UN 46 008 0 à Yo Fe Re am. He Pn ESEL ALOE AS Wee in fons rer FA ok Note AO об Pe NA MAI de MANE уве a И. й A nett Мар Bathe e ere ме EEE AN Mine A им бони ом ана a à € a EU 4 Em овчины ве, » ONE M MON A ANI AT AG LAN A бей, La A, mt ADA bu BENE ne и EVA AR MON cn ek Ua ETAPE ne ni Ah EAE VA A A MED AD lee SI 18 Ai MAL ua ION ARR RQ enn A te AE вет éd д mm NE A PA Ra Ms Dr AAN EL EEE NA nm teed AA НИЯ EC PTE Pa REN nA ORL be Rm лид nn, A nd Is A A A киев еже nm не мои eden e, A nn a ta Tn nl ofa EN ar en af, Nan ON WEA a ann am an hast gents ATED A mn ani CAP mnt mt adh nll rm PE en aR A nn eit and dont mel at wn! pation An it POM ene ne mtn AN a e A ee UE En dah oe tn О dt ann ot вр mimi enter A MAN EN RENE вол у ель a боны COUT ры Пя ВИ ae un Pla A MLA д ob Ale OMe A NOS EEE namen DL te in NA A Bek att AI Meyer A dds AAA fads пресное he une, AA Mee Tee Se Ser pha tnt Bae Cor re rat er licita. LAN joo 2. г Au, ANA toed hoor, AAA ees DCE PET Ma 07 IMAN A Аа, em. Cn AD DA ra ema eked, AA RA hy tem. ar AAA EE Eee TETE DDR NE Mc ne D LEITER SE er ann À ai Vd CI id ma U EA EE NES is y im О ве AAA Ra. En nn on, Am a ном es Mam ae DOTE te Dan x са, Pa nn Taten me A D nn DCE TE) > N een es NANA RH om nan TE ETES Verla mas oe Ann nh Poe Stns han, an mas rea dee in anne AA cm nm Da ae VOL. 33 1991 MALACOLOGIA iernational Journal of Malacology Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift ——- Vol. Vol. Vol. Vol. Vol. Vol. Vol. Publication dates ‚No . 1-2 ut a2 . 1-2 E 2 42 19 January 1988 28 June 1988 16 Dec. 1988 1 Aug. 1989 29 Dec. 1989 28 May 1990 7 June 1991 MALACOLOGIA, VOL. 33 CONTENTS ULRICH BIELEFELD Histological Observation of Gonads and Digestive Gland in Starving Dreissena polymorpha (BivalVia) =. ae TU MARCO BODON & FOLCO GIUSTI The Genus Moitessieria in the Island of Sardinia and in Italy. New Data on the Systematics of Moitessieria and Paladilhia (Prosobran- chia: Hydrobiidae) (Studies on the Sardinian and Corsican Malaco- TAU N ee CR ee THERESE D. BRACKENBURY & C. C. APPLETON Morphology of the Mature Spermatozoon of Bulinus tropicus (Krauss, 1848). (Gastropoda®: Planorbidae) иене L. M. COOK Fluctuations and Immobility in Age-Structured Snail Populations ....... ROBERT T. DILLON, JR. Karyotypic Evolution in Pleurocerid Snails Il. Pleurocera, Goniobasis, ENG et Yon. atin eens E EER ee Rian EN KENNETH C. EMBERTON The Genitalic, Allozymic and Conchological Evolution of the Tribe Mesodontini (Pulmonata: Stylommatophora: Polygyridae) .............. RICHARD S. HOUBRICK Systematic Review and Functional Morphology of the Mangrove Snails Terebralia and Telescopium (Potamididae; Prosobranchia) ............ C. DAVID ROLLO Endogenous and Exogenous Regulation of Activity in Deroceras retic- ulatum, A Weather-Sensitive Terrestrial Slug ........................... BARRY ROTH Tropical “Physiognomy” of a Land Snail Faunule From the Eocene of SO аи eae MARTIN SPRUNG Costs of Reproduction: A Study on Metabolic Requirements of the Gonads and Fecundity of the Bivalve Dreissena polymorpha .......... MARTIN SPRUNG & JOST BORCHERDING Physiological and Morphometric Changes in Dreissena polymorpha (Mollusca; Bivalvia) During a Starvation Period ......................... MICHELE K. SURBEY & C. DAVID ROLLO Physiological and Behavioural Compensation for Food Quality and Quantity in the Slug Lehmannia marginata ............................. RITA TRIEBSKORN The Impact of Molluscicides on Enzyme Activities in the Hepatopan- creas of Deroceras reticulatum (Müller)... neces. J. P. POINTIER, J. L. TOFFART & M. LEFEVRE Life Tables of Freshwater Snails of the Genus Biomphalaria (B. glabrata, B. alexandrina, B. straminea) and of One of its Competitors Melanoides tUbérculata Under Laboratony{Conditions.:..2...... нонььь 31 273 55 339 71 289 199 281 63 179 193 255 43 JANET R. VOIGHT Morphological Variation in Octopod Specimens: Reassessing the As- sumption of Preservation-Induced Deformation ......................... 241 G. THOMAS WATTERS Utilization of a Simple Morphospace by Polyplacophorans and Its Evo- lutionary Implications .................................................... 221 OL. 33, NO. 1-2 er as | SEP 11 1991 HARVARD UNIVERSITY IMALACOLOGIA ernational Journal of Malacology Revista Internacional de Malacologia ie 4 Br); LAS y _ Journal International de Malacologie 4 | _ Международный Журнал Малакологии Internationale Malakologische Zeitschrift MALACOLOGIA Editor-in-Chief: GEORGE M. DAVIS Editorial and Subscription Offices: Department of Malacology The Academy of Natural Sciences of Philadelphia 1900 Benjamin Franklin Parkway Philadelphia, Pennsylvania 19103-1195, U.S.A. Co-Editors: EUGENE COAN CAROL JONES California Academy of Sciences Payson, AZ San Francisco, CA Assistant Managing Editor: CARYL HESTERMAN Associate Editors: BE ВО ANNE GISMANN University of Michigan Maadi Ann Arbor Egypt of which (also serving as editors) are: KENNETH J. BOSS JAMES NYBAKKEN, President Museum of Comparative Zoology Moss Landing Marine Laboratory Cambridge, Massachusetts California JOHN BURCH, President-Elect CLYDE: FEO ROPER MELBOURNE R. CARRIKER Smithsonian Institution Washington, D.C. W. D. RUSSELL-HUNTER Syracuse University, New York SHI-KUEI WU | University of Colorado Museum, Bo University of Delaware, Lewes GEORGE M. DAVIS Secretary and Treasurer CAROLE S. HICKMAN, Vice-President University of California, Berkeley ee Participating Members EDMUND GITTENBERGER JACKIE L. VAN GOETHEM Secretary, UNITAS MALACOLOGICA Treasurer, UNITAS MALACOLOGI Rijksmuseum van Natuurlijke Koninklijk Belgisch Instituut Historie voor Natuurwetenschappen Leiden, Netherlands Brussel, Belgium я Emeritus Members J. FRANCIS ALLEN, Emerita ROBERT ROBERTSON j Environmental Protection Agency The Academy of Natural Sciences — Washington, D.C. Philadelphia, Pennsylvania j ELMER G. BERRY, Ä NORMAN F. SOHL Germantown, Maryland U.S. Geological Survey Reston, Virginia Copyright © 1991 by the Institute of Malacology J. A. ALLEN Marine Biological Station Millport, United Kingdom E. E. BINDER Muséum d'Histoire Naturelle Geneve, Switzerland A. J. CAIN University of Liverpool United Kingdom P. CALOW University of Sheffield United Kingdom A. H. CLARKE, Jr. Portland, Texas, U.S.A. B. C. CLARKE University of Nottingham United Kingdom R. DILLON College of Charleston SC, U.S.A. C. J. DUNCAN University of Liverpool United Kingdom V. FRETTER University of Reading United Kingdom E. GITTENBERGER Rijksmuseum van Natuurlijke Historie Leiden, Netherlands F. GIUSTI Universita di Siena, Italy A. N. GOLIKOV Zoological Institute Leningrad, U.S.S.R. S. J. GOULD Harvard University Cambridge, Mass., U.S.A. 1991 EDITORIAL BOARD A. V. GROSSU Universitatea Bucuresti Romania T. HABE Tokai University Shimizu, Japan R. HANLON Marine Biomedical Institute Galveston, Texas, U.S.A. J. A. HENDRICKSON, Jr. Academy of Natural Sciences Philadelphia, PA, U.S.A. K. E. HOAGLAND Association of Systematics Collections Washington, DC, U.S.A. B. HUBENDICK Naturhistoriska Museet Góteborg, Sweden S. HUNT University of Lancaster United Kingdom R. JANSSEN Forschungsinstitut Senckenberg, Frankfurt am Main, Germany R. N. KILBURN Natal Museum Pietermaritzburg, South Africa M. A. KLAPPENBACH Museo Nacional de Historia Natural Montevideo, Uruguay J. KNUDSEN Zoologisk Institut & Museum Kobenhavn, Denmark A. J. KOHN University of Washington Seattle, U.S.A. A. LUCAS Faculté des Sciences Brest, France C. MEIER-BROOK Tropenmedizinisches Institut Tubingen, Germany H. K. MIENIS Hebrew University of Jerusalem Israel J. E. MORTON The University Auckland, New Zealand J. J. MURRAY, Jr. University of Virginia Charlottesville, U.S.A. R. NATARAJAN Marine Biological Station Porto Novo, India J. OKLAND University of Oslo Norway T. OKUTANI University of Fisheries Tokyo, Japan W. L. PARAENSE Instituto Oswaldo Cruz, Rio de Janeiro Brazil J. J. PARODIZ Carnegie Museum Pittsburgh, U.S.A. W. F. PONDER Australian Museum Sydney В. D. PURCHON Chelsea College of Science & Technology London, United Kingdom OZ. У. Academia Sinica Qingdao, People’s Republic of China N. W. RUNHAM University College of North Wales Bangor, United Kingdom S. G. SEGERSTRALE Institute of Marine Research Helsinki, Finland F. STARMUHLNER Zoologisches Institut der Universitat Wien, Austria У. |. STAROBOGATOV Zoological Institute Leningrad, U.S.S.R. W. STREIFF Université de Caen France J. STUARDO Universidad de Chile Valparaiso S. TILLIER Muséum National d'Histoire Naturelle Paris, France R. D. TURNER Harvard University Cambridge, Mass., U.S.A. J. A. VAN EEDEN Potchefstroom University South Africa N. H. VERDONK Rijksuniversiteit Utrecht, Netherlands B. R. WILSON Dept. Conservation and Land Management Netherlands, Western Australia H. ZEISSLER Leipzig, Germany A. ZILCH Forschungsinstitut Senckenberg Frankfurt am Main, Germany MALACOLOGIA, 1991, 33(1-2): 1-30 THE GENUS MOITESSIERIA IN THE ISLAND OF SARDINIA AND IN ITALY. NEW DATA ON THE SYSTEMATICS OF MOITESSIERIA AND PALADILHIA (PROSOBRANCHIA: HYDROBIIDAE) (STUDIES ON THE SARDINIAN AND CORSICAN MALACOFAUNA, IX)'? Marco Bodon & Folco Giusti Dipartimento di Biologia Evolutiva, Universita di Siena, Via Mattioli 4, 1-53100 Siena, Italy ABSTRACT The recent discovery of the genus Moitessieria in subterranean waters of the island of Sar- dinia provided the opportunity to increase knowledge of the genus in Italy and to revise all the populations known to live in western Liguria (Ligurian and Maritime Alps, Italy). Shell (shape and teleoconch microsculpture) and anatomical characters have been studied in many Italian pop- ulations and compared with those of some populations of the following French species: M. simoniana (De Charpentier), M. massoti Bourguignat, M. rayi (Bourguignat), M. juvenisanguis Boeters & Gittenberger, and M. lescherae Boeters. Shell structure was also compared with the following French species: M. vitrea (Draparnaud), M. puteana Coutagne, M. rollandiana Bour- guignat, and M. locardi Coutagne. The results suggest that the Ligurian populations belong to M. simoniana, and