responses of young? Urosalpinx cinerea from Ocean City (VIMS-37), to effluents from prey compartments Percent total responses to - Total en Crass- Mytilus Control possible ostrea responses Not fed 322 Fed Balanus 263 Fed Crass- ostrea 194 Fed Mytilus 124 young = hatched, or removed from eggs in protoconch stage in laboratory, length ca. 1 mm, but no greater than 6 mm. but not at all to the 2 other groups. After about a week the responses of the 2 diet groups to the prey species each was offered did not differ significantly: the tests conducted on days 12-15 showed that ingestive conditioning was ap- parently complete. Statistical analysis showed that predators fed upon either diet responded to the ingested species Significantly (P< 0.005); unfed con- trols, though they continued to select effluents from barnacles, did not do so as pronouncedly (P < 0.025) as did the group fed on barnacles. Having experimentally induced a state of ingestive conditioning in the predators, reversal was attempted by changing diets of both predator groups. The results of olfactory tests conducted during the second controlled diet suggest that there may be a tendency on the part of juvenile U. cinerea to resist a secondary re- versal. Those which had gone through diet sequence Balanus-Balanus-Crass- ostrea responded to “correct,” i.e., last, prey (Crassostrea) significantly (P < 0.005) on days 8-25, but not on the days 32 and 44. Those which had had diet sequence Balanus-Crassostrea-Balanus did not express a significant preference at first, but did (P< 0.005) later, and for the “correct” prey, Balanus. In both series, predator activity rate was low. The last of the 3 selected age groups to be examined weretherecently hatched young (maximum size of 6 mm). Results of these experiments (VIMS-37) are pre- sented in Table 4. Two questions were asked by these experiments: first, do “naive” Urosal- pinx exhibit positive orientational re- sponses to prey effluents, and if so are these made selectively? Second, are very young snails subject to ingestive conditioning? Hence in the first examination of VIMS-37 results it should be deter- mined whether distribution of responses amongst the 3 prey and 1 control com- partments differed significantly from chance ( = 25% to each). Since total responses observed were 59, chance alone would predict 14.75 responses to each of 4 compartments. In the case of 2 of these, responses did not differ significantly from chance (12 toCrasso- strea, 14 to Mytilus). But in the other 2 (29 to Balanus, only 4 to control) the difference is highly significant (sum of chi-squares = 21.098, degree offreedom =*9, P <=0.00N: In considering the extent to which the response distribution of unfed snails differs from that of diet groups, it is convenient to compare statistically each combination of pairs (NF x BF, NF x OF, NF x MF, where NF = not fed, BF = barnacle-fed, OF = oyster-fed, MF = mussel-fed). Computation of 300 L. WOOD chi-square contingency tests for each pair revealed that all pairs differed sig- nificantly (P < 0.001) in the response of their components except one, NF x OF. In other words, these 2 groups made essentially similar responses to the same prey effluents, indicating that the olfactory behavior of the OF group had not been altered by feeding them Crassostrea. Both groups responded most frequently to Balanus effluents, as did also BF young. Indeed, re- sponse frequency of the BF group to preferred prey was greatly increased during the controlled diet. Ofall groups, only MF young failed to choose ¿n majoris the barnacle effluent compartment; in- explicably, they also did not respond very often to Mytilus, but rather to the con- trol compartment. It will be noted that only a minor fraction of test animals made responses of any kind in these tests. Mean re- sponse ratio was 24.8% for the 4 groups, considerably less than those seen in Similar experiments with juvenile and adult Urosalpinx. In fact, failure of so many of the young snails to respond to any prey odors whatsoever is remi- niscent of experiments with unfed con- trol groups (p 303-305). Thus it may be asked whether those predators of the above series that were given oppor- tunities to feed upon a Single species did in fact ingest their tissues. Since careful records were kept, where possible, of direct evidence of attack and ingestion, this question can be answered positively for most diet groups whose behavior has been discussed. The ex- ception, for the following reason, is in the case of the barnacle-fed groups. It is true that Urosalpinx leaves tangible evidence of its attack upon most prey: a small, slightly conical hole bored through an exposed valve. These holes can be counted and their ratio to total number of dead prey calculated. If a hole is bored through compartmental plates of barnacles, this can also be taken as direct evidence of predation. But if the snail adopts the more efficient mode of entry through the opercular aperture, it can kill and cleanabarnacle without leaving a clue. This can be especially damaging to quantitative analysis when other predators suchas the flatworm Stylochus ellipticus (Girard), which get into experimental chambers as larvae and which also leave no marks of attack (Wood & Deibel, unpubl.), can account for a considerable but unknown fraction of dead barnacles. It is there- fore necessary to go to indirect evidence for ingestion of barnacles. Inthe present work, growth rates (where available) of barnacle-fed predators were compared to those of unfed controls, and a signifi- cant difference between groups was ac- cepted as evidence of feeding. With these qualifications in mind, evi- dence of ingestion can be reviewed for each experimental group: IFR-3b (adults). No growth obser- vations were made of barnacle-feeders; in situ observations of both barnacle and oyster feeding groups confirmed that attacks were in progress. At the end of the feeding period, 21 dead oysters were counted, each with at least 1 bored hoie, while more than 100 barnacles had been killed and cleaned, probably by Urosal- pinx cinerea. VIMS-36a,b (juveniles). Barnacle- feeders showed significant growth during the first 20 days on diet, as compared to that of unfed controls, whose size- class frequency distribution did not change and was indistinguishable there- fore from that of the originally collected sample. Bored holes in compartmental plates of barnacles were very rare. Oyster-feeders similarly increased in size during the 20 days, and left behind them many bored valves. During the second reversal of diet, in situ obser- vations confirmed feeding, but no quanti- tative evidence was recorded. VIMS-37 (young). Though precise in- formation is lacking, repeated micro- scopic examination of dead pelecypod prey showed that they were being per- forated and consumed. Whether the youngest snails perforated the larger PREY SELECTION BY UROSALPINX 301 prey organisms is not known, as pre- dators of several size classes were kept together. This much was observed: the higher mortality rates among young predators were in the 2 pelecypod diet groups. These results are consistent with those from my previous attempts to rear Urosalpinx upon single-species diets: success has been achieved only by feeding them small Balanus. Es- pecially in the case of oysters was great difficulty experienced. While very small Mytilus were available in great quantity, young Crassostrea spat, produced in the laboratory by artificial fertilization pro- cedures, were harder to obtain at the time of the year when the experiment was performed. Therefore, the oysters frequently seemed too large and thick- Shelled for the young predators; many living C. virginica were seen with partial holes bored in their shells, suggesting the premature death of young borers. On the basis of experiments reported here, there is good reason to believe that most predators given opportunities to feed upon single prey species did in fact complete successful attacks. The only group about which serious doubt exists is oyster-fed young, yet some of these were observed consuming prey. 4. Responses of Conditioned Snails to Odor of One Species at a Time a. Introduction A fundamental question raised by in- gestive conditioning experiments con- cerns the frequency with which olfacto- metric responses are _ elicited by effluents from prey species to which predators have not been conditioned. We have seen that a predator, given a choice between 2 prey effluents, will with increasing frequency select that to which it has been conditioned. But what if the other, “unconditioned,” effluent is the only one present in the olfactometer: will the predator respond at all? While failure to respond cannot be taken directly as evidence of failure to per- ceive, such a conclusion would be strongly favored. In previous experiments, it had not been possible to obtain adult or even juvenile predators that had not had access to one or more of the major prey species. In experiment VIMS-36a, for example, barnacles were the native food of both experimental groups. For several years I had searched the U. S. East Coast for a population of Urosalpinx cinerea for which neither barnacles nor oysters constituted normal prey, so that this experiment could be done. To my chagrin, such a population was located by a student (Roberts, pers. comm.) in a subtidal Zostera bed in front of my own home. b. Materials and Methods Experimental animals (VIMS-38) were collected in April 1966 from subtidal Zosteva beds adjacent to Locust Point, Saddlers Neck, Gloucester County, Virginia, in the weakly estuarine North- west Severn River. They were mostly adults, though their modal size class was 9-10 mm (previous observation had confirmed the small size of adult individuals inthis population). Numerous examinations during the summers of 1965 and 1966 convinced me that the primary natural prey of the Locust Point population was the slipper limpet Crepi- dula convexa Say, whichoccurred there in great abundance. Other potential prey Species were the barnacle Balanus im- provisus, the gastropods Anomia (= Cavolinia) simplex and Bittium sp., and a small epiphytic mussel (probably Amygdalum papyria Conrad), but these were either (in the case of Balanus im- provisus) rare in occurrence or there was little or no field evidence of attack by U. cinerea. In brief, the Locust Point predator population had apparently not been con- ditioned to either of the chief prey species hitherto discussed in this work. Nearly 400 U. cinerea were collected, brought to the laboratory, and tested in the olfactometer within hours. After the initial test, they were randomly assigned 302 L. WOOD Control 80 Oysters 60 40 Correct Responses (Percent Total Responses) 3-6 9-12 15-21 28-34 41-48 55-70 Duration of Exposure to Controlled Diet (Days) FIG. 10. Prey selection responses during pro- gressive ingestive conditioning of Locust Point adults (VIMS-38) to barnacles and oysters. The predators’ natural diet did not include either of these prey animals. The “control” group of predators was maintained throughout the experiment upon the eelgrass Zostera and its many accompanying epiphytic inverte- brates, chiefly Crepidula convexa. The curves showing experimental responses of the barna- cle- and oyster-fed snails are reproduced in FIGS. 11 and 12, together with additional cor- ollary information. to 3 diet groups: young Balanus spp., cultured Crassostrea spat, and a control consisting of material from the natural habitat (Zostera plus epiphytic flora and fauna). Barnacle and oyster diet speci- mens were obtained and treated as de- scribed above in experiment VIMS-36a. Detailed feeding observations were re- corded throughout the experiment. In other respects, experimental methods were as already described, except that all 3 predator groups were tested against each of the 2 prey species separately instead of simultaneously. c. Results and Discussion Initial olfactometer tests, made on the day of collection, confirmed sup- positions made above about the natural prey of Locust Point predators: of the 384 tested, less than 1% responded to either barnacle or oyster effluents, but 5.5% selected one or another of the “control” compartments. It should be recalled at this point that “control” Barnacles Barnacles water consisted of seawater pumped from the York River estuary and through a modified controlled conditions system (Wood, 1965a) before flowing into the several olfactometer compartments from a common reservoir. The main seawater pump intake was situated about 40 m from the edge of an extensive Zostera community, the fauna of which was markedly similar to that at Locust Point. Hence it is not unreasonable to assume that external metabolites pro- duced by the nearby Zostera community, and pumped into the laboratory system, were quite similar to those of the pre- dator population’s original habitat. Such an interpretation was further supported by subsequent tests of the 3 controlled diet groups, partly shown in Fig. 10. The control group, maintained in an aquarium with Zostera and associated biota, did not exhibit a typically dramatic increase in “correct” responses to con- trol compartments since the ratio of “correct” responses in the first time period (3-6 days) was already 63%. In the terminal period (55-70 days) it was 92%, and intervening ratios fluctuated between 78 and 100%. Both oyster- feeders and barnacle-feeders, on the other hand, did exhibit a pronounced incremental tendency, commencing with 9% “correct” responses in the first period and terminating with 70 and 83% “correct” responses, respectively. The “control” responses to seawater (not illustrated) of BF and OF diet groups showed the opposite, decreasing, ten- dency: barnacle-feeders started with a 90% response ratio and terminated with 12%; oyster-feeders began with 86% and ended with 29%. Of considerable significance, in my opinion, is the fact that in the case of both diet groups, increase in “correctness” of response was primarily at the expense of “control” (Zostera community) responses, and not “incorrect” prey effluent responses, again suggesting that barnacles and oysters were not part of the Locust Point snails’ natural diet. Evidence of actual ingestion of prey PREY SELECTION BY UROSALPINX 303 during the experiment is offered in Figs. 11 and 12, in the form of dotted vertical bars which denote the ratio between the number of predators observed actually feeding and the total number of predators living in the aquaria. This ratio is corrected for variations in duration of the periods of observation. Curves from Fig. 10 are superimposed upon Figs. 11 (barnacle-feeders) and 12 (oyster-feeders), respectively, so that “correct” olfactory response ratios can be compared to feeding activity and also to rate of oviposition (open vertical bars), the latter process being commonly accepted as an index of health in captive invertebrates. Two factors are ap- parent in these results. (1) Feeding activity was greater and its onset was quicker in the oyster-feeding (Fig. 12) than in the barnacle-feeding (Fig. 11) group. This is reflected in both the more rapid conditioning rate of oyster- feeders (compare both response curves in Fig. 10) and in their greater fertility. (2) In both groups, feeding and ovi- position increased together to a peak in the 3rd time period (days 15-21) and then declined as the summer season waned, а trend widely observed in natural populations of Urosalpinx cine- rea (Carriker, 1955) and often seen in our laboratory populations. The more rapid conditioning of U. cinerea to oyster effluents, observed in both phases of experiment VIMS-36a (Fig. 9) and in the present experiment (VIMS-38), is thought to be due toa difference in required attack techniques, which resulted in the more rapid com- mencement of feeding upon young oyster Spat than upon barnacles. As has been stated above, an experienced predator can successfully complete an attack upon a barnacle in less than an hour by simply inserting its proboscis between the barnacle’s opercular plates. Boring a hole through thickened compartmental plates of Balanus, on the other hand, may require more time and energy than boring a hole through the thin valve of an oyster spat of the same or even Slightly greater basal diameter. Ex- amination of attacked barnacles revealed that Locust Point U. cinerea apparently did not often employ the opercular entry method, but instead bored holes through Opercular or compartmental plates. Frequently these holes were incom- pletely bored at the time of observation. Whether or not U. cinerea chemo- receptor surfaces are actually sensitized by ingestive conditioning is a question that must await application of electro- physiological techniques (commenced summer 1967), but experiment VIMS-38 Supplies circumstantial evidence that some kind of sensitization may occur. The alternative explanation of these results requires that a quasi-rational “decision,” in favor of one prey effluent Over another, be made as a function of some ganglionic process when the snail arrives at the point-of-choice in the olfactometer. 5. Effects of Long-term Exposure to Prey Odors a. Introduction The concept of “olfactory conditioning” was placed in the literature by Thorpe & Jones (1937). It is now reasonable to ask whether ingestive conditioning is not really the same thing as that which Thorpe & Jones described in insects. In addition to theoretical considerations, a quite practical reason exists for making a clear distinction between the 2 types of conditioning: if predatory gastropods are influenced by what they smell, might not their responses to later olfactometer tests be modified by exposure during an earlier test to concentrated prey effluent in the selected compartment? Two ex- perimental approaches were adopted. The first was a straightforward experi- 304 L. WOOD 100 Feedings observed / population ыы 0 Egg capsules deposited / population ao 2 © 80 = n 22 Е of 5 Ger Oe Poe ® © a a = © ae ‘eons с E o o о? 40 © © © = 20 ‚= D Lire D 28-34 41-48 55-70 Duration of Diet (Days) FIG. 11. Feeding and egg-laying activity of Locust Point adults during period of ingestive con- ditioning to barnacles (VIMS-38). Dotted bars represent the ratio between the number of snails actually feeding and the total number present in the group. Open bars represent a similar ratio for egg capsules deposited. In both cases, figures are corrected for varying time periods covered by the observations. The superimposed curve for “correct” responses is from FIG. 10. 100 Feedings observed / population = 1.20 © = 80 0 Egg capsules deposited / population > © a 7) 1.00 о 5 г o © = Zn 60 = 5 x so UI 12) o © o Е ¡O В Е 8 = 5 © .40 = о = = a 20 o = 205 T 2 0 3 ых Е E-= ] Eo. Lito mp at 350 712 15-21 28-34 41 — 48 55—70 Duration of Diet (Days) FIG. 12. Feeding and egg-laying activity of Locust Point adults during period of ingestive con- ditioning to oysters (VIMS-38). Bars represent activities as described in caption for FIG. 11 on the opposite page. The superimposed curve for “correct” response is from FIG. 10. In com- paring FIG. 12 to FIG. 11 (opposite), note correspondence in each case between onset of repro- ductive and feeding activities and the rapidity of ingestive conditioning as indicated by the shape of the response curve. PREY SELECTION BY UROSALPINX 305 mental design which will be described immediately below; the second was a post-hoc statistical analysis presented later in this section. b. Materials and Methods Experimental subjects for VIMS-36b were Nobska juveniles collected on 30 November 1963 and maintained without visible food and at ambient salinity and temperature until the beginning of the experiment. When controlled diet aquaria for VIMS-36a were set up and that experiment started, perforated poly- ethylene boxes containing samples of Nobska predators were placed in the aquaria with prey (or inanempty control tank) and with the freely moving West Haven predators. When the latter were tested for response to prey effluents, so were the Nobska animals, under the same conditions. c. Results and Discussion As shown in Table 5, there was no Significant difference between those predators exposed to the concentrated odor of Balanus, their natural food, and those exposed to seawater alone. Those exposed to odor of Crassostrea made slightly fewer (statistically not signifi- cant) responses to Balanus effluents, but made no responses to oyster effluents or to controls. Comparison of these re- sults with those for unfed control juve- niles (pp 298-299) of VIMS-36a indicate that no essential difference exists be- tween responses of that group and any olfactory exposure group: all 3 groups, in fact, displayed a signficiant preference for barnacle effluents, a consistency which is not surprising since none of them was known to have fed prior to the experiment upon prey other than Balanus. 6. Effects of Short-term Exposure in Olfactometer a. Introduction There are at least 2 ways of assessing the influence of early olfactometric choices upon those made later in a continuing series of tests. First, there could be a simple correlation between initial and later choices. But a better method is to determine degrees of choice consistency in a series of tests. Does the choice in Test 1 positively affect the choice expressed in Test 2, etc., or is observed consistency (to be dis- tinguished from preference) due entirely to chance? These questions are not just hypothetical, as Chew & Eisler (1958) and Chew (1960) reported without ex- planation that “Ocenebra” japonica tended to repeat its initial experimental prey choice, regardless of the species attacked in its natural habitat, and Putnam (1962) reported that Aleochara, a coleopteran insect, repeated its initial choices of either right or left arms of a Y-tube. Should such a tendency be operating in Urosalpinx cinerea, it would be best to know about it before inter- preting prey Selection results. b. Materials and Methods The only test sequence of sufficient length available for the present analysis was OI-3, which will be described in greater detail in another paper. Suffice it here that the experiment was run with standard techniques (described under General Methods, p 283 et seq.), but at a graded series of controlled tempera- 306 L. WOOD TABLE 5. effluents of one prey species? Effects of exposure on juvenile? Urosalpinx cinerea from Nobska (VIMS-36c) to Exposed an to effluents from Balanus eburneus 17 Crassostrea virginica seawater alone а 6-15 mm b Prey in nature = Balanus balanoides tures. The temperature serieS was carried out in sequential order, the lowest temperature tests being done first, and soon. The controlled tempera- tures employed were 16, 20, 25, and 30°C. The prey used were Crassostrea vir- ginica, Balanus eburneus and Brachi- dontes exustus, all from Nassau Sound. Predators were collected at West Haven, Shark Shoal, and Nassau Sound. Samples of 20 animals from each habitat were tested in a series of 14 runs; those which failed to make criterion response in at least 10 of 14 runs were eliminated from the analysis. Since responses to the 2 pelecypod species did not differ signifi- cantly, these were lumped together and predator’s selections were therefore coded “b” for “barnacle,” and “p” for “pelecypod.” Individual predators were numbered for identification and their responses recorded at the end of each run. As the response ration b/b+p differed signficantly between the first 2 (at 16°C) and 12 subsequent runs, it would not be proper to subject the group responses to consistency analysis; rather, the consistency of individuals’ responses should be examined. The method by which this was accomplished is de- scribed below. An empirical measure of consistency was derived from the series, Percent total responses (pooled) to Balanus Crassostrea Total Controls possible (2 responses compartm. ) bbbb, pppp, pbbb, bbbp, bppp, pppb, bpbb, bbpb, pbpp, ppbp, bbpp, ppbb, bppb, pbbp, bpbp, pbpb, where run sequences of 4 selections are arbitrarily ranked from top to bottom in order of decreasing consistency. Con- sistency has 2 components: (1) pre- dominance of one prey type over another; (2) number of runs of either р or b, i.e., the number of times a prey “prefer- ence” is reversed during a sequence. Preference for a prey type can be ex- pressed as a simple ratio, P = x/x+y where x is the predominant choice ina sequence, and y is the other. Since consistency decreases as the number of runs (r) increases, this element can be introduced as C = P/r, where C is an index of consistency. Now let us apply the index C to the ranked series of sequences shown above. Selection Sequence E E C bbbb, pppp 1. 000 1 1. 000 pbbb, bbbp, bppp, pppb .750 2 . 375 bpbb, bbpb, pbpp, ppbp .750 3 . 250 bbpp, ppbb - 500 2 . 250 bppb, pbbp . 500 3 . 167 bpbp, pbpb .500 4 „125 P = Preference; r = runs; C = Consistency PREY SELECTION BY UROSALPINX 307 In a sequence of 14 selections (re- sulting from a total of 14 runs in ex- periment OI-3), the range limits of C can be shown to be 0.036 and 1.00, the former indicating least consistency: bpbpbpbpbpbpbp or pbpbpbpbpbpbpb, and the latter indicating perfect consist- ency: bbbbbbbbbbbbbb or pppppppppppppp. As the length of a sequence increases, the lower limit of C can approach, but never reach, zero. The index (C) was applied to all sequences in experiment OI-3, of length of 10 or greater (41 out of 60 choice sequences, one for each snail), so that values of C derived therefrom could be compared statistically to the values of C derived from application of the index to analogous chance sequences. These chance sequences were obtained from a statistician’s “random gener- ator,” a small plywood box from which protruded a clear plastic runway just large enough inside to accommodate a single file of marbles. In normal usage, the box is partly filled with marbles of 2 colors (in equal numbers), shaken, and then placed on its side, to permit marbles to roll down the plastic runway, where they are counted. In the present case, the “population” of marbles was deliberately biased, so that each population contained blue marbles in direct proportion to the number of choices of preferred prey selected, and red marbles in proportion to choices of non-preferred prey ina given predator sample. Thus the analog for the predator sample from the West Haven contained 88% blue marbles and 12% red marbles, since those predators had made, in experiment OI-3, a total of 162 choices for preferred prey and 23 choices for non-preferred prey. The function of this deliberate bias was to eliminate the effect of preference from the analysis of consistency. A table of random numbers, which has a roughly even distribution of odd and even numbers, yields sequences whose con- 50 40 = [= 8 OO & > 2 20 o = o pue u 10 O | 2 >; 4 5 6 Values for Consistency (pooled data, in numbered data classes) FIG. 13. Correspondence of expected and ob- served values of the consistency index C. Ex- pected values were derived from chance se- quences of marbles of 2 different colors in a statistical “random generator.” Observed values were derived from prey selection se- quences by Urosalpinx in Series OI-3. The congruence depicted above suggests strongly that internal consistency within a sequence of prey selection does not differ significantly from chance. sistency differs significantly from that shown by Urosalpinx cinerea, but this dif- ference is not primarily due to predator attributes, but rather to the decrease in the numerator of the expression C=P/r, produced by the non-preference of ran- dom numbers for either odds or evens. The analogous nature of the marble sequences was carried one step further: the length of each predator’s sequence of prey selections (between 10 and 14) was duplicated inits corresponding mar- ble sequence. In this way, 4 analogous series of marble color sequences were produced to simulate each prey choice sequence. The values of C for these sequences were compared to the value of C for the actual prey selection sequences. Significance of difference between the 2 members of each sequence pair (proto- type and analog) was determined by t- test. 308 L. WOOD c. Results and Discussion The distribution of values of the con- sistency index C in the prey selection and marble color sequences showed no Significant difference; the frequency distribution curves for data classes can be seen in Fig. 13 and their close correspondence noted. In other words, consistency of preda- tors’ responses in OI-3 did not differ from chance (assuming a biased sample, with barnacles preferable to pelecypods). This, in turn, suggests strongly that no one response, whether in the beginning or near the end of a Sequence, necessarily influenced subsequent responses. If this proposition is acceptable, responses to prey effluents should be regarded as an expression of what might be called “latent” preference, either conditioned or natural, and not the result of olfactory conditioning from a previous exposure to the same effluents in the olfactometer. 7. Summary of Ingestive Conditioning Experiments Evidence has been offered in this chapter which confirms the general hypothesis that expressed preferences for specific prey by Urosalpinx cinerea are at least partially the result of ingestive conditioning. This process has been defined as a modification of the predator’s responses to prey effluents induced by maintenance upon single- species diets. Results of conditioning experiments with adult Locust Point predators in experiment VIMS-38 were markedly Similar to those given in Fig. 10 (VIMS- 36a) for juveniles for the period after reversal of diets. In both cases, necessary conditioning periods were longer than the time required for initial conditioning of juveniles. Further, snails fed oysters in experiment VIMS-38 were conditioned more rapidly than those fed barnacles which suggests some carry- over from one molluscan prey (Crepi- dula, in native habitat) to another (Crassostrea, in the laboratory). Third, terminal results in both the adult and second-reversal juvenile experiments ranged around 80% correct responses (as contrasted with 100% for the first diet in VIMS-36a, Fig. 10). Finally, responses by both barnacle- and oyster- conditioned Urosalpinx to “control” con- ditions (i.e., odor similar to that of original habitat, which did not include these 2 species) remained dispro- portionately high in comparison to such responses in previous conditioning ex- periments, which suggests that, as adults, the Locust Point predators were rather resistant to diet changes. These considerations lead me to be- lieve that ingestive conditioning is Similar to imprinting, a term proposed by Lorenz (1935) to explain his adoption by anseriform young as their surrogate mother, and since reported widely by investigators of avian behavior. Two recent reports (Burghardt & Hess, 1966; Burghardt, 1966) described “food im- printing” in vertebrates, a process ap- parently identical to that described here as ingestive conditioning. The first paper dealt with conditioning of snapping turtles (Chelydra serpentina) to a preference for 1 of 3 types of food. The second mentioned the possibility that imprinting might account for differential responses of young garter snakes (Thamnophis sirtalis sirtalis) to extracts of food organisms. Another generalization for which evi- dence has begun to accumulate concerns the original question of statistical preference: it appears that members of the genus Balanus produce an effluent which is, under most conditions, more attractive to Urosalpinx than are those of any of the pelecypod species tested to date. The specific exception to this generalization is provided by the in- gestive conditioning experiments. That is, barnacles are more attractive than pelecypods, except to juvenile and adult predators which have been kept ona diet of Crassostrea virginica. It should be kept in mind, in con- PREY SELECTION BY UROSALPINX 309 Sidering theoretical aspects of this problem, that predators are discrimi- nating the effluents from the various prey species, andby olfactory cues alone. Hence it is the olfactory apparatus (in the general sense) which is in some way modified by ingestive experience. The fact of discrimination permits the in- ference that prey species not only differ in the chemistry of their effluents, but also of their tissues. The very nature of the demonstrated phenomenon of in- gestive conditioning, in fact, tends to reject an hypothesis of strictly quantitative response to prey effluents as proposed by Blake (1960), at least when it refers to predators conditioned to prey from differing phyla. Hence the investigator is led to comparisons of the chemical composition of the prey, and of both tissues and effluents. These comparisons have been commenced and will be presented in a later paper. Preliminary results already suggest that ammonia may bea generalized attractant to which unconditioned predators are sensitive. This idea has also been proposed by Blake (1961) and is entirely consonant with both circumstantial and available experimental evidence. VI. GENERAL DISCUSSION 1. Revision of Preference Concept The concept of prey preference, dis- cussed in the Introduction (pp 271-272), needs redefinition. First, it has been shown that selection of specific prey, in either natural or experimental situ- ations, varies with a complex of factors. Thus it is pointless to propose a genetically fixed prey preference. It has also been shown that a demonstrated preference can be altered by changing the predator’s diet. Therefore, the idea of chemoreceptor “types” specific to effluents of given species is no longer relevant. It is more fruitful to isolate those factors which influence individual prey selection under known conditions, and from such an analysis to suggest general hypotheses about the predator’s behavior toward its prey. Evidence offered in this work suggests that the most important factors, all mutually interacting, are as follows: 1. Co-existence of predator and prey within given intertidal zones, either in common response to the same environ- mental factors, or because the predator is attracted to that zone by the prey. 2. Relative population densities of the prey species within restricted local areas inhabited by the predator. 3. Recent ingestive experience of the predator. Other factors, shown to affect attrac- tance results in experiments (Blake, 1960), must be regarded as less im- portant in natural habitats. 1 refer to quantitative aspects of metabolite pro- duction, either in time by individual prey organisms (e.g., changes in meta- bolic rate), or in space by distribution of individuals per unit area. It is likely that metabolic variations due to local environmental factors would be common to all prey present. Spatial concen- trations of one species would increase probability of attack from a predator, but it is not easy to Separate spatial and chemical causes. Analaysis of a mixed prey habitat at West Haven (p 281, Fig. 3) showed the number of attacks upon barnacles and mussels, respectively, to be proportional to the number of individuals of each species present, and not to their biomass. Another factor not considered before is the area occupied by a prey individual relative tothose of its neighbors. Connell (1961, and pers. comm. 1965) suggested that larger prey organisms will cover greater surface areas and therefore be more susceptible to attack, by chance alone. But, had this been the case in the West Haven habitat, observed attacks upon the larger mussels would have significantly exceeded calculated ratios, which they did not. An alternative ex- planation is now available: ingestive conditioning of predators is in direct proportion to the number of prey in- 310 L. WOOD dividuals present, and which would there- fore be capable of affecting the preda- tors’ behavior. Such speculative thoughts do not seem extraordinary when one considers the dilemma imposed by field information, which indicates that hypotheses of ol- faction need not be invoked by explain prey selection, and experimental results, which indicate rather elegantly developed olfactory capabilities. The extent to which olfaction operates in nature as a Selection factor must be regarded as problematic, simply because direct information is not available. There is little question that under the conditions of certain types of field experiments, such as those reported by Carriker (1955), Urosalpinx can and does orient to food odor sources and can follow current-borne attractants to a source when currents approach recti- linearity and there is a suitable inter- vening substrate. After all, U. cinerea is a bilateral animal belonging to a phylum whose orientational chemo- Sensitivity has been well documented (Blake, 1960; Kohn, 1961; and the present work). But natural situations approximating those in which the field experiments were done are probably rare; as we have seen, most intertidal populations of U. cinerea live amongst a multiplicity of prey species (Figs. 2 and 3). Those which do not (Nobska, Ocean City, Wrightsville) have im- mediate access to dense substrates of prey of a single species; in fact, their feet rarely touch solid rock but instead stay nearly always (when on outward, exposed surfaces) upon prey. The snails feed upoh prey, move around onit, copu- late, and, finally, deposit egg capsules either upon or near it. Why, then, has U. cinerea evolved a capacity for ol- factory discrimination, demonstrated in the laboratory, if it does not use this capacity in its natural milieu? Of course, there is no evidence that it does not. The crucial question is whether coexistence of predator and a specific prey results from independent factors (exposure, temperature, turbu- lence, etc.) or from the predator’s directed and purposeful movement toward that prey. The experiment with Wrightsville predators, in which snails from each of 2 clearly separated zones (barnacle and oyster) were tested, failed to distinguish cause and effect. That is, were snails in a zone initially because of an a priori preference for its sessile fauna, or had they become conditioned to the fauna through ingestion, having arrived in the zone as the result of other factors? On the basis of field infor- mation alone, it is not possible to answer. But ingestive conditioning experiments lead me to choose the latter alternative, at least in interpreting the Wrightsville experiments (p 295). In support of this choice, several field observations may be cited. While nothing is known of the effect of conditioning in multiple, mixed prey habitats, it is apparently of importance in “single species” habitats such as Nobska and Ocean City, or in the rocky intertidal habitats described by Fischer-Piette (1935), Moore (1938), and Fretter & Graham (1962). At Nobska, Balanus balanoides is overwhelmingly dominant, and Urosal- pinx cinerea apparently fails to take advantage of seasonal opportunities to feed upon regular but usually sparse sets of Mytilus edulis. In spring, 1963, for example, Balanus and Mytilus, both in great abundance and both recently set, were present in a density ratio of about 3:1. Following analysis of the mixed-prey habitat at West Haven, about a quarter of the attacks at Nobska should have been upon mussels, but none at all were seen; all Urosalpinx observed, both adults and juveniles, were feeding upon the dominant barnacles. At Ocean City, where, because of peculiarities of zonation, barnacles intergrade with mussels only upon under- sides of rocks, and where U. cinerea is in effect conzonal with M. edulis, few attacks upon barnacles have ever been reported, despite the fact that in the PREY SELECTION BY UROSALPINX 311 upper intergrade zone, predators have access to numerous B. balanoides. Thus we see here the exact inverse of the Nobska situation. The case cited by Fischer-Piette is one in which separation of prey was in time rather than space. He described a shore community of Balanus balan- oides located on a point at St. Lunaire (France). The barnacles were preyed upon by a population of “Purpura” lapillus, а muricid gastropod eco- logically analogous to the American Urosalpinx cinerea and conspecific with “Thais” lapillus of the U. S. northeast Coast.