those from Sardinia are closely related to M. massoti. The latter is herein recognized as a species anatomically and conchologically well distinguished from the “M. si- moniana group of forms.” The anatomical study (body, genitalia and radula) of the different Moitessieria species herein considered, and of the type species of genus Paladilhia (P. pleuro- toma Bourguignat), provided results that do not justify the inclusion of both genera in a family, Moitessieriidae, distinct from the Hydrobiidae. Even if Moitessieria and Paladilhia differ in many details, the scheme of their genitalia is basically the same and corresponds to that in the Hydrobiidae. Moitessieriidae is consequently recognized as a junior synonym of Hydrobiidae. The presence in Sardinia of Moitessieria, a genus untill now considered proper to southwestern Europe (Spain, southern France, northwestern Italy), is interpreted as additional evidence the Sardo-Corsican complex migrated from the southwestern border of Palaeo-Europe by mi- croplate drift during the middle Tertiary. Key words: Hydrobiidae, Moitessieria, Paladilhia, France, Italy, Sardinia, systematics, bioge- ography. INTRODUCTION The census program for the stygobiont spe- cies of Gastropoda in Italy has led to the col- lection of many taxa, some of which are un- described and others new to the Italian fauna. These undescribed taxa are at present under study. This paper is devoted to the genus Moitessieria Bourguignat, 1863, reported to date in only a few sites of northwestern Italy (Ligurian Alps) (Bodon, 1980; Giusti & Pez- zoli, 1982a, 1982b; Boato et al., 1985; Bodon & Pezzoli, 1986). Recent research has also revealed its presence in other parts of north- western Italy (Maritime Alps) and in Sardinia. Careful anatomical study of the Italian popu- lations, certain species living in southern France, and of French Paladilhia pleurotoma Bourguignat, has furnished new data en- abling the revision of the status of the family Moitessieriidae Bourguignat, 1863. The latter is here recognized to be a junior synonym of Hydrobiidae Troschel, 1857. MATERIAL AND METHODS Collecting Empty shells and whole specimens were collected by selecting variable amounts of sediment from the inside of springs and sub- terranean streams (locality data follows spe- ‘Research supported by a CNR grant (Gruppo di Biologia Naturalistica) and by МР! 40% and MPI 60% grants. “This paper is dedicated to the Sardinian people as a sign of appreciation of the many manifest and hidden beauties of their island and in gratitude for their kindness and hospitality. 2 BODON & GIUSTI cies descriptions) and immediately plunged in 95% ethanol. Anatomy Unrelaxed material, preserved in 75% eth- anol, was studied by optical microscopy (Wild M5A). Bodies were isolated after crushing the shell, then dissected using very thin, pointed watchmaker's forceps. Images of entire bod- ies or isolated portions of the genitalia were drawn using a Wild camera lucida. Critical point drying, gold sputter-coating and scan- ning electron microscopy (Philips 505), in search of magnified details of the genitalia of females was also attempted without results. Radulae were manually extracted from buccal bulbs then washed in pure 75% ethanol, mounted on copper blocks with electrocon- ductive glue, sputter-coated with gold and photographed using a Philips 505 SEM. Shell Morphology All the parameters, maximum shell height, last whorl height, last whorl diameter (mouth excluded), mouth height, were merasured in variable numbers of shells from various local- ities (for numbers and locality data see cap- tions to Figs. 6-7), using a Wild M5A micro- scope and a millimetric lens. The first three parameters were utilized for biometric analy- sis to demonstrate differences in shell shape. Mouth parameters were not considered be- cause of the wide variability in shape and di- mensions of the peristome (more or less re- flexed and sinuous). At the same time, the number of spiral ridges on a portion of the last whorl just above the umbilicus and crossing a perpendicular line (relative to the spiral ridges) 100 рт in length, was counted to de- termine possible differences in the external shell ornamentation in the various popula- tions (for numbers and localities, see caption to Fig. 8). Entire shells and shell surface de- tails were photographed under optical and scanning electron microscopes using the pro- cedures described above for the radula. SYSTEMATIC ACCOUNT Hydrobiidae Troschel, 1857 Hydrobiinae Troschel, 1857 Moitessieria Bourguignat, 1863 Type species: Paludina simoniana De Charpentier, 1848. Moitessieria simoniana (De Charpentier, 1848) Paludina simoniana De Charpentier, in De Saint Simon, 1848: 39. Type locality: alluvium of the Garonne, above Toulouse, France. Paludina simoniana—Küster, 1852: 58-59, pl. 11, figs. 9-10. Moitessieria simoniana—Bourguignat, 1863: 15: 440—444. Moitessiera simoniana—Coutagne, 1883: 130, 144-145. Moitessieria simoniana—Coutagne, 1884: 107- 108, pl. 3, fig. 7. Paludinella (Moitessieria) simoniana—Westerlund, 1886: 49. Moitessieria saint-simoni (sic!)—Germain, 1931: 660—662, fig. 726 [partim]. Moitessieria sp.—Boeters, 1973: 64-65, figs. 2-4 [partim ?]. : Moitessieria simoniana,—Bodon, 1980: 1-5, pl. 1, figs. 1-8. Moitessieria cf. simoniana—Giusti & Pezzoli, 1982a: 438. Moitessieria cf. simoniana—Giusti & Pezzoli, 1982b: 463. Moitessieria cf. simoniana—Boato et al., 1985: 258-259. Moitessieria cf. simoniana—Bodon & Pezzoli, 1986: 307, pl. 1, figs. 17-18. Description of Italian Material Shell (Figs. 1F-L, 2A-E, ЗА, В) very small, elongate cylindro-conical in shape, glossy and transparent when fresh. Spire of 5.5-6.5 convex whorls; sutures deep. External sur- face of protoconch (as seen by SEM) with very thin spiral ridges and very finely pitted; external surface of teleoconch with close, thin spiral and radial ridges giving overall net-like microsculpture, with squares of variable di- mensions and shape. Spiral ridges always more evident than radial ridges; both kinds of ridges may be reduced or absent. Teleoconch surface interrupted in the spaces between two adjacent spiral ridges or inside the squares by pits of highly variable diameter (Fig. 2C, E). Portions of teleoconch surface sometimes show rows of wide cup-like pits (Fig. 3A, B); 3.5-6.7 (79 specimens exam- ined) spiral ridges or rows of pits cross a per- pendicul bar line of 0.1 mm on surface of last whorl just over umbilicus. Umbilical opening narrow; aperture ovate, peristome simple, sub- acute, little reflexed in correspondence with columellar area and with sinuous external margin. Dimensions—Shell height: 1.14—2.10 тт; shell diameter: 0.43—0.65 mm; mouth height: 0.35—0.49 mm. MOITESSIERIA AND PALADILHIA SYSTEMATICS FIGURE 1. Shells of Moitessieria simoniana (De Charpentier) from France (A-E) and Italy (Liguria) (F-L) and of M. massoti Bourguignat from France (M-Q). A-D: floods of the river Garonne at Toulouse (Haute Garonne), from Paulucci Coll.; E: Resourgence de Moulis (Moulis, St. Girons, Ariège); F—I: cave of Rio di Nava no. 911 Li (Pornassio, Imperia); L: spring of the waterworks of Buggio (Pigna, Imperia); M-Q: Fontaine Saline de Fouradade, near Tautavel (Pyrénées Orientales). BODON & GIUSTI FIGURE 2. Shell and details of shell surface in Moitessieria simoniana (De Charpentier) from the cave of the Rio di Nava (Pornassio, Imperia, Italy). A: an entire shell (scale bar = 1 mm); B: the protoconch of the same shell (scale bar = 0.1 mm); C: detail of the shell surface of the same shell (scale bar = 0.1 mm); D: shell surface near the umbilicus of the same shell (scale bar = 0.1 mm); E: shell surface at the end of the last whorl of another shell from the same site (scale bar = 0.1 mm). Note in C and E areas with rows of pits. MOITESSIERIA AND PALADILHIA SYSTEMATICS 5 FIGURE 3. Details of shell surface of a specimen of a peculiar population of Moitessieria simoniana (De Charpentier) collected in the cave “Тапа da Fontana do Boro” no. 220 Li, (Toirano, Savona, Italy) (А-В) and of a specimen of Moitessieria rollandiana Bourguignat collected near Montpellier (Hérault, France) (C—D). A: shell surface near the end portion of the last whorl; note the distinct pits that cover almost all the shell surface near the mouth; B: detail of the surface of the penultimate whorl; note the pits that are large and deep near the suture and small in the centre; C: shell surface near the end portion of the last whorl; note the pits that cover entire the surface of the shell and that merge into grooves near the mouth; D: detail of the last whorl surface (scale bars = 0.1 mm). 6 BODON & GIUSTI Operculum corneous, thin, oligogyrous, with subcentral nucleus (Fig. 4C). Genital tract of males (7 specimens exam- ined) with testis, slender convoluted spermi- duct, pear-shaped prostatic gland, slender vas deferens, and copulatory organ (penis) (Fig. 4B, D, E). Prostate gland scarcely bulg- ing into mantle cavity; vas deferens arises from anterior end of prostate, crosses dorsal wall of body and enters penis, running down its length to the tip. Penis colourless, con- tained inside pallial cavity, sinuous and often bent upon itself (Fig. 4B, D); of conical-de- pressed shape and more or less elongate; sides slightly corrugated, tip pointed; an ap- parently non-glandular excrescence on right side about half way along penis. Genital tract of females (3 specimens ex- amined) (Fig. 4 A, F) with ovary, renal oviduct and pallial oviduct. Renal oviduct widens slightly near distal end and gives rise to a vesicle of variable width situated near initial portion of small loop. Short, narrowed portion of oviduct connects vesicle to short, widened portion which meets duct of bursa copulatrix. Vesicle content non-refracting, whitish mate- rial of different appearance from sperm mass. Refractivity is visible in next widened portion of oviduct, which presumably acts as seminal receptacle. Bursa copulatrix sac-like in shape and with long duct arising from posterior por- tion. Anterior part of bursa adheres to albu- men gland. Short canal arises where bursa copulatrix duct joins oviduct entering uterine complex level with albumen gland. Seminal groove seems to run all along ventral side of uterine complex. Genital opening lies close to distal apex of uterus. Presence or absence of gonopericardial duct impossible to ascertain; thin tissue stripes at initial portion of oviduct loop may be gonopericardial duct residues: Other anatomical characters show in Fig- ure 4A-F. Body unpigmented; pigmented eyes ab- sent. Stomach lacking caecal diverticulum. Initial portion of intestine lying far from style- sac. Intestine forming only one loop near stomach; rectal portion of intestine long. Anus lying close to pallial margin; pallial tentacle absent. Ctenidium absent. Osphradium oval or reniform. Radula (Fig. 5B—D) taenioglossate; central teeth with usual butterfly-shaped hydrobiid structure; apex V-shaped, cutting edge with 15-17 denticles, central longer than lateral (which progressively diminish); long, slender lateral wings on each side. Internal side of base of each lateral wing with long, pointed basal cusp. Body of tooth extends downwards to form plough-shaped structure wedged in curved apex of next central tooth. Lateral teeth with elongate body and widened apex having row of 13-15 denticles, central denti- cle longer. Inner marginal teeth rake-shaped; elongated apex with long row of more than 20 denticles on anterior side. Outer marginal teeth similar in shape to inner marginal teeth but apex smaller, spoon-shaped, with many small thin denticles on latero-posterior side. Collecting Sites To the sites already known in the literature (Ligurian Alps: Bodon, 1980; Boato et al., 1985), we add the following in the Maritime Alps (northwestern Italy): Piedmont: springs of Cialombard, on the right side of the river Gesso (Valdieri, province of Cuneo, 2 shells); Liguria: spring on the road between Calvo and Torri, Bevera Valley (Ventimiglia, prov- ince of Imperia, 1 shell). Specimens complete with soft parts were collected in the spring of the cave of Rio di Nava (province of Imperia). Shells examined are kept in the collection of M. Bodon; dissected material in the collec- tions of M. Bodon and F. Giusti. Discussion Comparison of the shell characters of the Italian populations with the best of the classic descriptions of typical Moitessieria simoniana by Küster (1852) (according to Bourguignat, 1863, based on typical materials sent by A. De Saint Simon to J. De Charpentier: “testa minutissima, rimata, cylindracea, nitida, vit- rea, obsoletissime striata, lineis spiralibus densissimis cincta; spira elata, obtusa, an- fractibus 8 convexis; apertura ovata; peristo- mate patulo, subacuto, margine columellari reflexiusculo”) and with the characters de- duced from analysis of topotypical shell ma- terials kept in the Paulucci Collection (Figs. 1A-D, 5A-C) revealed almost total coinci- dence. The Italian material has a generally more widely spaced microsculpture, and the last whorl is a little higher in relation to shell height. Such differences nevertheless appear to be of little value when other populations from France recognized to belong to M. simo- niana (cf. Figs. 1E, 6, 8; Table 1) are consid- ered. The different number of whorls (8 according MOITESSIERIA AND PALADILHIA SYSTEMATICS 7 05mm FIGURE 4. Moitessieria simoniana (De Charpentier) from the spring of the cave of the Rio di Nava (Pornassio, Imperia, Italy) (A-F), and from the Resourgence de Moulis (Moulis, St. Girons, Ariege, France) (G-M). A: female (on the left) and a male (on the right) both removed from shell. Note the absence of ocular spots and of pigment and the oblique packing of the fecal pellets; B, G: male with the pallial cavity open. Note the penis bent upon itself; C: operculum (outer side); D, H: penes; E: genital tract and pallial organs of male; F: genital tract of female; |: operculum (outer side); |: stomach and pallial organs of male; M: genital tract of a young female. Explanation of the symbols used in Figs. 4, 9, 12, 16: ag, albumen gland; bc, bursa copulatrix; c, ctenidium; cg, capsule gland; ex, penial excrescence; fp, fecal pellets; he, digestive gland (= hepatopancreas); ho, hook; i, intestine; |, oviduct loop; od, oviduct; op, operculum; os, osphradium; ov, ovary; pe, penis; pr, prostate gland; pt, pallial tentacle; pw, posterior wall of pallial cavity; sp, spermiduct; sr, seminal receptacle; st, stomach; sy, style sac; te, tentacles; ts, testis; ut, uterus; vd, vas deferens. 8 BODON & GIUSTI y ws, N | x - Be » FIGURE 5. Details of shell surface of a specimen of Moitessieria simoniana (De Charpentier) collected in the floods of the river Garonne at Toulouse (Haute Garonne, France) (Paulucci Coll.) (A,C); central portion of the radula in two specimens of Moitessieria simoniana (De Charpentier) from the cave of the Rio di Nava (Pornassio, Imperia, Italy) (B,D); details of shell surface of a specimen of Paladilhia pleurotoma Bourguignat collected in the Source du Lez (Prades le Lez, Hérault, France) (E-F). A: shell surface near the umbilicus (scale bar = 0.1 mm); C: detail of shell surface (scale bar = 0.1 mm); B: the central teeth are clearly visible (scale bar = 10 pm); D: central teeth (tooth shape is distorted for technical reasons) (scale bar = 10 um); E: the surface of the protoconch and first whorls is almost smooth (scale bar = 0.1 mm); F: detail of shell surface near the umbilicus; most of the surface is smooth; a few areas have lines of very thin ridges (arrow) (scale bar = 0.1 mm). MOITESSIERIA AND PALADILHIA SYSTEMATICS 9 H/b 4.0 3.5 3.0 2,9 2.0 1.0 1.5 20 h/b FIGURE 6. Biometric analysis of the shells of M. simoniana (De Charpentier) from France and Italy (1-4) and of M. massoti Bourguignat from France (5). H, shell max. height; h, last whorl height; b, last whorl diameter (mouth excluded). The central point in each figure represents the medium value, the perimeter includes the extreme values in each population. 1: floods of river Garonne at Toulouse (Haute Garonne, France), from Paulucci Collection (6 specimens); 2: well at the thermal baths of Ginoles (Aude, France) (22 specimens); 3: Resourgence de Moulis (Moulis, St. Girons, Ariége, France) (2 specimens); 4: cave of the Rio di Nava no. 911 Li (Pornassio, Imperia, Italy) (30 specimens); 5: Fontaine Saline de Fouradade, near Tautavel (Pyrénées Orientales) (30 specimens). 10 BODON & GIUSTI H/b 4.0 3.5 3.0 25 2.0 1.0 1.5 2.0 h/b FIGURE 7. Biometric analysis of the shells of M. cf. massoti Bourguignat from Sardinia (1-6) and of M. massoti Bourguignat from France (7). H, max. height; h, last whorl height; b, last whorl diameter (mouth excluded). The central point in each figure represents the medium value, the perimeter includes the extreme value in each population. 1: Cave “Cane Gortoe” in Siniscola (Nuoro) (17 specimens); 2: spring on the right side of the river Cedrino upstream from Orosei (Nuoro) (15 specimens); 3: spring Su Cologone, S. Giovanni (Oliena, Nuoro) (3 specimens); 4: spring in the park of the Marquis of Laconi at Laconi (Nuoro) (20 specimens); 5: spring of the cave “Зи Anzu” near the convent of $. Giovanni (Dorgali, Nuoro) (1 specimen); 6: spring below the cave “Su Marmuri” near Ulassai (Nuoro) (17 specimens); 7: Fontaine Saline de Fouradade, near Tautavel (Pyrénées Orientales, France) (30 specimens). | MOITESSIERIA AND PALADILHIA SYSTEMATICS 11 to Küster) is an error due to Kuster’s incorrect way of numbering whorls (cf. Kuster, 1852: pl. 11, fig. 10). The absence of marked differences be- tween Italian and French populations was con- firmed by anatomical research carried out on French material (Resourgence de Moulis, Ar- iége, ex H. Boeters Coll., and material col- lected by M. Bodon) (Fig. 4G-M) and on a series of specimens collected in an Italian lo- cality (spring of the cave of Rio di Nava, prov- ince of Imperia, Liguria) (Fig. 4A-F). The fe- male genital tracts were identical. A small difference was, however, noted in the males: the excrescence that lies on the right side of the penis appears to lie nearer to the penis tip in the French males. Although the pictures of the radular teeth of French specimens are not particularly clear, it seems possible to exclude any difference in tooth structure and number of apical denticles. Because no typical material was avail- able for anatomical examination, it was im- possible to compare the Italian material with the two French “species” considered in the past to belong to the “M. simoniana group of forms” by Coutagne (1883). Two of these “forms” were later made junior synonyms of M. simoniana by Germain (1931): M. fagoti Coutagne (1883) (type locality: alluvium of the Garonne at Toulouse) and M. bourguignati Coutagne (1883) (type locality: alluvium of the Garonne at Toulouse). Comparison was made with another of the Moitessieria species considered by Coutagne (1883) to belong to the “M. simoniana group of forms”: M. massoti Bourguignat, 1863 (type locality: saline spring of Fouradade, near Tau- tavel, Pyrénées Orientales, France). The an- atomical and radular study of this population (Figs. 9A-F, 10A-D) allowed us to ascertain complete coincidence in the female genital tract and radula but a constant difference in the structure of the penis. M. massoti has a longer, slender penis totally without lateral ex- crescences (6 specimens examined) (Fig. 9F). The shell of this population is very similar to that of French and Italian material of M. simo- niana but seems to be distinguished by a somewhat more cylindrical shape, a higher height/width ratio of the last whorl and by closer spiral ridges on the whorls (Figs. 1M-Q, 6-8, 11A-C). Although the differences in the penis may not be of great importance (penis shape sometimes varies inside populations of other species of the Hydrobiidae), for the time being we prefer to consider M. massoti as a distinct taxon in the hope of obtaining more French populations of