* During a period of 4 years, the predators were never seen attacking the scattered Mytilus edulis. At the end of this 4-year period, the mussels began to wax while barnacles correspondingly waned. Eventually, mussels were dominant, but for 2 years (1930-31) “Purpura” was not observed to attack mussels at all, and restricted its pre- dations to the diminishing Balanus population (note the similarity between this behavior and that of barnacle-con- ditioned U. cinerea). Not until the “Purpura” found themselves in small areas which they had cleared of living barnacles, and were surrounded by mussels, did they begin feeding upon the now numerous mussels (near the end of 1931). Fischer-Piette (1935: 167) attributed the reluctance ofthe gastro- pods to change prey to the relative ease with which they could obtain nourish- ment from the barnacles, rather than to differences in ectocrine attractiveness (“loi du moindre effort ... ayant plus de facilité a sucer des Cirripèdes qu’a percer des Moules. Mais l’explication * Thiele (1931: 298) lists Nucella lapillus as valid name for Linné’s Purpura lapillus. Clench (1947: 86), however, maintains that “Roding did not intend the species now com- monly known as Thais lapillus to be in- cluded in his subgenus Nucella” and that “this latter name should either be abandoned or else associated with Cantharus. ” scientifique reste a trouver.”).** Another relevant observation was made by Moore (1938) on British “Pur- pura” lapillus. He found that young of that species were frequently collected from undersides of rocks inlower inter- tidal zones, where they were feeding upon the polychaet Spirorbis. When he brought young snails in his laboratory, he found they continued to feed upon Spirorbis, and not upon small barnacles to which they were also given access. He found, how- ever, that as they matured they changed to a barnacle diet. He did not indicate whether this change was linked directly to maturation. Fretter & Graham (1962) expressed the opinion that “Ocenebra,” another predatory muricid, retained its prefer- ence for its natural diet after being re- moved to captivity, citing Orton’s (1929) studies of gastropod predation. The re- tention of a natural preference by Conus (Kohn, 1959) has already been mentioned. 2. Mechanism of Ingestive Conditioning There are few papers concerning the effect of experience in modifying gastro- pod behavior. Fischer-Piette (1935), whose observations of “Purpura” lapillus have already been examined above, cited a comment by Pelseneer (1924), who spoke of the “case of Natica (feeding upon Donax and Tellina) which had profited (tiré parti) from the experience acquired from its initially fruitless efforts, this being the criterion generally accepted for the existence of in- telligence.” (auth. transl.) Fisher- Piette also recounted the difficulties “P.” lapillus apparently had in adjusting to the task of penetrating the valves of its new prey (Mytilus edulis). Experienced P. lapillus entered barnacles through the opercular aperture without boring a hole, but such a technique was not **“Law of least effort ... as it is easier for them to suck cirripedes than to pierce mus- sels. But the scientific explanation remains to be found.” (Edit. transl. ) 312 L. WOOD applicable to mussels. Eventually, Fischer-Piette reports, the gastropods made the adjustment successfully, though he seemed reluctant to ascribe this ad- justment to individual learning. His comment (p 165) is worth quoting. “Nous aurions done sous les yeux une sorte d’apprentissage des Pourpres dans leur facon de se nourrir au dépens des Moules. Mais la notion d’éducation individuelle ne suffit pas dans le cas présent.”* He goes on to state that while mem- bers of the first generation may have had to learn, individually, how to per- forate mussel shells, subsequent generations “know right away how to perforate without error, without passing through the stage of apprenticeship.” (auth. transl.) I have made similar observations of Urosalpinx cinerea’s patterns of preda- tor behavior in its attacks upon mussels and barnacles, and preliminary results suggest a degree of behavior modification as a result of experience, or a kind of trial and error learning. Before Fischer-Piette, Garth & Mitchell (1926) used a T-maze to in- stitute a conditioned (= Pavlovian) re- sponse in a land snail, Rumina decollata L. These investigators stated that the learned response was retained after a 30-day period. They also cited the work of Thompson (1916), who found evidence that another snail, Physa gyrina Say, could modify its behavior by forming an association between 2 stimuli. Bullock & Horridge (1965) reviewed briefly the problem of gastropod learning and implied that such early investigations of classical conditioning in gastropods were wide of their mark, a notion with which I thoroughly agree. In Bullock’s words (Bullock & Horridge, 1965: 1344): “It *“We thus seem to have, in Purpura, a sort of apprenticeship as to the manner in which it nourishes itself at the expense of the mus- sels. But the concept of individual learning does not cover the case.” (Edit. transl. ) seems probable that we will have no adequate idea of gastropod capacities until tests are used that are natural and meaningful to the species.” The investigators cited above were apparently dealing with associative learning, defined by Thorpe (1956) as establishment of bonds between dis- crete stimuli and units of behavior. Associative learning depends, at least to some degree, upon reinforcement, or reward, a type of which is referred to variously as “drive reduction” or “appetite satisfaction.” Clearly, if ingestion of prey tissue were accompanied simultaneously by olfactory stimulation, and if perceived stimuli were produced only by the species being a.tacked, criteria for associative learning would be satisfied. Repetition of this experience often enough or long enough (about 2 weeks in the case cited in Fig. 10, p 302) would hypothetically result in a state of ingestive con- ditioning. It has been shown that simple ex- posure to prey effluents will not con- dition Urosalpinx cinerea: ingestion of tissue is required. This implies either reinforcement or the physical transfer of some cue from prey to predator. While I hesitate to ascend to a more speculative plane, it is en- tirely possible that future studies may show that ingestive conditioning is linked with changes in nucleic acid composition of the predator’s chemoreceptor surfaces and these in turn may be in- fluenced by composition of free amino acid pools or patterns of amino acids obtained from metabolic breakdown of ingested prey proteins. A preliminary attempt has already been made (auth. unpubl. data) to compare intracellular free amino acids of 2 groups of preda- tors, one of which had been allowed to feed on barnacles, the other upon oysters, for about 2 weeks. No signficant differ - ences were found by chromatographic analysis; nonetheless, comparisons of specific anatomical areas thought to be involved in chemoreception, such as the PREY SELECTION BY UROSALPINX 313 osphradium or propodial groove (an- terior pedal gland of Fretter & Graham, 1962) may yield positive results. In summary, it has been suggested that the phenomenon of ingestive con- ditioning is plausible in the light of 2 general lines of evidence, one from be- havioral studies and the other from field observations. The next task is to attempt to fit ingestive conditioning into an adaptive context: that is, of what value is such a mechanism to the predator? 3. Adaptive Aspects of Ingestive Con- ditioning Carriker (1957) points out that newly hatched Urosalpinx cinerea must be able to identify, locate, and penetrate food organisms. He demonstrated their ability to perform the second of these functions under his experimental con- ditions: in a typical experiment, in 1 hour, 72% (67) of the young predators had moved away from their starting places and either to slowly-moving effluents from 10,000 young clams (88%) or to a seawater control (12%). The response ratio of 72% is much greater than that reported from the present ex- periment with young U. cinerea; it is possible that in Carriker’s studies the young gastropods were in better con- dition, because they probably had been feeding in the interim between hatching and testing (pers. comm. from Dr. Carriker). However, this interpretation goes only part way in rationalizing dis- crepancies between the 2 activity rates (see percentage “none,” Table 5). It is possible that differences in response criteria may furnish the remainder of the explanation. With this qualification, I propose that the low response ratios reported in this paper for very young U. cinerea were not an accident but indicated immaturity of behavioral apparatus necessary for their direct orientation to prey effluents. Such an interpretation accords with ex- perimental results and also with the general hypothesis of ingestive con- ditioning. The adaptive value of such a necessity for maturation should be ap- parent: when protoconch stage snails emerge from egg capsules, they are probably best equipped for simple radu- lary browsing. In natural habitats (and in the nursery environment provided for his animals by Carriker), the sub- strate would be covered with hydroids, various matted algae, and other minute encruSting biota, any or all of which could provide nutrition for the young, 1-mm snail. (It is noteworthy that young predators in VIMS-37 were denied such a substrate but were kept in scrubbed plastic containers pending testing; this may have contributed to observed mortality rates.) Eventually, the newly hatched would rasp away at young barnacles or serpulids and for the first time ingest genuine prey tissue, an activity simultaneously accompanied by exposure to effluents from sur- rounding prey. lfinitial, and subsequent, contacts with large prey occurred ina Single-species habitat, the tendency to respond to that species would be rein- forced (viz. Wrightsville, Nobska, Ocean City); if not, it would not (any of the other southern habitats, including Shark Shoal). Meanwhile, responses to a general attractant such as ammonia would be reinforced, regardless of prey species. The same adaptive mechanism was suggested in the parasitic insect Neme- vitis canescens’ studied by Thorpe (Thorpe & Jones, 1937; Thorpe, 1938, 1939): М. canescens in Europe para- sitizes only larvae of the genus Ephestia (meal moth), but could be reared arti- ficially by insertion into larvae of the wax moth Meliphora grisella. While Nemeritis adults reared upon Meliphora still preferred their “normal” host, they gave signficant olfactometric responses to odors from the “abnormal” host, Meliphora. To make an analogous comparison, the results of experiment VIMS-37 (Table 4) showed that responses to prey effluents preferred by “naive” Urosalpinx cine- 314 L. WOOD yea were heightened after a diet of that prey (Balanus), and so were responses to another prey, second in order of preference in tests (Crassostrea). Such an adaptive mechanism shouldbe of great selective value to a predator (or parasite). Ideally, a predator main- tains a condition of stasis with its prey, but occasionally factors extrinsic to the predator-prey coaction will seriously disturb equilibria. Such a condition of disequilibrium was described in the paper by Fischer-Piette (1935). Should the prey population be completely elimi- nated under such conditions, the predator population must change its diet or it may not survive. Equilibrium between predator and prey may also be viewed as the interacting evolution of attack and defense mecha- nisms. The predator will be most successful if it can utilize a variety of food species, but opposed to this is the probability that continued attacks upon a single prey type will permit improvement of attack efficiency. Though at this time quantitative infor- mation is not available, impressions suggested by preliminary observations are: (1)attacktechniques differ radically from one prey type to another, and (2) individual U. cinerea change and im- prove attack techniques with experience (see Fretter € Graham, 1962, for a discussion of other observations of this kind). Hence on purely deductive grounds, the ideal situation for a preda- tor would be as follows: 1. To possess basic apparatus for efficient attack upon a variety of prey species. 2. To be unrestricted genetically as to prey selection, i.e., to have no innate preference. 3. To possess behavioral mechanisms for concentrating upon one prey type so that attack procedures become more efficient through practice. But of course prey species are also evolving. To have survived, they must have developed mechanical defenses against attack, resorted to camouflage, or increased reproductive capacities so much that large resident predator popu- lations can be economically supported. Let us examine the 3 major prey species of Urosalpinx cinerea and de- termine the degree to which the above adaptations have in reality been effected. As to mechanical defenses, in only 1 case may such have evolved. Carriker (1955) and Hancock (1959) have both reported that heavier-shelled, older oysters are not as attractive to U. cinerea as are younger ones, and this may be a “defensive adaptation.” In the case of barnacles, it depends upon the experience of the predator. At West Haven or Nobska, where Balanus balanoides had been dominant for a long time (longer at Nobska), an adult barnacle may be approached, penetrated (between opercular plates), and cleaned out by native U. cinerea within 20 minutes (pers. obs.). Predators inexperienced with barnacles, however, drill a hole through or between compartmental plates, re- quiring several hours. Mussels (Mytilus edulis) can be bored with apparent ease, but are more efficiently penetrated at valve intersections, though this con- tention is not easy to support with quantitative data. Camouflage, in this predator-prey re- lationship, should mean chemical unde- tectability, since it is by chemosensory means that prey is located. In this con- text, the evolution of Urosalpinx cine- rea may have outstripped that of its prey. For potential prey to be chemically “invisible” to U. cinerea, one or more of several conditions would have to apply, on the basis of available evidence: (1) the prey individual could be already dead; (2) it could be old, slowly growing or both; (3) it could be starved; (4) the species, during its evolutionary develop- ment, could “elect” not to produce an attractant; but since the probable at- tractants in this case are waste products, such a course seems unlikely. It is in the light of this point that the irony of the proposed ammonia attractant is most clearly revealed: a material which, if PREY SELECTION BY UROSALPINX retained in the prey, would be lethal in dilute concentrations, is excreted, only to be detected, in dilute concentrations, by receptors of a predator, also lethal. To avoid ammonia production, prey species would be required either to excrete all excess nitrogen as amino acids, short-circuiting normal de- amination processes, or as urea (an adaptation found in animals not blessed with abundant water). There is no evi- dence, certainly, that either process is occurring. A more effective defense, evolved by all of U. cinerea’s major prey, is high reproductive capacity. The enormous potential of the oyster is well-known: it has been estimated that each pair of mature Crassostrea virginica may shed more than a billion gametes into the water during a Spawning season. Mus- sels (Mytilus edulis) are similarly adapted. In 8 years of observing Ocean City populations, I have never noted even a slight reduction: the mussel zone is replenished, year after year, by tens of millions of young, despite predations of abundant Urosalpinx. Barnacles may be less able than the other 2 major prey species to repopulate denuded areas under combined pressures of predation or diseases and mechanical destruction from winter ice or storms. The intertidal habitat described by Fischer- Piette (1935) is one example, while Moore (1958) has discussed others. Moore points out that despite low rates of larval production per individual (he estimates a few thousand), there are so many adult individuals that within 1 km of the shore, barnacle larvae may be present in nn in numbers as high as 108-101 per km of shore. Such extremely high numbers seem doubtful, despite a personal observation of an overwhelming set of Balanus balanoides along the New England shoreline in May 1963. These early spring sets repopulate rocky surfaces which have been scoured free of barnacles at the time of winter ice. In any case, during the years in which I have monitored 315 the Nobska populations of U. cinerea and B. balanoides, in no instance has an entire barnacle population on any boulder been consumed by resident U. cinerea. In conclusion, it appears that Uro- salpinx cinerea is a highly, successful predator, not only because it satisfies ideal criteria proposed in this dis- cussion, but also because its prey has evolved techniques of survival which do not normally thwart its predations. U. cinerea can apparently perceive and lo- cate any ammonotelic, sessile inverte- brate within its range; it has evolved a variety of attack techniques which in each case are adapted to mechanical defenses evolved by its prey; and while it can and apparently does develop transient predilections for specific prey, it is not held rigidly to suchpreferences by genetic limitation. Its success as a competitor was noted by Cole (1942), who observed how long established popu- lations of Nucella lapillus and “Ocene- bra” erinacea on British east coast oyster beds were unable to compete with recently introduced Urosalpinx cinerea. While U. cinerea populations increased, N. lapillus declined, and apparently O. erinacea disappeared en- tirely from at least one area (Blackwater Estuary). It is therefore not surprising that, in the United States, U. cinerea is the only common, intertidal, predatory gastropod within its continuous middle Atlantic range: “Thats” lapillus inter- grades with it in the north, and Thais haemostoma floridana and Murex ful- vescens Sowerby can be found in limited numbers in isolated habitats to the south, but within its own intertidal do- main, from Cape Cod to the north coast of Florida, Urosalpinx cinerea (Say) is clearly dominant. ACKNOWLEDGEMENTS Parts of this work were done at the U.S. Fishand Wildlife Service Biological Laboratory, then at Annapolis, Maryland, 316 L. WOOD Messrs. John Glude and James Engle, in charge; Institute of Fisheries Re- search of the University of North Caro- lina, Morehead City, Dr. A. Chestnut, Director; Alligator Harbor Marine Laboratory of Florida State University, Dr. C. B. Metz, then Assistant Director; and the Virginia Institute of Marine Science, Dr. W. J. Hargis, Jr., Director. The facilities and assistance lent me by these institutions are gratefully acknow- ledged. It is especially a pleasure to affirm my gratitude to Professors R. P. Korf, E. C. Raney, and J. M. Anderson of Cornell University for their material help. Much of whatever clarity and continuity the reader may have found in this paper is the result of extraordinarily careful editing by Mrs. Anne Gismann. Deep gratitude goes to my wife and my mother for their unfailing moral and material support. Finally, let me mention that the work would have never been done at all without the support, critical vision, and warm encouragement of Dr. M. R. Carriker. LITERATURE CITED BAKER, В. B., 1951, Interesting shells from the Delmarva Peninsula. Nautilus, 64: 73-77. BLAKE, J. W., 1958, A biotic factor influencing the gastropod Urosalpinx cinerea in its choice of prey. M. A. Thesis, Univ. North Carolina, 35 p. 1960, Oxygen consumption of bivalve prey and their attractiveness to the gastropod, Urosalpinx cinerea. Limnol. & Oceanogr., 5: 273-280. 1961, Preliminary character- ization of oyster metabolites attractive to the predatory gastropod Urosalpinx cinerea. Ph.D. Thesis, Univ. North Carolina. Univ. Microfilms, Ann Arbor, Mich. (Diss. Abstr. 62-3117), 46 р. BROGDEN, W. J., 1951, Animal studies of learning, р 568-612. In: 5.5. STEVENS (Ed.), Handbook of experi- mental psychology. Wiley, New York. BULLOCK, T. H., 1953, Predator and escape responses of some intertidal gastropods in presence of starfish. Behaviour, 5: 130-140. BULLOCK, T. H. € HORRIDGE, G. A., 1965, Structure and function in the nervous system of invertebrates. Freeman, San Francisco. 2 vols., 1719:p: BURGHARDT, G. M., 1966, Stimulus control of the prey attack response in naive garter snakes. Psychonom. Sci., 4: 37-38. BURGHARDT, G. M. & HESS, E. H., 1966, Food imprinting in the snapping turtle, Chelydra serpentina. Science, 151: 108-109. BUTLER, P. A., 1953, The southern oyster drill. Conv. Papers, Natl. Shellf. Assoc., 1953: 67-75. CARR, W. Е. 5., 1967, Chemoreception in the mud snail, Nassarius obsoletus. I. Properties of stimulatory sub- stances extracted from shrimp. Il Identification of stimulatory sub- stances. Biol. Bull., 133: 90-105; 106- 127. CARRIKER, M. R., 1943, On the struc- ture and function of the proboscis in the common oyster drill, Urosalpinx cinerea Say. J. Morph., 73: 441- 506. 1954, Seasonal vertical move- ments of oyster drills (Urosalpinx cinerea). Proc. natl. Shellf. Assoc., 45: 190-198. 1955, Critical review of biology and control of oyster drills Urosalpinx and Eupleura. U. 5. Fish and Wildlife Serv., Spec. Sci. Rep.-Fish. 148, 150p. 1957, Preliminary study of be- havior of newly hatched oyster drills, Urosalpinx cinerea. J. Elisha Mitchell Sci. Soc., 73: 328-351. 1961, Comparative functional morphology of boring mechanisms in gastropods. Amer. Zool., 1: 263-266. CARRIKER, 'M.. R., SCOTT, Da Bese MARTIN, G. N., 1963, Deminerali- zation mechanism of boring gastro- pods, p 55-89. In: R. F. SOGNNAES PREY SELECTION BY UROSALPINX 317 (Ed.), Mechanisms of hard tissue de- struction. Amer. Assoc. Adv. Sci., Washington, D. C., Publ. 75. CHEW, K., 1960, Study of food prefer- ence and rate of feeding of Japanese oyster drill, Ocinebra japonica (Dun- ker). U. S. Fish & Wildlife Serv., Spec. Sci. Rep. -Fish. 365, 27 p. CHEW;.4K,; «& «EISLER, В., 1958, А preliminary study of the feeding habits of the Japanese oyster drill, Ocinebra japonica. J. Fish. Res. Bd. Can, 15: 529-535. CLENCH, W. J., 1947, The genera Pur- pura and Thais inthe Western Atlantic. Johnsonia, 2(23): 61-91. COLE, H. A., 1942, The American whelk tingle Urosalpinx on British oyster beds. J. mar. Biol. Assoc. U. K., 25: 477-508. CONNELL, J. H., 1961, Effects of competition, predation by Thais lapil- lus, and other factors on natural popu- lations of the barnacle Balanus bala- noides. Ecol. Monogr., 31: 61-104. DAVENPORT, D., 1950, Studies in the physiology of commensalism. I. The polynoid genus Arctonoe. Biol. Bull., 98: 81-93. 1966, Theexperimentalanaly- sis of behavior in symbiosis, p 381- 429. In: S. M. HENRY (Ed.) Sym- biosis: V. 1, Association of micro- organisms, plants, and marine organisms. Academic Press, New York. DOTY, M. S., 1957, Rocky intertidal surfaces, p 535-585. sn Jive Wi. HEDGEPETH (Ed.), Treatise on ma- rine ecology and paleoecology. Geol. Soc. Amer., Mem. 67. Vol. 1, Ecology. EKMAN, S., 1953, Zoogeography of the sea. Sidgwick Jackson, London, 417 р. ENGLE, J. B., 1942, Growth of oyster drill Urosalpinx cinerea Say feeding upon four different íood animals. Anat. Rec., 84: 505 (Abstr.). FEDERIGHI, H., 1931, Salinity death points of the oyster drill snail Uro- salpinx cinerea Say. Ecology, 12: 246-353. FISCHER-PIETTE, E., 1935, Histoire d’une mouliére: observations sur une phase de déséquilibre faunique. Bull. Biol. France et Belgique, 69(2): 153- 147. FRETTER, V. & GRAHAM, A., 1962, British prosobranch molluscs, their functional anatomy and ecology. Ray Soc., Lond., 755 p. GALTSOFF, P. S., PRYTHERCH, Н. Е. & ENGLE, J. B., 1937, Natural his- tory and methods of controlling the common oyster drills (Urosalpinx cinerea Say and Eupleura caudata Say). Us S. Bur. “FISH.” Circ) 29722 ав. GARTH, Т.В. .& MITCHELL, М. P., 1926, The learning curve of a land snail. J. comp. Psychol., 6: 103-113. GIBSON, D. G., Ш, 1964, Mating be- havior in Littorina planaxis Philippi. Veliger, 7(2): 134-139. HANCOCK, D. A., 1959, The biology and control of the American whelk tingle Urosalpinx cinerea (Say) on English oyster beds. Gr. Brit., Min. Agr. Fish. Food, Fish. Invest., Ser. 2, 22(10): 1-66. HANKS, J. E., 1957, The rate of feeding of the common oyster drill, Uro- salpinx cinerea (Say) at controlled water temperatures. Biol. Bull., 122: 330-335. HASKIN, H. H., 1940, The role of chemotropism in food selection by the oyster drill, Urosalpinx cinerea Say. Anat. Rec., 78: 95 (Abstr.). 1950, Selection of food by common oyster drill Urosalpinx cin- erea Say. Proc. natl. Shellf. Assoc., 1950: 62-68. HATHAWAY, ВБ. R.& WOODBURN, К. D., 1961, Studies on the crown conch Melongena corona Gmelin. Bull. mar. Sci. Gulf Carib., 11: 45-65. HODGSON, E. S., 1955, Problems in invertebrate chemoreception. Quart. Rev. Biol., 30: 331-347. HUTCHINS, L. W., 1947, The basis for temperature zonation in geographical distribution. Ecol. Monogr., 17: 325- 335. JENNER, С. E., 1959, Aggregation and schooling in the marine snail, Nas- 318 L. WOOD sarius obsolatus, р 183. In: M. SEARS (Ed.), Int. Oceanogr. Congr., Preprints, Amer. Assoc. Adv. Sci., Washington, D. C. KLEEREKOPER, H., 1961, The role of chemical sense in the orientation of Petromyzon marinus. Amer. Zool., 1: 456 (Abstr.). KOHN, A. J., 1959, Conus in Hawaii. 47-90. 1961, Chemoreception in gas- tropod molluscs. Amer. Zool., 1: 291-308. LORENZ, K., 1935, Der Kumpen in der Umwelt des Vogels. J. f. Ornithol. 83: 137-213, 289-413. LUCAS, C. E., 1955, bolites in the sea. Biol. Oceanogr., Res., 3: 139-148. MOORE, H. B., 1938, The biology of Purpura lapillus. Part Ш. Life his- tory and relation to environmental factors. J. mar. Biol. Assoc. U. K., 23: 67-74. 1958, Marine ecology. Wiley, New York, 493 p. MORGAN, C. T., 1951, The psycho- physiology of learning, p 758-788. т: 5. 5. STEVENS (Ed.), Handbook of experimental psychology. Wiley, New York. ORTON, J. H., 1929, Habitats and feed- ing habits of Ocinebra erinacea. Nature (Ser. 2), 124: 370-371. PELSENEER, P., 1924, Comment man- gent divers Gastéropodes aquatiques. Ann. Soc. roy. zool. de Belgique, 55: 41 (n.v.). PETERS, R. S., 1964, Function of the cephalic tentacles in Littorina plan- axis Philippi. Veliger, 7(2): 143- 148. PUTNAM, C. D., 1962, Thenon-random behavior of Aleochara bilineata Gyll (Coleoptera: Staphylinidae) in a Y- maze with neither reward nor punish- ment in either arm. Anim. Behav., 10: 118-125. The ecology of Ecol. Monogr., 29: External meta- Papers in mar. Suppl., Deep- Sea SASTRY, A. N. & MENZEL, R. W., 1962, Influence of hosts on the be- havior of the commensal crab Pinno- theres maculatus Say. Biol. Bull., 123: 388-395. SNEDECOR, G. W., 1956, Statistical Methods. 5th Edition. Iowa .State Univ. Press. STEPHENSON, T. A. & STEPHENSON, A., 1949, The universal features of zonation between tide-marks on rocky coasts. J. Ecol., 9: 289-305. 1952, Lifebetweentide-marks in North America. II. Northern Florida and the Carolinas. Jbid., 40: 1-49. THIELE,’ 37, 193% Handbuch der systematischen Weichtierkunde. Vol. 1, 771 p. Reprinted in 1963 by A. Asher & Co., Amsterdam, by per- mission of Gustav Fischer, Stuttgart. THORPE, W. H., 1938, Further experi- ments on olfactory conditioning in a parasitic insect. The nature of the conditioning process. Roy. Soc. (Lon- don), Proc., B., 126: 370-397. 1939, Further studies in pre- imaginal conditioning ininsects. Zbid., 127: 424-433. 1956, Learning and instinct in animals. Harvard Univ. Press, Cam- bridge, Mass., 439 p. THORPE, W. H. & JONES, F. G. W., 1937, Olfactory conditioning in a parasitic insect and its relation to the problem of host selection. Roy. Soc. (London), Proc., B., 124: 56-81. WELLS, H. W. & GRAY, I. Е., 1960, Some oceanic subtidal oyster popu- lations. Nautilus, 73: 139-146. WOOD, L., 1965a, A controlled con- ditions system (CCS) for continuously flowing seawater. Limnol. € Oceanogr., 10: 475-477. 1965b, Physiological and eco- logical aspects of prey selection by the marine gastropod, Urosalpinx cin- етеа (Say). Ph.D. Dissertation, Cornell University. Publ. No. 66-4684, Diss. Abstr., 27: 10. PREY SELECTION OF UROSALPINX RESUMEN ASPECTOS FISIOLOGICOS Y ECOLOGICOS SOBRE SELECCION DE PRESA POR EL GASTROPODO MARINA UROSALPINX CINEREA (PROSOBRANCHIA: MURICIDAE) Langley Wood Relaciones fisiolôgicas, ecolôgicas y de comportamiento entre el molusco rapaz Urosalpinx cinerea (Say) у su presa, en la costa oriental de los Estados Unidos, fueron estudiadas en el laboratorio y en ambiente naturales. Por la relativa frecuencia de los ataques naturales, y estudios oftalmométricos realizados en laboratorio, se comprobó que los percebes, Balanus spp., son preferidos por U.cinerea a cualquier otra presa mayor de la zona entre mareas, ostras, mejillones, etc. Esta preferencia se destacó en observaciones sobre 4416 caracoles en 11 habitats diferentes de la zona de mareas, en la contínua distribución de la especie de Massachusetts hasta el norte de Florida. La preferencia estadística no está fijada geneticamente: factores ecológicos pueden contarse para la selección de la presa en tales lugares. Uno de ellos es la mutua zonacion de predator y presa, y otro es la abundancia relativa de una especifica presa en la zona. Se introduce, con evidencia experimental, el concepto de «condicio namiento ingestivo, por el cual la tendencia de la especie rapaz de responder a las emanaciones de la presa, aumenta después de haber ingerido tejidos vivos de ella. En U. cinerea juveniles, se produjo una reversión parcial de tal acondicionamiento, por retorno a sus dietas originales. La operación de este proceso se comprobó experimentalmente en caracoles de habitats con presas únicas y otras múltiples. La tendencia estadística a preferir Balanus en lugar de ostras, fue parcialmente confirmada al experimentar con jóvenes, que se acondicionan a Balanus con mayor facilidad, lo que no siempre sucede con los adultos, cuyas divergentes dietas naturales fueron difíciles de cambiar. El aspecto evolutivo de estas relaciones se discute con preferencia particular a el valor adaptivo de la rapacidad de ingestión acondicionada. Restricción a una sóla presa podría tener efectos desoperativos, y es ventajoso para la especie rapaz ser capaz de alimentarse con más de una especie de presa. Sin embargo, diferentes técnicas de ataque son usadas para la penetración eficiente de cada presa, y en apariencia son adquiridas individualmente. Concentración en una especie única aumenta la eficiencia del ataque. El mecanismo aqui descripto como acondiciona- miento ingestivo produce concentrada influencia, sin la irreversibilidad, o fijación genética de la especificidad de la presa. ABCTPAKT ФИЗИОЛОГИЧЕСКИЙ И ЭКОЛОГИЧЕСКИЙ АСПЕКТЫ ВЫБОРА ЖЕРТВЫ МОРСКИМИ ГАСТРОПОДАМИ UROSALPINX CINEREA (PROSOBRANCHIA: MURICIDAE) KH.) ВУУД В лабораторных условиях и в поле, на литорали восточного побережья С. Ш. А. изучались поведенческие, Физиологические и экологические взаимоотношения между хищными моллюсками Uro- salpinx cinerea (Say). B природе, при исследовании относительной частоты нападения моллюсков на свою жертву и ответной их обонятельной реакции в лабораторных условиях оказалось, что балянусы (Balanus spp.) значительно более привлекательны Для Ir 319 320 L. WOOD cinerea чем большинство любых других их жертв Ha литорали, как например, устрицы и мидии. Полевые данные об этом предпочтении основываются на прямых наблюдениях над большим количеством (4,416 экз.) моллюсков из 11 мест обитания на литорали, в области непрерывного распространения U. cinerea, OT Массачузетса до северной Флориды. Эти статистические данные показывают, что предпочтение балянусов моллюсками не закреплено наследственно; так, по полевым наблюдениям экологические Факторы на литорали могут влиять на выбор жертвы моллюсками. Одним из таких Факторов служит степень совпадения зоны обитания жертвы и хищника; другим является относительное обилие на литорали данного вида жертвы. Также нельзя не учитывать при полевых наблюдениях роль внешних метаболитов, приводящих хищника к жертве. Чтобы ввести представление об условиях заглатывания, при которых у хищника наблюдается тенденция реагировать на влияние данного вида жертвы и которая ‘увеличивается после заглатывания им живых тканей этого вида, -автором приводятся полученные экспериментальные данные. Условия заглатывания у молоди U. cinerea частично меняются в обратном направлении, через возвращение их к исходной диете. Данные о действии этого процесса в природе были получены из экспериментов с улитками из мест обитания одиночных и многих экземпляров особей жертвы. Статистические данные о тенденции U. cinerea выбирать баля- нусов, предпочитая их устрицам, были частично подтверждены экспериментами с молодью и ювенильными стадиями, наиболее легко приспосабливающихся к балянусам; это не всегда бывает со взросдыми Формами, разнообразную естественную диету которых бывает трудно восстановить в эксперименте. В работе обсуждается взаимоотношения хищника и жертвы в эволюционном аспекте, с особым вниманием к адаптивному значению для хищника условий заглатывания. Ограничение лишь одним видом-жертвой было бы невыгодным для моллюска так как хищник получает определенные преимущества, если может питаться более разнообразно - более, чем одним видом жертвы. Хищниками употребляются различные способы нападения на жертву и это ведет к более эффективному захвату различных видов жертв и это, видимо свойственно различным особям U. cinerea. Описанный здесь механизм заглатывания обеспечивает такое концентрированное влияние при нарушении наследственно- закрепланной приуроченности (специфичности) жертвы моллюска. MALACOLOGIA, 1968, 6(3): 321-367 CULTURING ONCOMELANIA SNAILS (PROSOBRANCHIA: HYDROBIIDAE) FOR STUDIES OF ORIENTAL SCHISTOSOMIASIS! Henry van der Schalie and George M. Davis2 Museum and Department of Zoology University of Michigan, Ann Arbor, Michigan 48104, U.S. A. ABSTRACT Six types of vivaria were tested to determine the ecological conditions in which the 4 subspecies of Oncomelania hupensis thrived best in the laboratory. Efficient procedures were established to assure optimal conditions for: adult survivorship, production of young per female per unit time, survival of young and rapid growth of the young. Guided by an extensive survey of the literature and past experience, aquater- raria were established in the following containers: aquaria, cylindrical glass (battery) jars, large and medium sized shallow clay pots, plastic trays and Petri dishes. Success in culturing Oncomelania hinged on providing 2 distinct environments: (1) one where adult mortality was minimal and productivity optimal; and (2) another where young would grow rapidly without stunting and with low mortality. Conditions in the “medium clay pot,” with 5 males and 5 females, proved superior for both adult survival and production of young. Its superiority is due to the fact that a few females are highly productive in a limited volume where a soil-filter paper-water ratio apparently is optimal. Because of the small volume of the environment, more productive culture units can be used in place of larger, more cumbersome, less productive culture types. Finite monthly rates of adult mortality were about 2% during the first year; mortality had not reached 50% at close to 2 years and was usually considerably less. For pro- ducing young it is recommended that, at 24 months of age, females of all sub- species except Oncomelania hupensis quadrasi be replaced by young females. Those of the latter subspecies should be replaced at 12-14 months since they showed a marked increase in mortality in the second year of adult life. In other vivaria rapidly increasing rates of mortality, or finite rates of 12% or greater per month, indicated unfavorable culture conditions. The poorest cultures were in the aquarium, large clay pot and battery jar. The aquarium was extremely awkward to handle because of its bulk and the snails did not develop well in it. The large clay pot showed excessive mortality of young snails. The battery jar was characterized by high adult mortality rates and extreme erosion of the shell. The greatest yield of juveniles occurred in the medium clay pots where, depending upon the subspecies, 3-11 young hatched per female per month for the productive months, a rate about twice that of other cultures. Although pro- duction was sporadic, all subspecies reproduced each month of the year. lThe investigation was sponsored (in part) by the Commission on Parasitic Diseases of the Armed Forces Epidemiological Board and was supported (in part) by the U. S. Army Medical Research and Development Command, and (in part) by a research grant (5 T1 AI 41) from the National Institute of Allergy and Infectious Diseases, U. S. Public Health Service. 2Current address: 406 Medical Laboratory, U. S. Army Medical Command, Japan, APO San Francisco, California 96343. (321) 322 VAN DER SCHALIE AND DAVIS In medium clay pots, mortality of young in the parental culture was 4-10%, but generally 5%. In other vivaria it was much higher; in large clay pots, for instance, from 42-78%. The deaths occurred within 1-2 weeks of hatching. As discussed by us (1965), optimal rates of growth occurred in Petri dish cultures with 1-2 snails per dish at the 2.0-2.5 whorl stage. The logarithmic growth phase was over in 5-9 weeks, while full growth was obtained within 8-13 weeks, depending on the subspecies. A maximum of 20% mortality occurred over the 8-9 week growing period. In culturing Oncomelania the following factors are important. The soil, both a source of food and a substrate for depositing eggs, should be fine textured, high in calcium content, and should support a dense flora of diatoms. This flora, along with the attendant decomposers, provides an adequate source of food; the only other supplement is filter paper, the classic food additive. Water should be neutral to slightly alkaline (pH 7. 0-7. 6) and devoid of chlor- ine or other toxic agents, such as copper ions. Room level light intensity was found adequate for good survival of adults and production of young; constant light tended to increase the rates of mortality over prolonged periods of time (1 1/2 to 2 years), and also caused mortalities on account of excessive proliferation of algae. Optimal rates of growth for young occurred in light (130-160 ft. candles) cycled 10-12 hours per day. Productivity decreases and mortality rates increase with increased snail density. Daily maintenance is necessary to assure optimal conditions. Biotic factors most destructive to snail cultures were mold, worms (oligochaetes) and mites. A model is presented on the number of medium clay pots, Petri dishes, space and labor necessary to raise 500 snails per month of each of the 4 subspecies of Oncomelania hupensis. INTRODUCTION A number of papers have been published on various aspects of the natural history and laboratory culture of Oncomelania Gredler (1881). Many of them are useful in that they give information on such isolated phenomena as rearing and maintaining these snails; the time it takes for eggs to hatch; the optimal temperature for survival or pro- duction of young; the most suitable foods, etc. However, data usually are lacking which would enable one to pre- dict what culture conditions are necessary to produce a definite number of snails of uniform size and age within a designated time. Predictions of that kind are possible only when one knows the effects of various culture conditions on the mortality, natality, survival and growth of young. During the past 4 yearsit waspossible to develop the methods here presented. These methods are designed to assist those involved in experiments which require large numbers of specimens of each of the 4 so-called species of Oncomelania. To meet these demands, the various culture conditions described in the literature were alsotested. These tests have been undertaken not only to find the most efficient procedures but also to discover the conditions which will provide large numbers of snails allyear round for experimental purposes. The information has been organized to pro- vide in a comparative way the results obtained in the various types of culture and with an emphasis on the following categories: (1) mortality rates offieldandlabo- ratory-reared snails; (2) production of young per female per month; (3) the survivorship of newly- hatched snails in various en- vironments; CULTURING ONCOMELANIA TABLE 1. Types of vivaria used by various workers for culturing Oncomelania Vivarium Type 1) Large scale reconstruction of the environment 2) Aquaria 3) Battery Jars 4) Plastic Trays 5) Clay Flower Pots 6) Petri Dishes (4) conditions snails; and providing growth and survival of young Author Vogel, 1948 Stunkard, 1946 Ward et al. , 1947 DeWitt, 1951, 1952 Pesigan et al. , 1958 Bauman et al. , 1948 Moose & Williams, personal communication, 1965 Davis, 1967 Moose & Williams, 1961-62 van der Schalie & Davis, 1965 Davis, 1967 Sugiura, 1933 Williams, personal communi- cation, (1952) Wagner & Wong, 1956 Wong & Wagner, 1956 McMullen, 1949 Sandground & Moore, 1955 Otori et al. , 1956 Komiya et al. , 1959 van der Schalie & Davis, 1965 optimal important. Dimensions 8х 287 х 12” 9257416784102 12” diam. , 18” high 8” diam. , 10” high DURAS 352 TDR o 257 32 cm x 24 cm 5” diam. , 1.25” high 6” diam. , 1.75” high 10 cm diam. 15 cm diam. 9 cm diam. 9 cm diam. 9 cm diam. Note on Nomenclature 323 Success in hybridizing the 4 so-called (5) the culture type which provides minimal mortality as well as maximum production of young. For those involved inthe technological aspects of chemotherapy, etc., a model is presented which shows the type and number of cultures needed to raise 500 snails per month of eachof the “species” of Oncomelania. The model is designed to assist in procuring the facilities necessary with regard to equipment, space, and the manpower needed for rearing that number of snails. Based on culturing experience the recom- mended procedures take into account ease of handling as well as yield, both of which are considered to be equally species of Oncomelania and in obtaining fertile hybrids was reviewed by Davis et al. (1965). Burch (1964), after a cytological study of both the parents and the hybrids concluded that the 4 species were no more than geographic popu- lations or races of the same species. After additional morphological studies, Davis (1967) considered these 4 groups as subspecies of Oncomelania hupensis. Consequently, this subspecies concept will be applied throughout this work. HISTORICAL Reports (Table 1) on methods for maintaining Oncomelania in the labo- 324 VAN DER SCHALIE AND DAVIS ratory have appeared over a period of about 30 years. A review of this earlier work is important since those earlier papers lay a foundation upon which successful culture work has been made possible. The procedures that have proven successful in our laboratory were derived in part from the information obtained from many of the published papers. For purposes of review the data can be most easily organized under 2 headings: i.e. the physical and the biotic factors which are involved in culturing Oncomelania. A review by Ritchie (1955) should be consulted since his report also pertains, in part, to culture technique. Reports of the 406th Medical Laboratory (1951-64) contain a wealth of isolated facts concerning the biology of Oncomelania. Physical Factors 1. Culture Types In most cases cultures were designed to simulate as nearly as possible the amphibious natural environment of On- comelania snails. Vogel (1948) referred to such cultures as aquaterraria; they are generally characterized as a con- tainer with a bank of soil that slopes into a reservoir of aerated water. The kinds of aquaterraria used by various investigators vary, but usually 6 major types (Table 1) can be recognized. Active aeration was generally used only where there was a large reservoir of water, such as in aquaria or plastic trays. Battery (cylindrical glass) jars used as vivaria were constructed in various ways; Bauman et al. (1948) had a mud bank with a water reservoir, while Moose and Williams (1965, person- al communication) preferred to use these jars for holding large numbers (200 per jar) of snails for 6 to 7 months and covered their bottoms with moist filter paper. Davis (1967) modified this latter arrangement by adding a glass plate which was placed on the bottom with a mound of soil to encourage egg laying. Most of these vivaria were covered with glass plates with a crack left open for ventilation; battery jars were usually covered with cheesecloth or glass plates. Moose & Williams, (1961-62) coveredthe plastic trays with snug fitting lids per- forated with many small holes, while van der Schalie & Davis (1965) used plexiglass covers which were drilled in several places to permit ventilation. The soil banks were generally arranged to project one to several inches above the water in such a way that the emergent soil accounted for 1/3-1/2 of the area of the container. The reservoirs were generally one to several inches deep. The use of Petri dish cultures (see Table 1) for maintaining Oncomel- ania was recently discussed by van der Schalie & Davis (1965). Sandground & Moore (1955) used 10-15 cm Petridishes in which they constructed a sloping soil bank and small reservoirs of water. For food they used filter paper strips im- pregnated with sodium alginate. Komiya et al. (1959) used 9 cm dishes in which some had a sloping soil bank (“good for adults”) while others had a flattened soil mass covered by a Sheet of water (“good for young”); they used cultured diatoms and rice powder for food. Van der Schalie & Davis (1965) emphasized that such cultures were best for rearing young snails to maturity and were not suitable as vivaria for maintaining adults or encouraging production of young. 2. Substrata The several types used and mentioned in earlier accounts have been recorded in Table 2. While the substrate is generally described as a soil bank, the nature of the soil has seldom been further characterized. Wagner & Wong (1956) sterilized a mixture of soil made of 2 parts soil, 1 part gravel, and 1 part sand. In our laboratory (van der Schalie & Davis, 1965), it was found unnecessary to sterilize soil and our cultures were seldom plagued by mold or algal contamination. We found that a high incidence of mold accompained CULTURING ONCOMELANIA 325 TABLE 2. Substrates used in vivaria for culturing Oncomelania Author Sugiura, 1933 Substrate Soil from the habitat of Oncomelania hupensis nosophora dead leaves and sticks placed on the soil Stunkard, 1946 Mud Ward et al. , 1947 Bauman et al. , 1948 DeWitt, 1952 Sandground & Moore, 1955 Loam Mud, moss, small sticks Mud and sand Sandy loam sprinkled over with decaying vegetation Wong & Wagner, 1956 Wagner & Wong, 1956 Komiya et al. , 1959 Moose & Williams, 1961-62 Mixture of 2 parts soil, 1 part gravel, 1 part sand; sterilized; pieces of brick added Clayey-sandy soil Gravel lightly covered with sterile loam that was passed through a U. S. #80 screen. Soil from the habitat of Oncomelania hupensis nosophora van der Schalie & Davis, 1965 Non-sterile soil from the habitat of Pomatiopsis cincin- natiensis, a North American snail related to Oncome- lania. Soil alkaline; sand 40-70%; silt 13-42%; clay 7-24%. TABLE 3. Soil analysis of 34 colonies of Oncomelania hupensis quadrasi as presented by Pesigan et al. , 1958 Chemical Radical Range in ppm Average ppm 0. 8-20 .9 P 12. 5-100 90.8 K 45-185 90.4 Ca 250-5000 1286.8 NH3 0. 5-2. 5 0.85 Mg 0. 5-5. 0 35 Mn 0. 5-5. 0 345 Al 0. 5-50 23.0 NO» 1-5 015 Fe 0. 5-25 15. 0 504 50-250 85. 2 ei 25-100 51.0 average: 6.6 range: 4.6-7.2 soil sterilization. It appeared that soil texture per se is not critical for the survival of Oncome- lania hupensis nosophora (Ishii & Tsuda, 1951). Oncomelania were maintained successfully (van der Schalie & Davis, 1965) under conditions where sand varied from 40-69%, silt from 13-42% and clay from 7-24%. Wagner & Wong (1956) had success rearing Oncomelania on a mixture of 50% soil, 25% gravel and 25% sand. Hosaka et al. (1953) noted, however, better growth of О. h. noso- phora on sandy and pebbly soil. Pesigan et al. (1958) found Oncomel- ania hupensis quadrasi surviving in the field on sandy loam; fine sandy loam; silty clayloam; siltloam; andclay loam. They also found that soil chemistry had “nothing to do with the distribution of Oncomelania in the Palo area” in the 326 VAN DER SCHALIE AND DAVIS TABLE 4. The physical conditions for culturing Oncomelania Author Stunkard, 1946 NS* Ward et al., 1947 water Abbott, 1948 NS DeWitt, 1952 Moose & Williams, 1961-62 van der Schalie & Davis, 1965 water *Not stated Philippines. Table 3 containsthe results of their chemical analyses of soils from snail habitats. Areas whichlookedlikely to support snails, but in fact did not, were tested and the chemistry was not found to be significantly different. Komiya (1964) simply reiterated their findings. However, the fine grain soil particles are important since these snails use the substrate for food, sites of egg deposition and for encapsulating the egg with a jacket of fine soil (Sugiura, 1933; Abbott, 1946). 3. Water Water from a wide variety of sources can be used in cultures with success. DeWitt (1952) used tap-water which had stood a few days. Moose & Williams (1961-62) used water dechlorinated by Type of water used Filtered Potomac River Tap water allowed to stand a few days Dechlorinated tap water Ber Boiled, filtered pond Temperature of culture °C pH of water NS . 2-8. 260-270 6.8-7.8 O. h. quadrasi 240-270 O. h. позорйота O. h. hupensis 160-240 269-280 220 1,2 240+ 20 18 mg sodium thiosulfate per 1 1/2 gal- lons. Table 4 gives a list ofthe sources of water used by various workers (when mentioned) as well as the pH ranges. Stunkard (1946) statedthatthese snails lived equally well and reproduced in water with a pH range from 6.0-7.0 or 7.3-7.7. Neutral or alkaline water is recommended to avoid undue erosion of the shell. Wagner & Wong (1956) in- dicated that the water levels in vivaria did affect egg laying. They found that snail production was best in clay pot vivaria filled to 1/3 capacity with water. 4. Light Oncomelania avoids direct sunlight or strong, direct light rays (Abbott, 1948; Kawamoto, 1952; Pesigan et al., 1958; Komiya et al., 1959; Moose € Williams, 1961-62). Ward et al. (1947) used CULTURING ONCOMELANIA 327 TABLE 5. Foods provided for Oncomelania in laboratory cultures Author Sugiura, 1933 Paper; raw or boiled cucumber; cabbage; decayed Substances provided as food leaves; pieces of wood. Stunkard, 1946 Ward et al. , 1947 Bauman et al. , 1948 McMullen, 1949 Leaves smeared with yeast and diatoms. Coconut shells; fresh maple leaves; palm fronds. Nipa fronds, coconut husks, water plants. Filter paper. DeWitt, 1952 Sandground & Moore, 1955 Otori et al. , 1956 Wong € Wagner, 1956 Komiya et al. , 1959 Moose et al. , 1962 fluorescent lights to supplement ordinary room lights. Wagner & Moore (1956) and Chi & Wagner (1957) maintained cultures under a Single, 20-watt fluorescent tube 9”-10” above the cultures. Wagner (1954- 55) reported that there was a definite trend towards greater production of young under constant light. Van der Schalie & Davis (1965) showed that growth of young was excellent in Petri dishes maintained under 100-150ft. candles constantly supplied by a 40-watt, white, “cool,” fluorescent tube suspended 10” above the cultures. Growth was not appreciably diminished when the expo- sure to light was halved to 10-12 hours daily, whereas constant light stimulated too great a growth of algae on the non- sterile soil used. Cultures appear to thrive best under cycled artificial light, indirect sunlight or normal room-level daylight. 5. Temperature Table 4 gives the temperatures at Decaying vegetation and powdered commercial fish food. Sodium alginate. Filter paper, decayed leaves, straw. Filter paper and dried maple leaf. Rice powder and diatoms. Rice cereal which Oncomelania cultures have been maintained. DeWitt (1952) asserted that the 4 “species” of Oncomelania repro- duced throughout the year at tempera- tures between 260 С and 28°C. Wagner & Wong (1956) and Chi & Wagner (1957) reported that the greatest production of young occurred at 26° C for O. h. noso- phora, O. h. quadrasi and O. h. formo- sana. They stated that at this tempera- ture the snail had an intermediate rate of mortality. Van der Schalie € Getz (1963) tested the response of Oncomelania to ther- mal gradients and found that the average temperature preference was: О. h. hupensis, 21% С; О. h. formo- sana, 23° С; 0. h. nosophora, 25° С; O. h. quadrasi, 26% C. Van der Schalie € Davis (1965) found that the growth rates for the 4 subspecies of On- comelania hupensis were greater at 259 C + 29 C (under constant or fluctuating light) than those pre- viously reported. 328 VAN DER SCHALIE AND DAVIS Biotic Factors 1. Food The materials that have been pro- vided as food for Oncomelania are listed in Table 5. Sugiura (1933) stated that fecal material from snails in the field contained vegetable matter such as decayed leaves, roots of water plants and decayed pieces of wood. Mao (1958) found that the foods of Oncomelania in the field were Gramineae, diatoms, ferns. Dazo € Moreno (1962) stated that O. h. quadrasi “appears to be a herbi- vore, its diet consists mainly of green algae and diatoms.” In the laboratory, Ward et al. (1947) found that Oncomelania ingested dead vegetable matter, e.g. “water-logged maple leaves, twigs, husks and shells of coconuts, and palm fronds.” They stated that detritus and mud provide food. McMullen (1949) found that filter paper served as food, and more recent research at the Loma Linda University, Loma Linda, California (Wagner, 1954-1955) showed that snails survived longer on filter paper when compared with other substrates such as leaves, soil, fish food or wood. Winkler & Wagner (1959) discuss the physiological basis for filter paper digestion. Van der Schalie & Davis (1965) maintained Oncomelania solely on soil which supported a high level of green algae and diatom pro- duction. They found that young snails grew on such a substrate at rates higher than those described anywhere in the literature for laboratory-reared snails. 2. Density of Adults per Culture Abbott (1948) stated that an aquater- rarium 12” in diameter should contain no more than 100 specimens. Moose & Williams (1961-62) usually placed 50 specimens in their plastic tray vivaria. Wagner & Wong (1956) and Chi& Wagner (1957) varied snail density in clay saucers from single pairs to 5 females and 3 males per vivarium. Komiya et al. (1959) recommended putting 8-10 adults in Petri dish cultures. The dimensions for the above vivaria are givenin Table 1. To date, data have been lacking to in- dicate the optimum density in a given culture or environment at which one could expect least mortality and the greatest production of young. What data are available tend to indicate that cultures perform better when they are maintained with smaller numbers of snails. 3. Production of Young per Female The number of young produced per female varies withtime and experimental conditions as the compilation (Table 6) shows. These data provide the order of magnitude one can expect for average production of young per female per unit time. It is known that females mated only once can produce young for 2 years but not in the 3rd year (406th Medical General Laboratory, 1952). This trend was also reported by Chi & Wagner, 1957. Not all females will produce in a uniform manner while some will not produce at all. It is reported by the 406th Medical General Laboratory (1952) that of 114 individually isolated mature females, with 6.5 whorls or larger, only 89% produced young. The varying fecundity of different females was dis- cussed in the above reports. Since young are produced sporadically, it would seem important that observations be made over a 9-11 month period, to obtain more reliable averages. Pesigan et al. (1958) recorded an average of 8.26 days between egg laying for “Oncomelania quadrasi,” but the most common interval was 4 days. Chi & Wagner (1957) re- ported that the observed time lapse between 2 periods of egg-laying varied from 1 to 55 days. Ishii & Tsuda (1951, Japan) stated that egg laying for Oncomelania hupensis nosophora began in the laboratory in May and ended in August. Reports at the 406th Medical Laboratory in Japan (1952) indi- cated that peak hatching inthe laboratory (0.33/female/day) for O. h. nosophora occurred in April (eggs laid in March) and then declined to nil in September. Vari- CULTURING ONCOMELANIA 329 TABLE 6. Production of young under laboratory conditions as reported in the literature Average pro- Subspecies of} Length of |Initial No. Е Actual % 4 : : duction of Special Oncomelania | observations | females ser females not TER Author hupensis (months) |observed SE producing 2 female per day single pair of snails, female mated only once formosana Wagner, 406 Med. Lab. 1952 Chi & Wagner, 1957 nosophora single pair of snails, females mated only once Chi & Wagner, 1957 single pair of snails, females mated only once Pesigan et alí 1958 Water level fluctuating *NS = Not stated TABLE 7. Incubation period for eggs of Oncomelania Most common Subspecies of time lapse Total Oncomelania Author fr. egg laying range of Temperature hupensis to hatching variation °C Days formosana Chi & Wagner, 1957 позорйота NS Ishii & Tsuda, 1951 NS Otori et al., 1956 20-25 24-29 Chi & Wagner, 1957 Abbott, 1946 McMullen, 1947 Chi & Wagner, 1957 quadvasi NS * NS = Not stated 330 VAN DER SCHALIE AND DAVIS ous authors have confirmed that most or all “species” of Oncomelania pro- duced young in the laboratory the year round (DeWitt, 1952; Wagner & Moore, 1956; Chi & Wagner, 1957; Davis, 1967). 4. Sites for Egg Deposition The eggs of Oncomelania are laid singly, covered in a soil jacket by the snail and deposited on soil or other objects such as sticks or rocks. This type of egg laying is a characteristic of the subfamily Pomatiopsinae (Davis, 1967). Sugiura (1933) found that eggs of O. h. nosophora were deposited on the sides of his clay pot vivaria and also on dead leaves or sticks in the pots. Ritchie et al. (1951), noted that O. h. nosophora preferred soil as a site for laying eggs. Wagner & Wong (1956) found that in their vivaria O. h. noso- phora and O. h. quadrasi laid over 71% of the eggs above the water line. They noted that eggs of O. h. nosophora were laid largely in soil, while those of O. h. quadrasi were laid predominantly on objects provided, such as brick and sticks. 5. Hatching of Eggs and Incubation Period The average time (Table 7) for eggsto hatch tends to vary. Otori et al. (1956) showed that increased temperature shortens this period. As shown in Table 7, the range may vary from 20-30 days; some eggs may not hatch until 40 days after being laid. Otori et al. (1956) observed that 85- 90% of the eggs of Oncomelania hupensis nosophora hatched. Pesigan et al. (1958) found that on the whole 88% of the eggs of O. h. quadrasi did hatch, although in some experiments as many as 96% hatched. 6. Longevity of Adults In terms of field conditions, Sugiura (1933) showed that individual Oncome- lania hupensis nosophora may survive for 5 years. McMullen et al. (1951) did not recognize any peak mortality for this snail over a 2-year study period in the field. Li (1953) recorded peak mortality for O. h. formosana following a time of greatest reproductive activity and also indicated that most O. h. formosana live about 1 year. McMullen (1947) stated that O. h. quadrasi had a life span of at least 1 year in the field; Pesigan et al. (1958) found that after reaching maturity an average female lives 65.8 days (total age 7-8 months). In the laboratory Ritchie (1955) reported that Oncomelania hupensis nosophora can survive about 5 years. Wagner & Wong (1956) found an average of 40% mortality of adult snails in 1 year when they maintained them in medium clay pots under varying experi- mental conditions. Davis (1967) showed that, in the laboratory, mortality rates of field snails, obtained when they were about 1 year old, depended upon the culture conditions to which they were subjected. 0. h. formosana had a finite death rate of 12% per month when maintained in plastic tray vivaria under 150 ft. candles of light cycled 10-12 hours per day. When O. h. formosana was maintained in battery jar vivaria they had significantly greater rates of mortality; their mortality increased with time and 50% of the snails were dead within 4 months. 7. Habits and Survival of Young Pesigan et al. (1958) stated: “As has long been recognized, the newly hatched snails are aquatic and remain so for some time.” These authors presented data for Oncomelania hupensis quadrasi showing that during the first week after hatching very few snails are found out of water. After 2 weeks, 75-80% were to be found out of water and this per- centage remained fairly constant there- after; this behaviour also was verified with field observations. They also stated that about 50% of the young snails die by the end of the aquatic stage. However, van der Schalie & Davis (1965) found that among young snails of taxa of 331 CULTURING ONCOMELANIA JIOAIOSOI oqui 3119915 дэзем 04 Ixau d11IS «С 1499X9 JEIJ [10S Ч5тр jo лэзаэо ит punoul e OJUL peyjoours (w3 gg moge) [10$ (xeqowuerp 46 *2-2) "ysıp ul AJfeayusd paoe¡d punou e 110$ IIOAIOSOI ojut Suidojs 9123$ IPIM «5 в 14ээхэ 31 1108 Joded 191184 0094 "WO 05 JO SSOUNOIYF этапор uo SIS91I [105 yum эзета sselo 008 Ye pado]s 110$ SHIPUOY asueyoxe snooses лоу $[елл91 -Ul леаЗэл Je POT[IAP sse[sixeld as Е ee «S’LX «IT п эЗелэле 095 «G'T GL «TT ysıp 11394 rensey ael {xo7eq Ul se 93814 $5819 Br del Aaoyeq ul se 93814 55819 «GL °O-«S '0 SI9A09 Á[9A1] -u9 oye[d $5819 oyejd ssers Е SUOISUSULIA uorFeTIJu9A за -iod 04 рэцээло$ 7 uodo эзрэ euo ‘э381а SSI) aıenbs ut UMIIBAIA AlOAIOSOY DIUDJAWOIUO JO DOULUSJULICU PUB Зитлвэл JUOLOIJJO лоу ÁJOJBIOQB] INO UL P9ISDI BIABALA ‘8 ATAVL keaL 95а YSid 11384 104 Ато wipe 304 AEI9 93187 Jef Areyeg umıaenbvy UINTABATA jo odAL 332 VAN DER SCHALIE AND DAVIS CULTURING ONCOMELANIA 333 all “Oncomelania” in Petri dish vivaria, at the 2.0-2.5 whorl stage, a mortality of only about 10-20% occurred until the snails reached adult size (60 days). 8. Growth Rates of Young This topic was reviewed by van der Schalie & Davis (1965) in a paper per- taining to growth and stunting in On- comelania. They concluded that, in the laboratory, optimal rates of growth occur in Petri dish vivaria under fluctuating light (150 ft candles cycled 12 hours per day) with 1 or 2 snails per container. In- creasing snail density (i.e., 5or 10 young per Petridishor 40-50 young per plastic tray) caused a retardation in growth and also suppressed the development of the sex organs. Growth rates of 0.3 - 0.45 mm per week for the first 8 weeks indi- cated unfavorable laboratory conditions, while 0.6 - 0.65 mm per week were opti- mal. A review of the literature per- taining to growth is found in that paper. It was also shown that snails grown under optimal conditions are mature and will commence egg laying in the 9th - 11th week after hatching. FIGS. MATERIALS AND METHODS 1. Sources of Snails Oncomelania hupensis nosophora was collected from the Yamanashi Pre- fecture, Japan, an endemic area for Schistosoma japonicum. O.h. quadrasi was sent from Palo, Leyte, inthe Philip- pines. O. h. formosana was sent from Taiwan, and O. h. hupensis was provided by Dr. H. Vogel from his laboratory in the “Institut für Tropenkrankheit,” Hamburg, Germany. Е 2. Culture Types Utilized Six kinds of vivaria were initially tested; aquaria (Fig. 1), a plastic tray (Figs. 3, 7, 8), unglazed shallow large (Figs. 9-11) and medium diameter clay pots (Figs. 5, 6), battery jars (Figs. 2, 4) and Petri dishes (Figs. 12-14); their dimensions and other particulars are given in Table 8. Active aeration was not used where the volume of water in the cultures was small (Table 8) (battery jars, medium clay pots, Petri № —8 FIG. 1. Aquarium with aeration hose, filter paper strip and snails on the soil. FIG. 2. Battery jar vivaria; snails have climbed up on sides of jar. FIG. 3. Plastic tray container with aeration tube and plexiglass cover drilled with numerous holes to permit ventilation. FIG. 4. Top view of battery jar vivarium with a rectangular glass plate sup- porting a mound of mud; glass plate rests on filter paper. Note that snails are on the sides of the jar. FIG. 5. Four medium clay pots shown as they are usually housed, in a large plastic tray containing water, which allows handling them as a group and main- tains proper water levels for the unglazed clay pots. FIG. 6. Top view of a medium clay pot with filter paper strips placed in the culture to replace the filter paper which had developed a pronounced mold. FIG. 7. Plastic tray vivarium just prior to clearing its water reservoir of accumulated silt; snails are on the exposed soil surface. FIG. 8. Plastic tray not heat-treated for worms; hence soil shows churned up condition due to oligochaete activity. Two snails can be seen on the filter paper. 334 VAN DER SCHALIE AND DAVIS CULTURING ONCOMELANIA 335 dishes) or where there was little depth of water in the reservoir (large clay pots) as gaseous exchange was adequate. Active aeration was used where the depth of water in the reservoir was great as in the aquarium and plastic tray types. The (unglazed) medium clay pots were kept in large plastic trays, 4 per tray, (Fig. 5) that were half filled with water. The dimensions of the trays are 17’’ x15” x 3”. Glazed pots should work as well, and then the water in the large plastic trays would not be necessary. In our work it was easier to handle 4 of the clay pots as a unit than singly. The large clay pots were also unglazed and maintained on a constant flow water table (Fig. 9) with the water entering the table at 29° C and leaving at 28° C. 3. Soil The unsterilized soil used was obtained from the habitat, in Michigan, U.S. A., of Pomatiopsis cincinnatiensis, an am- phibious hydrobiid snail related to On- comelania. The soil was alkaline and supported high yields of naturally oc- curring green algae and diatoms. Van der Schalie & Getz (1962: 207-209) indicated that the soil textures in various habitats of Pomatiopsis varied widely. The soil used in this study was analyzed chemically by the Cooperative Extension Service of the U. S. Department of Agriculture at Michigan State University (see Table 9). Qualitative observations indicated that of all the organic matter present in the surface soil, decaying organic material was more abundant than any other organic component; living diatoms were very abundant but did not approach the density of decaying organic substances; green algae were less abundant than diatoms but slightly more prominent than the blue-green algae. Quantitative estimates of algal density were made in early July and August, 1962, for 36 quadrats (each 0.25 ft. Square) of river bank inhabited by Pomatiopsis cincinnatiensis. Uniform FIGS. 9-14 FIG. 9. Water table with numerous large unglazed clay pot vivaria; with glazed pots the water table would not be necessary. A single, large, unglazed clay pot could also be kept in a plastic tray like the one shown in Fig. 5. FIG. 10. Top view of large clay pot used for temporary storage of 100-200 Oncomelania. FIG. 11. View of a large clay pot which was not heat-treated to kill worms; note piles of worm castings over the surface of the soil. FIG. 12. Shelving units used to house Petri dish cultures. These cultures are maintained for 8-9 weeks on these shelves under 10-12 hours light daily provided by one fluorescent tube per shelf. FIG. 13. Four Petri dishcultures with lids removed to showthe arrangement of adult snails on the soil. FIG. 14. Four Petri dish cultures at different stages of algal development; Upper left: fresh and newly established dish; upper right: algal growth moder- ate (dark cap on soil) and soil spreading out towards side of dish (about 4 weeks); lower left: algal growth moderate and soil almost entirely covered by algae, the algal masses then begin to accumulate in water (5 weeks under constant light); lower right: algae excessive with soiland water clogged with algae (con- stant light, 5-7 weeks). Snails develop best when the algal growth is maintained at a moderate level. 336 20 р 2,800,000 19 e JULY 18 о AUGUST DEAD Ш —— LIVING NO. DIATOM CELLS IN UNITS OF 100,000 N wo aA ao on © © ee 05 10 15 20 25 30 35 40 45 DISTANCE IN FEET FIG. 15. Mean numbers of living and dead diatoms per gram of dried soil collected from the surface of the bank of the River Raisin. Samples were taken from the top of the bank down to the edge of the river. soil samples of the surface (1 mm depth) of the river bank were diluted and aliquots placed on slide counting chambers. Only whole diatom frustules were counted; filaments of green and blue-green algae were counted as 1, regardless of whether there were 4, 10, or more cells per filament. Results were recorded in numbers per gram of dried soil. As shown in Fig. 15 diatom density increased as one sampled from the top of the bank down to the river edge. Typically, there were many more dead than living diatoms (average ratio 7.7:1). VAN DER SCHALIE AND DAVIS —— GREEN ALGAE — — BLUE GREEN ALGAE @ JULY о AUGUST NO. ALGAE IN UNITS OF 1000 Top of river bank 0.5 10 15 2.0 25 3.0 35 4.0 DISTANCE IN FEET FIG. 16. Mean numbers of green and blue- green algae per gram of dried soil. Samples taken as in Fig. 15: note absence of green al- gae from samples in August. TABLE 9. Chemical analysis of air dried surface soil from Bank of River Raisin below Manchester, Mich. , U. 5. A., used in our cultures. Chemical radical % organic material Range in ppm 2540-3470 112-312 24-40 1. 5-50 25-75 < 600 Chlorides < 80 Soluble salt (K)* and total exchange capacity of this soil: K=18-42 (very low to low);* the ex- change capacity (mill equivalents per 100 gms soil) = 14. 4-20. 0. *K=conductivity value expressed in mho x 10-5 obtained from a 1:2 soil/water mixture. CULTURING ONCOMELANIA 337 Living diatoms were as abundant as 100-450 thousand per gram. In other areas studied on the same river bank, densities of 300-600 thousand per gram have been found. As many as 3 million dead frustules per gram were calculated. Diatoms appeared in every sample studied. In August the numbers of diatoms and other algae were significantly less (Figs. 15, 16). Green and blue-green algae seemed to be scattered irregularly; only half the aliquots studied contained blue-green algae, while 40% contained green. Peak densities of 83,000 per gram for green and 30,000 per gram for blue-greens were uncommon. Densities of bothtypes, where found, were generally below 20,000 per gram. In the August sample no green algae were found and densities of the blue-greens were much less than in July. Although the counting methods are somewhat crude and samples were not extensive in time or space, the results show the order of magnitude of micro- flora supported by the soil utilized in our experiments and in the routine maintenance of Oncomelania. The cultures when fully prepared were heated in an oven at 60° C for 2 hours. This baking eliminated difficulties often caused by earthworms (oligochaetes) since inunheated cultures the soil usually became totally disrupted with piles of worm castings, and the soil alsobecame badly churned up (Figs. 8, 11). While sterile soil was also tested, more than 80% of the cultures in our laboratory became either heavily infested with mold in a month or the snails suffered ex- cessive rates of mortality. Asa result, nonsterile soil has been used con- tinuously and with success. 4, Light Petri dish cultures were maintained under light (100-150 ft. candles) provided by a 40 watt, white, “cool,” fluorescent tube suspended 10” above the cultures (as before, van der Schalie & Davis, 1965). The light was cycled 10-12 hours per day. Large clay pots and battery jars were provided with normal daylight combined with regular overhead lights (50-80 ft. candles during the day; room level light). Medium clay pots were maintained either in room level light, or constant light of 70-100 ft. candles. Plastic tray cultures were placed in room level light, fluctuating artificial light (130- 160 ft. candles for 10-12 hours during the day in a windowless room) or con- stant artificial light of 130-160 ft. candles. 5. Temperature Temperatures in the Petri dish cultures were 23% - 270 С. In large clay pots the reservoir temperature was 28° while the ambient temperature was 26° + 10 С. Temperatures in the other cultures varied from 22° to 27° C with an average of 24° C. 6. Water Boiled filtered pond-water which had a pH of 7.3 was added initially to the cultures; water levels were then subse- quently restored with distilled water. 7. Food It was assumed that the natural soil with its high concentration of algae and decayed organic matter provided food energy in all cultures, especially those under light. Filter paper was used as a food additive in all medium clay pot cultures and in all other cultures main- tained in room-level light. 8. Routine Maintenance Water levels were checked daily inall cultures, at which time snails which had climbed up on the side of the containers were knocked down (Oncomelania has a pronounced negative geotropism). Filter paper was added to cultures when needed. 338 VAN DER SCHALIE AND DAVIS All cultures except medium clay pots were individually serviced every 2 weeks to remove and record dead adults, to record living and dead young snails as well as their whorl counts; further, reservoirs which had become clogged with soil particles were cleaned, and occasional cultures which were badly invaded by mold or in which the soil was much eroded were replaced. Medium clay pots were checked as above once only each month. ff, in daily maintenance, a culture was found to be moldy or infected with mites, that culture was changed immediately. In all cases where cultures were changed, the old culture was also maintained for another month to obtain all young which might hatch from eggs laid prior to the change. Young snails removed from parental culture were immediately placed into Petri dish cultures (2 young per dish), where they would grow to maturity in 8-9 weeks. These cultures were not removed from the shelves for that whole period, though water was added when needed and, after 2 weeks, constant watch was maintained to knock down young which had climbed up to the glass covers. Large vivaria, such as the large clay pots or plastic trays, were not changed unless excessive amounts of algae or mold had invaded the culture. Such cultures often remained in good condition into their 3rd year. The medium clay pots maintained under both room level and constant light averaged one change per culture every 8 months, although some had to be changed every 3rd month. Under both lighting con- ditions a prime reason for this rate of change was the utilization of the soil by the snails, as well as the wear on the soil brought about by handling during maintenance, which caused erosion and mechanical soil breakdown. Mites were often a problem and were responsible for 15% of all culture changes. Mites pro- liferate rapidly and tend to overrun a culture. Excessive mold and algae also necessitated changing cultures. In the medium clay pots black mold frequently developed on the filter paper. In such cases it was not necessary to change the culture but only the infected portion of the filter paper. 