Moitessieria to study. Comparison was also made with material of “M. simoniana lescherae” Boeters (1981) (type locality: Gave d’Alcay, near Tardets- Sorholus, Pyrénées Atlantiques, France), which has a partially despiralized last whorl. The scarcity of living specimens (only one adult and one young female) did not allow enough anatomical data to be collected to reach a valid conclusion (Fig. 12D-G). Nev- ertheless, the presence in both specimens examined of an operculum having a distinct inner hook and the closely packed spiral ridges of the shell microsculpture similar to those in M. massoti (cf. Fig. 8) can be inter- preted as evidence of specific differentiation from M. simoniana. Moitessieria cf. massoti Bourguignat, 1863 Description of the Sardinian Material Shell (Figs. 11D-E, 13A-P, 14A-D, 15A- D) very small, elongate, cylindro-conical or conical in shape, glossy and transparent when fresh. Spire of 5-7 convex whorls; su- tures of variable depth. External surface of protoconch (as seen by SEM) with very thin spiral ridges and very finely pitted; that of te- leoconch with close thin spiral and radial ridges that form a net-like microsculpture hav- ing squares of variable dimensions and shape. Spiral ridges always more evident than radial ones (except in areas of arrested shell growth), but both kinds of ridges some- times reduced. Pits (usually indistinct) in the space between two adjacent spiral ridges or inside squares (more evident by optical mi- croscopy). Number of spiral ridges crossing a 0.1 mm perpendicular line on surface of last whorl (just above the umbilicus): 7.0-11.7 (65 specimens examined). Umbilicus very narrow to fairly wide. Aperture ovate and peristome simple, acute, slightly reflexed in columellar area, with sinuous external margin. Dimensions—height: 1.43—2.11 mm; diam- eter: 0.45-0.82 mm; mouth height: 0.36— 0.66 mm. Operculum corneous, thin, oligogyrous with subcentral nucleus (Fig. 16B, |, N). Genital tract of males (3 specimens exam- ined from three populations) (Fig. 16A, D-F, M, P) with testis, slender and convoluted spermiduct, oval or reniform prostatic gland, slender vas deferens and copulatory organ (penis). Prostate scarcely overlays mantle 12 BODON & GIUSTI = 8 9 10 11 12 w > al © NS o o | 4 28 O о MOITESSIERIA AND PALADILHIA SYSTEMATICS 13 cavity; vas deferens arises from anterior end, crosses dorsal wall of body and enters penis, running along penis to distal tip. Penis colour- less, simple, elongated, elliptical in transverse section, contained inside pallial cavity, and of- ten bent upon itself; sides slightly corrugated near base, tip pointed. No excrescences on penis sides (Fig. 16E, F, M). Genital tract of females (4 specimens ex- amined) (Fig. 16 C, H, O) with ovary, a slen- der renal oviduct and a pallial oviduct. Renal oviduct widens a little near loop where swell- ings (containing refractive material) appear to act as seminal receptacles. True sac-like seminal receptacle absent. Bursa copulatrix duct arising at end of loop; bursa copulatrix oblong, sac-like in shape, its long canal ter- minating in posterior portion. Anterior portion of bursa adhering to albumen gland. Distal renal oviduct running a short distance (from where bursa copulatrix canal arises) before entering uterine complex level with albumen gland. Seminal groove seems to run all along ventral side of uterine complex (= pallial ovi- duct); small uterine complex (albumen gland plus capsule gland). Radula (Figs. 17A-D, 18A-F) taenioglos- sate; central tooth with usual butterfly-shaped hydrobiid structure; apex V-shaped, cutting edge with 11-15 denticles, central one longer than laterals (which progressively diminish); long, slender lateral wing on each side. Inter- nal side of base of each lateral wing with long, pointed basal cusp. Body of tooth extends downwards in a plough-like structure wedged in curved apex of central tooth of next row. FIGURE 8. Number (N) of spiral ridges on a portion of the last whorl surface (just above the umbilicus) crossing a perpendicular line (relative to the spiral ridges) 100 ¡um in length, in the different Moitessieria populations. The extremities of the bar correspond to the extreme limit values; the central point represents medium values and the wider portion includes the standard deviation. Wider black wider bars are based on the data of more than 9 specimens. A: M. simoniana (De Charpentier) from Italy (Piedmont and Liguria) (1-14), from France (15-24) and from spain (25). 1: springs of Cialombard, on the right bank of the river Gesso (Valdieri, Cuneo); 2: floods of the river Tanaro at Garessio (Cuneo); 3: cave “dell’Orso” No. 118 Pi (Ormea, Cuneo); cave of the Rio di Nava no. 911 Li (Pornassio, Imperia); 5: Tana da Fontana do Boro no. 220 Li (Toirano, Savona); 6: springs near Nasino, in Val Pennavaira (Nasino, Savona); 7: floods of the stream Pennavaira at Nasino (Savona); 8: spring near the bridge downstream from the bridge “del Carpe”, in Val Pennavaira (Castelbianco, Savona); 9: Fontana Calda, in Val Pennavaira (Zuccarello, Savona); 10: spring of the waterworks of Taggia (Taggia, Imperia); 11: spring of the waterworks of Buggio (Pigna, Imperia); 12: floods of river Nervia upstream from Dolceacqua (Dolceacqua, Imperia); 13: spring at the bridge of Fanghetto, Val Roia (Olivetta S. Michele, Imperia); 14: spring beside the road between Calvo and Torri, Valley of the Bevera river (Ventimiglia, Imperia); 15: floods of river Loup at Villeneuve Loubet (Alpes Maritimes); 16: floods of river Siagne, near Pegomas (Alpes Maritimes); 17: thermal spring in the Parc Thermal de Salut at Bagnéres de Bigorre (Hautes Pyrénées); 18: floods of the river Garonne at Toulouse (Haute Garonne), materials of the Paulucci Coll.; 19: Resourgence de Moulis, St. Girons (Ariege); 20: underground river of Labouiche, Foix (Ariege); 21: spring of the stream of Ginoles (Aude); 22: floods of the stream of Ginoles near the thermal baths (Aude); 23: well near the spring of the stream of Ginoles (Aude); 24: well at the thermal baths of Ginoles (Aude); 25: floods of the river Manol, west of Vilafant (Catalonia). B: M. massoti (Bourguignat) From France (1) and M. cf. massoti from Sardinia (2-7). 1: Fontaine Saline de Fouradade near Tautavel (Pyrénées Orientales); 2: cave “Cane Gortoe” in Siniscola (Nuoro); 3: spring on the right side of the river Cedrino upstream from Orosei (Nuoro); 4: spring “Su Cologone,” S. Giovanni (Oliena, Nuoro); 5: spring of the cave “Su Anzu” near the convent of S. Giovanni (Dorgali, Nuoro); 6: spring in the Park of the Marquis of Laconi, at Laconi (Nuoro); 7: spring below the cave “Su Marmuri” near Ulassai (Nuoro). C: M. lescherae Boeters from the underground waters of the Gave d'Alcay near Tardets-Sorholus (Pyrénées Atlantiques, France). D: M. vitrea (Draparnaud) from the floods of the river Rhöne near Les Angles (Gard, France). E: M. puteana Coutagne. 1: floods of the river Siage near St. Jean (Alpes Maritimes, France); 2: spring at “La Cascade near Grasse (Alpes Maritimes, France). F: M. juvenisanguis Boeters & Gittenberger from the well near the spring of the stream of Ginoles (Aude, France). G: M. rayi (Locard) from the Cave of Bèze near Bèze (Côte d'Or, France). H: M. rollandiana Bourguignat. 1: floods of the river Lez at Castelnau near Montpellier (Herault, France); 2: floods of the river Vidourle downstream from Sauve (Gard, France). |: M. locardi Coutagne from France. 1: floods of the river Durance near Orgon (Bouches du Rhône); 2: floods of the river Rhöne near Les Angles (Gard); 3: Source du Vivier, N.D. de Vaucluse near Auribeau sur Siagne (Alpes Maritimes). 14 BODON & GIUSTI TABLE 1. Shell parameters of the populations of Moitessieria referred to in Figs. 6-7. H = shell max. height (mm); h = last whorl height (mm); b = last whorl diameter (mouth excluded) (mm); x = mean; s.d. = standard deviation; n = number of specimens. The numbers of the single populations correspond to those in the captions of Figs. 6 and 7. pit X = sid. Population (range) M. simoniana 1 (fig. 5) 1.83 + 0.19 (1.53 — 2.04) France and 2 (fig. 5) 1.92 + 0.17 Italy (1.64 — 2.30) 3 (fig. 5) ST TÍ (1.70 — 1.84) 4 (fig. 5) 1.72 + 0.14 (1.48 — 2.02) M. massoti 5 (fig. 5) 1.51 =0.11 France 7 (fig. 6) (1.311570) M. cf. massoti 1 (fig. 6) 1.77 = 0.20 (1:43 — 2.11) Sardinia 2 (fig. 6) 1755011 (1.59 — 1.93) 3 (fig. 6) 1.72 (1:69 1.77) 4 (fig. 6) 17322012 (1.54 — 1.95) 5 (fig. 6) 1.94 6 (fig. 6) 1.92 + 0.11 (1.642511) Parameters sh AID ESO: Х = Sid: (range) (range) n 0.78 + 0.06 0.53 + 0.04 6 (0.68 — 0.86) (0.47 — 0.58) 0.84 + 0.06 0.56 + 0.04 22 (0.74 — 0.94) (0.47 — 0.62) 0.81 0:53 2 (0.79 — 0.84) (0.53 — 0.54) 0.79 + 0.05 0.50 + 0.04 30 (0.66 — 0.89) (0.40 — 0.60) 0.74 + 0.04 0.46 + 0.02 30 (0.67 — 0.82) (0.41 — 0.51) 0.92 + 0.06 0.71 + 0.04 ИИ (0.81 — 1.02) (0.63 — 0.75) 0.78 + 0.04 0.56 + 0.02 15 (0.71 — 0.90) (0.53 — 0.62) 0.77 0.51 3 (0.76 — 0.79) (0.50 — 0.54) 0.76 + 0.05 0:51 = 0:03 20 (0.67 — 0.86) (0.44 — 0.55) 0.84 0.55 1 0.81 = 0.03 0.54 = 0.03 17 (0.75 — 0.85) (0.47 — 0.57) Lateral teeth with elongated body and wid- ened apex provided with row of 10-15 denti- cles, central denticle longer. Inner marginal teeth rake shaped, elongated, their apex with long row of 18—26 denticles on anterior side. Outer marginal teeth similar in shape to inner marginal teeth but apex smaller, spoon- shaped, with 14-18 small thin denticles on latero-posterior side. Other anatomical characters shown in Fig- ure 16. Body almost totally unpigmented; traces of pigment visible in a few areas of visceral sac. Tentacles lacking ocular spots. Stomach lack- ing caecal diverticulum. Initial portion of intes- tine forming distinct loop distant from style sac, then turning to run near it. Rectal portion of intestine running length of pallial walls, sometimes forming sinuosities but not second loop. Anus close to pallial margin. Pallial mar- gin lacking pallial tentacle. Ctenidium absent. Osphradium oval or reniform. Collecting Sites on the Island of Sardinia Spring