9. Adult Density and Sex Ratio The densities of snails in terms of their culture type are listed in Tables 10 and 11. The medium clay pots usually contained 5 females and 5 males. In other vivaria the females accounted for 45% to 51% of the snails. The sex of the snails was easily and rapidly de- termined by the method of Williams as adopted by Wong & Wagner (1954). Their technique takes advantage of the fact that Oncomelania is dioecious and the verge of the male canreadily be observed even at a very young stage. EXPERIMENTS A. Survivorship 1. Mortality of Stock Snails Initially Received In order to simplify matters, snails received from the field as well as the Oncomelania hupensis hupensis pro- vided by Dr. Vogel from his laboratory are referred to throughout the paper as field snails. Each of the 4 sub- species was received at different times and established in culture as shown in Table 10, where they are listed alpha- betically (column 1) and then according to vivarium type (column 2). The age of the snails was unknown but for reasons stated by Davis (1967) it was assumed that the snails had a minimal age of 7-8 months. Only those with varix formation were placed in culture since that con- dition indicates maturity and cessation of growth. None of the snails that showed extreme wear or shell erosion were used. The average rates of mortality are shown in Figs. 17-20 plotted on a semi- logarithmic scale, The data here given for Oncomelania hupensis formosana 339 SUTAIT 119$ SIIRUS эмо$ UJIM SUIMUIJUOO = 09 pvop 919m STIBUS [fe 99UIS PAJEUIWII9Y = L ‘saanqyno = E 05 9 I, GZ LZ LT 9 09 08 Le x 9 I ae 08 ст = EL L ЧА 007 El < à IT & 30 = 6 € he a) O el т ый à 91 с L Oo FA GT G 09 -ч ew * * 09 E * * 0) 4 is 1Z g 09 ST ет (syjuoun) (% syuow) 3 5141 Surureyye UTAIT SUIATT | OUI Je saanJ[n9o % pue STIEUS STIeus sa1Inimo OJH 9ANIMO JO OT 9,08 FU mo y (£10y810q8'] pue platy) зээлпоз 9pISJNO шодлу рэлтэоэл sTreus Jo зэлизто jo dn-398 433991 1598919 [9A9] WOOY = WOOY queysuog = *ysuog Suryeugoy]y = ‘лэму Ga SIT 65 OLT 8 905 Ст SOT el TE LT OTE Ta DET 61 SL 05 ES 05 СТ 05 OT GE (syyuow) aanyma adÁ3 iad sjreus aanyınd JO jo ‘ou ЭТ эзелэлу э3Звлэлу ‘JUSTT v GLY $ & T 5 ji G T T T Г L 7 OT OTS 819 SOT EIG OTS VES GL posn SIIeuS Te4OL 89 СТ OT 908 речовох jou pey ANTEIION x AVL], 95а = Ld 104 Авто untpew = dON 304 AEI9 эзлет = аэт Jef 19M = PA umıaenby = y :odAL WNIIBAIA "TIVLOL ıspıpomb DAOY dosou sısuadny “TOV Ld woog чот wooy fa DUDSOUAOL sısuadny DIUD]IULODUO a1An3mo jo setoedsqns soxnJpno р “ON ‘OT ATAVL 340 VAN DER SCHALIE AND DAVIS 20 = > nn PER CENT ADULTS SURVIVING a) 02 4 6 DADA ADAN VIVARIUM TYPE LIGHT CONDITION Room Level Room Level Alternating Light O Battery Jar O Large Clay Pot + Plastic Tray 22 4 26 28 30 32 34 (365 838 80 MONTHS IN CULTURE FIG. 17. Survival of field collected Oncomelania hupensis formosana in 3 different environ- ments. maintained in plastic tray vivaria are taken directly from Davis (1967); those for this same subspecies maintained in battery jars were obtained by further expanding those provided by him. A complete survey of mortality is shown graphically in Figs. 17-20, while the basic data are given in Table 10, which presents (1) the average length of life of a culture type, i.e., the period after which all of the snails had died in50% of the cultures; (2) the greatest length of culture life for a vivarium type; (3) when, on the average, 50% and 10% of the snails were still living in each of the vivarium types. Exponential death rates over the total culture period are inthe minority (linear plottings; Fig. 18, plastic tray; Fig. 20, battery jar); irregularly changing rates were more commonly encountered (Figs. 17, 19, except battery jar, 20). Trends in mortality are, therefore, more readily grasped from observing the changes in mortality rates recorded in Figs. 17-20. It is readily seen that Oncomelania hupensis nosophora did not survive as well as the other subspecies in the vivaria types tested. Whether this was due to these snails being older than originally supposed or because of the detrimental effects of the vivaria types tested is not known. However, one should note that relatively few cultures were used in establishing this subspecies (Table 10). Whatever the reason, O. h. nosophora reacted to the environments in such a way as to call for discussion apart from O. h. formosana and O. h. quadrasi, which reacted quite similarly to the environments provided. Since comparatively few O. h. hupensis were available it was not possible to test the survival of this subspecies in vivaria CULTURING ONCOMELANIA 341 VIVARIUM TYPE e Plastic Tray O Medium Clay Pot + Medium Clay Pot LIGHT CONDITION Room Level Room Level Constant <2 40 w <> PER CENT ADULTS SURVIVIN г o 10 9 8 7 6 5 4 0 2 4 8 10 2 4 16 18 20 MONTHS IN CULTURE FIG. 18. Survival of Oncomelania hupensis hupensis supplied from Dr. Vogel’s laboratory in 3 types of environment; note high survival in medium clay pots. other than those shown. However, its reaction in the plastic tray (Fig. 18) together with the data on productivity in that environment (Table 12) suggested that it responded similarly to O. h. nosophora. Optimal survivorship for Oncomelania hupensis hupensis occurred in the medium clay pot vivaria (Fig. 18). When this trend was recognized, 6 O. h. noso- phora, then 18 months in culture in several plastic tray vivaria, were placed in medium clay pots at room-level light. In their former environment the snails had mortality rates roughly Similar to those shown in Fig. 19, and the snails surviving at 18 months repre- sented about 1% of the original population. Data on this small population of O. h. nosophora are not included in either Table 10 or Fig. 19 because those cultures had not been uniformly treated and some snails had been removed for various experiments. Consequently, continuous life table data were inter- rupted. In the medium clay pots the Oncomel- ania hupensis nosophora suffered no mortality for 8 months (total calculated VIVARIUM TYPE @ Bottery Jar © Large Clay Pot + Aquarium x Plastic Tray e PER CENT ADULTS SURVIVING 0 2 4 6 O ZA 20 MONTHS IN CULTURE FIG. 19. Survival of field collected Oncome- lania hupensis nosophora in 4 types of vivaria in room level light. age of 33 months), dropped below 50% mortality after 18 months (calculated age of 43 months) and were all dead at 20 months or a minimal calculated age of about 3 1/2 years. In both large clay pot and plastic tray vivaria Oncomelania hupensis formo- sana (Fig. 17) ando. h. quadrasiinlarge clay pots (Fig. 20) still had 1% of their initial population after 3 years in the laboratory; their calculated ages then exceeded 3 1/2 years. In these environ- ments the exponential rates of mortality changed at various times (e.g., 4 months, large clay pot, Fig. 17) yet during the total life of the culture’s populations the overall rates of mortality were low with finite rates of 8% or less per month. From initial data with field snails it appeared that vivaria were unsuitable if rates of mortality increased markedly with time as shown by the curves for all vivaria (Fig. 19) and that for aquaria in Fig. 20. Increasing rates of mor- 342 VAN DER SCHALIE AND DAVIS — > mn 5-17-21) PER CENT ADULTS SURVIVING 14 16 VIVARIUM TYPE e Battery Jar O Large Clay Pot + Aquarium x Plastic Tray 20 MONTHS IN CULTURE FIG. 20. Survival of field collected Oncomelania hupensis quadrasi in the same 4 types of vivaria in room level light. tality yield curves that are not linear on the semi-log plots but bend toward the time axis. In the experiments with such increasing mortality rates, less than 1% of the field adults survived 24 months in the laboratory. Likewise, vivaria appeared very unsuitable if the finite death rate was 17% per month or greater, e.g. plastic tray, Fig. 18 (17%); battery jar, Fig. 20 (35%). It was evident that the battery jar vivarium and the aquarium were very poor. In the former, Oncomelania hupensis formosana and O. h. quadrasi had extremely low survival, in fact the lowest average culture life, as is apparent from Table 10. Inthis environ- ment, more than in any other, the shells became rapidly eroded. The aquarium was inefficient, not only because of poor survivorship, but we also found, as did Sandground & Moore (1955), that it was the most cumbersome to maintain: it takes much more space than the other cultures, it is difficult to move, to inspect for young and dead snails, to clean and service. The greatest length of life attained in this laboratory with field collected snails was, up to the date of writing: Oncomelania hupensis formosana, 4 years (some were still living); O. h. quadrasi, 3.5 years (all dead); O. h. nosophora, 3.5 years (all dead); О. h. hupensis, 2 years (and over 50% still living). Under the conditions tested, the average duration of life inthe laboratory 343 CULTURING ONCOMELANIA peyover jou AyıTeJıou %OG = x AVL] 95“ = Ld SULATT [9A9] моо = WOOY 3904 Азто WNIPON = dON IIMS S[IBUS owos se SUMULJUO) = 09 10815409 = *4ysuod 104 Аето э8летл = GOT реэр SIIRUS [fe Se payeurwas} = J, :5эли то SUIJPUI9I[Y = °лэму :14317 Jef Azoyeg = га :odÂL UMIIBALA | 69 | 6991 TVLOL ar 9 09 Ga T *ysuop $5 L 09 05 v "1071 * 91 09 er I wooy * * 09 6 I 6 "35909 de se 09 OT Z wooy ıspıponb LT СТ L * * 09 * * 02 voy dosou [Be y re 3 * * 09 * * 09 sısuadny GG 9 09 “IVY LG СТ 09 OSE wooy * * 09 ‘1SU09 * * 09 wooy LT G 09 WOOY PI G L wooy DUDSOWMAO[ (syyuow) | (syyuow) |3uryLım Jo = nn (syyuou) aanyma posn sısuadny Sutra | Sua | oumge | se ый us od4} 10d sjreus ne sıreus oe piuvjamooug sjreus STIeus | зэли по Sa so oanyma JO jo ‘ou у N [P10], emma Eo OT %0S ' OJIT SÍBIDAY эЗвлэлу satoadsqng yI3u9] 1502919 PIO SUJUOUI $ *Z-0 'Z DIUDJIMOIUN рэхзэл ÁJOYBIOQB] JO зэлизто Jo dn-38$ ‘тт ATIVL 344 VAN DER SCHALIE AND DAVIS PER CENT ADULTS SURVIVING > a Oa OOo 0 2 4 6 8: 125 714 VIVARIUM TYPE LIGHT CONDITION A Medium Clay Pot Room A Medium Clay Pot Constant + Large Clay Pot Room x Plastic Tray Alternating O Plastic Tray Room @ Battery Jar Room 18 1 20 22 24 26 BONNE MONTHS IN CULTURE FIG. 21. Survival of laboratory reared Oncomelania hupensis formosana in various environ- mental conditions; note high survival in medium clay pot. was 5-7 months or minimal ages of 12-14 months for all but Oncomelania hupensis hupensis, which had suffered only 10% mortality after 18 months in culture. It is evident from the response of O. h. hupensis and the initial results with O. h. nosophora that the average duration of life would have been greatly increased, in general, by maintenance in medium clay pots. 2. Mortality of Laboratory Reared Snails First generation laboratory reared snails were placed in culture at a known age of 2.0-2.5 months as listed in Table LL: The average rates of mortality are plotted in Figs. 21-23. With the exception of medium clay pots the density of snails per culture was lower than that of the “field” snails. Table 11 lists: (1) average life of a culture type, (2) greatest length of life for a culturetype, and (3) at what month 50% and 10% of the snails were still living. Optimal survivorship for all species was obtained in the medium clay pots, as was previously suggested onthe basis of data from field Oncomelania hupensis hupensis and O. h. nosophora: 50% mortality had not occurred after 14-18 months in culture (compare with field PER CENT ADULTS SURVIVING CULTURING ONCOMELANIA Ó O à О i Q D © O © © 30 x SK RO o — PORN A tt de FT 345 LIGHT CONDITION VIVARIUM TYPE ® Large Clay Pot Room X Plastic Tray Alternating о Plastic Tray Constant + Medium Clay Pot Constant A Medium Clay Pot Room 16 8 20 22 24 26 28 MONTHS IN CULTURE MIG.) 22. under varying light conditions. snails). In fact, 60% or more of On- comelania hupensis quadrasi were alive at the end of 14-18 months while 80% or more of the other taxa were surviving. The finite rate of mortality was less than 5% per month, generally about 2% per month. Mortality rates in battery jar vivaria for laboratory reared Oncomelania hupensis formosana (Fig. 21) confirm that this environment is not suitable for maintenance when optimal survivorship Survival of laboratory reared Oncomelania hupensis quadrasi in 3 types of vivaria is required. Plastic tray and large clay pots provided intermediate survivorship with finite rates of mortality between 5% and 12%. At this point optimal survival is defined in terms of a finite rate of mortality less than 5% per month for at least 2 years, and intermediate survival as a rate between 5% and 12% per month, while in a very unsuitable environment, poor survival rates of 13% or more are found. О. h. formosana appeared to have better survival in the 346 VAN DER SCHALIE AND DAVIS LIGHT CONDITION o Constant e Room Level 20 > с nn 00 PER CENT ADULTS SURVIVING VIVARIUM TYPE LIGHT CONDITION O Medium Clay Pot Room Level + Medium Clay Pot Constant e Plastic Tray Constant eee\e oo 0 2 4 6 8 ОИ ЛЕСУ 18 2073922 MONTHS IN CULTURE FIG. 23. A. Survival of laboratory reared Oncomelania hupensis hupensis in medium clay pots under varying light conditions; B. Oncomelania hupensis позорйота under simi- lar environments. plastic trays than in large clay pots, as also observed for field snails, but not so pronouncedly, while O. h. quadrasi survived better in large clay pots than in plastic trays. It was evident that the increased duration of light was accompanied by increased rates of mortality. Thistrend can be seen by comparing survivorship curves for Oncomelania hupensis formo- sana maintained under room level or alternating light in plastic tray vivaria (Fig. 21); by the curves for O. h. quadrasi in plastic trays under alter- nating and constant light (Fig. 22); and generally by the survivorship curves for snails in medium clay pots under room level light or constant light (Figs. 6, 22-23). Although spot checks on male and female mortality in cultures other than the medium clay pot showed essentially equal rates of mortality, female On- comelania hupensis quadrasi had notice- ably greater rates of mortality in the medium clay pots with 50% dead in the 16th or 19th month (Figs. 31, 32). O. h. hupensis females in medium clay pots under constant light also had a greater rate of mortality than the males with 20% dead at 14 months and no males dead. B. Productivity 1. Output of young in relation to sub- species and culture media Productivity was calculated in terms of young produced per female per month (y/£/m) and measured in terms of living and dead young snails removed from parental culture, not of eggs laid. In all cases involving medium clay pots, exact rates of female mortality were recorded. For all other cultures, rates of female mortality were determined by periodic checks on the cultures; it wasfound that, in these, female mortality rates were Similar to those of males. In Table 12 the average number of young obtained per field-collected fe- male per month is listed for each sub- Species, and in decreasing order. Only months when young were hatched are considered (column 6). These figures represent the actual output per female during periods of production. More meaningful to a culture program is the average y/f/m over a 12 month period, where non-productive months are averaged (Table 12, column 7). Com- parison of our figures for the monthly output per female with the average daily output cited from the literature in Table 6 shows that the levels of productivity obtained in our laboratory werefar below those recorded by previous authors. These lower levels may be due, in part, to the longer periods of time our snails were in cultureandtothe greater number of snails in our cultures. However that may be, our results concerning reproduction involve (1) the 347 CULTURING ONCOMELANIA *P9799/[09 PJOIJ SE 03 рэллэуэл oq [JIM Axoyeaoge] ayJ ит SeTOadsqns в JUNYSITALISO хоз pesh SIDINOS 9PISJNO WOAF рэлтэрэл STIBUS ПУ x AB17 911514 = Ld 104 Авто unıpaN = dOMW Ззацеилхэз[\у = *191V 304 Ae]o эзает = dol 14815905 = ‘15409 ael Aroyeg = га [9A9] оо = WOOy ‘JUSIT umıaenby = y :э94А3 UMIIBALA 61 ‘0 86 *0 wooy У LS 20 6S °0 wooy Ld TS *0 28 *0 wooy fa 86 °0 $7 “€ wooy dot 1SDAponb all: 00 °0 00*0 60*0 85 *0 ус "0 87 ‘0 LT*0 03 *0 24014050 0 00 *0 00 ‘0 00*0 G €T°0 GT°S T6°S G G9 °T 80° 98 ‘SG sısuadny OT 90 *T 59 *0 88 *0 68 79 GG “104 V Ld 61 LPT 6£ °0 66 °T 78 OL 174 wooy fa 09 6° 70 ‘1 $5 "$ Tp S6 65 wooy dOT DUDSOWAOf STIOYM (FZ-E1 syquow) | (¿1-1 syjuowu) Атао syyuow (syyuow) % ‘ON И 95-02 reak reak эл14эпрола лоу p9UySI[(8]S9 ax à DECITRE pug то} ST 10] эЗелэлу SEM paonpoad ozom | 348171 ra ed reyusaed ит aanımna au 3unoÁ yorum ed Jo sa19odsqns peop 3uno4 % (w/J/A) цотопрола jo yy3uoT Зитлир SYJUOJA SUOT}IPUOD Алозвлочет SUIAIVA Aapun зэтемоу рэзээПоэ APTOLJ JO soJea uoryonpoadey ‘ст ATAVL 348 VAN DER SCHALIE AND DAVIS effect of the environment on repro- duction for each subspecies and (2) the different potential of each subspecies for laying eggs in the laboratory. In column 4 of Table 12 one sees which of the taxa produced more con- tinually than the others in terms of months in which young hatched. In decreasing order of fecundity were 0. h. formosana, O. h. quadrasi, О. h. noso- phora (especially based on results inthe plastic tray cultures) and О. h. hupen- sis. In column 6 of Table 12, peak repro- ductive potential is recorded in terms of y/f/m for productive months only. The greatest potential was reached by O. h. hupensis in medium clay pots with 5.26 y/f/m. Considering all environments provided, O. h. formosana followed by О. h. quadrasi showed higher repro- ductive potential than О. h. nosophora (no production in large clay pots) and O. h. hupensis (no production in plastic trays). As will be shown later, pro- ductivity of all subspecies studied here is heightened in medium clay pots. For O. h. hupensis an indication of this Situation is given in Table 12 by its relatively high productivity in this culture type, while it produced no young in plastic trays. Reproductive potential in medium clay pots will, therefore, be discussed later under laboratory reared snails, where medium clay pots were tested with each subspecies. When productivity for each subspecies is considered for 2 years (Table 12, columns 7, 8) in the various environ- ments (medium clay pots excluded), it is evident that again O. h. formosana is the most fecund followed by O. h. quadrasi, O. h. nosophora, O. h. hupen- sis. The effect of the environment on pro- duction of young, when it can be com- pared, points to peak production in large clay pots; 2nd best productivity was in battery jars and 3rd best in plastic trays. Corresponding data for laboratory reared females are listed in Table 13. As already noted for “field” O. h. hupensis, the medium clay pot culture was superior, for the other subspecies also, to the other culture types tested. Generally, young were produced in a greater percentage of the months in culture (column 4), except for O. h. quadrasi in large clay pots and plastic trays. However, consideringthe average y/f/m for productive months only, (column 6) it was clear that superior productivity occurred in medium clay pots, at either room level or constant light. Production was generally higher than in field collected snails, but only in medium clay pots was it of the same order of magnitude as that cited by other workers (Table 6). The gradient in reproductive potential outlined for the various subspecies is further confirmed on the basis of multi- plication in the medium clay pots; yield in productive months was highest in Oncomelania hupensis formosana (Table 13, column 6) followed by O. h. quadrasi, O. h. nosophora and O. h. hupensis and the same pattern holds true for pro- ductivity over a 2 year period (columns 7 and 8) even though, in the first year, O. h. quadrasi had the greatest output of young. The lowest productivity in terms of percentage of months where young were produced and of y/f/m oc- curred in battery jars and plastic trays under constant light. There was no Significant difference between the number of juveniles hatched in medium clay pots in room level light or constant light when Oncomelania hupensis hupensis, О. h. nosophora and O. h. quadrasi are compared (Table 13, column 6). Significantly more young were produced, however, by O. h. formo- sana in constant light (11.10 vs. 6.48). Productivity increased in the second year for O. h. formosana and declined only in the case of O. h. quadrasi and O. h. hupensis under constant light. With the latter subspecies, one notes, by comparing Tables 12 and 13, that production for field specimens declined in the 2nd year, when kept under con- 349 AVI} 9135 4 = Ld CULTURING ONCOMELANIA [9A9] мооч = WOOY 309 AÁe]o 9318] = 407 10815409 = *]Ssu0) 304 Авто WnIpoN = TON SUTJEULOYIY = ‘лэу ‘JUSTT ael Aaoyeg = fq :9dAL UMIIBALA T ale 8 9F °S | 95 °S 96 °S GZ 36 $5 “TOV v 00*0 $8 *0 87 °S 55 LT v O) El 07 °S Se ‘y 60°S 85 68 $5 wooy G 98 °P 85 °S 6S *8 05 09 el wooy 3 08*€ 60 “ET Tp 01 05 06 8T *ysuop uspaponb 0 00 *0 78 "0 $896 GG 8T 2 ‘su09 8 94 *8 84"? LT°S 8T 68 ST wooy OT 90 °S cel 18:39 st Oot 8T ‘1su09 poy qosou ee v 64 *9 Т8 *0 ТО °5 PI TL OT 9 00 *0 TG 8Т "$ PI LS 8 sısuadny $8 65 9 wooy G $8 144 “IOV 9 78 95 wooy 84 El 36 wooy 3 00T 8T wooy Е 001 ST *ISU0J vupsowAaof STIOUM (p&-£1 syjuow) | (5т-Т syyuou) ÂTuO syjuoux (syyuoun) % "ON sisuadny O Ce reok reok aAtjonpoad рэЧ$ 435 dk ns Susan ye aanınd puz 10] IST 10] JOJ OSVIOAY SEM poonpoad элэм jusrT 2: ut Y reygusaed ut oxnyno DUI SunoÁ yorum eid Jo sor9odsqns реэр 3uno4 Y, (w/7/4) uorJonpoad jo ч13чэт SulInp SyJUON SUOT}IpUOD ÁXOJB1OQB] SUIÁIBA LOpUN SAJEUAJ рэлеэл ÁJOJB OB] JO SAJEA чоцопролаэч "EI ATAVL 350 VAN DER SCHALIE AND DAVIS AVERAGE YOUNG PRODUCED PER FEMALE PER MONTH 5 an =] PER CENT FEMALES SURVIVING UA 1G) В 10 I2 1401601800 2022 MONTHS IN CULTURE = 22 3 = 20 100 5 a 18 90 tad = 16 ® 80 2 ES = aii 70 = A 60 © ree] ws S 10 50 = 3 = Ев 40 к = = 6 30 5 > „4 20 & = 2 10 = 0 0 0 2 4 6 8 10) 12) 14) 16 ВВ 120822 MONTHS IN CULTURE FIGS. 24-32. The average number of young produced per female each month in medium clay pot cultures; each subspecies is shown under both constant and room level light. Data involve snails reared in our laboratory, excepting Fig. 26 which gives the production by females re- ceived from Dr. Vogel’s laboratory. shown. previous month. FIG. 24. Oncomelania hupensis room level light. nosophoya; stant light. The decline in fecundity for Oncomelania hupensis quadrasi is probably associated with the increasing rate of female mortality in the 2nd year (Figs. 31-32). 2. Sporadic Nature of Reproduction Young are not produced at a steady rate but sporadically. In any given culture numerous young will appear for 1 or 2 months, followed by a period when few or none hatch. Figures 24-32 show graphically the average number of y/f/m hatched each month in medium clay pots. Also shown are the per- centages of females surviving each month. On these graphs, since eggs may take at least 30-40 days to hatch, the production for any given month resulted from eggs layed by females alive in the previous month. From the data in Figs. 24-32 one can assess that the highest average pro- duction reached in any 1 month was 40 y/f/m for Oncomelania hupensis quadrasi, 22.5 y/f/m for O. h. formo- sana, 18 y/f/m for O. h. nosophora and The percentage of females surviving each month is also Production in any month is calculated on the basis of the eggs laid by females in the FIG. 25. Oncomelania hupensis formosana; constant light. 11 y/f/m for O. h. hupensis. Since one cannot predict general levels of productivity for a single culture, data for Oncomelania hupensis quadrasi in a medium clay pot in constant light are inconclusive. An example for such unpredictability is given by 2 medium clay pots, which were made up in the same way at the same time and main- tained side by side. Each received 5 males and 5 females of O. h. hupensis chosen at random froma large population of snails sent by Dr. Vogel. After 3 months 1 culture had produced young at 0.5/f/day, while the other produced none. 3. Adult Density, Light and the Pro- duction of Young A small scale experiment was set up using 2 sets of 4 medium clay pots and Oncomelania hupensis formosana. In one set, a single female was placed in each pot with 3 males; 2 pots were held under constant light and 2 in room level light. The second set of pots was set up similarly except that 2 females and 3 males were established in each CULTURING ONCOMELANIA — > PER CENT FEMALES SURVIVING AVERAGE YOUNG PRODUCED PER FEMALE PER MONTH D wo 2 мо + © co o 0 16 18 20 o mM > 6’ 78. 10 12.14 MONTHS IN CULTURE FIG. 26. Oncomelania hupensis hupensis from Dr. Vogel’s laboratory; room level and con- stant light. culture. The cultures were checked every 2 weeks for 12 months; young were re- moved and recorded, dead females and males were removed and replaced. The results are shown in Table 14. The yield in y/f/m of O. h. formosana in this experiment for a 1-year period can be compared with that for 10 snails per pot (5 males and 5 females) for the first year (Table 13, column 7). The production in these latter averaged 7.5 y/1/m (average of productivities at con- stant and room-level light environment). In this experiment, the overall average was 14 y/f/m for 1 and 2 females in both light and room-level environments. Generally then, productivity isincreased when the density of snails is lower in the same given volume and environment. Too few cultures were used with 1 and with 2 females to provide data that would accurately indicate in which case productivity per female had been greater. With constant light (Table 14) the average output in y/f/m certainly was the same. The difference in output for snails at room-level light (18 y/f/m against 28 y/f/m) may only reflect the difference in an individual snail’s capacity to pro- duce young: the greatest potential re- 351 = = 210 O 100 = 9 90 = 8 80 Е ыы > a 1 70 Е a 6 60 « a =] S 5 50 = a i 24 40 — = = pa] = 3 30 = = 2 202 S = 1 10 pr] =0 0 0.2 4 6 В 10 12 14. 161 18 20 MONTHS IN CULTURE FIG. 27. Oncomelania hupensis hupensis; laboratory reared under room level light. Е 210 100 pr 9 90 = 8 Q 80 = m 7 10 = ce => a 6 60 > a = = = = 40 = = = > 2 20 => ыы = 1 10 [re =z 0 0 CAMERA 16. 18 20 MONTHS IN CULTURE FIG. 28. Oncomelania hupensis hupensis; laboratory reared under constant light. corded so far in this laboratory for a 1- year period was that of 31.45 y/f/m in culture 7 of this experiment. Observations in this experiment further indicated that greater fecundity and lower rates of mortality occurred in room-level light. The data for cultures 3, 4, 7, 8 show that, in the first year, females at low density (2 females, 3 males) can produce about 18- 352 20 O 100 AVERAGE YOUNG PRODUCED PER FEMALE PER MONTH 5 сл > PER CENT FEMALES SURVIVING 024 6 8 0 1219 1618) 20122 MONTHS IN CULTURE FIG. 29. Oncomelania hupensis formosana; room level light. 30 28 Е 26 = 24 & 22 a. Е 20 100 18 ? 90 Zr 80 2 = > 2 14 70 = = ” 2 12 60 Y = = = 10 50 = = 8 40 = 3 S = 6 30 © la us = 4 20) = 2 10 0 0 0 2 4 6 8 10 12 14 16 18 20 22 MONTHS IN CULTURE FIG. 31. Oncomelania hupensis quadrasi; room level light. 28 y/f/m, i.e., at levels of magnitude corresponding to those reported by Chi & Wagner. However, the role of light is problematic. In the larger scale experiments presented in Table 13, one notes that this same subspecies showed VAN DER SCHALIE AND DAVIS in > 520 100 pe 18 90 2 16 Q 180 = < > 514 70 5 512 60 4 a = = 10 50 ы = 8 40 Е É 6 30 = 24 20 tu Z 2 10 5 = 0 0 0 42:44 26 ¿8 A0 ¿12 14 6 18002022 MONTHS IN CULTURE FIG. 30. Oncomelania hupensis nosophora; room level light. + BR MR KR BR N n © ons À à > + DD QQ = o soo s PER CENT FEMALES SURVIVING CA > +O — = г a o © AVERAGE YOUNG PRODUCED PER FEMALE PER MONTH — o an a 8 40 6 30 4 20 2 10 0 0 O24? II 1012 ae eae MONTHS IN CULTURE FIG. 32. Oncomelania hupensis constant light. quadvasi; greater productivity in constant light, when the number of y/f/m was 9.95 in the first year. Thisfigure is comparable to those here obtained for constant light: 5.03-5.33 (Table 14), rather than to the higher values obtained in room level CULTURING ONCOMELANIA 353 TABLE 14. Production of young by 1 or 2 female Oncomelania hupensis formosana per Medium Clay Pot Vivarium*, for 12 months, under different conditions of lighting Constant Light Room Level Light 1 female per culture No. Culture females 1 2 Average 5 6 Average *Each pot also contained 3 males. light: 18-28. 4. Seasonal ductivity Periodicity in Pro- The percentages of young produced each month of the year in medium clay pots both at room level and constant light with a period of 18 months are shown in Figs. 33 and 34. One should keep in mind that all cultures were not started simultan- eously. Some cultures existed for the whole 18 month period while others were only 4 or 5 months old. No seasonal rhythm was apparent. The marked fluctuations recorded throughout the year mostly reflect culture changes due to culture deterioration, or initiation of new cultures, with the usual month lag before young hatch. C. Survivorship and Growth of Young 1. Survivorship Data on mortality of young snails were derived from both the number of young which died in the parental culture before their routine removal to Petri dish cultures and from the number of young No. females dying Average y/f/m Culture Average which died in the Petri dish cultures. The former were just hatched, since about 95% of them had 2.0-2.5 whorls, i.e., hatching size. After hatching, the snails generally added 0.5 whorls per week in the parental culture for a period of at least 4 weeks. Very few dead with 3.0-3.5 whorls were found in the parental culture. Young snails were maintained in the Petri dish cultures for 8-9 weeks. At the end of that period very few deaths had occurred among snails with 4-6 whorls. Most of the snails which had died in the Petri dish vivaria had but 2.5-3.0 whorls, which indicated that they had died shortly after being placed in culture, i.e., at 2-4 weeks of age. Tables 12 and 13 showthe percentages of young which perished in each vivarium type used for producing young. It is evident that the culture type greatly affects the number of young surviving past the time of hatching. Especially noticable is the high rate of mortality in large clay pots where 42-78% of the young die upon the soil bank, especially towards the back wall of the bank, away from the water. In other vivaria mor- 354 VAN DER SCHALIE AND DAVIS tality was 10% or less; in the medium clay pots it was 4-5% with the exception of an average 9%for Oncomelania hupen- sis nosophora. In Petri dish vivaria, van der Schalie & Davis (1965) found that mortality was 10% or less when the algae were kept in check. In the experiments devoted to the study of growth rates, all con- ditions were closely observed and con- trolled. One would expect higher mor- tality rates in the routine handling of large numbers of snails because: (1) in rapid handling of thousands of snails the young tend to be treated a little more roughly in transfer to Petri dishes; (2) in routine maintenance, an individual Petri dish does not usually receive the attention assuring optimal care. Records kept for several years on Oncomelania hupensis formosana and O. h. quadrasi show that routinely handled snails suffer a 16-20% mortality in Petri dishes when PER CENT TOTAL YOUNG PRODUCED Lal ME Ам MONTHS OF YEAR лом ONCOMELANIA HUPENSIS FORMOSANA —— — ONCOMELANIA HUPENSIS NOSOPHORA isolated at 2.0-3.5 whorls but only 8-10% mortality when isolated at 4.0-5.0 whorls. Considering all sources of mortality from the age of hatching to 10 months, the percentage of snails surviving is shown in Fig. 35. These generalized graphs are based on the use of medium clay pots for producing young, on an assumed 5 or 10% mortality, depending on the subspecies, inthe parental culture and on a 20% mortality in the Petri dishes. 2. Growth In 1965 we have already discussed factors favoring optimal growth for Oncomelania hupensis formosana and those which retarded growth. In this ONCOMELANIA HUPENSIS HUPENSIS PER CENT TOTAL YOUNG PRODUCED ГОР. МА М. Л ОВО MONTHS OF YEAR FIGS. 33-34. Monthly percentages of young from all vivaria for each of the subspecies. FIG. 33. Young produced each month by On- comelania hupensis formosana and O. h. no- sophora. FIG. 34. Young produced each month by On- comelania hupensis hupensis and O. h. quad- vast. CULTURING ONCOMELANIA 355 — Mortality of adults in medium clay pots Mortality of snails in Petri dish culture Mortality of newly hatched snails 30 (encom HUPENSIS FORMOSANA 20 ONCOMELANIA HUPENSIS HUPENSIS PER CENT YOUNG SURVIVING nm ONCOMELANIA HUPENSIS NOSOPHORA ———— ONCOMELANIA HUPENSIS QUADRASI 0 1 2 3 4 5 6 7 8 Sing 10 MONTHS AFTER EGGS HATCHED FIG. 35. The percentage of snails surviving from the time of hatching up to 10 months. Mortality of newly hatched snails (0-2 weeks) occurred in parental, medium clay pot cul- tures; mortality of snails isolated in Petri dish cultures generally occurred from 2 to 4 weeks. After 8-9 weeks in Petri dish cul- ture, snails were placed in medium clay pots. section, discussion on growthis extended to cover: (1) growth rates for young of that subspecies maintained in plastic trays under constant or room level light; (2) growth rates for the 4 subspecies in Petri dish cultures under constant light. a. Growth in Plastic Trays Two plastic tray vivaria were es- tablished. One culture was placed under constant light (about 150 ft. candles) while the other was maintained in room level light. Forty snails, at the 2.0- 2.5 whorl stage, were placed into each culture. Every 3rd day a subsample of 20 snails was chosen from each culture; their lengths were measured to 0.01 mm, using a Nippon Kogaku sliding ocular micrometer, and they were then returned to their respective cultures; theresults are shown in Fig. 36. In the initial 8 weeks the snails under constant light grew 0.47 mm per week, while those at room level light grew only 0.22 mm per week. b. Growth in Petri Dishes Specimens of each of the 4 subspecies studied were placed singly into Petri dish cultures at the 2.0-2.5 whorls stage. The cultures were maintained under LENGTH IN mm © Room Level Light @ Constant Light 0 1 2 3 4 5 6 7 8 3 WEEKS IN PLASTIC TRAY CULTURES FIG. 36. Growth rates for Oncomelania hu- pensis formosana maintained in plastic trays under constant light and room level light (40 snails each). constant light and measurements of shell length were made every 3 days; the data are plotted in Figs. 37-40. In all, 25 specimens of each sex of Oncomelania hupensis formosana were measured (data from Davis, 1967); 19 females and 18 males of O. h. hupensis; 23 females and 30 males of O. h. nosophora; and 20 female and male specimens of O. h. quadrasi. Calculations from growth curves (Figs. 37-40 and Table 15) indicate that growth rates for all the subspecies, except Oncomelania hupensis quadrasi, exceed 0.60 mm per week for the first 8 weeks. In the logarithmic phase of growth, the length of shell increases at rates greater than 0.40 mm per week for О. h. quadrasi, and at about 0.70 mm per week for the other subspecies. O. h. formosana grows most rapidly, followed ру O. h. позорйота, O. h. hupensis and О. h. quadrasi in that order. In nature, Oncomelania hupensis hu- pensis has the largest shell followed in decreasing order of size by O. h. noso- phora, О. h. formosana and 0. h. quad- yasi. These differences in size are maintained in laboratory reared snails. In Figs. 37-40 one notes that sexual di- morphism is more pronounced in the smaller subspecies at earlier periods 356 VAN DER SCHALIE AND DAVIS LENGTH IN mm LENGTH IN mm ATA oom A AC PEL WEEKS IN PETRI DISH CULTURES 0 VE OR RE TTC ale ea El WEEKS IN PETRI DISH CULTURES LENGTH IN mm LENGTH iN mm Ot 2 За A UL WEEKS IN PETRI DISH CULTURES NAVA OC В TN A WEEKS IN PETRI DISH CULTURES FIGS. 37-40. Growthrates for the 4 subspecies of Oncomelania hupensis, reared singly in Petri dishes under constant light. FIG. 37. Oncomelania hupensis formosana FIG. 38. Oncomelania hupensis hupensis. (from Davis, 1967). FIG. 39. Oncomelania hupensis nosophora. FIG. 40. Oncomelania hupensis quadrasi. CULTURING ONCOMELANIA 357 TABLE 15. Comparative growth rates of the subspecies of Oncomelania hupensis in Petri Dish Cultures under constant light Subspecies Growth rate Relative growth Termination of : of over first rate between logarithmic Maximal growth Oncomelania 8 weeks in 3 and 6 weeks in phase approached hupensis mm per week mm per week (weeks) (weeks) formosana hupensis nosophora quadvasi of development than in the larger ones; sexual dimorphism throughout growth is most pronounced in O. h. quadrasi (Fig. 40). The graphs start with initial meas- urements of 0.6-0.7 mm, the size of the young snails generally found in routine isolation from parental cultures. Young snails, measured directly upon hatching, have a minimum length of 0.5 mm and take about a week to attain the size of 0.7 mm. DISCUSSION 1. Vivaria Each of the culture conditions dis- cussed in this paper is extremely com- plex with regard to the interactions of such variables as adult population density, volume of vivarium, soil and water surface area, lighting andtemper- ature, etc. Shifts in these variables markedly affect productivity and sur- vivorship. In the successful culture of Oncomelania an environment must be satisfactory on 2 counts, i.e., it must provide conditions (1) where adult sur- vival is optimal and where production of young is the highest, and (2) where the young have rapid growth with low mortality as well. We have found that no single vivarium type simultaneously provides these 2 points and that the 2 aspects must be separated and handled in different ways. Our data show that among the vivaria tested for maintaining adults and en- couraging the production of young, the medium clay pot provides the best en- vironment. Easy handling must also be a major criterion for rearing large numbers of Oncomelania; for efficient manipulation the medium clay pot again proved to be superior to the other types tested since it was most rapidly prepared, maintained and can be most easily surveyed for the recovery of young snails. It takes less space than the bulky aquaria, plastictray or battery jars. No active aeration is needed as there is adequate gaseous exchange in the little water used in the pot. In the medium clay pot the greatest numbers of young are produced from fewer females, which show exceptionally good survival. The use of a separate container for the young, the Petri dish, has proved to be an efficient method of rearing. In the parental culture overcrowding usually inhibits the growth of the young snails very markedly. When relatively low numbers of juveniles (40) are placed in plastic tray vivaria, their growthrate is extremely slow, attaining only 1/3 - 2/3 the rate in Petri dish vivaria. If as few as 15-20 young were reared in the plastic trays the problem of in- creased bulk (i.e., of many trays), the necessity for active aeration and in- creased time in maintenance would make the process laborious and inefficient. 358 VAN DER SCHALIE AND DAVIS In any case, data are not available to state whether comparative, i.e., optimal, rates of growth would have occurred with 20 young in a plastic tray. The aquarium proved to be the poorest of the vivaria types tested, especially in terms of manipulation. The very bulk of an aquarium makes it the hardest con- tainer to clean; also, on account of its depth, there is considerable difficulty in handling the adults and finding or removing the young. In the large clay pots, survival of adults was of an inter- mediate type. The relatively low pro- duction of young and their very poor survival after hatching makes this culture unacceptable as a vivarium for the production of young. We have found the large clay pot (Fig. 10) to be an excellent vivarium for the temporary housing of 200-300 snails which are to be used up in experiments. Active aera- tion is not necessary as only a shallow reservoir of water is used and the cul- ture is easily maintained when one 1$ not concerned with the raising of young. The battery jar does not encourage production of young and showed excessive rates of adult mortality. This kind of vivarium is best used to keep large numbers of snails (100) for short periods of time (4-5 months) as it is easy to maintain, requires little space, and no soil substrate is necessary. Survival of adults in the plastic tray was of an intermediate degree and gen- erally the yield of young was low. Aside from the necessity for active aeration because of the depth of water in the reservoir, it can be efficiently main- tained but not as quickly and easily handled and used in laboratory studies as the medium clay pot. The use of large numbers of these cultures for any purpose presents the problem of supplying numerous air outlets for active aeration as well as the need for ex- tensive shelf space. This culture type is not suited for holding large numbers of snails for experiments. 2. Survivorship Pesigan et al. (1958), using the aquarium type vivarium, found in labo- ratory experiments that only 50% of the young of Oncomelania hupensis quadrasi survived past the aquatic stage, i.e., 20 days, and only 21% survived 70 days. Over the past 3 years, we found that, by using the medium clay pot, 70% or more of several thousand young survived for 90 days. When the large clay pot vivaria were used the survival dropped to levels of about 20% at the end of 70 days. The higher mortality of newly hatched Snails in the large clay pots appeared to be associated with the areas of egg deposition. Eggs were quite frequently deposited near the soil-brick interface of the retaining wall, an area farthest removed from water. Most of the dead young were found in this area and it is thought that the distance to the water may have been a main factor in the sur- vival of the young. The good survival of young in our culturing methods is attributed, in part, to thinning them out in Petridishvivaria when they are about 1-2 weeks old. This type of culture appears to afford a pair of snails an adequate and balanced en- vironment with respect to food energy, volume of container, and soil-water ratio. When 15 female Oncomelania hupensis quadrasi were maintained in medium clay pots (5 per pot) they had anaverage finite rate of mortality of roughly 3%per month for the first 14 months after maturity (total age 16.5 months) and thereafter an increasing rate of mor- tality with 50% dead at an average 18 months after maturity (total age 20.5 months). Pesigan etal. (1958) calculated from field data for this subspecies that females, on the average, lived 65.8 days after reaching maturity, i.e., for a total age of about 5 months. Considering the other subspecies, the CULTURING ONCOMELANIA 359 finite rates of female mortality in medium clay pots were about 2% per month for 16-18 months (total age 18.5- 20.5 months). Inother vivaria suchrates were much larger, usually much in excess of 9% per month over a 24 month period. Generally, any environment where the finite rates of mortality exceed 12% per month is unsuitable for maintenance. Where the finite rates are 2 - 9% per month and the rates begin to noticeably increase before 15 months, one suspects a deteriorating condition in the culture. Population density affects survival. When 250 or more adults were placed in large clay pots (Tables 10, 11), initial rates of mortality were exceedingly high (Figs. 17, 19). However, when 200 or less were placed in culture, the rates of mortality were less pre- cipitous (Figs. 20, 21, 22). Based on area alone and on the fact that optimal survival was obtained when 10 adults were maintained in the medium clay pot (for all the 4 subspecies), it is estimated that, in the large clay pot, optimal survival would probably occur when 60 adults are maintained in it. As shown in Fig. 22, when only 43 Oncomelania hupensis quadrasi were kept in the large clay pot, survivor- ship, while not as excellent as that in medium clay pots, was extremely good. Lowering the snail density in battery jars from 168 to 77 per vivarium did not appreciably changethe mortality rate of Oncomelania hupensis formosana (Tables 10, 11; Figs. 17, 21). Lighting seemed to affect rates of mortality more than density, as far as survivorship in plastic trays was con- cerned. Increasing daily exposure to light was correlated withincreased rates of mortality. 