near pond in park of Marquis of La- coni (Laconi, province of Nuoro), 5 specimens + many shells; spring below cave “Su Mar- muri” near Ulassai (Province of Nuoro; spring arises near the road to the cave on a preci- pice), 1 specimen + many shells; spring of cave “Su Anzu,” near convent of S. Giovanni (Dorgali, Province of Nuoro), 2 shells; spring “Su Cologone,” S. Giovanni (Oliena, province of Nuoro), 4 shells; spring on right side of River Cedrino, about 3 km upstream from Orosei (province of Nuoro), 1 specimen + many shells; subterranean stream in cave “Cane Gortoe” of Siniscola (province of Nuoro), 8 specimens + many shells. Shells are in collection of M. Bodon; dissected ma- terial in collections of M. Bodon and F. Giusti. Discussion These Moitessieria populations living in the subterranean waters of karstic systems in eastern Sardinia are considered to belong to a unique taxon, possibly conspecific with M. massoti (Bourguignat, 1863: 439—440, pl. 21, figs. 1-5. Type locality: saline spring of Fouradade, near Tautavel, Pyrénées Orien- tales; for other references cf.: Coutagne, 1883, 1884; Westerlund, 1886; Germain, MOITESSIERIA AND PALADILHIA SYSTEMATICS 15 FIGURE 9. Moitessieria massoti Bourguignat, Fontaine Saline de Fouradade near Tautavel (Pyrénées Orientales, France) (A-F) and M. rayi (Bourguignat), Grotte de la Beze near Beze (Cote d'Or, France) (G-M). A: female with the pallial cavity open. Note the absence of ocular spots and the presence of pigment traces in a few areas of the visceral sac; B, L: genital tract of female; C, M: operculum (outer side); D: intestine, osphradium and prostatic gland of male; E: stomach; F, H: penes; G: female with the pallial cavity open. Note the absence of ocular spots and the presence of pigment traces in a few areas of the visceral sac; |: stomach, intestine, prostate gland, ctenidium and osphradium of male. For explanation of symbols see Fig. 4. 16 BODON & GIUSTI FIGURE 10. Radula in specimens of Moitessieria massoti Bourguignat collected in the Fontaine Saline de Fouradade, near Tautavel (Pyrénées Orientales, France). A: central teeth (C); B: apices of some lateral (L), inner marginal (FM) and outer marginal (SM) teeth; C: apices of some lateral (L), inner marginal (FM) and outer marginal teeth (SM); D: some outer marginal teeth (scale bars = 10 um). MOITESSIERIA AND PALADILHIA SYSTEMATICS 17 FIGURE 11. Shell and details of shell surface in a specimen of Moitessieria massoti Bourguignat collected in the Fontaine Saline de Fouradade, Tautavel (Pyrénées Orientales, France) (A-C) and of Moitessieria cf. massoti Bourguignat collected in the cave “Cane Gortoe” (Siniscola, Nuoro, Sardinia) (D-E). A: an entire shell (scale bar = 1 mm); B: shell surface near the umbilicus (Scale bar = 0.1 mm); C: detail of the shell surface (scale bar = 0.1 mm); D: shell surface near the umbilicus (scale bar = 0.1 mm); E: detail of the shell surface (scale bar = 0.1 mm). 18 BODON & GIUSTI sy FIGURE 12. Moitessieria juvenisanguis Boeters & Gittenberger, well near the spring of the stream of Ginoles (Aude, France) (A-C), Moitessieria lescherae Boeters, underground waters of the Gave d'Alcay near Tardets-Sorholus (Pyrénées Atlantiques, France) (D-G) and Paladilhia pleurotoma Bourguignat, Source du Lez, near Prades le Lez (Hérault, France) (H-M). A, |: female removed from shell. Note the absence of ocular spots and pigment; B, D, M: stomach; C, F, L: genital tract and pallial organs of female; E: female with the pallial cavity open. Note the absence of ocular spots and the presence of pigment traces in parts of the visceral sac; G, H: operculum (frontal and lateral views). For explanation of symbols see Fig. 4. à MOITESSIERIA AND PALADILHIA SYSTEMATICS 19 FIGURE 13. Shells of Moitessieria cf. massoti Bourguignat from Sardinia. A-C: spring in the Park of the Marquis of Laconi at Laconi (Nuoro); D-E: spring below the cave “Su Marmuri” near Ulassai (Nuoro); F-G spring “Su Cologone”, S. Giovanni (Oliena, Nuoro); H-L: spring on the right side of the river Cedrino upstream from Orosei (Nuoro); M-P: cave “Cane Gortoe” in Siniscola (Nuoro). 20 BODON & GIUSTI FIGURE 14. Shell and details of shell surface in a specimen of Moitessieria cf. massoti Bourguignat collected in the spring in the Park of the Marquis of Laconi (Laconi, Nuoro, Sardinia). A: an entire shell (scale bar = 1 mm); В: detail of the shell surface (scale bar = 0.1 mm); С: shell surface near the umbilicus (scale bar = 0.1 mm); D: detail of the shell surface near the umbilicus (scale bar = 10 рт). MOITESSIERIA AND PALADILHIA SYSTEMATICS 21 FIGURE 15. Shell and details of shell surface in a specimen of Moitessieria cf. massoti Bourguignat collected in the spring on the right bank of the river Cedrino (Nuoro, Sardinia). A: an entire shell (scale bar = 1 mm); В: the protoconch and first whorls (scale bar = 0.1 тт); С: detail of the shell surface (scale bar = 0.1 mm); D: shell surface near the umbilicus (scale bar = 0.1 mm). 22 BODON & GIUSTI FIGURE 16. Moitessieria cf. massoti Bourguignat, Sardinia (Italy). A-E: spring in the Park of the Marquis of Laconi at Laconi (Nuoro); Е: spring below the cave “Su Marmuri” near Ulassai (Nuoro); G-I: spring on the right side of the river Cedrino upstream from Orosei (Nuoro); L-P; Cave “Cane Gortoe” in Siniscola (Nuoro). A: male. Note the absence of ocular spots and the presence of pigment traces in parts of the visceral sac; В, |, N: operculum (outer side); С, H: genital tract and pallial organs of female; D, P: stomach, intestine (note the oblique packing of the fecal pellets), prostate and osphradium of a male; E, F, M: penis; G, L: female with the pallial cavity partly open. Note the absence of ocular spots and the presence of pigment traces in a few areas of the visceral sac; O: genital tract of two females. For explanation of symbols see Fig. 4. MOITESSIERIA AND PALADILHIA SYSTEMATICS 23 FIGURE 17. Radula in specimens of Moitessieria cf. massoti Bourguignat collected in the cave “Cane Gortoe” (Siniscola, Nuoro, Sardinia). A: central (C) and lateral teeth (L); B: a different view of central teeth; C: lateral (L), inner marginal (FM) and outer marginal (SM) teeth; D: apices of a series of outer marginal teeth (scale bars = 10 um). 24 BODON & GIUSTI FIGURE 18. Radula in specimens of Moitessieria cf. massoti Bourguignat collected in the spring in the Park of the Marquis of Laconi (Laconi, Nuoro, Sardinia) (A-C) and in the spring on the right bank of the Cedrino River (about 3 km from Orosei, Nuoro, Sardinia) (D—F). A: central teeth (C) and apices of the lateral teeth (L); B: apices of some inner marginal (FM) and outer marginal (SM) teeth; C: two outer marginal teeth; D: central portion of a radula: central tooth (C), and lateral teeth (L); E: apices of some lateral (L) and inner marginal (FM) teeth; F: apices of two outer marginal teeth (SM) seen from their internal side (scale bars = 10 um). MOITESSIERIA AND PALADILHIA SYSTEMATICS 25 1931, in part, as “M. saint-simonr’), as sug- gested by the fact they share the diagnostic characters of M. massoti (cf. discussion of M. simoniana): slender, elongate penis lacking any lateral excrescence; shell surface micro- sculpture with closely packed spiral ridges (Fig. 8). Radula and shell shape provide no clues. In fact, the radula has no distinguishing feature, and shell shape could be taken to suggest relationships with more than one of the continental Moitessieria species (Figs. 6, 7; Table 1). The northern-most population (Siniscola in the Albo Mount area) is the most peculiar in shell shape. Its clearly conical shell distin- guishes it from M. simoniana and typical M. massoti and is reminiscent of M. juvenisan- guis Boeters & Gittenberger (1980: figs. 1-2), of M. ollieri Altimira (1960: fig. 1) (cf. also Boeters, 1988: pl. 1, fig. 3) and particularly of M. puteana Coutagne (1883). There is little or no anatomical data about these species (Fig. 12A-C) but, like M. simoniana, they have less closely packed spiral ridges on the shell sur- face (cf. Fig. 8 and Boeters, 1988: pl. 1, fig. 3). The four populations living in the central- southern sites (Laconi, Ulassai, Oliena, Dor- gali) have a cylindroconical shell similar to that of typical M. simoniana and M. massoti. The only population found near Orosei, half way between the northern and the central- southern sites, has a conical shell that is more similar to that of the four central- southern populations but appears be a link between them and the northern populations. Notwithstanding this at least apparent clinal variation (from more to less conical shape, moving south), we think that shell peculiarity may be a sign of genetic differentiation caused by isolation in independent karstic ar- eas, sufficient to postulate the existence of distinct subspecific taxa (at least in the case of the Siniscola population). The present scarcity of material and thus of a statistically valid quantity of morphological data invites prudence. We attributed all the Sardinian populations to M. massoti only by comparison, because they differ from the French population of Bourguignat’s species by virtue of a propor- tionally shorter last whorl (as witnessed by the smaller value of the height/width ratio of the last whorl). Although this character is appar- ently of little significance, it is constant, sug- gesting a certain degree of differentiation. Finally, we recall that another French spe- cies, M. lescherae Boeters (1981), has a mi- crosculpture pattern with closely packed ridges similar to that of Sardinian populations (cf. Fig. 8). This species is nevertheless Clearly distinct by virtue of the swollen last whori of the shell and the operculum with an internal hook. MOITESSIERIA AND PALADILHIA: THEIR RELATIONSHIPS AND FAMILIAR STATUS All the species described herein, M. simo- niana, M. massoti, M. cf. massoti, have basi- cally similar genital tracts in the females. Our analysis of other French species, M. le- scherae Boeters, 1981 (Fig. 12D-G), M. rayi (Locard, 