3. Productivity It is expected that slightly different rates of reproduction and survivor- Ship should occur in different labo- ratories, even where using the same techniques and Similar set-ups, on account of such variables as: different origins of snails, divergencies in temperatures that might be difficult to regulate, and of maintenance care that may vary markedly. Nevertheless, one can expect the same trends to occur. Thus, in experiments involving only 1 or 2 females per medium clay pot (Table 14) our data essentially agree withthose of Chi & Wagner (1957), who found that greater production occurred at a low density of females per medium clay pot. Whereas, with 5 males and 5 females per medium clay pot under constant light, Oncomelania hupensis formosana produced 10 young/f/m in the first year, i.e., 600 snails per culture in 12 months; with 2 females and 3 males per culture under room level light, production was about 27 y/f/m, or 648 snails per culture in 12 months, as shown in Table 14. Perhaps the yield could be increased above those levels without increasing the number of cultures by having 3 females and 2-3 males per culture. High levels of production can be ex- pected from the medium clay pots in the 2nd year as well as in the 1st year, except for Oncomelania hupensis quadrasi. Because female mortality in- creases in the 2nd year and production declines, it is advisable to set up new breeding stocks for this subspecies after 14 months. 4. Growth Growth in Petri dish cultures was more rapid and uniform than in any of the culture conditions tested as shown in this paper (Table 15) and previously (van der Schalie & Davis, 1965). Only the growth rates of Oncomelania hupensis quadrasi were comparatively slow and more closely equivalent to the rates previously reported for that subspecies, i.e., 0.20-0.25 mm per week (McMullen, 1947; Сы € Wagner, 1957; Pesigan et al., 1958). We recorded average rates of 0.36 and 0.45 mm per week for 360 VAN DER SCHALIE AND DAVIS male and female O. h. quadrasi, re- spectively, and of about 0.6 and 0.7 mm per week for the males and females of the other subspecies. According to Chi & Wagner (1957) О. h. quadrasi reached up to 6 mm in length. In our laboratory the snail rarely exceeded 5mm. Data by Pesigan et al. (1958) also show growth culminating at 5mm. In our laboratory, full growth occurred in 11-12 weeks (Fig. 40), whereas, according to Chi & Wagner and Pesigan et al., it took 20 and 27 weeks, respectively, for full growth to occur. Pesigan et al. (1958) pointed out that О. h. quadrasi males reach sexual maturity at a smaller size thanfemales: 3.5 mm as opposed to 3.7 mm. Hence, as shown in the growth curves for the males and females of that subspecies, (Fig. 40) sexually mature snails can be removed from Petri dish cultures at 7-8 weeks, by which time the snails are about 84% maximum size. In 1965 we reported that Oncomelania hupensis formosana males also reached sexual maturity at a smaller size than females, and that sexual maturity was a function of size, not of age. Our histo- logical data (1965) showed, however, that at 96% growth only 29% of the snails were fully mature, 47% being almost mature. Hence, O. h. formosana should be maintained in Petri dish cultures for 9 weeks (about 98% full growth, Fig. 37) to assure full sexual maturity of all the snails. We suggest that Oncomelania hupensis hupensis be left in culture for 10 weeks and O. h. nosophora for 9 weeks, when they have attained 90-95% total growth (Figs. 38, 39). We have foundthat snails of these subspecies at these sizes serve well in establishing new cultures for propagation of stock and are suitable for all types of experimental work. 5. Light as a Factor “Room level light” as described inthis paper is sufficient for maintaining adults with minimal rates of mortality. Young are produced equally well in room level light or constant light with one exception: Oncomelania hupensis formosana ap- peared to produce more young in constant light (70-100 ft. candles) when 5 males and 5 females are kept per medium clay pot. The pilot experiment with 2females and 3 males per medium clay pot showed, however, greater production in room level light. As a result the uniform use of room level light is recommended. Stronger constant light (150 ft. candles) was associated with increased rates of adult mortality. One of the mainfactors involved is the rapid proliferation of algae on the soil. Algal mats soon cover the soil and, if the culture is not changed, snail mortality increases. We encountered the same problem when using constant light over the Petri dish cultures. Within 4 weeks algae overran the cultures, causing increased mortality (van der Schalie € Davis, 1965). When light was provided for 10-12 hours per day nearly the same rates of growth were obtained, but algal growth was held in check, and mortality decreased. The rate of algal growth varied markedly with the different soil collections brought in from the river bank. If algal growth was slight (algae covering not more than 1/4 soil area) in 5 weeks, the duration of light was increased. We attempted to maintain a moderate algal growth in the dishes up to the 8th week. 6. A Model for Rearing 500 Snails of each Subspecies to Maturity each Month To assure 500 living snails of each subspecies in the medium clay pot vivaria at the end of 8 weeks, one needs a total production of 666 newly hatched young of each subspecies, assuming a minimum mortality of 25% of the young up to 8 weeks. Table 16 liststhe number of medium clay pots needed (with 5 males and 5 females per culture) to produce this number at the rate of production listed in Table 13 for the first 12 months. CULTURING ONCOMELANIA 361 TABLE 16. Medium Clay Pot Cultures needed to produce at least 500 snails per month for each subspecies of Oncomelania hupensis Gee adi ae ubspecies Rare rate Minimum ен No. trays needed of : cultures to house 4 : of production - No. of Oncomelania in’ vie needed with It Medium Clay hupensis у 5 females/culture war: Pots each fovmosana hupensis nosophora quadvasi TABLE 17. Purpose of shelving H. betw. shelves Shelving space required to house the Medium Clay Pot and Petri Dish Cultures needed to yield 500 young of each subspecies of Oncomelania hupensis per month Holding trays with 4 MCP per tray Holding Petri dish cultures The suggested number of snails and cultures is greater than the minimum requirement because (1) it is better to overshoot in production a little; (2) more cultures help to compensate for the sporadic production that occurs in individual vivaria; (3) the number of cultures was adjusted to multiples of 4 so that each plastic tray, which ac- commodates 4 medium clay pots, would contain only 1 subspecies. With 666 young produced monthly by each of the 4 subspecies, facilities must be available to house 2664 young monthly for at least 2 months. As each Petri dish takes 2 young, space is needed to house 1332 Petri dishes in the lst month, and since the snails take 8 weeks (2 months) to mature, it is necessary to 52 Petri dishes Capacity No. Units Suggested per shelf needed arrangments Optional 9 units each with 6 shelves in a tier (see Fig. 12) have an additional 1332 Petri dishes and space to house the production of the second month; thereafter sets of dishes are rotated. Table 17 shows the shelf space needed to house both plastic trays with the medium clay pots and Petri dishes. Although the plastic tray housing the 4 clay pots is but 3” high, the recommended distance between shelves is 10” to allow for easy inspection of the cultures, for adding water and knocking down the snails daily, and for removing 1 clay pot with- out having to remove the whole tray. The length of shelf holding the Petri dishes (52”) is determined by the length of the light fixture holding the 47” long fluorescent tube. The width used (16”) is recommended for ease of maintenance 362 VAN DER SCHALIE AND DAVIS as well as for providing adequate il- lumination for all the dishes from the Single, centrally placed tube 8” above the dishes on each shelf. We have found it convenient to use 6-shelf units for rapid maintenance and efficient use of space. To handle this level of production, 2 full time assistants are needed. 7. Necessity for Routine Maintenance Daily routine maintenance is abso- lutely necessary for keeping snail cultures of Oncomelania productive and rates of mortality low. Aside from knocking snails from the sides of the vivaria to keep them from dying by desiccation, water levels tend to fluctuate markedly in cultures depending on atmospheric conditions. Mold can appear, spread rapidly, and within а week cause excessive mortality of adults and young. ACKNOWLEDGEMENTS Several technical assistants devoted many hours to the exacting routines of maintaining cultures of Oncomelania and collecting data used inthis paper. For this assistance we are indebted to Berton Roffman, Robert Wakefield, and Andrew and Gerald Bratton. We are especially indebted to those who made the study possible by providing snails from field and laboratory: LTC John W. Moose and Mr. James E. Williams of the U. S. Army’s 406th Medical Laboratory in Japan; Dr. Robert E. Kuntz, formerly of the Naval Medical Research Unit No. 2 in Taiwan; the Late Dr. Т.Р. Pesigan of the Department of Health in the Philip- pines, and Dr. H. Vogel of the “Institut für Tropenkrankheiten” in Hamburg, Germany. We acknowledge the sug- gestions and advice of Mr. James E. Williams of the 406th Medical Laboratory in establishing aquarium type vivaria in our laboratory. This program itself was sponsored by the Commission on Parasitic Diseases (an affiliate of the Armed Forces Epidemiological Board); their continued support has made these studies possible. LITERATURE CITED ABBOTT, В. T., 1946, The egg and breeding habits of Oncomelania quad- rasi Moellendorff, the schistosomiasis snail of the Philippines. Occ. Paps. Mollusks, Harvard Coll., 1 (6): 41-48. 1948, Handbook of medically important mollusks of the Orient and the Western Pacific. Bull. Mus. comp. Zool. Harvard, 100 (3): 245-328. BAUMAN, P. M., BENNETT, H. J. & INGALLS, J. W.,Jr., 1948, The mol- luscan intermediate host and schisto- somiasis japonica. II. Observations on the productivity and rate of emergence of cercariae of Schistosoma japonicum from the molluscan intermediate host, Oncomelania quadrasi. Amer. J. trop. Med., 28(4): 567-575. BURCH, J. B., 1964, Cytotaxonomy of the genus Oncomelania, intermediate hosts of Schistosoma japonicum. Bull. Amer. malacol. Union., No. 31: 28-29. CHI, L. W. & WAGNER, E. D., 1957, Studies on reproduction and growth of Oncomelania quadrasi, O. nosophora, and O. formosana, snail hosts of Schistosoma japonicum. Amer. J. trop. Med. & Hyg., 6(5): 949-960. DAVIS, G. M., 1967, The systematic relationship of Pomatiopsis lapidaria and Oncomelania hupensis formosana (Prosobranchia: Hydrobiidae). Mala- cologia, 6(1-2): 1-143. DAVIS, G. M., MOOSE, J. W. & WILLIAMS, J. E., 1965, Abnormal development in a hybrid Oncomelania (Gastropoda: Hydrobiidae). Ibid., 2(2): 209-217. DAZO, B. C. & MORENO, R. G., 1962, Studies on the food and feeding habits of Oncomelania quadrasi, the snail intermediate host of Schistosoma japonicum in the Philippines. Trans. Amer. micr. Soc., 81(4): 341-347. DEWITT, W. B., 1951, Twodevices use- ful for maintaining aquaterraria. Tur- CULTURING ONCOMELANIA 363 tox News, 29: 58-59. 1952, Pomatiopsis lapidaria, its occurrence in the Washington, D.C. area and its laboratory rearing in comparison to that of Oncomelania spp. J. Parasit., 38(4): 321-326. GREDLER, V., 1881, Zur Conchylien- Fauna von China. Jb. Dsch. Malakoz. Ges., 8: 110-132. HOSAKA, Y., IJIMA, T. & NASAYAMA, S., 1953, Biological studies on On- comelania nosophora. Jap. J. Parasit. 2: 95. ISHII, N. & TSUDA, E., 1951, Possibility on (sic) the spreading of Oncomelania nosophora, the intermediate snail host of Schistosoma japonicum, in other areas besides its own habitats. Yokohama Med. Bull., 2: 366-375. KAWAMOTO, S., 1952, On the photo- phobotaxis of Oncomelania nosophora. Med. € Biol., 23: 76-79. KOMIYA, Y., 1964, A survey on (sic) the habitat of Oncomelania snails, the intermediate host of Schistosoma japonicum in the Philippines and For- mosa. Jap. J. med. Sci. & Biol., 17(4): 195-210. KOMIYA, Y., KUNIKO, K. & KOYAMA, C., 1959, A simple breeding method for Oncomelania using a Petri dish. Jap. J. Parasit., 8(5): 721-724. LI, S. Y., 1953, Studies on schisto- somiasis japonica in Formosa. Ш. The bionomics of Oncomelania formo- sana, a molluscan intermediate host of of Schistosoma japonicum. Amer. J. Hyg., 57: 30-45. МАО, С. P., 1958, Researchon schisto- somiasis japonica in China. Amer. J. trop. Med. & Hyg., 7(1): 58-62. McMULLEN, D. B., 1947, The control of schistosomiasis japonica, 1. Obser- vations on the habits, ecology, and life cycle of Oncomelania quadrasi, the molluscan intermediate host of Schistosoma japonicum in the Philip- pine Islands. Amer. J. Hyg., 45(3): 259-273. 1949, A plate method of screening chemicals as molluscicides. J. Parasit., 35(Suppl.): 28. McMULLEN, D. B., KOMIYAMA, S. € ENDO-ITABASHI, T., 1951, Ob- servations on the habits, ecology, and life cycle of Oncomelania nosophora, the molluscan intermediate host of Schistosoma japonicum in Japan. Amer. J. Hyg., 54: 402-415. MEDICAL GENERAL LABORATORY (406), U. S. Army Medical Command, Japan. APO San Francisco, 96343. Reports for 1951-1964. MOOSE, J. W. & WILLIAMS, J. E., 1961-62, Medical General Laboratory (406), U. S. Army Medical Command, Japan, Professional Reports. 1965, personal communication. MOOSE, J. W., WILLIAMS, J. E. & FLESHMAN, P., 1962, Rice cerealas sustenance for rearing oncomelanid snails in the laboratory. J. Parasit., 48(1): 68. OTORI, Y., RITCHIE, L. 5; € HUNTER, а. W., Ш, 1956, The incubation period of the eggs of Oncomelania nosophora. Amer. J. trop. Med. & Hyg., 5: 559- 561. PESIGAN, T. P., HAIRSTON, N. G., JAUREGUI, J. J., GARCIA, E. G., SANTOS, A. T., SANTOS, B. C. & BESA, A. A., 1958, Studies on Schisto- soma japonicum infections in the Philippines. Bull. Wld Hlth Org., 18: 481-578. RITCHIE, L. S., 1955, The biology and control of the amphibious snails that serve as intermediate hosts for Schistosoma japonicum. Amer. J. trop. Med. & Hyg., 4(3): 426-441. RITCHIE, %. S., HUNTER, G. W., ME OTORI, Y., 1951, Observations on the laying and incubation of eggs of Oncomelania nosophora. J. Parasit., 37 (5/2): 16-17, abstract. SANDGROUND, J. H. & MOORE, D. V., 1955, Notes on the rearing of On- comelania spp. inthe laboratory. /bid., 41(1): 109-113. STUNKARD, H. W., 1946, Possible snail hosts of human schistosomiasis in the United States. Ibid., 32: 539-552. SUGIURA, S., 1933, Studies on biology of Oncomelania nosophora (Robson), 364 VAN DER SCHALIE AND DAVIS an intermediate host of Schistosoma japonicum. Mittingn. pathol. Inst. 4. mediz. Fakultät, Niigata, 31: 1-18. VAN DER SCHALIE, H. & DAVIS, G.M., 1965, Growthand stunting in Oncomel- ania (Gastropoda: Hydrobiidae). Malacologia, 3(1): 81-102. VAN DER SCHALIE, H. & GETZ, L.L., 1962, Distribution and natural history of the snail Pomatiopsis cincin- natiensis (Lea). Amer. Midl. Nat., 68(1): 203-231. 1963, Comparison of tempera- ture and moisture responses of the snail genera Pomatiopsis and On- comelania. Ecology, 44(1): 73-83. VOGEL, H., 1948, Uber eine Dauer- zucht von Oncomelania hupensis and Infektionsversuche mit Bilharzia japonica. Z. f. Parasitenk., 14: 70- 91. WARD, P. A., TRAVIS, D. € RUE, R.E., 1947, Methods of establishing and maintaining snails in the laboratory. Bull. natl. Inst. Hlth, 189: 70-80. WAGNER, E. D., 1954-55, Annual pro- gress report to the commission on parasitic diseases of the U. S. Armed Forces Epidemiological Board, Loma Linda University, Loma Linda, Calif., UNS Ay * WAGNER, Е. D. & MOORE, B., 1956, Effects of water level fluctuation on egg laying in Oncomelania nosophora and Oncomelania quadrasi. Amer. J. trop. Med. & Hyg., 5: 553-558. WAGNER, E. D. & WONG, L. W., 1956, Some factors influencing egg laying in Oncomelania nosophora and On- comelania quadrasi, intermediate hosts of Schistosoma japonicum. Ibid., 5(3): 544-561. WILLIAMS, J. E., 1952, personal com- munication. WINKLER, L. R. & WAGNER, E. D., 1959, Filter paper digestion by the crystalline style in Oncomelania. Trans. Amer. micr. Soc., 98(3): 262- 268. WONG, L. W. & WAGNER, E. D., 1954, A rapid method of sexing snails, On- comelania nosophora. Ibid., 73: 66-67. 1956, Some effects of ultra- violet radiation on Oncomelania noso- phora and Oncomelania quadrasi, snail intermediate hosts of Schistosoma japonicum. Ibid., 75(2): 204-210. *Permission to cite this reference in the text and bibliography was given by the authors. RESUMEN CULTIVO DE ONCOMELANIA (PROSOBRANCHIA: HYDROBIIDAE) PARA ESTUDIOS DE LA ESQUISTOSOMIASIS ORIENTAL H. van der Schalie & G. M. Davis Se probaron seis tipos de vivarios para determinar en cuales condiciones las cuatro subespecies de Oncomelania hupensis proliferan mejor en laboratorio. Se establecieron los procedimientos eficientes que aseguran las condiciones óptimas para: sobrevi- vencia de adultos, mayor producción de cria por cada hembra en unidad de tiempo, sobrevivencia de jovenes y desarrollo rápido de los mismos. Guiados por nuestra adquirida experiencia y extensiva consulta de la bibliografía, se formaron acua-terrarios en los siguientes recipientes: acuarios, jarros cilíndricos (de baterías), macetas de arcilla grandes y medianas, bandejas plásticas y cápsulas de Petri. Se obtuvo buen resultado de cultivo en dos distintos ambientes: (1) uno en el cual la mortalidad de adultos fue mínima y la produción de crias óptima: y (2) otro en el cual los jóvenes crecieron sin impedimento, rápidamente, y con baja mortalidad. Las condiciones en macetas medianas, conteniendo 5 hembras y 5 machos, resultaron superiores en cuanto a la sobrevivencia de adultos y producción de cria, debido al CULTURING ONCOMELANIA hecho que unas pocas hembras fueron muy productivas en un volúmen limitado, donde la proporción suelo-filtro-papel-agua parece ser óptima. Siendo los ambientes de pequeño volúmen, se pueden usar más unidades productivas, en lugar de los ambientes grandes que son mas dificiles de manejar y producen menos. El por- centaje de mortalidad de adultos fue de 2% durante el primer año, al término de dos años no alcanzó al 50% sino que fue menor. Para producir crias se recomienda que a los dos años de edad, las hembras de todas las subespecies, excepto O. h. quadrasi, sean reemplazadas por otras mas jovenes; las de la subespecie mencionada deben ser reemplazadas entre los 12 y 14 meses, proque mostraron un aumento en mortalidad en el segundo año de vida adulta. En otros vivarios, el porcentaje de mortalidad aumentó rápidamente, 12% o más por mes, indicando condiciones de cultivo desfavorables. Más pobres aún fueron los acuarios, las macetas grandes y los jarros cilíndricos. El acuario era muy incomodo de manejar por su mayor tamaño y los caracoles no se desarrollan bien. Las macetas grandes mostraron mortalidad excesiva de las crias. Y los jarros de baterías se caracterizaron por la gran mortalidad de adultos y la corrosión extreme de las conchillas. El mayor número de crias se produjo en las macetas medianas donde cada mes, dependiendo de la subespecie, la puesta por hembra era de 3-11, casi el doble de la de los otros cultivos. Aunque la producción era esporadica, todas las subespecies se reprodujeron cada mes del año. Como ya fue discutido por nosotros (1945) proporciones optimas de crecimiento fueron en cápsulas de Petri, conunoodos caracoles por cápsula de 2 a 2 1/2 anfractos cada uno. La fase del crecimiento logarítmico se completó en 5-9 semanas, mientras que desarrollo completo llevo de 8 a 13 semanas, segun las subespecies. Una mortalidad máxima del 20% ocurrió entre la 84 y 9% semana. Para cultivar Oncomelania, los siguientes factores son de importancia: el suelo, tanto como fuente de alimento y como sustrato para la deposición de los huevos, debe ser de textura fina, con alto contenido de calcio, y debe mantener una flora densa de diatomeas. Esta flora junto con agentes de descomposición, provee una adecuada fuente alimenticia; el único suplemento es papel de filtro, adimento aditivo ya clasico. El agua debe ser neutral o ligeramente alcalina (pH 70-76) y libre de clorina u otros gases toxicos como iones de cobre. La intensidad luminosa de una habitación común fue adecuada para la sobrevivencia de adultos y produción de jovenes: luz constante tiende a aumentar la mortalidad por la excesiva proliferación de algas. Condiciones óptimas de crecimiento para las crias es una intensidad luminosa de 130 a 160 bujías, en ciclo de 10-12 horas diarias. Productividad decrece y mortalidad crece con el aumento de densidad en caracoles en cultivo. Mantenimiento diario es necesario para condiciones óptimas. Factores bióticos más destructivos fueron los mohos, lombrices y gorgojos. Se ofrece un modelo para trabajo, con el número de macetas medianasm cápsulas de Petri, espacio, y labor necesaria para criar 500 caracoles por mes de cada una de las cuatro subespecies de O. hupensis. ABCTPAKT КУЛЬТИВИРОВАНИЕ УЛИТОК ONCOMELANIA (PROSOBRANCHIA: HYDROBIIDAE) ДЛЯ ИЗУЧЕНИЯ ВОСТОЧНОГО ШИСТОЗОМИАЗИСА Г. ВАН ДЕР ШАЛИ И ДЖ. М. ДЕВИС Для определения экологических условий, необходимых для наилучшего развития в лабораторных условиях 4 подвидов Oncomelania hupensis были испробованы 6 Типов различных вивариев для их содержания. Были приннты эффективные меры, чтобы создать оптимальные условия для выживаемости взрослых, 365 366 VAN DER SCHALIE AND DAVIS продукции молоди Ha 1 самку и в определенное время, выживаемости молоди и ее быстрого роста. На основании изучения обширной литературы и имеющегося опыта, в различных контейнерах были созданы различные акватеррарии: в аквариумах (в стеклянных цилиндрах или батарейных банках), в крупных и средней величины мелких глиняных горшках, на пластикатовых подносах и в чашках Петри. Успех культивирования Oncomelania зависел от установления 2 различных сред обитания: 1) где смертность взрослых была бы минимальной, а продуктивность молоди оптимальной, и 2) где молодь росла бы быстро, без нарушений и с малой смертностью. Условия в "среднем глиняном горшке; где содержались 5 самцов и 5 самок оказались наилучшими, как для выживания взрослых, так и для продукции молоди. Их хорошие качества объяснялись тем, что самки, живущие в малом количестье в ограниченном объеме имеют высокую продуктивность при оптимальном соотношении условий-почва-Фильтровальная бумага- вода. Благодаря малому объему места их обитания, можно более успешно употреблять малые ёмкости вместо крупных и громоздких менее продуктивных типов культур. Темп отмирания взрослых моллюсков втечение первого года составлял всего около 2%, не достигая 50% к концу двух лет, обычно же была значительно меньше. Для успешной продукции молоди рекомендуется, чтобы в возрасте 24 месяцев самки всех подвидов, кроме Oncomelania hupensis quadrasi заменялись более молодыми самками. Что касается указанного выше подвида, его самки должны заменяться в возрасте 12-14 месяцев, т.к. у них наблюдается заметное увеличение смертности на втором году жизни у взрослых особей. Наблюдаемое быстрое увеличение скорости утмирания до 12% и больше P месяц в других вивариях указывает на неблагоприятные условия содержания моллюсков. Самые обедненные культуры наблюдались в аквариумах, больших глиняных горшках и стеклянных цилиндрах (батарейных банках). Аквариумы оказались исключительно неудобными для ухода за культурами из-за их большого объёма и моллюски развивались в них плохо. Большие глиняные горшки давали чрезмерную смертность молоди. В стеклянных цилиндрах наблюдалась большая скорость утмирания взрослых и очень сильная эрозия раковин. Наибольший урожай молоди наблюдался в глиняных горшках средней величины, где, в зависимости от подвида, нарождалось 3-11 экз. молоди на 1 самку в месяц, в наиболее продуктивные месяца т.е. наблюдался вдвое больший темп их продукции, чем в других вивариях. Хотя продукция молоди была спорадической, все подвиды размножались каждый месяц года. В средних глиняных горшках отмирание молоди в исходных (материнских) культурах была 4-10%, обычно же составляла 5%. В других вивариях отмирание было гораздо выше; в крупных глиняных горшках, например, отмирание составляло 42-78%. Смертность наступала в первые 1-2 недели после отрождения. Как указывалось авторами (1965) оптимальная скорость роста наблюдалась в культурах на чашках Петри, с 1-2 моллюсками в каждой; прирост раковины при этом составлял 2-2.5 оборота. CULTURING ONCOMELANIA Начальная Фаза роста занимает 5-9 недель, в TO время, как весь период роста обнимает 8-13 недель, в зависимости от подвида. Максимальная смертность в 20% наблюдалась после 8- 9 недель роста. При культивировании Oncomelania важны следующие факторы: почва, как источник пищи и как субстрат для откладки яиц должна иметь тонкую структуру, высокое содержание кальция и высокую плотность поселения Флоры диатомей. Эта Флора, вместе с сопровождающими ее потребителями, составляет нео- бходимый источник пищи для моллюсков. Другим единственным классическим добавком к пище служит Фильтровальная бумага. Вода должна быть нейтральной или слабо щелочной (pH 7.0-7.6). Вода не должна содержать хлора или других токсических веществ, таких, как ионы меди. Степень (сила) комнатного освещения вполне достаточна для хорошей выживаемости взрослых и для продукции молоди; вместе с тем постоянная освещенность имеет тенденцию увеличивать скорость отмирания моллюоков втечение длительного периода времени (1.5-2 года), а также способствует смертности из-за чрезмерного развития водо- рослей. Оптимальная скорость роста молоди наблюдается при свете, силой 130-160 свечей, действующем 10-12 часов в день. Продуктивность моллюсков уменьшается, а темп отмирания возрастает по мере увеличения плотности поселений моллюсков. Для создания для них оптимальных условий необходим ежедневный уход. Биотическими факторами, наиболее разрушительными для культур моллюсков служат олигохеты и клещи. Была выработана типовая модель культуры моллюсков, созданная на целом ряде глиняных горшков средней величины и чашек Петри, определено пространство и количество труда, необходимого для'’поддержания культуры из 500 улиток в месяц для каждого из четырех подвидов Oncomelania hupensis. 367 vee NAAA ARS IA Reed “Misco EN a ae neha (Raa bee РЕ z M A Sr г y d sy Wis Race ty Bree 2 Ness - т No er SEAN O ea 9 vi ae > 27 EL er ¡cai e u» | paws 4 TOR #1 ea Re ena i ANS nape ers и ttre Ne N A À mn | ident Lions » us | u re 7 Tt PARI OM có Konad 3 Г ee, pred ee Men | EURE ALL CEE a CA “NOT ¡AE E I NZ Au. f 4 oie ee Py SNA tel ” OR Fade лы, A | A EE бра GR AN O ie” м B { \ | 4 A 4 ait № Bere: yee yh? on PSS as i i ЗИ } } > 00 if 324 Лим FT LAA } р Date | : ROK + LIME [1 1 A À ye ETA PIS y 1 45 s = Bi alt FAT hy 4 Sa i UT ERA м5 Г. ARRET" NI rus < 5.63 We Ad i i af ER Y р dy Ч а к, ежи {$ A iar A 4 Gi NIE Ad +. 1248 +4 + Va! fd = er = < wth sh # ян = Te a 4 и 1 e | à A A) : ase Sat er We af a у At E a ae x SA ALES Vy) PROG AA TO & si hate ИИ" 4 | i s pia nut up) fr dc > - 7 + i x M 3 a eee, PAU yo So Sk ЮУ ЭХ | J р x L Ñ > ” er Pr NE ит. Shae wee TE) E Na E NE NEN. 5 IHR В ve 7” Gris eee? «30859 & zn ‘ ve ¡RA SHARE Masia: rs, A CES . 3 nur ene ос»! F ож SEA a wi E . +1 Eu $ > ¿EAS # HER Jun AO. jui A MN >: : ы TAR Sl ART‘ | | APARTA A OA Y Wily. | iR | 4 ME NT ie gH UA MELIA MR | TN ae APM A 1 ' 4 As Moro | b Уже y | Pr 4 ls м HAN 7... RAS x | uit CALA A 7 DATE dr yee pu a ys iu “PR Y ER AN, ek UL wee! ent rd о gr autre ie =. | Aa Y partes У" tial Oe ee PTT 4 M 2 ива ANT SIREN AVS di: PY | Jun: xt, COOTER tait TEL ACL, О À sn Lo la LES MS anes hogy me : un o A ec. a oN «A : = A Ar 14, bre Kipa Teil аки st à Ed AMA KU ye lai AS Re eel BAT № № Ри WR AD UR ЧА re 2 x г 4 dre ey. LE Bia elt LOS od MALACOLOGIA, 1968, 6(3): 369-377 STUDIES ON SLUGS OF ARABLE GROUND I. SAMPLING METHODS Р. J. HUNTER Agricultural Research Council Unit of Insect Physiology School of Agriculture, University of Newcastle upon Tyne, U.K.1 ABSTRACT A number of methods used in the past for estimating slug populations were evaluated on a plot in Northern England. The species involved were Agriolimax reticulatus (Müller), Avion hortensis Férussac and Milax budapestensis (Hazay). Soil Sampling was the only method togive unbiased information onthe species composition and age distribution of populations. Slugs were extracted from soil by: (1) Soil washing, the most efficient method, (with 100% recovery of all stages except for the recently hatched juveniles, which had a recovery rate from 63-86%) that also showed the presence of eggs (91-100% recovery), and by (2) Flooding, a process less laborious and time consuming than soil washing, but less efficient in extracting slugs (88-92% recovery) and not capable of extracting eggs. A population estimate by the Marking-release-vecapture method was com- pared with a simultaneous estimate by soil sampling. The method was less convenient than soil sampling and tended to under-estimate population density. Duthoit’s baiting method of estimating populations was evaluatedin relation to the actual density of slugs in the area and the elements of weather that affect slug activity (temperature and humidity). Since population density was the only factor significantly influencing the extent of damage to baits, the latter can be taken as a measure of the relative population density. By the Night-searching method, the proportion of the surface dwelling and light coloured Agriolimax reticulatus in populations was overestimated. The mean weight of slugs from these samples was greater than that in comparable soil samples. Trapping under sack-shelters also gave a high estimate of the proportion of Agriolimax reticulatus. More slugs were trapped in damp cloudy weather than when it was sunny and dry. It was concluded that, for most purposes, the most efficient sampling method would be a small soil sample to establish species composition and age distri- bution, plus a large number of baits to measure the relative density of popu- lations. Slug populations are difficult to esti- mate since their distribution tends to be aggregated (Hunter, 1966) and many of the commonest species live underground. In previous studies, a number of attempts to estimate the relative density of these animals have been made using baits (Duthoit, 1961), traps (Getz, 1959), or by counting the number of active individuals at night (Bett, 1960). Their absolute density (i.e., the total number of slugs in a given area) has been estimated in- directly by a system of marking, re- lease and recapture (Johnson, 1964), lPresent address: National Agricultural Advisory Service, Brooklands Avenue, Cambridge, U.K. (369) 370 P. J. HUNTER TABLE 1. Recovery of slugs and eggs, previously added to 1 cubic foot of soil, by soil washing Species Wt. of over 12.5 mg % Agriolimax veticulatus 15 100 Arion hortensis 15 100 Milax budapestensis 5 100 Весоуегу Вес O и Bere Wt. of | we. of under 12.5 mg | 12.5 mg Recovery % *The operator did not know the number added until after the experiment. and directly by taking samples of soil and later extracting the slugs from these (South, 1964). In this study the latter direct method was used for 2 1/4 years while taking a series of samples for a study of the biology of Agriolimax reticulatus (Müller), Arion hortensis Ferussac, and Milax budapestensis (Hazay) on an arable plot in Northumber- land. The opportunity was taken to compare the results from these samples with further estimates from marking- release-recapture, baiting, trapping and night-searching. SOIL SAMPLING Sampling Technique The size of sampling unit used andthe depth to which sampling is taken will depend on the distribution and abundance of the slugs on the plot. In the present study, soil cores 4 inches indiameter by 1 foot deep (approx. 10 x 30 cm) yielded too few slugs to provide sufficient bio- logical material (a test sample of 20 of these cores produced only 5 slugs, 2 of which were damaged by the sampling tool). Units of 6” and 8” square were difficult to excavate to the required depth, so that the unit finally adopted was of 1 cubic foot of soil. There had been only shallow cultivations on the plot for some time so that slugs were rarely found below 1 foot deep. Extraction Technique Various unsuccessful attempts (South, 1964) have been made to extract slugs from undisturbed soil by using repel- lants. There is no record of attempts to expel slugs from soil т situ by electrical methods, but this technique does not seem entirely satisfactory even for those animals for which it has been used (Satchell, 1955). The only extraction methods attempted in the present study were therefore those applicable to soil samples brought into the laboratory prior to extraction. (i) Soil Washing (Salt & Hollick, 1944; Raw, 1951). Soil was broken down with a water jet on a bank of sieves (3 meshes, 10 meshes and 30 meshestothe inch). The sieves with their residues were then agitated in magnesium sulphate solution of at least 1.17 specific gravity. All organic material rose to the surface so that both slugs and eggs could be picked off. The recovery rate was tested by adding a number of slugs and eggs to slug-free samples of soil and later extracting them by soil washing. Of the older stages (over 12.5 mg in weight), 100% were recovered but considerably fewer of the newly hatched slugs (63- 86%; Table 1). Presumably the small slugs that were not recovered were either missed or destroyed by the force of the water jet. All slugs lost weight during the process, Agriolimax reti- culatus losing an average of 34% oftheir weight, Avion hortensis 33% and Milax budapestensis 27%. This reduction can probably be attributed to loss of slime during the severe washing and ex- SAMPLING OF SLUGS ON ARABLE GROUND 371 TABLE 2. Recovery of slugs present in 16 cubic feet of soil by flooding Nos. Species Extracted not extracted* Nos. Success % Nos. damaged** 78 49 49 Agriolimax reticulatus Avion hortensis Milax budapestensis * Found later by soil washing. 87.5 89. 1 **Slugs damaged during digging of samples and not included in the calculation of % success. TABLE 3. Comparison of population estimates for a 1350 sq. ft. plot by the marking-release- recapture and by the soil sampling methods Nos. marked and released Species Ben Nos. Agriolimax reticulatus Arion hortensis Milax budapestensis traction processes. All Milax budapes- tensis and most Agriolimax reticulatus eggs were recovered, but, being much less robust, only a few Arion hortensis eggs survived the soil washing process. (ii) Flooding. South (1964) was ableto extract slugs by the “cold water pro- cess”, in which soil is slowly immersed in water, thus forcing slugs inside to creep to the surface. This technique was modified for large scale arable Sampling by using plastic bowls of 15” diameter fitting into dustbins of the same circumference. The bowls had holes in the bottom so that soil placed in them could be flooded from below. The bins were filled with water up to the base level of the bowls. The water level was raised 1/2” every 12 hours so as to immerse the soil in 4-5 days. Slugs were forced to move upwards as the water level rose and could be easily picked off from the soil surface or the lid of Marking-Release-Recapture recovered |recaptured unmarked Soil Sampling Nos. Population estimate Population estimate marked 5,364 + 2,611 6,683 + 4,374 1,800 +1,807 10,958 + 3,645 2,970 + 3,358 5,693 + 1,836 the bin. Sixteen sampling units of 1 cubic foot each were tested for recovery rate by this method. After extraction by flooding,the soil was washed (Table 2). A high recovery rate, 92% for Agrio- limax veticulatus, 88% for Arion hor- tensis and 89% for Milax budapestensis, was achieved for all 3 species. MARKING-RELEASE-RECAPTURE This method was tested in November 1964, and results were compared with those from a soil sample taken at the same time. A collection of 300 Agrio-, limax reticulatus, 300 Arion hortensis - and 180 Milax budapestensis was made from sites adjacent to the area to be sampled. In this type of experiment it is usual to capture animals from the population to be estimated, butthis would have disturbed the other observations being made (Hunter, 1968a). The slugs : 372 Data for the evaluation of the bait- ing method:* numbers of wheat TABLE 4. grains damaged weekly in relation to slug density, temperature and humidity Temper. in °C Hours Grains humidity | damaged Rh tp D © Nr OS 13. 40 12. 50 42 33 158 14. 90 20 24 11.37 36 31 14. 14 29 9 15. 56 20 0 = 14.13 15 0 17.51 52 3 10. 23 20 1 217 16.55 36 16 14.53 81 27 13. 01 16 19 212 13.10 70 53 11.71 42 51 277 11.51 58 53 9.00 36 60 8. llo 8. 0. llo 2. 6% 4 1 0. 4. 4. *The partial regression coefficients (b!) are as follows: bl Y1.23 = 0.833 bl Y2.13 =-0. 213 bl Y3.12 =-0. 230 P. J: HUNTER were marked by feeding them on jelly containing neutral red dye (South, 1964). This dye stains the digestive gland a deep pink colour that can be seen through the skin. Radioactive tracers have been used to mark slugs (Johnson, 1964) but neutral red is more convenient to use and gives satisfactory results. The marked slugs were released at random in an area of 1,350 sq. ft. and the recapture samples taken were collected on the 4th, 5th and 6th nights after release. The population of each species was estimated (Table 3) using the formula suggested by Bailey (1951, 1952), which was adapted to take into account the unusual capture method, i.e. x =salun+ 2) m +1 where a = number captured, marked and re- leased u = number of unmarked slugs re- covered m = number of marked recaptured slugs X = number in the population before the addition of marked slugs A direct maximum likelihood estimate of the variance of x is given by au(u +m + 1) (m + 1)2 (m + 2) and the 95% confidence limits can be estimated as 1.96 \variance. The above population estimate was compared with one from a soil sample taken at the same time. Six units, each of 1 cu. ft. of soil were taken in stratified random fashion and the slugs were extracted by flooding. The number of slugs in each unit was multiplied by 1.1 to approximately correct for loss during extraction. The mean numbers/ cubic foot andtheir 95% confidence limits were then multiplied by 1,350 to give direct population estimates for the total area of release (Table 3). Comparison of the 2 types of estimate indicates that marking-release-recapture was fairly accurate for Agriolimax reticulatus but considerably underestimated the num- SAMPLING OF SLUGS ON ARABLE GROUND 373 TABLE 5. Comparison of the species composition of night-searching and soil samples, Novem- ber, 1964 T ] Night-searching Soil Sample Species х2* р Agriolimax veticulatus 27.84 AO (OTOL Arion hortensis 46 47.42 106.5 |< 0.001 Milax budapestensis 24 24,74 16.0 |< 0.001 = = —- 97 100. 