1883) (Fig. 9G-M), and M. ju- venisanguis Boeters & Gittenberger (1980) (Fig. 12A-C) confirms that Moitessieria is characterized by a female genital tract lacking a distinct sac-like seminal receptacle, this function being undertaken by small widened portions of the oviduct. These portions are clearly recognizable by their refractivity and are situated level with the oviduct loop, just before the point of origin of the bursa copul- atrix duct. The diagrams of the female geni- talia reproduced by Boeters (1973) for Moi- tessieria sp. and by Bernasconi (1984a, 1984b) for M. rollandiana, M. vitrea (i.e. M. lineolata, cf. Boeters, 1969) and M. rayi ap- pear to be incorrect. These authors did not detect the oviduct loop and interpreted its an- terior portion, which is often widened, as the distinct sac-like structure of a bursa copulatrix (cf. Boeters, 1973: fig. 4) or of a sac-like sem- inal receptacle (cf. Bernasconi, 1984a: figs. 4, 6, 1984b: fig. 4). The same happened for the female genital tract of Paladilhia. Boeters (1972; 1973: 65, fig. 6) described and repro- duced a scheme in which the oviduct appears to lack a loop, and two sac-like structures arise from the end of the oviduct. One of these structures, orientated outwardly (to- ward the gonopore), was interpreted as a bursa copulatrix and the other, orientated in- wardly, was considered to be a seminal re- ceptacle. This disposition was likened to that in Hyala vitrea (cf. Johansson, 1950) and was therefore different from that in the Hydrobi- idae. This was used to justify the inclusion of Moitessieria and Paladilhia in a distinct family for which the name Moitessieriidae Bourguig- nat, 1863, was available (cf. Boeters, 1972). Bernasconi (1984a,b) partially accepted Bo- eters’s opinion, although he considered the 26 BODON & GIUSTI scheme of the genital tract of females of Moi- tessieria and Paladilhia to be homologous to that of the Hydrobiidae, he reduces the family to the rank of subfamily (Moitessieriinae) of the Hydrobiidae. Careful study of genital tract structure in females of Paladilhia pleurotoma Bourguignat (1865) confirmed that the so- called “outwardly orientated bursa copulatrix” of Boeters is a slight swelling of the oviduct loop (Fig. 12L). The true seminal receptacle is a small, distinct, sac-like structure, which, in intact genital tracts, is orientated inwardly and lies on the actual bursa copulatrix, which at its turn is inwardly orientated. The bursa copul- atrix adheres so tightly to the posterior wall of the albumen gland that it is hard to distinguish the two structures. From the above, it follows that Moitessieria and Paladilhia have the same plan of the fe- male genital tract as the Hydrobiidae. When distinguishing the Moitessieriidae from the Hydrobiidae as a distinct family, Boeters (1972) also listed the following characters: absence of a gonopericardial duct and genital atrium and the anterior end of the uterine complex being far from the pallial margin. It is our opinion that none of these charac- ters can be utilized to justify distinct family status for the Moitessieriidae. Although we searched for the gonopericardial duct in all the females dissected, we were unable to ver- ify its presence. This does not necessarily mean that it does not exist. The specimens are, in fact, too small for a reduced duct to be detectable. Cases exist in the literature of lack of a gonopericardial duct (Rissoidae: Ponder, 1985; North American Hydrobiidae: Hershler, 1985). If it is proved to be absent in Moitess- Гепа and Paladilhia, this could be interpreted as an occasional phenomenon, arising here and there in the Truncatelloidea (= Ris- sooidea; cf. Ponder, 1988) by convergence. Excluding the possibility that the gonopericar- dial duct may only appear to be absent due to its transformation into a sperm tube to allow sperms to enter the oviduct (as hypothesized by Ponder, 1988, in some of the Pomatiop- sidae, a phenomenon to which phylogenetic and systematic importance could be given), we agree with Ponder (1985) and Hershler (1985) who ignored this duct in reconstructing their suprageneric classifications. Finally, as for the more posterior position of the uterine complex inside the pallial cavity, it is Our opinion that it is due to the elongation of shell and body. A similar explanation can also be advanced to explain the slightly looped or sometimes unlooped last portion of the intes- tine in the upper wall of the pallial cavity (in the shortest of the Moitessieria species, M. rayi, the female genital opening is closer to pallial margin and the last portion of the intes- tine is looped). We conclude that Moitessieri- idae Bourguignat, 1863, is a junior synonym of Hydrobiidae Troschel, 1857. Moitessieria and Paladilhia can be included in the subfam- ily Hydrobiinae extended to include some new North American genera recently described by Hershler & Longley (1986, 1987) because all their conchological, anatomical and radular characters correspond to those listed for the latter subfamily by Davis et al. (1982) (cf. Ta- ble 2). A noticeable number of characters distin- guishes Moitessieria from Paladilhia (cf. Ta- ble 2): a pallial tentacle and a seminal recep- tacle are absent in Moitessieria but present in Paladilhia; the central tooth of the radula shows one basal cusp on each side in Moi- tessieria, two or three in Paladilhia (Fig. 19A— D); the microsculpture of the teleoconch shows evident spiral ridges and/or spiral rows of pits in Moitessieria, absence of ridges or only faint spiral ridges in Paladilhia (Fig. 5E- F). All this suggests that the two groups fea- sibly represent distinct genera. The division of Paladilhia into the nomino- typical subgenus and P. (Spiralix) Boeters (1972) cannot be maintained. Our anatomical study of the type species of Spiralix, Lartetia rayi Locard (1883) (Fig. 9G-M), confirms Ber- nasconi’s (1984b) opinion that Spiralix should be included in the genus Moitessieria be- cause the anatomical characters of this spe- cies substantially agree with those of the other species of the genus. The main charac- ter utilized by Boeters (1972) to distinguish Spiralix from Paladilhia (s.s.), the presence of spiral microsculpture on the teleoconch sur- face, is not unique to Lartetia rayi, but is also occasionally found in the shells of the type species of Paladilhia, P. pleurotoma (Fig. 5F). In Lartetia rayi, however, the spiral microsculp- ture consists of rows of pits like those in the species of Moitessieria (cf. Bernasconi, 1984b: fig. 3). Before concluding it seems opportune to point out an apparent similarity between some species of a recently described North American (Texan) genus of the Hydrobiidae: Phreatodrobia Hershler & Longley (1986, 1987). Some of its species are similar to cer- tain species in the genus Moitessieria by vir- tue of radula structure, the opercular peg, the MOITESSIERIA AND PALADILHIA SYSTEMATICS 27 TABLE 2. Moitessieria and Paladilhia compared with some of the closest subfamilies of the Hydrobiinae. The scheme of the characters is modified from Davis et al. (1982). The Nymphophilinae (Taylor, 1966: 199) believed by Thompson (1979) to include some European genera placed by Radoman (1978) in the family Orientalinidae (= Horatiini Taylor, 1966: 175) has been considered even if in our opinion it is hardly distinguishable from the subfamily Hydrobiinae (cf. Giusti & Bodon, 1984, Giusti & Pezzoli, 1984). Character state groups 4 and 5 have also been considered despite the fact that their extreme variability renders them unsuitable for inferring affinities. Moreover, the screening of the same groups of character states is incomplete because many European genera have evidently not been considered. Groups of character states 2, 10, 13 have been modified to include a recently described genus (Phreatodrobia) included by Hershler & Longley (1986, 1987) amongst the Hydrobiinae. Character Hydrobiinae Nymphophilinae Moitessieria Paladilhia Female reproductive system 1.—Pallial oviduct and ventral channel 0 0 0 0 intercommunicating: oviduct joins ventral channel (0) —Spermathecal duct separated from pallial oviduct; oviduct joins albumen gland (1) 2. Duct of seminal receptacle 05 0 if 0 —joins oviduct directly (0) —joins via short sperm duct (1) (* : absence of a distinct seminal receptacle) 3. Pallial oviduct 1 1 1 1 —with 3 or 4 different tissue types (0) —with 2 types (1) Male reproductive system 4. Penis 1 0 1 1 —bilobed (0) —not bilobed (1) 5. Penis with lobe(s) 0,2 3 0,2 2 —convex edge (0) —concave edge (1) —no lobes (2) —glandular ridges (3) 6. Prostate 1 1 1 1 —posterior to end of mantle cavity (0) — overlays mantle cavity (1) Tentacles 7. Hypertrophy of cilia O, 1 0, 1 0 0 —none (0) —Hydrobia-like (1) —Spurwinkia-like (2) Radula 8. Central tooth 0 0 0 0 —Hydrobia-type (0) —Pomatiopsid-type (1) Stomach 9. Anterior lobe digestive gland 1 1 1 1 —present (0) —absent (1) Osphradium 10. —long (0) OMA 0, 1 0, 1 1 —short (1) Pallial tentacle 11. —present (0) 0, 1 1 1 0 —absent (1) Shell 12. Shape 0 O, 1 0 0 —ovate-conical, turreted (0) —globose, neritiform (1) 13. Embryonic 0,1 1 1 0 —Hydrobia-like (0) —Nymphophylus-like (1) 28 BODON & GIUSTI FIGURE 19. The radula in two specimens of Paladilhia pleurotoma Bourguignat collected in the Source du Lez (Prades le Lez, Hérault, France). A: central teeth; note the presence of 2-3 lateral cusps at the base of the lateral wings (there is only one differently placed lateral cusp on the internal side of the base of the lateral wings in Moitessieria); B: central (C), lateral (L), inner marginal (FM) and outer marginal (SM) teeth in another specimen; C: lateral (L) and inner marginal (FM) teeth; D: outer marginal teeth (scale bar = 10 um). MOITESSIERIA AND PALADILHIA SYSTEMATICS 29 simple penis and, particularly, the microsculp- ture of the teleoconch. The latter, particularly in P. coronae Heshler & Longley, 1987, closely resembles the microsculpture in some Moitessieria species. The presence of a dis- tinct