00 *2 x 2 contingency tests of the proportion of each species in samples. bers of Arion hortensis and Milax buda- pestensis (presumably because of the low recapture rate of these species - see section on night-searching). Soil sampling was the quickest and most con- venient of the 2 methods. BAITING The amount of damage to various baits has been used as a measure of the density of slug populations (Duthoit, 1961). In the present study Duthoit’s method was tested in the field between April 1964 and March 1965. Each bait consisted of 10 wheat grains laid on a piece of terylene netting measuring 6” x 4” (15 x 10 cm) and covered with a double thickness of sack- ing to prevent birds from eating the grains, the whole being fastened to the ground with wire clamps. Six of these baits were set up on an arable plot and readings were taken for a total of 16 weeks (not all consecutive). The total number of grains damaged was examined in relation to the following factors: (i) Direct estimates of the actual den- sity of slugs on the plot. It was only possible to take a soil sample every 4 weeks, i.e. there was not one for each week of the baiting experiment; therefore the baiting estimates were compared with estimates from the most recently taken soil sample. (ii) Mean weekly air temperature. (iii) Total number of hours per week when the relative humidity of the atmosphere exceeded 90%. The data (Table 4) were analysed by multiple regression. The partial re- gression coefficients of numbers of damaged grains on temperature and humidity were not significant (b’ =-0.21 and -0.23 respectively) but the co- efficient for regression on the actual slug density was significant (b’= 0.83). Thus in this case the population density was more important in determining the amount of damage to baits than factors affecting activity. NIGHT SEARCHING Barnes & Weil (1944a, b) investigated the activity of slugs by searching at night for a specified period (30 minutes) and counting the number of active slugs they observed. Bett (1960) used this method of collecting slugs for work on their life cycles. In the present study the method was tested by comparing the species composition and age distribution (measured in terms of weight) in the catches made by night searching and in soil samples of the same test area. The proportions of each of the 3 species in the 2 samples were subjected to 2 x 2 contingency tests. These tests were designed to show whether there was any significant difference between the ratio of the numbers of each species to the numbers of the other 2 species com- 374 P. J. HUNTER bined. Thus if Agriolimax reticulatus is represented by a, Avion hortensis by b and Milax budapestensis by c, tests were carried out as follows: Night- Soil Searching Sampling For Agriolimax a a reticulatus bte Dee For Arion b b hortensis AC ae For Milax с © budapestensis a +b a +b Significant differences were obtained for all 3 species (Table 5), there being a higher proportion of Agriolimax reti- culatus and a lower proportion of Arion hortensis and Milax budapestensis inthe night-searching sample than in the soil sample. Agriolimax reticulatus, being light coloured and surface dwelling, is clearly more likely than the other 2 species to be found by searching at night. The mean weight of the slugs caught at night was 189.7 mg for Agriolimax reticulatus, 150.5 mg for Arion hortensis and 401.1 mg for Milax budapestensis. Mean weights of slugs in the comparable soil sample (extracted by flooding) were 86.2 mg for Agriolimax reticulatus, 53.1 mg for Arion hortensis and 257.0 mgfor Milax budapestensis. Clearly, collecting at night is more likely to yield large slugs than small ones. TRAPPING Getz (1959) has used traps to collect TABLE 6. Number of slugs trapped under sacking under different weather conditions during a 10-day period in April 1963 Species Agriolimax veticulatus Arion hortensis Milax budapestensis 113 information on the biology and ecology of slugs. South (1964) has shown that this method is not reliable for col- lecting ecological data for slugs on grassland. For this reason a test was made of the efficiency of this method in producing data for slugs on arable land. Six sacks were laid on the ground at fixed points adjacent to the routine sampling plot, and slugs were collected from beneath them on 4 occasions during a 10-day period in April,1963. On 2 of these occasions the weather was sunny and dry, and on the other 2 it was cloudy and damp (Table 6). More slugs were trapped when it was cloudy and damp. Analysis of variance showed that the difference was significant (P= <0.001). The species composition in trap catches was compared with that in the routine April soil sample by contingency tests (Table 7). The proportion of Agriolimax reticulatus to other species in the trap catches was significantly greater than in the soil sample. The proportion of both Arion hortensis and of Milax budapestensis to the other 2 TABLE 7. Comparison of the species composition of trap samples and soil samples, April, 1963 Traps Species Agriolimax reticulatus Avion hortensis Milax budapestensis Totals : 328 61 135 25 68 12 31 8 «4 3 Soil Sample 1 17.3 59.7 49,4 24, 0 33413 23.1 4 40 27 1 *2 x 2 contingency tests of the proportion of each species in samples. SAMPLING OF SLUGS ON ARABLE GROUND 375 Species in the trap catches was signifi- cantly lower than in the soil sample. These differences were probably due to the fact that the surface-dwelling Agriolimax reticulatus was more likely to seek shelter under traps than the 2 subterranean species. DISCUSSION AND CONCLUSIONS The above data suggest that soil samp- ling is the only method giving accurate estimates of the total numbers, the species-composition and the age distri- bution of slugs in a given area. How- ever, itisatime consuming and laborious technique and is probably not suitable for workers who need to make frequent checks on the density of many popu- lations. Marking-release-recapture is also a time consuming technique without being as efficient as soil sampling. Night-searching and trapping have the disadvantage of yielding slugs that are not representative of the total popu- lation and also give no data during very dry or frosty weather when slugs are not active (Barnes € Weil, 1944a, b; Getz, 1959). While baiting also fails to give information on the structure of the population, it has an advantage over the latter 2 methods in that eachreading covers several days, i.e. evenin weather that does not favour slug activity some record is more likely to be produced. In spite of the dependence of slug activity on temperature and humidity (Hunter, 1968b), baiting gives a good indication of the relative population density. It would therefore appear that, for most purposes, greatest efficiency could be attained by (a) a small soil sample, with extraction of slugs by flooding, to give the species composition and age distribution of the population, and (b) a large number of baits to give the relative density of the population. ACKNOWLEDGEMENTS I am very grateful to Professor A. Milne of the School of Agriculture, University of Newcastle upon Tyne for his advice and encouragement throughout the study, and to Dr. A. South of Sir John Cass College, University of London for his advice at all stages of the work. I would like to thank the British Potato Marketing Board for a post-graduate studentship to finance this investigation. REFERENCES BAILEY, N. T. J., 1951, On estimating the size of mobile populations from recapture data. J. anim. Ecol., 38: 293-306. 1952, Improvements in the interpretation of recapture data. Bio- metrika, 21: 120-127. BARNES, Н. Е. & WEIL, J. W., 1944a, Slugs in gardens: their numbers, activities and distribution. Part 1. J. anim. Ecol., 13: 140-175. 1944b, Slugs in gardens: their numbers, activities and distribution. Part 2. /bid., 14: 71-105. BETT, J. A., 1960, The breeding seasons of slugs in gardens. Proc. zool. Soc. Lond., 135: 559-568. DUTHOIT, C. M., 1961, Assessing the activity of the field slug in cereals. Plant Pathology, 10: 165. GETZ, L. L., 1959, Notes onthe ecology of slugs: Arion circumscriptus, Dero- ceras reticulatus and D. leave. Amer. Midl. Natur., 61: 485-498. HUNTER, P. J., 1966, The distribution and abundance of slugs on an arable plot in Northumberland. J. anim. Ecol., 35: 543-557. 1968a, Studies on slugs of arable ground. U. Life cycles. Mala- cologia, 6(3): 379-389. 1968b, Studies on slugs of arable ground. Ш. Feeding habits. Malacologia, 6(3): 391-399. JOHNSON, C. G., 1964, Entomology Department ann. Rep. Rep. Rothamst. exp. Sta. U. K. for 1963: 146-160. RAW, F., 1951, The ecology of the garden chafer, Phyllopertha horticola L. with preliminary observations on control measures. Bull. ent. Res., 376 P. J. HUNTER 42: 605. SALT, G. & HOLLICK, F. S. J., 1944, SATCHELL, J. E., 1955, An electrical Studies on wireworm populations. I. А method of sampling earthworm popu- census of wireworms in pasture. Ann. lations. In: Kevan, D. К. McE. app. Biol., 31: 52. Soil Zoology. Butterworth’sSci. Publ., SOUTH, A., 1964, Estimation of slug London: 356-365. populations. /bid., 53: 251-259. RESUMEN ESTUDIOS SOBRE “BABOSAS” DEL SUELO ARABLE I. METODOS PARA SACAR MUESTRAS P. J. Hunter Se probaron varios métodos conocidos para estimar las poblaciones de “babosas”, en un solar del norte de Inglaterra. Las especies observadas fueron Agriolimax reticulatus (Muller), Arion hortensis (Fer.) y Milax budapestensis (Hazay). El único método que dió información imparcial sobre las especies y distribución de las poblaciones segun la edad, fue el de las Muestras de Suelo. Las babosas fueron extraídas del suelo por: (1) lavado del suelo, el método más eficiente (practicamente se recobraron el 100% de ejemplares, excepto de los juveniles de reciente eclosión que fue sólo en un 63-86%), el cual también indicó la presencia de huevos (91-100%), y por (2) inundación, proceso que consume menostiempo y trabajo, pero que es menos eficiente (recobrandose sólo 88-92% de ejemplares) y que no permite la extracción de huevos. El computo de población por marcado-suelta-recaptura, se comparó con el tomado simultaneamente de muestras de suelo; el primero fue menos conveniente y tendió a reducir el cálculo de densidad de población. El método de Duthoit, usando un cebo para captura, se estimó en relación a la densidad de babosas en el area y su influencia en la cantidad de cebo utilizado, asi como los elementos climáticos que afectan las actividades. En otro método, el de búsqueda nocturna, la proporción de habitantes de superficie y los Agriolimax reticulatus, de color claro fue sobrestimada, y el peso medio de las babosas en tales muestras resultó mayor al compararse con el calculado por muestras de suelo. También el uso de Trampas dio un cálculo muy alto para Agrio- limax, y mayor número de individuos fueron cazados asi en tiempo nublado y húmedo que en dias claros y secos. Puede concluirse que el método más eficiente sería una pequeña muestra de suelo, más una serie de capturas por cebo, para medir la densidad de la población y las edades. ABCTPAKT ИЗУЧЕНИЕ СЛИЗНЕЙ HA ПАХОТНЫХ ЗЕМЛЯХ I. МЕТОД СБОРА ПРОБ П. ДЖ. ХАНТЕР Сбор проб. Ряд методов, употреблявшихся в прошлом для определения популяций слизней были опробованы и оценены на участке земли в районе Новой Англии. Были исследованы следующие виды: Agriolimax reticulatus (Muller), Arion hortensis SAMPLING OF SLUGS ON ARABLE GROUND 377 Ferussac и Milax budapestensis (Hazay). Пробы почвы служили единственным методом получения объективной информации о видовом составе и распределении популяций слизней в почве. Моллюски извлекались из почвы различными способами: путем ее промывки, что является наиболее эффективным методом, который практически охватывает 100% особей, исключая только что народившуюся молодь; последняя охватывалась на 63-86%. Яйца учитываются этим методом на 91-100%. Метод отмучивания почв - менее трудоемкий и длительный, чем промывка, но менее эффективный в смысле извлечения слизней (88-92% охвата), не дающий яиц совсем. Популяция, оцененная методом сбора, пометок и повторного отлова, сравнивалась с TOM, которая была одновременно оценена почвенным методом. Этот метод оказался менее удобным, чем почвенный метод и видимо дает заниженные результаты оценки плотности популяции. Метод ириманок Дютуа для оценки популяций слизней сравнивался по полученным результатам с существующей в почве плотностью их поселений, с учетом условий погоды (температуры и влажности), влияющих на жизнедеятельность моллюсков. Поскольку плотность их популяций была единственным фактором, значительно влияющим на степень повреждения приманок, последние могли служить для измерения относительной плотности популяций слизней в почве. При помощи метода ночных поисков относительное количество в популяции обитающего на поверхности грунта и светло- окрашенного Agriolimax reticulatus была переоценена. Средний вес слизней в этих пробах был больше, чем в сравнимых C ними пробах почвы. Метод ловушек также дал завышенное относительное количество Agriolimax reticulatus. В сырую облачную погоду слизни ловились в большем количестве, чем в сухую и солнечную. Из всего указанного выше было сделано заключение, что для многих целей наиболее эффективным методом ‘учета слизней служит метод небольших почвенных проб, дающий возможность установить их видовой и возрастной состав и распределение. Чтобы оценить относительную плотность популяций необходимо параллельно употреблять большое количество приманок. dono Eta PLA MB TAO a DULITMAL у НАУ» as rar! 14 120 FU D Nam ve ne sue + \ fe [Pari 4 и ” AN у “| ide оон ОА ME Se weil CT MODS VEN TA REN, Ponte, Kits NG per Du Are ur RTE | ДА us у nee фи. UN sly eet aes FFE ATOM AU on "i II e OP LA" ra LD = 7 May UN Eu, < “4 } М N LE MS \ÿ ITA ie ur RN we Y | "be ANA $ Brad "5 PATENT, Y A + 2 M } + Ml ITA Lov a Southey 5 ыы ’ Y y fé ti PAL. RUN EN | ess 4 | y a 4 di à y 2 N и + (ur En 4 ’ AÑO aye ту ao UNAS ы 7 | ‘ Ra TN Alea ee FPL aa o ¿EA ! NUE NN RAGE < А er w 10 TN At AL SU ' de р if | wit Ati “E y Мины O И ий Ut: f ki ' я ‘ar ты - N vl у in: PAL A ri Ph PhO bei eo LC Lise DA ds en a LIL D A г, TP it ATOM : OP BND we Hebi | a PES I To re el awe a ди ION RE AGS Fr Audi o И \ EE A Ed С: N de \ ” RP x mr wh, (CU Le TL Pret ОТО Bi NT SN ENS Tits Тк N 4 У и: : wees 6 TITS FRAGA E A ee WEA Pe hak PL tic fai ne и je ch : dé ON ET" A thant т, ha Aha APT IE. NIBNEN i ye а net bil! ey ET eed AOA, СТО ' y \ р evr ATE HORI TS hey IS A TNT + # yf de oe 4 у ~ < SLL TE pent te À 4 AA] EN, à и A МОУ ORO y y wry | "st y cn “ + nel pout ia ha ES i a. MES Po ph À no" Ad sr à Lit ate | ee | | his / un ae, ae DE } OY ARS DANI сан Raa IL TR e aia OUT NES CE y у a i er à Ya ‘Наум: ЩИХ. Pore! ro RE ; у if BE ALL ER! Win i wah) 2 N YE | Аа BR DNS CAO O A SLT AU EA oye 2 ку а o! ES DOLO Misi at: cyl) VE "HRROON u RC NR MINA RN re RT Е, О eee Se AA SO Mary давай Фо od DNA NA y № ‘ FUN Sie ART T TNA LAINE TAA OT A Vee RATNOOH TS: ¿TIE wey а ke a His? ANAIS on к deren tal if NR M a MN Char ide wel sd ЧЕ AM. Ng | 2 ids } ‘ 1 № IN AO у р . M mr р ь (ee PAT \ To : Оса A o m Fu nen Ri Su | ( | y | - A UA N ARANA vw) К Oe os # MATA a Ma N MR e Winans | cpa UE уд ‘ow i 4 i 5 AAA 00190 р oh AU Я MUA: AR romani OA A aunt | ними АЛИ Pi о u aD ‚ IRL A UN № „о AAO eee A 1 ae LOHR у ИЕ, MALACOLOGIA, 1968, 6(3): 379-389 STUDIES ON SLUGS OF ARABLE GROUND Il, LIFE CYCLES P. J. Hunter Agricultural Research Council Unit of Insect Physiology School of Agriculture University of Newcastle upon Tyne, U. K.1 ABSTRACT The life-cycles of Agriolimax reticulatus (Müller), Arion hortensis Férussac and Milax budapestensis (Hazay) were investigated in a study based on routine sampling from an arable plot in Northern England. Agriolimax veticulatus had 2 generations a year, a spring generation hatching about May and an autumn generation hatching about late September. The latter generation took longer (7 months) to complete its life-cycle than the former (5 months). Most Avion hortensis had an annual life-cycle. Theyhatched about July, grew during the following 11 months, matured and laid eggs when about a year old. A few, however, hatched later and were not ready to lay eggs until after their 2nd winter, thus taking almost 2 years to complete their cycle. The develop- mental rate of all stages was found to depend on environmental conditions. In- dividuals laid an average of 64.5 eggs and died shortly after breeding. Most Milax budapestensis had a biennial life-cycle. They hatched between May and August and matured during their 2nd autumn and winter. A few, how- ever, hatched as early as April and laid eggs during their lst winter. Again the rate of development depended on the environment. Slugs collected from the field shortly before breeding laid an average of 23.5 eggs/individual and died soon afterwards. average of 32.5 eggs/individual. A sampling study of the slugs of an arable plot at Close House, Northumber - land, U. K. (Grid reference NZ127658), was conducted to determine their repro- ductive capacities, breeding seasons and generation intervals. The dominant species of slugs present were Agri0- limax reticulatus (Müller), Arion hor- tensis Ferussac and Milax budapes- tensis (Hazay). The sampling plot measured 15 x 25 yards (14 x 23 m) and was on a loam soil in a south-facing walled garden of about 2 1/2 acres. The plot remained uncultivated except for ploughing after the first year of sampling. Samples were taken every 4 weeks between January 1963 and March 1965. Slugs kept in field cultures from their young stages laid an Most sampling methods such as baiting, trapping and searching at night do not give representative data on species and age distribution of slug populations (Hunter, 1968) and since the latter were required in this study, only soil sampling was uSed to obtain slugs, and their eggs, for examination. Sampling units were of 1 cu. ft. of soil; 9 of these were taken in each sample during 1963 and 12 were taken thereafter. Between January 1963 and February 1964, slugs were extracted from samples by soil-washing and from February 1964 onwards, by flooding. Eggs were classified into 3 groups: no development, partly developed (when the rudimentary parts of the body were 1 Present address: National Agricultural Advisory Service, Brooklands Avenue, Cambridge, U. K. (379) 380 TABLE 1. P. J. HUNTER The total numbers of eggs, recently hatched juveniles and adult slugs from routine soil samples, 1,2 from an uncultivated arable plot in Northern England. Species and stage Agriolimax reticulatus Eggs - no development partly developed® 14 21110892 well developed? 47 69 0 Total 72 | 108; 105 Juveniles4 0 0 1 Adults® мии Arion hortensis Eggs - no development 0 0 0 partly developed 0 0 0 well developed 0 0 0 Total 0 0 0 Juveniles 4 AL 1 Adults 10 |10 2 Milax budapestensis Eggs - no development 12 0 2 partly developed 3 7 0 well developed 0 0 6 Total 15 7 8 Juveniles 0 0 Adults 4 5 62 |100 | 64 | 17 | 46 0 0 0 1 5 4 12 0.112560 0 0 8 | 21 0 0 0 3 3 0 0 2 1 6 0 0 | 13 | 22 9 4 0 0 3 | 13 7 lial IA ALO + 7 14 24 lSamples were of 9 cubic ft. of soil during 1963 and 12 cubic ft. during 1964/65. 2Samples were taken every 4 weeks, i.e. there were 13 per year; however sample xiii of 1963 was omitted. 3«Partly developed” eggs refer to stages I-III as described by Carrick (1938) “Well developed” eggs to stages IV-VI. 4 Juveniles” here refer toAgriolimax reticulatus and Arion hortensis of under 12.5 mg and Milax budapestensis of under 25 mg. 5“Adults” refers to mature slugs with sperm and ova in hermaphrodite duct (Bett, 1960). becoming differentiated: stages I-III of Carrick, 1938) and well developed (stages IV - VI of Carrick). Any young Agrio- limax reticulatus or Arion hortensis of under 12.5 mg and any Milax budapes- tensis of under 25 mg was regarded as having hatched “recently”. The weights of slugs extracted by soil-washing were adjusted to take into account the loss dur- ing that process. Slugs were considered “mature” if the hermaphrodite duct con- tained sperm or eggs (Bett, 1960). A number of experimental studies on Arion hortensis and Milax budapestensis are presented together with the data from soil samples. Such studies were not made for Agriolimax reticulatus since an investigation of this species had been undertaken by South (1964, 1965). SLUG LIFE CYCLES Table 1 (continued) 1964 LIFE CYCLES Agriolimax reticulatus Eggs were found in samples at all times of the year (Table1). Numbers fluctuated widely: very few eggs were found in July (sample vii of Tables and Figures) following a low density of adults in June and a very large number were found in November (sample xii, following a high density of adults in October). There were distinct peaks in numbers of recently hatched slugs (< 12.5 mg) in May and late September (samples v and x) of 1964 but no such peaks were apparent in 1963, probably because the soil- washing technique was not ex- tracting all slugs of the very young stages (Hunter, 1968). Smaller numbers of newly-hatched slugs continued to be found during the summer of 1964, but very few during winter. There were also peaks in the numbers of mature slugs around May and October of each year, while very few were found during June and February. The absence of mature slugs shortly after the breeding season suggests that, as in many other Mollusca (Comfort, 1957), the mature individuals lay their eggs and die within a short period of time. Arion hortensis Eggs Eggs were recovered from the routine samples between June and October (Table 1) but, since the eggs of this species are particularly fragile (Hunter, 1968), some may have been present at other times and destroyed during extraction. 381 382 Р. J. HUNTER Mean Temperature °C Mean Temperature °C МАМ Ap ae WN ПО 1964 1965 FIG. 1. Development rate of eggs at various times of the year. A. Arion hortensis; В. Milax budapestensis. Solid points: batch of eggs laid. End points of lines: first eggs in batch hatched. Open circles: mean monthly air tempera- ture. In order to ascertain the developmental rate of these eggs under field conditions, newly-laid batches were kept on damp filter paper in plastic pots out of doors. Whereas development took over 3 months during the winter, eggs laid during the summer hatched within a month (Fig. 1A). At constant temperatures in the laboratory, eggs required a minimum of 2 weeks at 20° C, 3 weeks at 15°C, 4 1/2 weeks at 10°C and 14 weeks at 5° C to hatch. Juveniles There was a peak in numbers of recently hatched individuals (<12.5 mg) around sample ix (early September) of both 1963 and 1964 (Table 1). However, the rise to this peak began later in 1963 (sample viii) than in 1964 (sample vii), probably because the protracted winter of 1962/63 delayed development. Small numbers of juveniles continued to be found during the autumn, winter and Spring, suggesting that some breeding occurs throughout the year. The growth rate of these slugs was determined from 2 groups of juveniles confined in the field as follows: a) In August 1963, 25 young slugs (mean weight 14 mg) were collected from the field and confined in cultures out of doors. Five plastic pots, 5” indiameter by 2 1/2” deep (approx. 13 x 6 1/2 cm), were filled to a depth of 2 ” with sifted soil. Gauze tops and bottoms prevented the slugs from escaping while allowing relatively free drainage. The pots were sunk into the soil and 5 slugs together with food (wheat grains, green and rotting vegetation) were added to each. Sacks were placed over the pots to prevent excessive evaporation and protect them from frost. During the winter of 1964/ 65 a layer of straw was added asfurther protection from frost. The slugs were weighed every 4 weeks (Fig. 2). They grew very quickly during their first autumn (mean weight 50mg in Novem- ber). The growth rate decreased during the winter (mean weight 60 mg in March) but increased again during the following spring (mean weight 170 mg in May). The first eggs were laid in June 1964 and, as in Agriolimax reticulatus, т- dividuals died shortly after breeding. The mean weight of the group began to fluctuate at that time because a sharp decrease in the weight of individuals occurred just before death. Some of the slugs continued laying until November and a further laying season began in the following March when they were SLUG LIFE CYCLES 100 Log Mean Weight of Slugs (mg) 10 в a хи Tee iii 1963 1964 LAR FIG. 2. Rate of growth and egg-laying of Avion hortensis in plastic pots. iva vi “Vile vil vi ix a xix Rei 383 200 s33q 1 1 | 1 | | | | | 1 | | | ! | I 1 U ! | 1 Г р I р 1 iii iv v м vii viii 1965 Mean growth rate is represented by curve, total eggs per month for all individuals by columns. TABLE 2. Mean weights, in mg, of batches of 10 slugs kept in terylene bags Arion Milax Dei hortensis budapestensis 1964- 1965 Batch | Batch about 18 months old. On the average, the slugs laid 64.5 eggs per individual. b) In August 1964, 20 young slugs (mean weight 7.1 mg) were collected from the field and kept in terylene net bags (10 slugs per bag) 4” in diameter by 12” deep (10 x 30 cm). The bags were sunk into the ground to a depth of 9” and filled with slug-free soil. Grow- ing vegetation and wheat grains were TABLE 3. Copulating pairs of slugs observed (January 1963 - December 1964) hortensis |budapestensis January February March April May June July August September October November December Hm O1 H O1 O © © © © À M vA N Nee Suen чЕ (NN ©. © Totals à added as food and the slugs were weighed after 3 and 6 months. On the average, these slugs grew slightly faster (approx. 120 mg by February - Table 2) than 384 P. J. HUNTER those in plastic pots (approx. 50 mg by February). This may have been due to differences in weather conditions in 1963 and 1964. Adults The abundance of mature individuals in the routine samples is shown in Table 1. Both in 1963 and 1964 there was a decrease in numbers during and after the peak egg-laying period of late June to August as slugs died after laying their eggs. Thereafter numbers rose again, suggesting that some slugs matured later and did not lay eggs until the following spring (as in the cultures above). In the field, copulating pairs were found in April, May, October, November and December (Table 3), pro- viding further evidence that some breed- ing occurred over most of the year. Copulation was not observed to last longer than 1 hour (in contrast to Milax budapestensis, which was seen in copula for over 24 hours). Three pairs of copulating Arion hortensis were kept in cultures until they laid their first eggs. The time between copulation and egg-laying varied considerably (3-13 weeks) probably because not all of the pairs were taken during their first copulation. Milax budapestensis Eggs Mainly newly-laid eggs were found in samples between December and January (Table 1) but by May many were well developed. The sharp fall in total numbers by June suggests that many hatched at that time. _ The presence of a few partly-developed eggs in winter indicates that some had been laid early enough to undergo a little development before the temperature fell. The de- velopment rate of eggs in the field was tested by placing newly-laid eggs in plastic pots out of doors, at intervals, between December 1964 and April 1965. The eggs laid at the beginning of winter took longer to develop (Fig. 1B), so that all of the eggs hatched in the following May and June. Eggs at constant tempera- tures hatched in 3 weeks at 209 С, 4 weeks at 15° C, 10 weeks at 10° C and 18 weeks at 7.59 C. Very little de- velopment occurred at 5° C and field soil temperatures are lower than that level from late November until March. All eggs laid during winter therefore tend to begin development together and most hatching is confined to a peak in spring. Juveniles Recently hatched juveniles (< 25 mg) were present in the routine samples between April and September. The peak in numbers was later in 1963 (July/ August) than in 1964 (June/July), probably due to the long winter of 1962/ 1963 (Table 1). The growth rate of these slugs was determined from various groups of con- fined slugs: a) Three groups (A, B and C) of young Milax budapestensis were collectedfrom the field and kept in plastic pot cultures out of doors (as Arion hortensis). They were weighed every 4 weeks and the logarithms of the mean weights of these groups were plotted against time (Fig. 3). Group A. In August 1963, 20 young slugs (mean weight 33 mg) were col- lected. By November these had reached an average weight of 138 mg. A very severe frost during late December killed all but one, which by mid-July weighed 396 mg. Group В. Between January and March 1964, 20 slugs, of approximately the same weight as Group A in December, were confined in cultures which were covered with additional sacks and straw to give further protection from frost. These slugs grew steadily through the spring and summer, and by November 1964 had an average weight of 793 mg. They began to lay eggs in December and lost weight, falling to a mean weight of 652 mg by early March 1965. Eggs continued to be found until July and a few in- dividuals (in poor condition) were still Log Mean Weight of Slugs (mg) SLUG LIFE CYCLES ‘ ЕК ЗН LS АМ 1963 1964 м м маме SEX xl 385 ' ' -0-- 9 „oe“ TOs. 0--0--0--0-_ KG UT SU A NAL NA VITE 1965 FIG. 3. Rate of growth and egg-laying of 3 groups of Milax budapestensis in plastic pots. Mean growth rates are represented by curves, total eggs per month for group B (20 slugs) by columns. alive in August. An average of 32.5 eggs per individual were laid. Group C. Five newly-hatched slugs, weighing 5-6 mg, were collected in May 1964. They reached a weight of 13-20 mg by June, but lost weight during their next month, probably due to a fall in the moisture content of the substrate in the culture. They then grew to 36-50 mg by mid-August, when 20 more slugs were collected. Five of these died during the next month, the remainder weighing an average of 128 mg in November and 163 mg in early March 1965. These were dissected in July and, although some weighed over 600 mg, none were mature. b) Two batches, each of 10 slugs, were collected from the field in June 1964 and kept in terylene bags (as used for Arion hortensis). The slugs were weighed after 2, 5 and 8 months (Table 2) and were found to grow at a slightly lower rate than those in plastic pots (Fig. 3). Adults The abundance of mature slugs in the routine samples is shown in Table 1. None of these slugs were found in July or early August (samples vii and viii) although some of the immature in- dividuals in these samples (not listed in table) weighed over 700 mg. In September, October and November, the numbers of mature slugs (some of which were under 300 mg) built up rapidly and then fell again during the winter and following spring, i.e. they did not live long after laying eggs. While collecting slugs at regular in- tervals throughout the year (both at night and during the day) for laboratory experiments, numerous pairs of copu- lating Milax budapestensis were noted (Table 3). Most of these were seen between September and December with occasional pairs between January and April, while none were detected between 386 P. J. HUNTER A 8 = т DER KES 8 = ey = = > Foy Syl des ee ан = a OT +” CA June 29 July 27 FIG. 4. Weight frequency histograms of slugs from routine samples in 1964. Weight classes in mgs starting from base line as follows: Agriolimax reticulatus and Arion hortensis; 0-24, 25-49, 50-74, 75-99, 100-149, 150-199, 200-249, 250-299, 300-399, < 400. Milax budapes- tensis; 0-49, 50-99, 100-149, 150-199, 200-299, 300-399, 400-499, 500-599, 600-699, < 700. May and August. Copulation in the field was often observed to last longer than 12 hours and at 5% C, in the laboratory, 1 pair remained ¿n copula for over 36 hours. Six pairs were dissected im- mediately after copulating and the spermatheca of each slug contained a spermatophore. Three slugs contained the disintegrating remains of a second spermatophore indicating that they had copulated twice. Two pairs, taken from the field in copula were observed to copulate again, one pair a week and the other a month after the first pairing. Pairs of slugs, collected at various times of year, were kept infield cultures, and periods of up to 15 weeks occurred between copulation and egg-laying. Pairs taken in the spring did not take as long to lay their first batch of eggs (as little as 5 weeks) as those taken inthe autumn, probably because the spring copulating pairs were less likely to have been taken during their first copulation. A further estimate of the egg-laying capacity of Milax budapestensis was made from 10 groups of 4 mature slugs kept in field cultures from mid-October 1964 until they died. Green vegetation and wheat grains were provided as food and the culture boxes (5” x 5” x 2 1/2”; approx. 13 x13x61/2cm) were covered with sacks to protect them from frost. The soil in the boxes was searched each fortnight for eggs and these were counted and removed. On the average these mature slugs laid 23.7 eggs/individual - considerably fewer than those in the growth rate cultures (Juveniles, Group B). Althoughthe slugs for this observation were collected be- fore the field egg-laying season and no eggs were laid in the boxes until 6 weeks after they were confined, it is possible that they had deposited some eggs be- fore being caught. However, the lower egg-laying rate in the boxes may also have been due to less favourable con- ditions of food or microclimate than in the bags. DISCUSSION AND CONCLUSIONS The generalised life cycles of the 3 Species were summarised by making weight-frequency histograms of data from samples; 1964 wastakenas typical and data for that year are presented in Fig. 4. These histograms, together with the above sections on separate SLUG LIFE CYCLES 387 LL pe ED A | + MR RE O Cd =] с) Aug. 24 Sept. 21 FIG. 4. continued. stages, suggested the following: Agriolimax reticulatus: There was some breeding activity throughout the year, but a distinct peak was apparent during spring and autumn. The slugs hatching in spring laid their eggs and died during the autumn and winter, and those hatching in autumn matured and died during the spring and summer of the following year. The spring generation developed more quickly (May to late September) than the autumn generation (late September to May). Arion hortensis: In the second half of the year there were 2 distinct gener- ations represented in the samples, (a) an older generation maturing during summer and autumn to die out in the first part of the following year, and (b) a new generation hatching from late June (sample vii) onwards and gradually growing in weight during the rest of the year. Most slugs of this gener- ation continued to grow in the spring of the following year to mature at about 1 year old, but a few which were late in hatching were not ready to mature before their second winter, i.e. they were almost 2 years old before breeding. Milax budapestensis: Inthe samples aíter May (sample v) 2 generations were : 1 de + Oct. 19 Nov. 16 Dec. 14 = 10 mature slugs === 10 immature slugs apparent, (a) an old generation repre- sented by a few slugs (over 700 mg by early August, sample viii) which matured in early autumn and died out in the winter, and (b) a new generation mostly hatching about May and growing in weight during the summer and autumn. Some of these matured during the winter and spring of their first year; but those that were unable to reach maturity be- fore their first spring did not seem to mature until the autumn of their second year (in spite of reaching weights of over 700 mg some months earlier). Since development of the eggs laid by the latter slugs was halted during the winter, it was 2 years before their cycle was completed. | In none of the 3 species was there a rigid life cycle followed by all in- dividuals. In Agriolimax reticulatus and Arion hortensis there seemed to be no intrinsic barrier to breeding at any time of year; the periods of highbreed- ing activity were probably regulated by the weather. However, there seemed to be some kind of barrier to the maturation of Milax budapestensis during the summer months: during June and July large individuals (over 700 mg) remained immature, but by October most Milax 388 P. J. HUNTER of over 400 mg were mature. Previous published work on the life cycles of slugs depended on night- searching (Barnes & Weil, 1944a, b; Bett, 1960) or trapping (Getz, 1959). These sampling methods probably gave unreliable life-cycle data since they tend to overestimate the proportion of large slugs in the population (Hunter, 1968). Thus the breeding seasons were possibly not recognised until some time after they occurred. However, it is still apparent that the life cycles of slugs vary considerably under different climatic conditions. Bett (1960) claimed that in Hertfordshire, central England, Agriolimax reticulatus has 2 generations a year, Arion hortensis 1 a year (with hatching in January-February) and Milax budapestensis 1 a year (with hatching during the Autumn). Getz (1959) re- ported that in Michigan, U.S.A., Dero- ceras reticulatum (=Agriolimax reticu- latus) was an annual species with only a few adults surviving the winter to lay eggs the next spring. Since the length of generation interval is the most important factor influencing the capacity of a population to increase in numbers, the density of a particular generation of slugs will depend mainly on the proportion of the previous gener - ation that matured during their first breeding season. Thus, the severe winter of 1962/63 impeded the development of all 3 species and the 1963 generations were small. A much greater proportion of slugs was able to continue develop- ment during the mild winter of 1963/64 and the subsequent generations were larger. Further research is required to es- tablish the reproductive controlling mechanisms of all 3 species and the effect on the life-cycles of the various climates in different regions of the world. ACKNOWLEDGEMENTS I am very grateful to Professor A. Milne, School of Agriculture, University of Newcastle upon Tyne, U.K. for en- couragement throughout this study and constructive criticism of the manuscript, and to Dr. A. South, Sir John Cass College, University of London, for helpful advice during all stages of the investi- gation. The work was supported by a grant from the British Potato Marketing Board. REFERENCES BARNES, H. F. & WEIL, J. W., 1944a, Slugs in gardens, their numbers, activities and distribution. Part 1. J. anim. Ecol., 13: 140-175. 