sac-like seminal receptacle in most Phreatodrobia species (apparently absent in P. coronae) and a usually planospiral, valva- toid, only rarely trochoid or conical shell (never elongated) seems nevertheless to ex- clude any close relationship. Worthy of note is the fact that similar teleoconch microsculpture is known in other genera of the Prosobranchia belonging to other families (Vanikoridae: Waren & Bouchet, 1988; Pelycidiidae: Ponder & Hall, 1983; Rissoidae: Ponder, 1985). Thus, this type of microsculpture is yet another of the many cases of convergence in gastropod shell characters. BIOGEOGRAPHY The discovery of Moitessieria species in northwestern Italy, although faunistically sig- nificant, is not particularly surprising because of the territorial continuity between northeast- ern Spain, south France and northwestern It- aly. However, the completely unexpected dis- covery of the genus in Sardinia is of special interest and deserves comment. The avail- able data on the geological evolution of the land bordering the western Mediterranean suggests that the Sardo-Corsican complex broke away from the southwestern border of the Palaeo-European Continent during the Oligocene-Lower Miocene (for a review, see Giusti & Manganelli, 1984). According to most reconstructions it lay between the present eastern Pyrenees and western Alps and was thus continuous with the present areas of dis- tribution of the genus Moitessieria. Because Moitessieria is obviously not capable of active dispersal via hypothetical ephemeral land bridges, or of passive dispersal, its presence in Sardinia provided collaborative zoological evidence of the validity of the above geolog- ical and historical reconstructions. ACKNOWLEDGEMENTS Our thanks to Dr. H. D. Boeters (Munchen, BDR) for providing spirit specimens of M. si- moniana collected in the Resourgence de Moulis and to Dr. R. Rouch of the Laboratoire Souterrain du CNRS (Moulis, France) for al- lowing us to enter the karstic cavity and col- lect more materials. An anonymous reviewer is also thanked for providing constructive comments on the original manuscript. LITERATURE CITED ALTIMIRA, C., 1960. Notas malacolögicas. Con- tribuciön al conocimiento de los moluscos ter- restres y de agua dulce de Cataluna. Miscellanea Zoologica, 1 (3): 9-15. BERNASCONI, R., 1984a. Hydrobides de France: Moitessieria, Bythiospeum et Hauffenia des de- partements Gard, Ain, Isere (Gasteropodes Prosobranches). Revue Suisse de Zoologie, 91 (1): 203-215. BERNASCONI, R., 1984b. Découverte du genre Moitessieria BGT (Mollusca Gastropoda Hydro- biidae) dans le Dijonnais (Cöte d’Or). Revue Su- isse de Zoologie, 91 (3): 687-697. BOATO, A., M. BODON & F. GIUSTI, 1985. Mol- luschi terrestri e d’acqua dolce delle Alpi Liguri. Lavori della Societa Italiana di Biogeografia, (n.s.) 9: 237-371. BODON, M., 1980. Segnalazione del genere Moi- tessieria Bourguignat in Italia (Gastropoda: Hy- drobioidea). Doriana, 5 (236): 1-5. BODON, M. & E. PEZZOLI, 1986. Nota preliminare sui molluschi ipogei del Piemonte e della Liguria. Atti del Convegno Internazionale sul Carso di Alta Montagna (Imperia, 30.1V.-4.V.1982), 2: 299-309. BOETERS, H. D., 1969. Die Hydrobiidae Badens, der Schweiz und der benachbarten franzö- sischen Departements, Nachtrag (Mollusca, Prosobranchia). Mittelungen des Badischen Landesvereins Naturkunde und Naturschutz e v Freiburg im Breigau., (n.f.) 10 (1): 175-177. BOETERS, H. D., 1972. Westeuropäische Moitess- ieriidae, 1. Spiralix n. subgen. (Prosobranchia). Archiv für Molluskenkunde, 102 (1/3): 99-106. BOETERS, H. D., 1973. Französische Rissoaceen- aufsammlungen von C. BOU. Annales de Spel- éologie, 28 (1): 63-67. BOETERS, H. D., 1981. Unbekannte westeuro- päische Prosobranchia, 2. Archiv für Mollusken- kunde, 111 (1/3): 55-61. BOETERS, H. D., 1988. Moitessieriidae und Hydro- biidae in Spanien und Portugal (Gastropoda: Prosobranchia). Archiv für Molluskenkunde, 118 (4/6): 181-261. BOETERS, H. D. 8 E. GITTENBERGER, 1980. Un- bekannte westeuropäische Prosobranchia, 4. Basteria, 44: 65-68. BOURGUIGNAT, M. J. R., 1863. Monographie du nouveau genre Frangais Moitessieria. In GUERRIN MENEVILLE, F. A., Revue et Magazine de Zoologie Pure et Appliquée, (ll) 15: 432-445, pls. 20-21. BOURGUIGNAT, M. J. R., 1865. Monographie du nouveau genre Francais Paladilhia. Paris: С. Е. Savy, Libraire-Editeur. Paris. 24 pp., 1 pl. 30 BODON & GIUSTI COUTAGNE, G., 1883-84. Revision sommaire du genre Moitessieria. La Feuillle des Jeunes Natu- ralistes, 13 (155): 130-132, 13 (156): 143-147, 14 (165): 105-110, pl. 3. DAVIS, G. M., M. MAZURKIEWICZ & M. MAN- DRACCHIA, 1982. Spurwinkia: morphology, sys- tematics and ecology of a new genus of North American marshlands Hydrobiidae (Mollusca: Gastropoda). Proceedings of the Academy of Natural Sciences of Philadelphia, 134: 143-177. DE SAINT SIMON, A., 1848. Miscellanées mala- cologiques. Premiére décade. 41 pp. Toulose. GERMAIN, L., 1931. Mollusques terrestres et fluvi- atiles (duexiéme partie). Faune de France, 22: 479-897, ix-xiv, pls. XIV-XXVI. GIUSTI, Е. 4 М. BODON 1984. Notulae malacolog- icae, XXXI. Nuove Hydrobiidae dell'Italia nord- occidentale (Gastropoda: Prosobranchia). Archiv für Molluskenkunde, 114: 157-181. GIUSTI, F. & G. MANGANELLI, 1984. Relation- ships between geological land evolution and present distribution of terrestrial gastropods in the western Mediterranean area. In SOLEM, A. & A. C. VAN BRUGGEN, eds., Worldwide snails. Biogeographical studies on non-marine Mol- lusca, pp. 70-92. Brill, Leiden. GIUSTI, F. & E. PEZZOLI, 1982a. Molluschi caver- nicoli italiani (Notulae malacologicae, XXVII). La- vori della Societa Italiana di Biogeografia, (n.s.) 7: 431-450. GIUSTI, F. & E. PEZZOLI, 1982b. Notes on the small Hydrobioidea in Italian subterranean wa- ters: catalogue, biogeography and some system- atic problems. Malacologia, 22: 463—468. GIUSTI, F. & E. PEZZOLI, 1984. Notulae Malaco- logicae, XXIX. Gli Hydrobiidae salmastri delle ac- que costiere italiane: primi cenni sulla sistemat- ica del gruppo e sui caratteri distintivi delle singole morfospecie. Lavori della Societa Mala- cologica Italiana (Atti del Simposio: Sistematica dei Prosobranchi del Mediterraneo—Bologna, 24-26 Settembre 1982), 21: 117-148. HERSHLER, R., 1985. Systematic revision of the Hydrobiidae (Gastropoda: Rissoacea) of the Cu- atro Cienegas Basin, Coahuila, Mexico. Malaco- logia, 26: 31-123. HERSHLER, R. & G. LONGLEY, 1986. Phreatic hydrobiids (Gastropoda: Prosobranchia) from the Edwards (Balcones fault zone) acquifer region, south-central Texas. Malacologia, 27: 127-172. HERSHLER, R. 8 G. LONGLEY, 1987. Phreatodro- bia coronae, a new species of cavesnail from southwestern Texas. The Nautilus, 101: 133- 139. JOHANSSON, J., 1950. Uber die wieblichen Ge- schlechtsorgane von Hyala vitrea, einer von dem Rissoa-Typus stark abweichenden Form der Gruppe Rissoacea. Arkiv for Zoologi, 42A (7): 1- 6. KUSTER, H. C., 1852. Paludina, Hydrocena und Valvata. In: MARTINI, F. H. W. & J. H. CHEM- NITZ, Systematisches Conchylien-Cabinet. For- gesetzt von Hofrath Dr. G. H. v. Schubert und Professor Dr. J. A. Wagner. In Verbindung mit Dr. Philippi, Dr. Pfeiffer und Dr. Dunker neu heraus- gegeben und vervollstandigt von Dr. H. C. Kuster, 2nd Edit., 1 (21): 96 pp., 14 pls. Nürn- berg. LOCARD, A., 1883. Contribution a la faune mala- cologique française Ill. Monographie du genre Lartetia. Annales de la Societé Linneenne de Lyon (N.S.), 29: 189-208, 1 PI. PONDER, W. F., 1985. A review of the genera of the Rissoidae (Mollusca: Mesogastropoda: Ris- soacea). Records of the Australian Museum, Supplement, 4: 1-221. PONDER, W. F., 1988. The Truncatelloidean (= Rissoacean) radiation. A preliminary phylogeny. In PONDER, W. F., ed., Prosobranch phylogeny. Malacological Review, Supplement 4: 129-164. PONDER, W. F. & S. J. HALL, 1983. Pelycidiidae, a new family of archaeogastropod molluscs. The Nautilus, 97: 30-35. RADOMAN, P., 1978. Neue vertreter der Gruppe Hydrobioidea von der Balkanhalbinsel. Archiv fur Molluskenkunde, 109 (1/3): 27-44. TAYLOR, D. W., 1966. A remarkable snail fauna from Coahuila, Mexico. The Veliger, 9: 152-228. THOMPSON F. G., 1979. The systematic relation- ships of the hydrobioid snail genus Nympho- philus Taylor 1966 and the status of the subfamily Nymphophilinae. Malacological Review, 12: 41— 49. WESTERLUND, C.A., 1886. Fauna der in paläar- ctischen Region lebeden Binnenconchylien. 6 Fam. Ampullaridae, Paludinidae, Hydrobiidae, Melanidae, Valvatidae & Neritidae, 156: 13 pp. Lund. WAREN, A. & P. BOUCHET, 1988. A new species of Vanikoridae from the western Mediterranean, with remarks on the northeast Atlantic species of the family. Bollettino Malacologico, 24: 73-100. Revised Ms. accepted 27 July 1990 MALACOLOGIA, 1991, 33(1-2): 31-42 HISTOLOGICAL OBSERVATION OF GONADS AND DIGESTIVE GLAND IN STARVING DREISSENA POLYMORPHA (BIVALVIA) Ulrich Bielefeld Zoologisches Institut der Universitat Heidelberg Im Neuenheimer Feld 230, D-6900 Heidelberg, Germany ABSTRACT In two groups of Dreissena polymorpha specimens, histological and cytological changes in the digestive gland and in the gonads were investigated during progressive starvation (30 days) and subsequent refeeding (5 days). One group, originating from a depth of 2 m (17.1°C), had already spawned; the other group, originating from a depth of 9 m (10.4°C), had not. Initially, cytological studies showed disturbed activity in the digestive gland of the 9 m group. After 10 days of starvation, reduced activity occurred in the digestive gland of both groups, although much stronger in the 2 m group. After 30 days of starvation, about 25% of the gland tissue was inactivated or degenerated in both groups. After 5 days of refeeding, regeneration of the gland was stronger in the 2 m group. During the experiment, 79% of the 2 m specimens and 6% of the 9 m specimens degraded the remaining gametes of their gonads. In the remaining specimens of both groups, the gonads degenerated only to minor extent. Haemocytes are involved in the