1944b, Slugs in gardens: their numbers, activities and distribution. Part 2. Ibid., 14: 71-105. BETT, J. A., 1960, The breeding seasons of slugs in gardens. Proc. zool. Soc. Lond., 135: 559-568. CARRICK, R., 1938, The life history and development of Agriolimax agrestis L. The grey field slug. Trans. R. Soc. Edinb., 59: 563-597. COMFORT, A., 1957, life in Molluscs. Lond., 32: 219-241. GETZ, L. L., 1959, Notes onthe ecology of slugs: Arion circumscriptus, Dero- сета; reticulatum and D. leave. Amer. Midl. Natur., 61: 485-498. HUNTER, P. J., 1968, Studies on slugs of arable ground. I. Sampling methods, Malacologia, 6(3): 369-377. SOUTH, A., 1964, Estimation of slug populations. Ann. appl. Biol. 53: 251- 259. The duration of Proc. malac. Soc. 1965, Biology and ecology of Agriolimax reticulatus (Miiller) and other slugs; spatial distribution, J. anim. Ecol., 34: 403-419. — ot E SLUG LIFE CYCLES RESUMEN ESTUDIOS SOBRE “BABOSAS” DEL SUELO ARABLE II. CICLO DE VIDA P. J. Hunter A. reticulatus tiene dos generaciones anuales, una primaveral haciendo eclosion para Mayo, y otra en Otofio, hacia Septiembre. La segunda toma mas tiempo para completar el ciclo (7 meses), que la primera (5 meses). A. hortensis en su mayoria tiene un ciclo anual. Nacen alrededor de Julio y crecen durante los once meses sigientes para desovar cuando llegan a tener un afio; algunos nacidos mas tarde no fueron capaces de desovar hasta el segundo invierno, asi que tomaron 2 años en completar el ciclo. Cada individuo pone un promedio de 64,5 huevos y mueren pronto después de la puesta. La mayoria de M. budapestensis tienen ciclo bienal; hacen eclosión entre Mayo y Agosto y maduran durante el segundo otoño e invierno. Los pocos nacidos temprano, en Abril, pueden desovar el primer invierno. Como en las otras especies, la rápidez de desarrollo depende de las condiciones ambientales. Depositan un promedio de 23,5 huevos por individuo, muriendo después. Individuos cultivados en el campo desde los primeros estados, llegaron a depositar un promedio de 32,5 huevos. ABCTPAKT ИЗУЧЕНИЕ СЛИЗНЕЙ HA ПАХОТНЫХ ЗЕМЛЯХ П. ЖИЗНЕННЫЙ ЦИКЛ П. ДЖ. ХАНТЕР Автором иссследовался жизненный цикл у Agriolimax reticulatus (Muller), Arion hortensis Ferussac и МИах budapestensis (Hazay); работа проводилась на основании изучения обычных проб, собранных на участках пахотной земли в Северной Англии. Agriolimax reticulatus имел 2 генерации в год,-весенняя, в мае, и осенняя - конце сентября. Последним особям требовалось больше времени (7 месяцев), чтобы завершить свой жизненный цикл, чем первым (5 месяцев). Большая часть Avion hortensis имели годовой жизненный цикл; они отрождались в июле, росли втечение последующих 11 месяцев, созревали и откладывали яйца в возрасте около 1 года. Некоторые, однако, рождались позже и до 2ой зимы своей жизни еще не были готовы откладывать яйца, т.е. употребляли для своего цикла около 2 лет. Было найдено, что скорость их развития на всех стадиях зависит от условий среды. Особи откладывали в среднем 64.5 яйца и погибали вскоре после размножения. Большинство Milax budapestensis имели двухлетний жизненный цикл. Они отрождались между маем и августом и созревали втечение второй осени своей жизни и зимой. Меньшая часть, однако отрождалась в апреле и откладывала яйца в первую зиму своей жизни. Но скорость их развития опять таки зависела от условий существования. Слизни, собранные в поле перед самым размножением, откладывали, в среднем по 23.5 яйца на 1 особь, 389 390 P. J. HUNTER вскоре после чего погибали. Слизни, содержавшиеся в полевых культурах, начиная с их ранния стадий, откладывали, в среднем по 32.5 яйца каждый. MALACOLOGIA, 1968, 6(3): 391-399 STUDIES ON SLUGS OF ARABLE GROUND Ш. FEEDING HABITS P. J. Hunter Agricultural Research Council Unit of Insect Physiology School of Agriculture University of Newcastle upon Tyne, U.K.1 ABSTRACT Observations and experiments were made to establish the feeding activity and preferences of the slugs Agriolimax reticulatus (Müller), Arion hortensis Férussac and Milax budapestensis (Hazay). Feeding activity, which has a bearing on control by baiting, was measured by the number of wheat grains damaged per unit time. A preliminary test showed that Arion hortensis fed more often than Agriolimax reticulatus and the latter more often than Milax budapestensis. The total weight of wheat eaten per unit time, however, was similar in all 3 species; i.e., the species that fed more frequently consumed less at each feed. Experiments in laboratory and field showed that feeding was dependent on temperature, being maximal at 20° C, al- though some feeding occurred even at very low temperatures: at justabove 0° C Agriolimax reticulatus was most and Milax budapestensis least active. It was also shown that activity was greater at 100% than at 95% relative humidity but did not depend on day length, and that a distinct nocturnal rhythm occurs, slugs feeding most in the early part of the night. Dissection of slugs from the field showed that the surface-dwelling Agriolimax reticulatus has a greater tendency to feed on green vegetation than the under- ground-dwelling Arion hortensis and Milax budapestensis. The ecology of slugs on an arable plot in Northumberland was investigated by a routine sampling study between January 1963 and March 1965 (Hunter, 1966, 1967, 1968a, b). Atthe sametime a number of observations and experi- ments on the feeding habits of these slugs were made to establish the relationship between their feeding and their general biology and ecology. Feed- ing habits are also important in that they have a direct bearing on the control of these pests by baiting. Work was confined to the 3 commonest species in the area, Agriolimax reticulatus (Müller), Arion hortensis Ferussac and Milax budapestensis (Hazay). FEEDING ACTIVITY The effect on feeding of the major environmental factors was studied in laboratory and field experiments. Activity was measured as the number of wheat grains damaged per unit time. This method gives an easily visible record of activity, so that accurate measurements can be quickly taken. A preliminary test was conducted to es- tablish the amount of each grain that each of the 3 species consumed at each feed and thistest wasfollowed by assess- ments of the effect of temperature, humidity, day-length and the nocturnal rhythm on feeding activity. Arion hor- lPresent address: National Agricultural Advisory Service, Brooklands Avenue, Cambridge, U.K. (391) 392 P. J. HUNTER TABLE 1. Relationship between number of grains damaged*, amount eaten from each grain and + total weight of wheat eaten Species Avion hortensis Agriolimax reticulatus Milax budapestensis Mean No. damaged per slug* Mean Wt. eaten per grain (mg)* Mean total Wt. eaten (mg) 137.4 124.5 * These means are significantly different (P = < 0.001) from each other (analysis of variance and Tukey’s test). * Over a total of 10 days. TABLE 2. Mean numbers of grains damaged at various constant temperatures by 10 slugs over 10 days Species Arion hortensis 2501825180 Agriolimax reticulatus | 0.80 | 1.65 | 1.85 Milax budapestensis 2.80 0.05 | 0.85 | 0.60 | 0.70 | Overall Mean 10.0 | 15.0 | 20.0 25.0 7.60 | 9.20! 9.60 4.78 3. 15 3, 20 | 3. 50 2.31 1. 90 | 2. 05 | 2.25 -- Means that do not differ significantly from each other (P =>0.05 by analysis of variance and Tukey’s test) are linked by underlining. -- The overall species means differ significantly from each other. tensis and Agriolimax reticulatus of over 200 mg and Milax budapestensis of over 400 mg were uSed in experiments. Preliminary Test Slugs were confined individually in glass tubes 2” high x 0.9” diameter. The bases of these tubes were sunk in damp vermiculite, an inert, granulate material, and the tops enclosed by gauze caps. Ten slugs of each species were subjected for 10 days to one of 5 sets of alternating temperatures, namely 25 and 200 C, 20 and 150 C, 15 and 10° C, 10 and 5° С, or 5 and O° С, the higher temperature in each case obtaining from 09:00 to 21:00 hrs Greenwich Mean Time and the lower from 21:00 to 09:00 hrs. The temperatures were maintained in cooled incubators which were accurate to approximately + 19 C (checked by maximum - minimum thermometers throughout the study). Each slug was given 3 wheat grains weighing 65+ 2.5 mg. Damaged grains were counted, air-dried and weighed after each 12-hour period. The data from this experiment pro- vided a number of points of information on slug activity: 1. Avion hortensis damaged more wheat grains than Agriolimax reticu- latus, which, in turn, damaged more than Milax budapestensis (Table 1). Thedata were subjected to analysis of variance which showed that there were significant differences in numbers of grains damaged by the 3 species; further FEEDING HABITS OF SLUGS 393 Mean no. of grains damaged 10 Mean Night Temperature °C FIG. 1. Relationship between mean night temperature and the mean number of grains damaged by 30 slugs (10 of each species) on 30 different occasions. a. Avion hortensis; b. Agriolimax reticulatus; c. Milax budapestensis. analysis of data by Tukey’s Test (cf. Snedecor, 1956) showed that each of the above differences was significant. 2. Milax budapestensis consumed more of each damaged grain than did Agriolimax reticulatus and the latter more than Avion hortensis. Again, by analysis variance and Tukey’s Test, these differences were significant. 3. The average weight of wheat eaten by individuals of the 3 species over 10 days was similar (Table 1) and analysis of variance showed that there were no Significant differences present. It would therefore seem that, although the 3 species feed at different frequencies, the actual amount eaten does not differ significantly. 4. There were no Significant differ- ences in the amount of grain eaten at each feed at the various temperatures. Thus, although temperature affects the frequency of feeding (See next section) it did not affect the extent of each feed. Effect of Temperature (a) The numbers of wheat grains damaged at constant temperatures of 0.5, 92,545, 10, 15,20 and 25° С (all + approx. 19 C) were compared. For this purpose, slugs were confined in plastic boxes (5” x 5” x 2 1/2”), with 10 slugs in each box. The boxes, with gauze tops and bases, were half filled with sand and were standing in 1/4” of water to keep the sand wet. The experi- ment ran for 20 days. Fifteen wheat grains were added to each box; damaged grains were counted and replaced daily. All slugs were renewed on the 10th day of the experiment, and any that died during the experiment were replaced immediately. 394 P. J. HUNTER Feeding activity of all 3 species in- creased as temperature rose to 20° C but showed a decrease at 25° С (Table 2). At low temperature (just above freezing point) Agriolimax reticulatus was the most active of the 3 species (as es- tablished by Mellanby, 1961) and Milax budapestensis the least active. (b) Nine terylene-net bags (4” in diameter x 12” deep) were half-filled with damp vermiculite and sunk (to the depth of the vermiculite surface) in the soil of an outdoor plot. There were 3 bags for each species and 10 slugs in each bag. Six sets of observations were taken between July and December 1963. Slugs were freshly collected for each experiment and kept in the bags for at least 2 days prior to the observations. Ten grains were added to each bag and during the following 5 days damaged grains were counted and replaced daily. The mean number of grains dam- aged daily by each species was plotted against the average night air temper- ature (18:00 - 06:00 hrs; Fig. 1). Significant regressions (P = 0.001) of feeding activity on temperature were obtained for the 3 species, i.e. under the above conditions, feeding was directly related to temperature. However, evenatamean night tempera- ture of 12°C, the feeding proportion of the population was only 50-60%. Effect of Humidity Controlled humidities (Winston & TABLE 3. Establishment of controlled hu- midity Added to lower Relative chamber of humidity desiccator produced Distilled water 100% Saturated soln. of potassium 98% permanganate Saturated soln. of sodium 95% sulphite 7 Bates, 1960) at 20° C were created in the upper chambers of 3 desiccators as shown in Table 3. The humidities were checked with hair hygrometers previously calibrated against a Gregory hygrometer. Three each of Arion hortensis, Agriolimax veticulatus and Milax budapestensis were placed in each desiccator at any one time, each individual being confined in a glass tube with gauze caps at either end. Three wheat grains were put in each tube and damaged grains were counted and replaced daily. The experi- ment comprised three 5-day ‘trials’; new slugs were used for each trial. Fewer grains were damaged at low humidities than high ones (Table 4). TABLE 4. Mean numbers of grains damaged under varying conditions of hu- midity at 200 C Relative Humidity Species Arion hortensis Agriolimax reticulatus Milax budapestensis All species The data were subjected to analysis of variance which showed that some signifi- cance (P =<0.05) canbeattachedto these differences. The restriction of slug activity by low humidity has been noted by Hughes & Kerkut (1956) and Kerkut & Taylor (1956), who suggested that activity is regulated by the haemolymph concen- tration, the latter being affected by humidity. Rozsa (1962) did not support this view in her claim that activity can be stimulated directly from “osmore- ceptors” (humidity receptors) in the foot. It is not known how often the humidity of the environment in the field is low enough to limit feeding while not low enough to kill the animal. Slugs lost weight in all experiments, FEEDING HABITS OF SLUGS 395 TABLE 5. Influence of Humidity on Weight loss* of slugs during a 5-day peri- od Relative Humidity (at 20° C) Species % % 36.50 | 43. 22 27.33 | 42. 50 22.58 | 25.60 Arion hortensis 24.72 Agriolimax reticulatus| 25. 47 Milax budapestensis *Mean losses calculated from 9 slugs each. the loss being the greater, the lower the humidity (Table 5). However, there was considerable variation between repli- cates, the weight loss decreasing and the number of damaged grains increasing in the 2nd and 3rd trials. Effect of Length of Daylight The effect of day length on the feeding activity of slugs was tested over a period of one week. For this purpose, 40 adults of each of the 3 species were confined individually in glass tubes out of doors, with 3 wheat grains placed in each tube. Damaged grains were counted and replaced at 22:00 and 10:00 hrs British Summer Time (B. S. T.). During the experiment natural darkness lasted from approximately 22:00 hrs - 04:00 hrs. Total darkness for this period and longer was contrived by covering the tubes with opaque boxes as shown in Table 6. TABLE 6. Establishment of controlled length of darkness Resulting hours of darkness No. of slugs of each species Time of day covered hrs.) B.S. Lo 22:00-04:00 22:00-10:00 16:00-10:00 continuously 10 10 The numbers of damaged grains are given in Table 7. Length of daylight had no significant effect (P =< 0.05) on the numbers of grains damaged. Slugs damaged significantly more grains during the 22:00 - 10:00 period than during the day (P =< 0.001) irrespective of when they were covered. These results confirm the work of Dainton (1954) who established that the nocturnal rhythm is maintained by falling tempera- ture at night. Effect of Nocturnal Rhythm Four groups of each of the 3 species (each group containing 10 slugs) were confined in plastic boxes outdoors. Ten wheat grains were added to each box and those damaged were counted andre- placed every 6 hours, at 01:00, 07:00, 13:00 and 19:00 hrs., B.S. T. The ex- periment took place over 5 consecutive days in October when darkness lasted from 19:00 - 07:00 hrs. Slugs were feeding most actively in the early part of the night and least actively during the day (Table 8). FEEDING PREFERENCES Between July and November 1963, slugs were extracted from 4-weekly routine samples by soil washing (Hunter, 1968a). The extraction was carried out immediately after taking each sampling unit and was completed within 2 hours of digging: since sampling took place during the day and most slugs feed at night (see above), it is unlikely that they would have fed while awaiting extraction in the laboratory. Slugs were dissected immediately and their gut contents were examined under alow power microscope. It was not possible to classify plants that had been eaten, but the type of vegetation (leafy green material, white stems and brown roots and soil) couldbe distinguished. For the purpose of statistical analysis, vegetation in the gut was classified into dominantly green or non-green material 396 P. J. HUNTER TABLE 7. The number of grains damaged by 10 slugs over 1 week under varying conditions of light and darkness Hours exposed to daylight Species Arion hortensis Agriolimax reticulatus Milax budapestensis Totals Night = total grains damaged between 22:00 and 10:00 hrs. B.S.T. Day = total grains damaged between 10:00 and 22:00 hrs. B.S.T. TABLE 8. Mean number of grains damaged* by 10 slugs throughout a 24-hr. period Species (40 each) Arion hortensis Agriolimax reticulatus Milax budapestensis 0. 30 Overall Mean * Out of 10 grains provided for each 6 hr. period TABLE 9. Numbers of slugs containing green and non-green material Agriolimax reticulatus Milax budapestensis 6 E > т то 9 1 56 50 26 2 8 53 52 19 ib 4 58 43 И Е = [= Month (1963) July August 21 September 20 October 14 November 24 `. Totals FEEDING HABITS OF SLUGS 397 Table 9). Analysis of variance showed that a significantly larger proportion of Agriolimax reticulatus (73.2%) contained dominantly green material than Arion hortensis (18.6%) or Milax budapestensis (15.8%). There was no significant vari- ation in the proportion of slugs eating green material during the 5 months of sampling, i.e. slugs were not obliged to eat more non-green vegetation when much of the surface vegetation died during autumn. DISCUSSION AND CONCLUSIONS The experiments on feeding activity have relevance to baiting as a method of slug control and as a method of esti- mating populations. There is clearly a close relationship between slug activity and the temperature and humidity of the environment. The greatest efficiency in control by baiting can be expected on warm, humid nights (as established by Webley, 1964, 1965). However, even at a mean night temperature of 120 С, when the substrate was wet (see experiment in terylene bags: points at highest temperature on Fig. 1), only a little over half the population was feeding. Thus, to be effective, poison baits must be applied over a considerable period of time. The effect of environment on activity, however, does not invalidate the results of bait trapping for population estimation (Hunter, 1968a). Although the amount of damage to baits cannot give accurate estimates of the size of populations, they can indicate the time and direction of changes of density. The observations on feeding prefer- ences showed that under normal con- ditions there was little competition for food. The 3 species ate a wide variety of food that was commonly available, so it would seem that there will be few situations in which an absolute shortage occurs. The same has been previously demonstrated by Boycott (1934) who con- cluded “that food has no influence either by its quality or quantity on the re- currence of our land Mollusca, excepting Testacellidae and such meagre habitats as shifting sand dunes”. Getz (1959) demonstrated in laboratory experiments that 3 species of slugs including Agrio- limax reticulatus accepted a wide range of plants as food. The present ob- servations further show that slugs do not move far to feed, i.e. species that live underground tend to feed underground and those that live on the surface are surface feeders. Agriolimax reticulatus which eats mostly green material, is mainly surface dwelling (Hunter, 1966); Arion hortensis, which eats less green food, is more often found under soil level; and Milax budapestensis, which eats still less green food, is found deeper under ground. ACKNOWLEDGEMENTS I am very grateful to Professor A. Milne of the School of Agriculture, University of Newcastle upon Tyne, U. K., for ‘supervision and encourage- ment during this study and to Dr. A. South of Sir John Cass College, Uni- versity of London, for advice throughout the investigation. The work was sup- ported by a post-graduate studentship from the British Potato Marketing Board. REFERENCES BOYCOTT, A. E., 1934, The habits of land Mollusca in Britain. J. Ecol., 22: 1-34, DAINTON, B. H., 1954, The activity of slugs. I. The induction of activity by changing temperatures. J. exp. Biol., 31: 165-187, GETZ, L. L., 1959, Notes onthe ecology of slugs: Arion circumscriptus, Dero- cevas veticulatum and D. laeve. American Midland Naturalist, 61: 485- 498. HUGHES, С. M. € KERKUT, G. A., 1956, Electrical activity in a slug ganglion in relation to the concentration of Locke solution. J. exp. Biol., 33: 282-294. 398 Р. J. HUNTER HUNTER, P. J., 1966, The distribution and abundance of slugs on an arable plot in Northumberland. J. anim. Ecol., 35: 543-557. 1967, The effect of culti- vations on slugs of arable ground. Plant Pathology, 16(4): 153-156. 1968a, Studies on sulgs of arable ground, I. Sampling Methods. Malacologia, 6(3): 379-377. 1968b, Studies on slugs of arable ground. Il. Life cycles. Mala- cologia, 6(3): 379-389. KERKUT, С. A. € TAYLOR, В. J. R., 1956, The sensitivity of the pedal ganglion of the slug to osmotic pres- sure changes. J. exp. Biol., 33: 493- 501. MELLANBY, K., 1961, Slugs at low temperatures. Nature, 189: 944. ROZSA, К. S., 1962, A reflex mechanism changing the activity in Gastropods upon osmotic effects. Proc. 5th Meet- ing, Hungarian biol. Soc. 1962. SNEDECOR, G. W., 1956, Statistical Methods. 5th Edition. Iowa State Univ. Press. WEBLEY, D., 1964, Slug activity in relation to weather. Ann. appl. Biol., 53: 407-414. 1965, Aspects of trapping slugs with metaldehyde andbran. Ann. appl. Biol., 56: 37-45. WINSTON, P. W. & BATES, D. H., 1960, Saturated solutions for the control of humidity in biological research. Ecology, 41: 232-237. RESUMEN ESTUDIOS SOBRE “BABOSAS” DEL SUELO ARABLE III. HABITOS ALIMENTICIOS P. J. Hunter Las funciones alimenticias en las mismas especies fueron medidas, y sus pre- ferencias controladas, por la cantidad de granos de trigo, usados como cebo, que averiaron por unidad de tiempo. Una prueba preliminar demostró que Arion hortensis come con mas frecuencia que Agriolimax reticulatus y la segunda mas que M. buda- pestensis, pero el total de trigo consumido fue similar en las 3 especies, desde que las que comen con mas frecuencia consumen menos por vez. indicaron que la actividad depende de la campo, asi como en el laboratorio, Experimentos en el temperatura, siendo máxima a los 20° C, aunque pueden alimentarse hasta en muy bajas temperatures: M. budapestensis. justo a 0°C Agriolimax reticulatus fue un poco más activo que Se notó también que la actividad alimenticia es mayor a 100% que a 95% de humedad relativa, pero que no depende de la duración diurna, sino que ocurre tambien en ritmos nocturnos, la mayoria alimentandose en las primeras horas de la noche. Disectando individuos silvestres se encontró que los que viven en superficie como A. reticulatus prefieren alimentarse de vegetación verde enuna mayor proporción que los que viven enterrados como Arion hortensis y Milax budapestensis. ABCTPAKT ИЗУЧЕНИЕ СЛИЗНЕЙ HA ПАХОТНЫХ ЗЕМЛЯХ II. ПИТАНИЕ П. ДЖ. ХАНТЕР Были проведены наблюдения и эксперименты для установления активности питания и предпочтения той или иной пищи слизнями FEEDING HABITS OF SLUGS 399 Agriolimax veticulatus (Muller), Arion hortensis Ferussac и Milax buda- pestensis (Hazay). Активность питания, контроль над которой осуществлялся при помощи приманок, измерялся количеством зерен пшеницы, поврежденных слизнями за единицу времени. Предварительные пробы показали, что Arion hortensis питается чаще, чем Agriolimax тенсша из, а последний-чаще, чем Milax budapestensis. Общий Bec съеденных зерен пшеницы за единицу времени был, однако сходным для всех трех видов, Т.е. виды, питавшиеся чаще потребляли меньше пищи за каждую еду. Эксперименты в лаборатории и в поле показали, что питание слизней зависело от температуры, будучи максимальным при 20°C, хотя некоторое питание наблюдалось даже при очень низкой температуре: почти при O°C Agriolimax reticulatus был более активен, а Milax budapestensis несолько менее активен. Было также показано, что активность питания была выше при 100%, чем при 95% относительной влажности, но не зависела от продолжительности дня, и что у них имеется хорошо-выраженный ночной ритм: слизни питаются больше в раннюю пору ночи. Вскрытия слизней, собранных в поле показали, что живущие на поверхности Agriolimax reticulatus имеют большую тенденцию питаться зеленой растительностью, чем живущие в под- поверхностном слое Avion hortensis и МИах budapestensis. р. MM i re AVE? ou Mer TE at Ñ Fes f Y, A "x т oun ‘ of } - | + Rr yt; AMO UALS ТА A mi h. a рыл} > iS Po. à poise IN vey in Г а BG | че И По И i 1 or bine » 142 Th 4 411194878 7 il sit: . { { йо ме) . ) в "a мы 14 Lie pile ‘ КМ? fev? Mä& PUT A eu, ed AAA 14 “Qi \ © AU ey м EN QAR A RCE с dl NU ds CO lea ’ E es yO! м ioe a "nié un Yo LA i i \ 2 у N у CYR AÑ \ . ‘ ^ р N I yo * Г A Ab de А y ah NA ENT ARTO An y vir E ме LT En J je El } lad AE) dar j 5 ¡EA A f M , va i у a гв 4 4 у ke! à tt OUR 4 f A Pe L y А. 175 2, 1 AES, e ae . Nid MATE | — Ha 1; ad п Fl У} vá fe \ al +) ; = y Ls Y pl ‘ i 0 A qua и’ Па tae ww ‘ A o A à y Alp nalts O р УМ t I р a MALACOLOGIA, 1968, 6(3): 401-410 THE LOCOMOTION OF THE FRESHWATER CLAM MARGARITIFERA MARGARITIFERA (UNIONACEA: MARGARITANIDAE) E. R. Trueman Zoology Department The University, Hull, England ABSTRACT The locomotion of Margaritifera mavgaritifeva (L.) was investigated by use of electronic recording devices and its habits are described with brief reference to the effect of shell shape (e.g., the presence of a pronounced pedal gape in fast water forms) on burrowing. The series of burrowing activities, termed the ‘digging cycle’, are very similar to those of other bivalves and are used both in burrowing and, with little modification, in locomotion over the surface of sand. The most obvious differences involve the more anterior orientation of the foot and the relative magnitude of contraction of the anterior and posterior retractor muscles. During the digging cycle the valves adduct to loosen the adjacent sand and to obtain a pedal anchorage through dilation of the foot by means of blood pressure. When on the surface of the sand, the valves are reopened exclusively by the hinge ligament; but, with more than 1/3 of the shell buried, the pressures de- rived from pedal retraction are used to supplement the ligament. Burrowing activity involves the integration of adduction and opening of the valves with pro- traction and retraction of the foot. Adduction produces a pedal anchorage, al- lowing the shell to be drawn down at retraction, while the shell is held firm in the sand by the opening thrust of the ligament (secondary or shell anchor) at pedal protraction. Bivalves living on a hard substrate have a basically similar locomotory pattern but without the occurrence of adduction, possibly because the pedal anchorage is not obtained in the same manner. INTRODUCTION The process of burrowing in the Bi- valvia has been recently studied in com- mon British littoral species using electronic recording techniques for de- tailed analysis of their activities (True- man, Brand & Davis, 1966a). Previous work, summarized by Morton (1964), who gives a full bibliography, in general lacks the precision which the more modern techniques can provide. Never- theless it indicates, together with the present observations, that burrowing by bivalves follows a common pattern of activity. Burrowing consists es- sentially of a series of step-like move- ments into the substrate, termed the ‘digging sequence’ by Ansell (1962) and more recently the ‘digging cycle’ (True- man et al., 1966a). Digging generally commences with the bivalve lying on its side extending its foot into the substrate to lift the shell erect before pulling it down into the sand in a succession of digging cycles. Theterm ‘digging period’ may be applied to describe this activity from the start of burrowing until the final position in the substrate is reached. The attention of the author was drawn to the freshwater clam Margaritifera margaritifera (L.) by Dr. В.М. С. Eagar, who pointed out that the form of its shell was similar to that of certain Carboniferous bivalves, e.g. the An- thracosidae. In some forms of Mar- gavitifera the shells have curved dorsal margins and inflected lower borders (IN, (401) 402 E. R. TRUEMAN 21cm] FIG. 1. Diagrams of Margaritifera margaritifera, reconstructed from filmed sequences of burrowing, (a) when buried and (b) moving over the surface of the sand. a: shows a sagittal section with the foot (F) extended in the active digging position, the surface of the sand (horizontal line), the extent of the mantle cavity (stippled) with gills hatched, and the inhalent (I) and exhalent (E) siphons. AA, PA, anterior and posterior adductor muscles. b: shows the position of the foot in the sand during locomotion across the surface of the sand (stippled), water ejection currents from the mantle cavity at adduction of the valves with the siphons closed producing a cavity (C) in the sand in front of the shell and over the upper surface of the foot. EL, position of electrode to record valve movement; IN, inflected ventral margin; M, outer mantle folds around pedal aperture. In both figures the direction and relative movement of the shell by anterior (A) and posterior (P) retraction is indicated (<—<) and the region of the foot forming the pedal anchor is shown by hatching. Fig. 1b) which are characteristically associated with relatively swiftly flowing water. In contrast, forms with straight hinge lines and more rounded lower borders are generally found in slowly moving water (Eagar, 1947, 1948). Al- though a comprehensive study of Mar- garitifera in relation to its ecology has been made by Hendelberg (1960), the current investigation of locomotion in this genus was initially carried out soas to attempt further elucidation of the relationship between shell shape and habitat. The findings concerning the latter were limited, but the work led on to a more complete understanding of the mechanics of burrowing in Bivalvia, which are described below principally in terms of Margaritifera. METHODS Observations of the digging period have been made by filming the activity of Margaritifera in a glass tank for subsequent analysis and by recording Simultaneously valve movements and the pressures developed inthe substrate. Recording valve movement involves the placing of a pair of fine wire electrodes (EL, Fig. 1b), one on each valve, be- tween the umbone andthe ventral margin. LOCOMOTION OF MARGARITIFERA 403 These are of very light wire and, with a loop stretching between the electrodes and the recording device, proved to be of little hindrance since the bivalve burrowed quite normally. A small oscillatory current (2 y amp, 25 kc/sec) is passed between the electrodes and any movement of the valves affected the impedance between them. A voltage, proportional to the change inimpedance, was fed to a pen recorder by A.C. coupling, which allowed any change of impedance to be recorded about a preset level. Thus opening or closing of the valves gave positive or negative swings of the pen respectively, while, if the valves remained still at any angle of gape, the preset level would be recorded (Fig. 3). The method of recording pressure in- volved the connection of a Sensitive Statham pressure transducer to a tube Opening beneath the sand near the foot of the burrowing bivalve. Interpretation of these recordings must always be re- lated to direct observations of the ani- mal’s activity since negative pressures are recorded either by the application of pressure or by withdrawal of the foot, while the ejection of water from the mantle cavity of a bivalve beneath the sand at adduction causes positive pres- sures. Once the recordings have been interpreted by observation they afford a ready means of continually recording the digging of bivalves even when completely invisible beneath the sand. A full account of these techniques is given by Hoggarth & Trueman (1967). HABITS Specimens of Margaritifera, of the arched morphological variety, were col- lected from the River Lune (at Crook of Lune, Lancashire, England), a fairly fast running salmon river, and were taken to the laboratories of the Zoology Department of the University of Hull for immediate investigation of burrowing habits. The clams were found, during a dry spell, in water 3 feet deep, almost completely buried in fine sand as in Fig. la, leaving about 2 cm of the posterior valve margin with the inhalent and ex- halent siphonal apertures exposed above the substrate. The same position was also taken up after burrowing in an aquarium. The angle at which Mar- garitifera burrows allows an un- interrupted exhalent current to flow from the postero-dorsal exhalent siphon. When specimens were taken to the laboratory and placed in an aquarium containing sand from the River Lune, a few burrowed immediately, but the majority actively moved across the surface of the sand for a day or so before digging down- wards. The position of the foot and its ex- tension from the valves anteriorly is shown in Fig. 1b when the clam was moving over the surface of sand. The foot was never observed to extend pos- teriorly of the inflection (IN) of the ventral margin of the valves at any stage of digging. In Margaritifera from a different habitat with more rounded ven- tral valve margins the foot was similarly extended, never being observed behind the mid-ventral region of the shell. The location of the extended foot in Margari- tifera is similar to that observed in other bivalves with elongated shells, e.g. Donax, whereas with more rounded valves, e.g. Cardium, the foot may ex- tend much more posteriorly. In the inflected forms there was also a pro- nounced anterior pedal gape even when the valves were completely closed mid- ventrally, through whichthe outer mantle fold (M) protruded at pedal extension. This pedal gape may possibly be associated with an extension of the foot for longer periods in those animals that live in faster waters, soasto ensure a better anchorage. For the other features of the shell associated with fast flowing water no function was ap- parent from these studies of burrowing. Movement over the sand took place by a series of digging cycles, identical in respect of sequence of activities with those observed when burrowing into the 404 E. R. TRUEMAN DILATION PROBING FOOT [pza] | SIPHON CLOSURE | | | RETRACTOR CONTRACTION WATER EJECTION hn GAPE © (degrees) 14 16 20sec 2 4 6 8 10 12 18 FIG. 2. Diagram of the analysis of a single digging cycle of Margaritifera margaritifera from film and recordings. The change of gape of the valves is taken from a single cycle when the bi- valve was moving over the surface of the sand and the probing action of the foot (mmm), its period of maximum dilation (===), the period of closure of the siphons, the contraction of the retractor muscles (AR, anterior; PR, posterior), and the period of water ejection are placed in the correct time sequence by observation of many digging cycles. The stages of the cycle (i-vi) are indicated. See text below for further information. sand, at intervals of about 1 1/2 minutes. Each cycle gave a forward movement of about 1/2 cm (in a specimen of 10 cm length). This surface locomotion con- tinued, sometimes rather spasmodically, until the clam began todig deeply. There was no apparent stimulation or other explanation for this change inbehaviour, but in the natural habitat locomotion over the surface of the sand, after being dislodged from the buried position, may be the means of finding more suitable conditions. EXPERIMENTAL RECORDINGS The detail of movements of Margari- tifera during the digging cycle have been elucidated by means of impedance and pressure recordings and analysis of film taken of complete digging cycles both from lateral and frontal aspects. The results, summarised in Fig. 2, indicate that the digging cycle is a closely in- tegrated series of movements of different regions of the body. This series is the same for all digging cycles of Margari- tifera and resembles those found in other species of bivalves, e.g. Gly- cymeris, Mercenaria, Ensis, Tellina, Donax (Trueman, 1966; Trueman et al., 1966a). The cycle, which is best under- stood by reference to Fig. 2, comprises the following 6 stages from left to right: (i) The foot makes a major probe downwards tending to raise the shell if pedal penetration is not easily achieved. This probe occurs in many bivalves, e.g. Tellina, Mactra and Ensis (True- man, 1966) and, although it has not been observed in Margaritifera, the stage is marked in the figure. Pedal probing appears to cease for approximately 5 seconds before adduction of the valves. (ii) Siphons close, preventing water passing out through their apertures during the next 2 stages. (iii) Adduction of the valves occurs rapidly within 0.25 seconds and corres- ponds to the onset of pedal dilation and the ejection of water from the mantle cavity. Pressure recordings of Mar- garitifera and other genera, e.g. Ensis, Donax, show that these are respectively brought about at adduction by the in- crease of fluid pressure in the haemo- coele and in the mantle cavity (Trueman, 1966). Dilation of the distal part of the foot secures the pedal anchorage. (iv) Contraction of the pedal retractor muscles, that of the anterior being followed by the posterior, imparts a LOCOMOTION OF MARGARITIFERA 405 +05 LS, P PRESSURE (cm) ch GAPE FIG. 3. Recordings of the pressure changes (upper records) in the sand and valve movements (Gape, lower records) during (a) locomotion over the surface of the sand and (b) burrowing into sand with 1/3 the shell under the surface. A pressure transducer is connected by tubing to near the foot of the Margaritifera beneath the sand, and the valve movements are recorded by means of electrodes attached to the valves connected to an impedance pneumograph, which was coupled (A.C.) to a pen recorder. a: shows adduction at A, corresponding to negative pressure in the sand caused by dilation and retraction of the foot. Successive pedal probes (P) commence after the valves have reopened, the thick line indicates period of closure of the siphons. b: with the specimen more deeply buried, negative pressure at adduction (A) becomes positive, due to the expulsion of water into the sand and many pedal probes (P) are recorded between the adductions of the valves. Negative pressure during the period of probing (O) is due to the re- traction of the foot associated with the further opening of the valves. Further information in the text below. rocking motion to the valves as indicated with loss of pedal anchorage. by the arrows (A and P) on Fig. 1. This (vi) Pedal probing recommences and sequence of contractions results in continues until the next cycle. This movement along the surface or into the stage has been previously termed the sand according to their relative magni- ‘static period’ (Trueman et al., 1966a) tude and the orientation of the foot. since the shell does not move except in During this phase of pedal retraction response to the downward pressure of the siphons reopen and pedal dilation is the foot during probing. reduced. Recordings of digging cycles when the (v) Relaxation of the adductor muscles, clam was either on the surface or slow reopening of the valves, together partially under the sand (Fig. 3a, b) show 406 E. R. TRUEMAN negative or positive pressure peaks (A) according to the depth of burial at stages iii and iv. The latter are probably caused by the ejection of water from the mantle cavity into the sand, whereas, when the animal is on the surface, this ejection is much more superficial so that the effect of pedal dilation is then seen as a negative pressure. The effect of water ejection is to loosen the sand in front of the shell immediately before pedal retraction takes place so as to facilitate shell movement (C, Fig. 1b). This loosening extends to the upper part of the foot as indicated by the arrows, but it does not affect its lower part, which must act as an anchor (Fig. 1, hatching). Work on Margaritifera and other bivalves (Trueman, 1966; True- man et al., 1966a) has demonstrated that pedal anchorage is obtained by the dilation of the foot brought about by a sudden increase in blood pressure in the pedal haemocoel that is derived from adduction of the valves. This pressure and anchorage are sustained by pedal retraction. Without an anchorage the foot would be pulled into the shell on contraction of the retractor muscles instead of the shell being pulled down. It is important that the loosening of the sand should only extend around the proxi- mal part of the foot and not affect the pedal anchorage. When Margaritifera is moving over the surface, the valves show much more movement than when it is beneath the sand, which has a damping effect. The movements are noticeable when probing (P, Fig. 3a) and immediately prior to adduction (A). A similar sharpincrease in gape has been observed in other bi- valves at the surface of the substrate, e.g. in Cardium, and may possibly be due to the relaxation of the ‘slow’ adductor muscle fibres before the ‘fast’ fibres contract. When the clam was buried more deeply, the surrounding sand pre- vented this additional gape. The cycle represented in Fig. 2 is of a Margaritifera moving over the surface of the sand and, apart from the relative magnitude of anterior and posterior re- traction, only differs materially from deeper burrowing by the increasing duration of the digging cycle with depth (Fig. 3b). Digging involves a greater interval between successive adductions, a longer static period, and consequently more probes by the foot. The longer time per cycle is probably related to the increasing difficulty that the foot may have to penetrate the substrate at greater depth (Trueman, Brand & Davis, 1966b). The rapid succession of probes (P) being made by the foot is clearly shown in Fig. 3a, where they may be observed to recommence only after the reopening of the valves. When burrowing on or near the surface, re- opening occurs in about 10 seconds, but after about 1/2 of the shell is beneath the sand a secondary opening movement (O) of the valves is necessary toachieve the full gape. Direct observations of this secondary opening movement in- dicate that it coincides with the closure of the siphons and the retraction of the foot. The negative pressure recorded (Fig. 3b) is probably caused by the withdrawal of the foot. Similar obser- vations made on Mercenaria (Ansell & Trueman, 1967) suggest that the con- traction of the pedal retractor muscles increases the hydrostatic pressure inthe haemocoele and that this pressure con- tributes to forcing the valves open. The function of the hinge ligament is to open the valves when the adductor muscles relax. When moving over the surface, the strength of the ligament (opening moment, Trueman, 1964) is sufficient to open the valves adequately; but, as the shell lies progressively deeper and the resistance to the opening of the valves becomes greater, the ligament requires supplementation. It appeared that the ligament ceases to be adequate for the full opening of the valves when more than 1/3 of the Shell is buried in fine sand. Deter- minations of the opening moment of the ligament, using the method described by Trueman (1954) show that a moment of LOCOMOTION OF MARGARITIFERA 407 27,600 g mm is available to open the valves of a Margaritifera 9.4 cm in length. This force is equivalent to a moment of 8.8 g mm/mm? in relation to the projected surface area of a valve. Since the ligament is inadequate for the opening of the valves when more than 1/3 is buried, the resistance of fine sand to opening is approximately 3 times this figure (26 © mm/mm2?). The shell of Mya was used to determine the resistance of marine substrates to the opening of the valves (Trueman, 1954) and it was shown that in fine sand the application of a moment of 20 g mm/mm2 would produce a gape of about 40. Allowing that Mya is more deeply buried than Margaritifera, these figures are comparable and suggest that the latter Species may well need to supplement the ligament when more than 1/3 ofthe shell is buried. The additional hydrostatic pressure would probably not be required if the valves reopened more rapidly after adduction, while the sand was still loosened by the water ejected from the mantle cavity, as occurs in Tellina and Ensis (Trueman et al., 1966a). DISCUSSION The digging cycles consist essentially of the repeated adduction and reopening of the valves, integrated with the re- traction and protraction of the foot. Ad- duction accomplishes 3 things: a) pedal anchorage, by dilation of the distal part of the foot; b) water ejection from the mantle cavity, thus loosening the sand adjacent to the shell; c) reduction in effective width of the shell (Fig. 4a). Functions b and c both facilitate move- ment of the shell into the substrate at retraction but a pedal anchor (P) is an essential prerequisite. After retraction the foot must be protracted before a further digging cycle can take place. Protraction is achieved during the static period (vi) largely by the intrinsic pedal musculature; it involves the contraction of the transverse pedal muscles and the relaxation of the retractor muscles (TM, FIG. 4. Diagrammatic transverse sections of a bivalve during burrowing into sand (stipple) to show the conditions (a) of pedal anchorage (P) after adduction and before retraction (stages iii-iv) and (b) of shell anchorage (S) during static period (vi). In (a) pedal an- chorage is established by the dilation of the foot (F; arrows) due to the in- crease of haemocoel (H) pressure at adduction (»—~<) by tension in ad- ductor muscles (AM). Adduction also causes the loosening of the sand ad- jacent to the valves (unstippled area) and places the ligament under maxi- mal strain (¢—»). In (b) the flat- tened foot probes downwards (ar- rows), the adductors (AM) are re- laxed, and the ligament (L) presses the valves (black) open against the adjacent sand (arrows). M, mantle cavity; RM, retractor muscles; TM, transverse pedal muscles. Fig. 4) in a manner similar to that described for Ensis and for members of the Tellinacea (Morton, 1964: Trueman et al., 1966a). When the shell is opened by the ligament, it is pressed against the sand forming a shell or secondary anchorage (S, Fig. 4b). This anchorage 408 E. R. TRUEMAN tends to prevent the animal from being pushed upwards as the foot pushes down. Drew (1907) and Pohlo (1963) have pre- viously observed that the valves of mem- bers of the Solenidae may grip the walls of the burrow during pedal protraction. The force with which the foot can probe is a function of the effectiveness of the shell anchorage, which in turn is related to the strength of the ligament. Probing forces in excess of the secondary anchor - age cause the shell to be pushed up- wards during probing as has been demon- strated in Cardium edule (Trueman et al., 1966a, b). When Margaritifera moves over the surface of the sandthere can beno Shell anchorage andthe strength of probing is limited to the weight of the animal. Whilst discussing the burrowing of worms, Clark (1964: 93) suggests that the fundamental method of burrowing used by all softbodied animals is the same. Part of the bodywall is dilated to form an anchor while the head is forced into the substrate by contraction of the circular muscles. The anterior end of the worm then dilates forming a new anchor while the body is drawn downwards by the contraction of the longitudinal muscles. These 2 anchor- ages correspond to the shell and pedal anchors of bivalves respectively, while the circular muscles are represented by the transverse pedal muscles and the longitudinal muscles by the retractors. Thus the burrowing movements of a bi- valve conform to Clark’s description. Bivalves have the advantage, however, of an additional fluid muscle system, the pallial system, by means of which powerful water jets assist penetration. Bivalvia are primitively adapted to shallow burrowing in soft, often un- stable, substrates (Morton, 1964). An important adaptation to a burrowing mode of life is the bivalved shell and the fluid-muscle system of the foot which permits the strength of adduction to be used in digging to anchor the foot. The pattern of the digging cycle is similar in all genera in which burrowing has been examined by the use of modern recording techniques (Trueman, 1966), and is retained by Margaritifera when moving over the surface of sand. Those bivalves, which have changed from the primitive infaunal to an epifaunal mode of life and can progress over a hard substrate, retain the rhythm of extension and foreshortening of the foot (Morton, 1964). Observations on the surface locomotion of Mytilus (by the author) and of Lasaea (Morton, 1960) indicate, however, that pedal movement in those genera is carried out without adduction of the valves. Extension of the foot involves only the intrinsic musculature, as does pedal probing during digging, and retraction is not preceded by dilation since pedal anchorage is obtained in a different manner on a hard substrate. It would appear from the present work that elimination of adduction from the locomotory cycle may be a consequence of movement over a hard rather than a soft substrate. Further detailed ob- servations of the locomotion of bivalves over hard surfaces, using the methods described in this article, would be of interest. REFERENCES ANSELL, A. D., 1962, Observations on burrowing in the Veneridae (Eula- mellibranchia). Biol. Bull., Woods Hole, 123: 521-530. ANSELL, A. D. & TRUEMAN, E. R., 1967, Burrowing in Mercenaria mer- cenaria (L.) (Bivalvia, Veneridae). J. exp. Biol., 46: 105-115. CLARK, R. B., 1964, metazoan evolution. don Press. DREW, G. A., 1907, The habits and movements of the razor clam, Ensis directus Con. Biol. Bull., Woods Hole, 12: 127-138. EAGAR, ВБ. M. C., 1947, A study of a non-marine lamellibranch succession in the Anthroconaia lenisulcata zone of the Yorkshire coal measures. Phil. Trans. B, 223: 1-54. Dynamics in Oxford, Claren- LOCOMOTION OF MARGARITIFERA 409 1948, Variation in shape of 104. shell with respect to ecological station. TRUEMAN, Е. R., 1954, The mechanism Proc. roy. Soc. Edin. B, 63: 130-148. of the opening of the valves of a HENDELBERG, J., 1960, Thefreshwater burrowing lamellibranch, Mya aren- pearl mussel, Margaritifera margari- aria. J. exp. Biol., 31: 291-305. tifera (L.). Rep. Inst. Freshwat. Res. 1964, Adaptive morphology in Drottningholm, 41: 149-171. paleoecological interpretation. In: HOGGARTH, К. В. € TRUEMAN, E. R., Approaches to Paleoecology, p 45-74. 1967, Techniques for recording the Imbrie, J. € Newell, N. W. (Eds.). activity of aquatic invertebrates. New York, Wiley. Nature, Lond., 213: 1050-1051. 1966, Bivalve mollusks: fluid MORTON, J. E., 1960, The responses dynamics of burrowing. Science, 152: and orientation of the bivalve Lasaea 523-525. rubra Montagu. J. marine biol. Soc. TRUEMAN, E. R., BRAND, A. R. & U. K., 39: 5-26. DAVIS, P., 1966a, The dynamics of 1964, Locomotion. т: Physi- burrowing of some common littoral ology of the Mollusca, Vol. I, p 383- bivalves. J. exp. Biol., 44: 469-492. 423. Wilbur, K. M. & Yonge, C. M. 1966b, The effect of substrate (Eds.). New York, Academic Press. and shell shape on the burrowing of POHLO, R. H., 1963, Morphology and some common bivalves. Proc. malac. mode of burrowing in Szliqua patula Soc. Lond., 37: 97-109. and Solen rosaceus. Veliger, 6: 98- RESUMEN LA LOCOMOCION DE LA ALMEJA DE AGUA DULCE MARGARITIFERA MARGARITIFERA (UNIONACEA: MARGARITIFERIDAE) E. R. Trueman Esta investigación se efectuó mediante registros electrónicos, y el comportamiento de Margaritifera margaritifera (L.) se describe con breve referencia a el efecto que la forma de las valvas ejerce enla actividad excavadora (por ejemplo, la presencia de una pronunciada brecha pedal en formas de aguas rápidas). La serie de funciones llamada “ciclo excavador” es muy similar a las de otros bivalvos (sus movimientos sirven tanto para la excavación propiamente dicha como para la locomoción sobre la superficie arenosa), y las diferencias más obvias se presentan en la orientacion, más anterior, del pié y la magnitud relativa de contracción de los músculos re- tractores, anteriores y posteriores. Durante el ciclo excavador, las valvas se contraen para que la arena circundante quede más suelta en el agua, y obtener asi anclaje pedal dilatanto el pié por presión circulatoria. Cuando la almeja esta sobre la arena las valvas se entreabren sólo por acción del ligamento, pero cuando estan enterradas hasta un tercio, entonces la presión de la retracción pedal reemplaza a la del ligamento. La actividad excavadora se integra con el entreabrir y cerrar de las valvas y protracción y retracción del pié. La aducción produce un anclaje pedal, permitiendo asi a la concha ser arrastrada en la retracción, hacia abajo, mientras está firme en la arena por el empuje apertural del ligamento durante la protracción pedal. Bivalvos que viven en substratos duros tienen basicamente una locomoción similar, pero sin que se produzca aducción, posiblemente porque el anclaje pedal no se obtiene en la misma forma. 410 Е. В. TRUEMAN АБСТРАКТ ДВИЖЕНИЕ ПРЕСНОВОДНЫХ МОЛЛЮСКОВ MARGARITIFERA MARGARITIFERA (UNIONACEA: MARGARITANIDAE) Bese Par ТРУМЕН Движение y Margaritifera margaritifera (L.) исследовалось при помощи электронных самописцев; в работе рассматривается также влияние формы раковины моллюска на его закапывание (при этом имеется ввиду наличие выдающегося ножного выступа у форм, обитающих в водах с быстрым движением). Ряд движений моллюска, при его зарывании, названные автором "циклом закапывания" в общем сходны с теми движениями, которые производят другие двустворчатые как при закапывании, так и при движении по поверхности песчаного грунта. Наибольшие различия заключаются в положении ноги, направленной больше вперед и в относительной силе сокращения переднего и заднего мускулов-ретракторов. Втечение цикла закапывания створки смыкаются, чтобы освободиться OT окружающего песка и чтобы образовать заякоривание моллюска при помощи расширения его ноги благодаря увеличению давления крови в ней. Если моллюск находится на поверхности грунта, то его створки при- открываются исключительно при помощи лигамента; но когда более 1/3 раковины уже погрузилось в грунт, давление, полученное при помощи сокращения ноги служит для усиления действия лигамента. Акт закапывания включает совместное действие смыкания и размыкания створок и вытягивание и сокращение ноги. Смыкание створок создает заякоривание при помощи ноги, позволяющее раковине быть втянутой вниз при сокращении, в то время как раковина крепко удерживается в песке при открывающем действии лигамента при втягивании ноги (вторичное или раковинное заякоривание). Двустворчатые, живущие на жестком грунте имеют сходные характеристики движения, но без смыкания створок, вследствие чего ножное заякоривание происходит иначе. INDEX TO SCIENTIFIC NAMES abyssinicus, Bulinus, 195 abyssinicus, Physopsis, 195 Acicula, 244 Aciculidae, 5, 244 Acleioprocta, 200, 228 Acmeidae, 5 acuminata, Albinula, 244 acuminata, Gastrocopta, 244, 246 aenigma, Deroceras, 254 Aeolidia, 224 exigua, 224 Aeolidioidea, 200 Aequipecten, 285 affinis, Herviella, 223-230 africanus, Bulinus, 190, 195 africanus, Physopsis, 190, 195 Agriolimax, 369-399 reticulatus, 369-399 albida, Herviella, 223-230 albilabris, Pupoides, 254 Albinula, 244, 246 acuminata, 244 ukrainica, 244 Alcyonidium, 205, 208 gelatinosum, 208 hirsutum, 208 verrilli, 208 Aleochara, 305 Amnicola, 5, 13, 14, 63, 134 integra, 63 lapidaria, 5 robusta, 134 sayana, 13 Amnicolidae, 4-6, 13 amoenula, Idaliella, 205 amoenula, Okenia, 205 Amygdalum, 301 papyria, 301 Anaspidea, 199 Ancylidae, 155, 167 Ancylus, 155, 156 burnupi, 157 equeefensis, 157 rivularis, 155, 156 tapirulus, 170 tasmanicus, 156 andrussovi, Caucasotachea, 244 angulifera, Vertigo, 244 angulifera, Vertilla, 244, 245 angustior, Vertigo, 245 Anisus, 195 coretus, 195 natalensis, 195 Anomia, 301 simplex, 301 Anoplodactylus, 216 brasiliensis, 216 antiqua, Hawaiia, 244, 246 antivertigo antivertigo, Vertigo, 245, 246 antivertigo callosa, Vertigo, 244 antivertigo, Vertigo antivertigo, 245, 246 Aplexa, 254 hypnorum, 254 Aplysia, 199, 200 modesta, 200 morio, 199, 200 Aplysiidae, 199 appressa, Lymnaea stagnalis, 66 arboreus, Zonitoides, 254 Archidoris, 201 arctatum, Mesodesma, 231, 234 areolata, Doriopsilla, 211 Arion, 369-399 hortensis, 369-399 armifera, Gastrocopta, 254 armigera campestris, Segmentina, 255 armigera, Planorbula, 253-265 Armina, 199, 213-217 californica, 215 columbiana, 215 convolvula, 215, 216 cuvieri, 215 digueti, 215 major, 216 mülleri, 215 natalensis, 216 semperi, 215 tigrina, 216 undulata, 199, 213-216 vancouverensis, 215 wattla, 199, 213-217 Arminoidea, 200 atropos, Dendrodoris, 209 australica, Ferrissia, 168-170 Australorbis, 112, 166 glabratus, 112, 166 balanoides, Balanus, 296, 306, 310, 311, 314, 315 Balanus, 271-320 balanoides, 296, 306, 310, 311, 314, 315 eburneus, 296, 306 improvisus, 301 baratariae, Corambella, 207 (411) 412 baratariae, Doridella, 207, 209 Basommatophora, 155, 175, 189 batava, Doridella, 209 bayeri bayeri, Candiella, 212 bayeri, Candiella bayeri, 212 bayeri misa, Candiella, 199, 211 bayeri misa, Tritonia, 199, 210-212, 217 bayeri, Tritonia bayeri, 212, 217 belokrysi, Pupilla, 245, 246 bicarinatus, Gyraulus, 195 binneyi, Pomatiopsis, 1, 2, 13, 14, 109, 110, 124, 134, 135 Biomphalaria, 112, 166, 167, 195 glabrata, 112, 166, 167 sudanica, 195 Bithiniinae, 6 Bithyniidae, 5,6 Bittium, 301 Blanfordia, 10, 14, 80, 81, 109, 110, 133, 134 formosana, 80, 81 japonica, 109, 110 Bollingeria, 245 pupoides, 245 bovis, Schistosoma, 185 Brachidontes, 276, 281, 306 exustus, 216, 281, 306 Bradybaenidae, 244 brasiliensis, Anoplodactylus, 216 buchi, Helix, 245 budapestensis, Milax, 369-399 Bulinus, 175-198 abyssinicus, 195 africanus, 190, 195 depressus, 176, 184, 185 forskalii, 190 guernei, 176 hemprichii depressus, 184 natalensis, 175-198 schackoi, 193 sericinus, 185, 193 tropicus, 175-198 truncatus, 175-198 truncatus sericinus, 193, 195, 196 truncatus truncatus, 195 burchi, Doridella, 199, 200, 205-210, 217 burchi, Herviella, 223-230 burnupi, Ancylus, 157 burnupi, Ferrissia, 155-174 burnupi, Gundlachia, 156 MALACOLOGIA Burnupia, 156, 161, 170 caffra, 161, 170 buryaki, Microstele, 244, 246 Bythinia, 5, 14 Caecilioides, 244 caffra, Burnupia, 161, 170 californica, Armina, 215 californica, Pomatiopsis, 1, 2, 13, 109, 110, 124, 134 callosa, Vertigo antivertigo, 244 Caloria, 216 occidentalis, 216 calumniosa, Gastrocopta, 245 calumniosa, Sinalbinula, 245 campestris, Planorbula, 253-265 campestris, Segmentina armigera, 255 Candiella, 199, 211 bayeri misa, 199, 211 bayeri bayeri, 212 canescens, Nemeritis, 313 Cantharus, 311 caperata, Stagnicola, Caracollina, 245 carambola, Corambella, 207 carambola, Doridella, 209 Cardium, 271, 403, 406, 408 edule, 408 Carychium, 244-245, 254 exiguum, 254 plicatum, 244 suevicum, 245 casertanum, Pisidium, Caspicyclotus, 245-247 praesieversi, 246, 247 caucasica, Eostrobilops, 244 caucasica, Microstele, 244, 246 caucasica strigata, Chondrula, 254, 262 254 244, 245 caucasica strigata, Mastus, 244 caucasica, Strobilops, 244, 246 caucasicus, Zootecus insularis, 246 Caucasotachea, 244, 245, 248 andrussovi, 244 fortangense, 244 maslovae, 245 Cavolinia, 301 simplex, 301 chacei, Oncomelania, 134 Chelydra, 308 serpentina, 308 Chilocyclus, 5 244, INDEX, VOL. VI 413 chiui, Oncomelania, 150 chiui, Oncomelania hupensis, 117, 133, 145-153 chiui, Tricula, 117, 133, 145, 150 Chondrula, 244-248 caucasica strigata, 244, 245 forcarti, 244, 246, 247 likharevi, 246 microtraga, 246 microtraga psedachica, 245 microtraga sunzhica, 245 tchetchenica, 245 chondrus, 244 christyi, Planorbula, 255 christyi, Segmentina, 255 Chthamalus, 276 cincinnatiensis, Cyclostoma, 5 cincinnatiensis, Pomatiopsis, 1-5, 13, 14, 66, 109-111, 118, 123-134, 325, cinerea cinerea, Urosalpinx, 269 cinerea follyensis, Urosalpinx, 269 cinerea, Urosalpinx, 267-320 cinerea, Urosalpinx cinerea, 269 Cionella, 254 lubrica, 254 circumstriatus, Gyraulus, 254 claror, Herviella, 223-230 Clausiliidae, 244, 247 Cleioprocta, 200, 228 clifdeni, Ferrissia, 156 Cochlicopa, 244 Cochlicopidae, 244 columbiana, Armina, 215 concentrica, Ervilia, 231-241 concentrica, Mesodesma, 231-241 Conus, ..271, 311 convexa, Crepidula, convolvula, Armina, Corambe, 206-209 evelinae, 207, 208 lucea, 208 pacifica, 206-208 sargassicola, 208 testudinaria, 207, 208 Corambella, 207 baratariae, 207 carambola, 207 depressa, 208 Corambidae, 200 coretus, Anisus, 195 301, 302 215-216 corona, Melongena, 271 costata, Strobilops, 244, 246 crassilabris, Planorbula, 262, 263 Crassostrea, 271-320 virginica, 271-320 crenimargo, Helicella, 245, 246 Crepidula, 271, 301, 302 convexa, 301, 302 fornicata, 271 cronkhitei, Discus, 254 Cryptobranchiata, 199 cuvieri, Armina, 215 Cyclostoma, 5 cincinnatiensis, 5 cylindrica, Truncatellina, 245 dalli, Fossaria, 254 Daudebardia, 245 dautzenbergi, Okenia, 204, 205 dautzenbergi, Okenia elegans, 205 deauratum, Mesodesma, 234 decollata, Rumina, 312 deflectus, Gyraulus, 254 Dendrodorididae, 200 Dendrodoris, 200, 209, 211, 217 atropos, 209 krebsii, 200, 209, 217 Dendronotoidea, 200 dentata, Truncatellina, 245 depressa, Corambella, 208 depressus, Bulinus, 176, 184, 185 depressus, Bulinus hemprichii, 184 derelicta, Doridigitata, 201 Deroceras, 254 aenigma, 254 Deroceras, 388 reticulatum, 388 digueti, Armina, 215 Discus, 254 cronkhitei, 254 Donax, 311, 403, 404 Dondice, 216 occidentalis, 216 Doridella, 199, 200, 205-210, 217 baratariae, 207, 209 batava, 209 burchi, 199, 200, 205-210, 217 carambola, 209 obscura, 207-209 steinbergae, 209 Dorididae, 199 Doridigitata, 201 derelicta, 201 414 MALACOLOGIA Doridinae, 199 Doridoidea, 199 Doriopsilla, 200, 209-211 ате аа, 211 leia, 211 pharpa, 200, 209-211 Doriopsis, 209 krebsii, 209 krebsii pallida, 209 Doris, 199, 201, 202, 217 januarii, 201 ocelligera, 202 verrucosa, 199, 201, 202, 217 Doto, 228 eburneus, Balanus, 296, 306 edule, Cardium, 408 edulis, Mytilus, 271, 292, 296, 299, 300, 310-315 edulis, Ostrea, 271 elatior, Vertigo, 254 electrina, Nesovitrea, 254 elegans dautzenbergi, Okenia, 205 elegans elegans, Okenia, 205 elegans, Okenia, 205 ellipticus, Stylochus, 300 Ellobiidae, 244 Elminius, 271 modestus, 271 Endodontidae, 244 Enidae, 244 Ensis, 404, 407 Eolidacea, 223, 228 Eostrobilops, 244 caucasica, 244 Ephestia, 313 equeefensis, Ancylus, 157 equeefensis, Ferrissia, 157 equeefensis, Gundlachia, 156 erinacea, Ocenebra, 315 Ervilia, 231-241 concentrica, 231-241 Euarminoidea, 200 Euconulus, 254 fulvus, 254 Eudoridoidea, 199 Euxina, 245, 246 somchetica, 245 tschetschenica, 246 Euxinophaedusa, 245, 246 steklovi, 246 volkovae, 246 evelinae, Corambe, 207, 208 evelinae, Herviella, 223-230 evelinae, Muessa, 223 evelinae, Okenia, 205 exacuous kansasensis, Promenetus, 254 exacuous, Promenetus, 262 exigua, Aeolidia, 224 exigua, Herviella, 223-230 exiguum, Carychium, 254 exilis, Stagnicola, 254 externa, Monacha, 244 exustus, Brachidontes, 276, 281, 306 Facalaninae, 200 farcimen, Gastrocopta, 244, 246 farcimen, Sinalbinula, 244 farsica, Quadriplicata, 244 Favorinidae, 200 Ferrissia, 155-174, 254 australica, 168-170 burnupi, 155-174 clifdeni, 156 equeefensis, 157 junodi, 156 rivularis, 168-170 parallela, 254 Shimekii, 170 tarda, 169-170 tenuis, 155, 157, 168-171 Ferussaciidae, 244 Fiona, 200, 216, 217 pinnata, 200, 216, 217 Fionidae, 200 fissidens, Gastropoda, 244, 246 fissidens, Sinalbinula, 244 floridana, Thais haemastoma, 271, 315 follyensis, Urosalpinx cinerea, 269 Fontelicella, 134 robusta, 134 forcarti, Chondrula, 244, 246, 247 forcarti, Mastus, 244, 247 formosana, Blanfordia, 80, 81 formosana, Oncomelania, 3, 145 formosana, Oncomelania hupensis, 1- 143, 149, 150, 327-360 fornicata, Crepidula, 271 forskalii, Bulinus, 190 fortangense, Caucasotachea, 244 Fossaria, 254 dalli, 254 fulvescens, Murex, 315 fulvus, Euconulus, 254 fuscus, Laevapex, 170 INDEX, VOL. VI 415 Gastrocopta, 244-248, 254 acuminata, 244, 246 armifeva, 254 calumniosa, 245 farcimen, 244, 246 fissidens, 244, 246 holzingeri, 254 magna, 244, 246 nouletiana, 244, 246 tappaniana, 254 ukrainica, 244, 246 zamankulense, 245 Gastropoda, 253 gelatinosum, Alcyonidium, 208 ghomfodensis, Scyllaea pelagica, 212 glabrata, Biomphalaria, 112, 166, 167 glabratus, Australorbis, 112, 166 Glycymeris, 404 Goniodorididae, 199 grisella, Meliphora, 313 guernei, Bulinus, 176 gumsiana, Zebrina, 244 Gundlachia, 155, 156, 169-171 burnupi, 156 equeefensis, 156 wautieri, 155, 170, 171 Gyraulus, 195, 254 bicarinatus, 195 civcumstriatus, 254 deflectus, 254 parvus, 254 gyrina, Physa, 312 haemastoma floridana, Thais, 271, 315 haematobium, Schistosoma, 155, 157, 171, 175-198 Haldemanina, 253, 263 hamva, Pleurobranchaea, 200, 201 hamva, Pleurobranchaea hedgpethi, 199 Hawaiia, 244, 246, 254 antiqua, 244, 246 minuscula, 254 hedgpethi hamva, Pleurobranchaea, 199 hedgpethi, Pleurobranchaea, 199, 201 Helisoma, 254 trivolvus, 254 Helicella, 245-248 crenimargo, 245, 246 libidinosa, 245 sunzhica, 245 Helicidae, 244 Helicodonta, 245 Helix, 112, 245, 247 buchi, 245 pomatia, 112 hemprichii depressus, Bulinus, 184 Herviella, 223-230 affinis, 223-230 albida, 223-230 burchi, 223-230 claror, 223-230 evelinae, 223-230 exigua, 223-230 mietta, 223-230 yatsui, 223-230 hinkleyi, Pomatiopsis, 20, 21 hirsutum, Alcyonidium, 208 Histiomena, 216 hohenackeri, Zebrina, 246 holzingeri, Gastrocopta, 254 Horatia, 14 hortensis, Arion, 369-399 hupensis chiui, Oncomelania, 117, 133, 145-153 hupensis formosana, Oncomelania, 1- 143, 149, 150, 327-360 hupensis hupensis, Oncomelania, 117, 129, 133, 149, 150 hupensis nosophora, Oncomelania, 4, 15, 87, 94, 95, 110, 111, 120, 129, 149, 150, 325-360 hupensis, Oncomelania, 3, 81, 97, 112, 118, 132, 133, 149, 321-367 hupensis, Oncomelania hupensis, 327- 4, 98, hupensis quadrasi, Oncomelania, 1,3, 15, 86, 110, 111, 120, 129, 134, 149, 150, 321-360 Hydrobia, 6, 13, 14, 23 ulvae, 6 Hydrobiidae, 1, 4-6, 12-14, 321 Hydrobiinae, 4, 13, 23 hypnorum, Aplexa, 254 Idaliella, 205 amoenula, 205 416 MALACOLOGIA iloktsuenensis, Paragonimus, 145 Imparietula, 245, 246 impexa, Okenia, 205 improvisus, Balanus, 301 insularis caucasicus, Zootecus, 244, 246 integra, Amnicola, 63 mtermedia, Quadriplicata, 245 Juminia, 245-248 iedereri, 245-247 pupoides, 245, 246 januarit, Doris, 201 januarit, Staurodoris, 201 japonica, Blanfordia, 109, 110 japonica, Ocenebra, 271, 291, 305 japonica, Tritonalia, 271 japonicum, Schistosoma, 3, 145-153, 333 jenksii, Planorbula, 262, 263 junodi, Ferrissia, 156 kansasensis, Promenetus exacuous, 254 karaganica, Pupilorcula, 244 Katayama, 81 nosophora, 81 krebsii, Dendrodoris, 200, 209, 217 krebsii, Doriopsis, 209 krebsii pallida, Doriopsis, 209 krebsii, Rhacodoris, 209 kristenseni, Moridilla, 228 labyrinthica, Strobilops, 254 Laevapex, 170 fuscus, 170 Laeviphaedusa, 245, 246 miocaenica, 246 lapidaria, Amnicola, 5 lapidaria, Oncomelania, 1-148, 151 lapillus, Nucella, 311, 315 lapillus, Purpura, 311 lapillus, Thais, 311, 315 Lasaea, 408 leat, Stenotrema, 254 lederi, Jaminia, 245-247 leia, Doriopsilla, 211 Lentorbis, 195 lepida, Vallonia, 244 libidinosa, Helicella, 245 likharevi, Chondrula, 246 Limacidae, 244 Limax, 245 lineata, Monodonta, 66 Lithoglyphus, 66, 71 naticoides, 66, 71 littorea, Littorina, Littoridina, 14 Littorina, 16, 27, 66, 70, 71, 289 littorea, "27, 66,10, и! planaxis, 289 lubrica, Cionella, 254 lucea, Corambe, 208 Lymnaea, 66, 195 natalensis, 195 stagnalis appressa, 66 Mactra, 404 magna, Gastrocopta, major, Armina, 216 Marciella, 223, 227, 228 mietta, 223, 227, 228 Margaritanidae, 401 Margaritifera, 401-410 margaritifera, 401-410 margaritifera, Margaritifera, marginata, Scyllaea pelagica, 212 marinus, Petromyzon, 285 maslovae, Caucasotachea, 245 Mastus, 244-248 caucasica strigata, 244 forcarti, 244, 247 mattheei, Schistosoma, 185 matyokini, Retowskia, 244, 246, 247 mediterranea, Okenia, 199, 205, 217 Melanoides, 195 tuberculatus, 195 Meliphora, 313 grisella, 313 Melongena, 271 corona, 271 27, 66, 70, 71 244, 246 Mercenaria, 404, 406, 407 Mesodesma, 231-241 arctatum, 231, 234 concentrica, 231-241 deauratum, 234 Microstele, 244-246 buryaki, 244, 246 caucasica, 244, 246 wenzi, 244, 246 microtraga, Chondrula, 246 microtraga psedachica, Chondrula, 245 microtraga sunzhica, Chondrula, 245 mietta, Herviella, 223-230 mietta, Marciella, 223, 227, 228 401-410 INDEX, VOL. VI 417 Milax, 369-399 budapestensis, 369-399 milium, Vertigo, 254 minuscula, Нашайа, 254 minutissimum, Punctum, 254 minutum, Opeas, 244, 246 miocaenica, Laeviphaedusa, 246 misa, Candiella bayeri, 199, 211 misa, Tritonia bayeri, 199, 210-212, 217 modesta, Aplysia, 200 modestus, Elminius, 271 mollis, Staurodoris verrucosa, 201 Monacha, 244, 245, 248 externa, 244 praeorientalis, 245 Monodonta, 66 lineata, 66 Moridilla, 228 kristenseni, 228 morio, Aplysia, 199, 200 morio, Уатта, 200 Muessa, 223 evelinae, 223 mülleri, Armina, 215 Murex, 315 fulvescens, 315 Muricidae, 267 muscorum, Pupilla, 254 mutabilis, Pupilla, 244, 246 Mya, 231, 272, 407 nitens, 231 Mytilus, 21,1212, 281% 282," 292, 296, 310-315, 408 edulis, 271, 292, 296, 299, 300, 310- 315 Nassarius, 289 obsoletus, 289 natalensis, Anisus, 195 natalensis, Armina, 216 natalensis, Bulinus, 175-198 natalensis, Lymnaea, 195 natalensis, Physa, 184 Natica, 311 naticoides, Lithoglyphus, 66, 71 Natricola, 134 robusta, 134 Negulus, 244-246 Nemeritis, 313 canescens, 313 Nesovitrea, 244, 246, 254 electrina, 254 petronella, 244, 246 nitens, Mya, 231 nosophora, Katayama, 81 nosophora, Oncomelania, 3, 146 nosophora, Oncomelania hupensis, 4, 15,81, 94, 95, 110,1 1207 129; 149, 150, 325-360 Notaspidea, 199 nouletiana, Gastrocopta, 244, 246 nouletiana, Sinalbinula, 244 Noumeaella, 223, 228 rehderi, 228 Nucetla, 211, 311,.315 lapillus, 311, 315 Nymphaea, 195 obscura, Doridella, 207-209 obsoletus, Nassarius, 289 obtusale, Pisidium, 254 occidentalis, Caloria, 216 occidentalis, Dondice, 216 ocelligera, Doris, 202 Ocenebra, 271, 272, 291, 305, 311, 315 evinacea, 315 japonica, 271, 291, 305 Okenia, 199-205, 217 amoenula, 205 dautzenbergi, elegans, 205 elegans dautzenbergi, 205 elegans elegans, 205 evelinae, 205 impexa, 205 204, 205 mediterranea, 199, 205, 217 plana, 204 sapelona, 199-205, 217 Oleacinidae, 244 Omalodiscus, 254 pattersoni, 254 Oncomelania, 1-148 chiui, 150 formosana, 3, 145 hupensis, 13181, 97, 112. 11977192 133, 149 117, 133, 145-153 1-143, 149, 150 4, 98, 117, 129, 133, 149, 150 hupensis nosophora, 4, 15, 87, 94, 95, 110, 111, 120, 129, 149, 150 hupensis chiui, hupensis formosana, hupensis hupensis, 418 MALACOLOGIA hupensis quadrasi, 1, 3, 15, 86, 110, 111, 120, 129, 134, 149, 150 nosophora, 3, 146 quadrasi, 3 Oncomelania, 321-367 hupensis, 321-367 hupensis quadrasi, hupensis nosophora, hupensis formosana, hupensis hupensis, Opeas, 244-246 246 minutum, 244, 246 Opisthobranchia, 223 orientalis, Scyllaea pelagica, 212 Ostrea, 271 edulis, 271 ovata, Vertigo, 254 ovatula, Vertigo, 244 Oxychilus, 245 Oxyloma, 254 pacifica, Corambe, Pagodulina, 246 Palisa, 228 papillata, 228 pallida, Doriopsis krebsii, 209 palustris, Stagnicola, 262 Paphia, 271 papillata, Palisa, 228 papyria, Amygdalum, 301 Papyrus, 195 Paragonimus, 145 iloktsuenensis, 145 parallela, Ferrissia, 254 Parmacella, 245 Parmacellidae, 244 parvus, Gyraulus, 254 pattersoni, Omalodiscus, 254 pelagica ghomfodensis, Scyllaea, 212 pelagica marginata, Scyllaea, 212 pelagica orientalis, Scyllaea, 212 pelagica, Scyllaea, 200, 210-213, 217 pelagica sinensis, Scyllaea, 212 Petromyzon, 285 marinus, 285 petronella, Nesovitrea, 244, 246 Pettancylus, 155, 169-171 171 Phanerobranchiata, 199 pharpa, Doriopsilla, 200, 209-211 Physa, 184, 254, 312 gyrina, 312 skinneri, 254 321-360 325-360 327-360 327-360 206-208 tropica, 184 natalensis, 184 Physopsis, 190, 195 africanus, 190, 195 abyssinicus, 195 pinnata, Fiona, 200, 206, 217 Pisidium, 254 casertanum, 254 obtusale, 254 plana, Okenia, 204 planaxis, Littorina, 289 Planorbidae, 175, 189, 253 Planorbula, 253-265 armigera, 253-265 campestris, 253-265 christyi, 255 crassilabris, 262, 263 jenksii, 262, 263 wheatleyi, 263 Pleurobranchacea, 199 Pleurobranchaea, 199, 201 hamva, 200, 201 199 hedgpethi, 199, 201 hedgpethi hamva, 199 Pleurobranchidae, 199 Pleurobranchinae, 199 Pleurophyllidia, 216 undulata, 216 Pleuroprocta, 228 plicatum, Carychium, 244 pliocenica, Retowskia schlaeflii, 245, 246 pomatia, Helix, 112 Pomatias, 244-248 vivulare, 244, 245 Pomatiasidae, 244 Pomatiopsidae, 4-6, 13 Pomatiopsinae, 1, 4-6, 10, 12, 13 Pomatiopsis, 1-143, 151, 325, 335 binneyi, 1, 2, 13, 14, 109, 110, 124, 134, 135 californica, 1, 2, 13, 109, 110, 124, 134 chacei, 134 cincinnatiensis, 1-5, 13, 14, 66, 109-111, 118, 123-134, hinkleyi, 20, 21 325, 335 hinkleyi, 20, 21 lapidaria, 1-143, 151 praelonga, 20 INDEX, VOL. VI 419 robusta, 134 Retowskia, 245-247 scalaris, 20, 21 matyokini, 244-247 Pontophaedusa, 245-247 schlaeflii pliocenica, 245, 246 praefuniculum, 246, 247 Rhacodoris, 209 Porostomata, 200 krebsii, 209 praefuniculum, Pontophaedusa, 246, Rissoidae, 4, 6, 23 247 Rissoinae, 6 praelonga, Pomatiopsis, 20 Rissoininae, 6 praeorientalis, Monacha, 245 rivulave, Pomatias, 244, 245 praesieversi, Caspicyclotus, 246, 247 rivularis, Ancylus, 155, 156 Promenetus, 254, 262 vivularis, Ferrissia, 168-170 exacuous, 262 robusta, Amnicola, 134 exacuous kansasensis, 254 robusta, Fontelicella, 134 umbilicatellus, 254, 262 robusta, Natricola, 134 Prosobranchia, 1, 267, 321 robusta, Pomatiopsis, 134 psedachica, Chondrula microtraga, Rumina, 312 245 decollata, 312 psedachica, Tropidomphalus, 245 sandbergeri, Vallonia, 244 pseudoverrucosa, Staurodoris, 203 sapelona, Okenia, 199-205, 217 pulchella, Vallonia, 245, 254 sargassicola, Corambe, 208 Punctum, 254 sayana, Amnicola, 13 minutissimum, 254 scalaris, Pomatiopsis, 20, 21 Pupilla, 244-247, 254 schackoi, Bulinus, 193 belokrysi, 245, 246 Schistosoma, 3, 145-153, 155, 157, 171, muscorum, 254 175-198, 333 mutabilis, 244, 246 bovis, 185 signata, 246 haematobium, 155, 157, 171, 175-198 signataeformis, 244, 246, 247 japonicum, 3, 145-153, 333 submuscorum, 246 mattheei, 185 triplicatoidea, 244 Schistosomophora, 81 Pupillidae, 244 quadrasi, 81 Pupilorcula, 244 schlaeflii pliocenica, Retowskia, 245, 246 karaganica, 244 Scyllaea, 200, 210-213, 217 Pupoides, 254 pelagica, 200, 210-213, 217 albilabris, 254 pelagica ghomfodensis, 212 pupoides, Bollingeria, 245 pelagica marginata, 212 pupoides, Jaminia, 245, 246 pelagica orientalis, 212 Purpura, 311, 312 pelagica sinensis, 212 lapillus, 311 Scyllaeidae, 200 pusilla, Vertigo, 246 Segmentina, 255 quadrasi, Oncomelania, 3 armigera campestris, 255 quadrasi, Oncomelania hupensis, 1, 3, christyi, 255 15, 86, 110, 111, 120, 129, Segmentorbis, 195 134, 149, 150, 321-360 semperi, Armina, 215 quadrasi, Schistosomophora, 81 sericinus, Bulinus, 185, 193 Quadriplicata, 244, 245 serpentina, Chelydra, 308 farsica, 244 Serrulina, 246, 247 intermedia, 245 sieversi, 246, 247 rehderi, Noumeaella, 228 shimekii, Ferrissia, 170 reticulatum, Deroceras, 388 sieversi, Serrulina, 246, 247 reticulatus, Agriolimax, 369-399 signata, Pupilla, 246 420 MALACOLOGIA signataeformis, Pupilla, simplex, Anomia, 301 simplex, Cavolinia, 301 Sinalbinula, 244, 245 calumniosa, 245 farcimen, 244 fisidens, 244 nouletiana, 244 sinensis, Scyllaea pelagica, 212 sirtalis sirtalis, Thamnophis, 308 sirtalis, Thamnophis sirtalis, 308 Skeneinae, 6 skinneri, Physa, 254 Somatogyrus, 5 somchetica, Euxina, 245 Sphaerium, 195 Spirorbis, 311 stagnalis appressa, Lymnaea, 66 Stagnicola, 254, 262 caperata, 254, 262 exilis, 254 palustris, 262 Staurodoris, 201 januarü, 201 pseudoverrucosa, 203 verrucosa, 201 verrucosa mollis, 201 steinbergae, Doridella, 209 steklovi, Euxinophaedusa, 246 Stenotrema, 254 leai, 254 strigata, Chondrula caucasica, 244, 245 strigata, Mastus caucasica, 244 Strobilops, 244-247, 254 caucasica, 244, 246 costata, 244, 246 labyrinthica, 254 ukrainica, 244, 246 Strobilopsidae, 244, 246 Stylochus, 300 ellipticus, 300 subcyclophorella, Vallonia, submuscorum, Pupilla, 246 Subulinidae, 244 Succineidae, 244 Suctoria, 199 sudanica, Biomphalaria, 195 suevicum, Carychium, 245 sunzhica, Chondrula microtraga, 245 sunzhica, Helicella, 245 244, 245 244, 246, 247 tapirulus, Ancylus, 170 tappaniana, Gastrocopta, 254 tarda, Ferrissia, 169, 170 tasmanicus, Ancylus, 156 tchetchenica, Chondrula, 245 Tellina, 311, 404, 407 tenuis, Ferrissia, 155, 157, 168-171 testudinaria, Corambe, 207, 208 Thais; 201, 311,315 lapillus, 311, 315 haemostoma floridana, 271, 315 Thamnophis, 308 sirtalis sirtalis, 308 tigrina, Armina, 216 Tomichia, 14, 109, 133, 135 ventricosa, 109 Tricula, 117, 133, 145, 150 chiui, 117, 133, 145, 150 Triculinae, 10 Trigonochlamididae, 244 triplicatoidea, Pupilla, 244 Tritonalia, 271 japonica, 271 Tritonia, 199, 210-212, 217 bayeri bayeri, 212, 217 bayeri misa, 199, 210-212, 217 Tritoniidae, 200 trivolvis, Helisoma, 254 tropicus, Bulinus, 175-198 tropica, Physa, 184 Tropidomphalus, 245 psedachica, 245 Truncatella, 5, 6, 12 Truncatellina, 244, 245 cylindrica, 245 dentata, 245 Truncatellinae, 4-6, 12, 14 truncatus, Bulinus, 175-198 truncatus, Bulinus sericinus, truncatus sericinus, Bulinus, tschetschenica, Euxina, 246 tuberculatus, Melanoides, 195 Typha, 134 ukrainica, Albinula, 244 ukrainica, Gastrocopta, ukrainica, Strobilops, ulvae, Hydrobia, 6 umbilicatellus, Promenetus, 254, 262 undulata, Armina, 199, 213-216 undulata, Pleurophyllidia, 216 Unionacea, 401 175-198 175-198 244, 246 244, 246 INDEX, VOL. VI Urosalpinx, 267-320 cinerea, 267-320 cinerea cinerea, 269 cinerea follyensis, 269 Vallonia, 244, 245, 254 cyclophorella, 254 gracilicosta, 254 lepida, 244 pulchella, 245, 254 sandbergeri, 244 subcyclophorella, Valloniidae, 244 vancouverensis, Aymina, 215 Varria, 200 morio, 200 ventricosa, Tomichia, 109 verrilli, Alcyonidium, 205, 208 verrucosa, Doris, 199, 201, 202, 217 verrucosa, Staurodoris, 201 verrucosa mollis, Staurodoris, 201 Vertigo, 244-246, 254 angulifera, 244 angustior, 245 antivertigo antivertigo, antivertigo callosa, 244 elatior, 254 milium, 254 ovata, 254 ovatula, 244 pusilla, 246 244, 245 245, 246 Vertigopsis, 246 Vertilla, 244, 245 angulifera, 244, 245 virginica, Crassostrea, 271-320 Vitrinidae, 244 Viviparus, 66 viviparus, 66 viviparus, Viviparus, 66 volkovae, Euxinophaedusa, 246 Watsonula, 155, 156, 170, 171 wautieri, 155, 156, 170, 171 wattla, Armina, 199, 213-217 wautieri, Gundlachia, 155, 170, 171 wautieri, Watsonula, 155, 156, 170, 171 wenzi, Microstele, 244, 246 wheatleyi, Planorbula, 263 yatsui, Herviella, 223-230 zamankulense, Gastrocopta, 245 Zebrina, 244-248 gumsiana, 244 hohenackeri, 246 Zonitidae, 244 Zonitoides, 254 arboreus, 254 Zootecus, 244-246 Zostrea, 301-302 421 18 NO ENT) RATE UE и эль } u "MM ape uf b WE re ar] d Arey PL on fee OS INIA к ORE- VE А ENS bx . | LBS à Lil VE 797 tri AU | de DU PET SELLER y de SMS Et LE 99 ии ise AO QE wy 2 к ope Pere MOUNTS: peon) и ке OTL we) NS A AS POE OTE COT о Аи и a pis Ris or mate fs OM amt 1er Ori al Ber. bre 4 bes e res) YO Se via eit Py) NT AU | | во” Аа и PES SSS мук een BOX: ¿JOE AGS: widowed Froth ged PLE SOX 208 wee” Sess ju зи 103 AVTODON ) i a PAC NL > ii ves "DONA. paola #08 ¿ION | ss BrS-ME eins beh , de q | LAN ) и NPA = А А rMoioatenor ddl PES. chtis aes 01, “À ы * à" LI E i" TER POS ahnen y PI к à у фа. ROUND mn PIS-PRE- .ewoatood wn Rh ew Ov ut WE SSR foe ATT Mar, i \ 4 р 5 / 1 10 Y р I 11d Mn 24 ay Ss \ Ai bi рых, м 4 rt, в А à Y / by 7 E N M Al ih 2; y M | N | a (de arc га >. à Ка af i rica la a nee de al > у EURE й fe Agni 1x Arch, a | a u 1 N rie WW: 4 ‘es want? “im a y Le у \ f 2 | И AE phi yaoi ‘ Pr tr: Ex к ir pl Мой 3 if vus, МОИ, 1 ar L р A wr. бе x un, AA nl VONT à des TEL PROFIS pr 1: er wtp LU x DY A ES M; и IDO ON 40 BOUND сот 1970 HU LL MIN 6 Lo in XP y æ AT ELLE PRET EEE > Fee ABAD DADS CUS LEE A kb ; E | = FE ew ore e Dm ——=---- OO ALA A a AA Dr 2 à A ee ae 2