resorption of the degenerating gonads. Key words: Dreissena polymorpha, starvation, gonad, digestive gland, ultrastructure. INTRODUCTION During the annual reproductive cycle, Dre- issena polymorpha has to cope with conditions when energy supply does not meet energy consumption. This investigation was designed to contribute to a better understanding of the adaptive response of Dreissena to this situa- tion. Thus, the effect of starvation in two groups of Dreissena, each in a different stage of their reproductive cycle, was examined. It is known that such stresses as shortage of food supplies or pollution induce cytological changes in various bivalve tissues (reviewed by Moore & Lowe, 1985). As target organs, the digestive gland and the gonads were chosen. Because the gland plays a central role in the digestion and resorption of food (reviewed by Morton, 1983) and the gonad is the organ with the highest energy demand (Bayne, 1976), both tissues should be affected by starvation. Furthermore, haemocytes have to be considered as they are an important func- tional link between the digestive gland and the gonads. (The various functions of haemo- cytes are reviewed by Cheng, 1981.) It is not the aim of this paper to give a com- plete description of the tissues. | will rather refer to good descriptions in the literature be- fore focusing on alterations induced by star- 31 vation. During the same experiment, physio- logical and morphometric data were obtained, which are described by Sprung & Borcherding (submitted). MATERIAL AND METHODS In August 1987, two groups of Dreissena polymorpha were collected from Fuhlinger See north of Cologne, Germany. The first group (2 m group) was taken from a depth of 2 m, where the water was fully oxygen-satu- rated and had a temperature of 17.1°C. The second group (9 m group) was obtained from a depth of 9 m at 10.4°C, where conditions were nearly anoxic. From previous studies, it was known that the 2 m population had spawned and the 9 m group had not. The specimens were kept in the laboratory under starvation conditions for 30 days and were then fed with a suspension of Chlamy- domonas reinhardii for 5 days. To avoid stress reactions to thermal adaptation (Wid- dows & Bayne, 1971) and to prevent spawn- ing in the 9 m group (Sprung, 1987), both groups were kept at their original environmen- tal temperatures with sufficient aeration. For 32 BIELEFELD TABLE 1. Fixation schedule of the experiment. The first figure is the number of all specimens in each batch. The second one designates the number of histologically identified females and the third the males. start 10 days 2 m group/17.1°C 3/3/0 3/1/0 9 m group/10.1°C 4/1/3 3/1/0 further details, see Sprung & Borcherding (submitted). The fixation schedule for histological exam- ination is given in Table 1. Small pieces of the visceral sac were fixed in 2% glutaraldehyde in 0.01 M sodium cacodylate buffer of pH 7.4. After rinsing in buffer, the tissues were fixed in 1% osmium ferrocyanide for several hours (Karnovsky, 1971) and stained with 1% ura- nyl acetate in maleate buffer. The tissue was dehydrated and embedded in Araldite. Ul- trathin sections were counterstained with al- kaline lead citrate and examined with a Zeiss EM 9. Semithin sections were stained with methylene blue-Azur Il (Richardson et al. 1960) and used for light microscopy. RESULTS Digestive Gland The digestive gland consists of numerous tubules, which are connected to the stomach by primary and secondary ducts. The epithe- lium of the tubules is made up of digestive cells, basophilic pyramidal cells, and flagel- late cells. The ultrastructure of these cells has been described by several authors (see Jan- ssen 1981a,b). During the experiment, the height of the predominant digestive cells and the number and form of their vacuoles changed. Furthermore, ultrastructural alter- ations in the basophilic pyramidal cells were observed. Thus, distinct phases in the epithe- lia of the tubules could be characterized, and these will be described below. Initial Observations Prior to Starvation: At the beginning, more than 90% of the tubules of both groups (Fig. 1) are in integer phase (Fig. 2). In this phase, the digestive cells are 40— 50 um high. They take up nutrients by means of microvilli and by endocytosis. In both groups, the cells are filled with translucent vacuoles and some residual bodies. Experimental conditions starvation refeeding 30 days 1 day 5 days 9/4/1 5/0/2 6/2/1 9/2/3 4/0/2 5/1/2 Only a few tubules are in the disintegrating phase (Fig. 1). Here, the digestive cells have pinched off the apical part into the lumen of the tubules (Fig. 3). The disintegrating phase is therefore characterized by a lower epithe- lium (about 20 um high). Differences Between the 2 m and 9 m Group Prior to Starvation: In the integer phase, the endocytic system in the digestive cells of the 2 m group is more extended (Fig. 5) than in the 9 m group. Compared to the 2 m group, the basophilic pyramidal cells of the 9 m group (Fig. 6) exhibit an uneven distribution of the rough endoplasmic reticulum (Fig. 8), ac- cumulation of Golgi vesicles (Fig. 9), and basal vacuoles with regular internal substruc- tures (Fig. 11). In the 9 m specimens, a high lipid content appears primarily in the cells of the gland ducts (Fig. 7). 10 Days of Starvation: In the 2 m group, the digestive gland is profoundly altered (Fig. 1): the number of tubules in the disintegrating phase increases while many tubules are in the condensed holding phase (Fig. 4). In this phase, the epithelium is lower (25-35 wm height) than in integer phase. The cytoplasm of the cells and the mitochondria are con- densed. In the digestive cells (Fig. 10), re- sorption continues, but the microvilli are smaller. The content of the vacuoles has changed, and the number of phagosomes and residual bodies has increased. The ba- sophilic pyramidal cells may contain basal vacuoles with conspicious substructures (Fig. 11), and the structure of the Golgi stacks may be disturbed (Fig. 12). In the condensed holding phase, only a few cells degenerate. In degenerating cells, the nucleus is dilated and the cytoplasm is elec- tron lucent. In the digestive cells, the mem- branes of the vacuoles disintegrate, thus shedding their content into the cytoplasm. The endoplasmic reticulum of the basophilic pyramidal cells (Fig. 14) appears as small 33 HISTOLOGY OF STARVING DREISSENA ‘(suey) aseud jueisueu pue '(Bap) aseyd Bunesauabap ‘(риоэ) aseyd Buipjoy pasuapuoo ‘(sip) aseyd бицелбазиер ‘(Baju!) aseyd лэбаци| :pasn зиоцелалаау “eBejueoied jenjoe эц} sajesıpuı 40} uo aunBy ayy "Alanıypadsaı ‘(yuejq) dnoib w z ay, pue (49e/q) ш 6 ay) 10} anjea ay) SAAIH suunjoo jo ле4 yoe3 ‘seseud 1ejnoiued ou} и! sajnqn] jo abejuao1ad ay, mous sieq ou} ‘Huljjas ¡ejueunadxa yoea 104 `бшрэз}эл pue иоцеллез Bulunp eydiowAjod euassialq JO sa¡nqn; pue|6 элцзэб!р ay) jo иоцоеа. ay | “| ‘914 SNVAL DIA ANOD SIG DALNI SNVAL DIA ANOD SIA DALNI INIAYAAAA SAV $ ONIGAHAAA AV I SNVAL DIA ANOD SIG DALNI SNVAL DIA ANOD SIA эм SNVAL DIA ANOD SIA DALNI NOILVAUVLS SAVA 0€ NOILVAAVLS SAV 01 LAVLS 34 BIELEFELD FIG. 2. Dreissena polymorpha, 2 m group, newly collected, digestive gland, semithin section of a tubule in integer phase with narrow lumen (L). Digestive cells (DC) filled with vacuoles of translucent content, baso- philic pyramidal cells (BC) and flagellate cells (FC) in crypts. LM 500 x, scale bar 20 um. FIG. 3. Dreissena polymorpha, 9 m group, 30 days starvation, digestive gland, semithin section of a tubule in disintegrating phase. Digestive cells (DC) shed the apical part into the lumen (L), apices of basophilic pyramidal cells reach freely the surface of the epithelium. Granulocytes (GC) in the interstitium, one has invaded the epithelium (arrow). LM 500 x, scale bar 20 um. FIG. 4. Dreissena polymorpha, 2 m group, 30 days starvation, digestive gland, semithin section of a tubule in condensed holding phase. The cells are condensed, the digestive cells (DC) are filled with late phago- somes. The lumen (L) is narrow, the outer surface rough, gaps in the epithelium (arrows) indicate degen- erated cells. LM 500 x, scale bar 20 um. vesicles, which are often devoid of ribo- somes. The Golgi complex consists of fewer cisternae, which continue to proliferate vesi- cles. The residual material of the cells is shed into the lumen of the tubule. Gaps in the dense epithelium indicate lost cells. The outer surface of the tubules is irregular, because the musculature surrounding the tubules has become located in grooves between protrud- ing parts of basophilic pyramidal cells. Haemocytes, which may contain large lipid droplets, contact the outer surface (Fig. 4) or penetrate into the epithelium. The digestive glands of the 9 m animals exhibit comparatively few changes. Most tu- bules are in the transient phase (Fig. 1). This phase shows features intermediate between the integer and the condensed holding phase. The epithelia contain light and dark staining cells adjacent to one another. 30 Days of Starvation: In both groups, the digestive gland is severely affected. About 75% of the tubules are still in the condensed holding phase, but 25% are either in the dis- integrating or in the degenerating phase (Fig. 1). The degenerating phase is characterized by the fact that, because most cells are de- generating cells, the tubules have lost their distinct contours. 1 Day of Refeeding: In the 9 m group, about 40% of the tubules have returned to the tran- sient phase. In the 2 m group, only a few tu- bules have changed (Fig. 1). 5 Days of Refeeding: Compared to the 9 m group, the 2 m group has regenerated further (Fig. 1), and some tubules have reestablished the structure and ultrastructure of the integer phase. Glycogen rosettes occur in the cyto- plasm of the epithelial cells and in haemo- cytes. Gonads The ultrastructure of the ovaries strongly resembles that of Mytilus (see Pipe, 1987); the testes of Dreissena are described by Max- well (1983). The state of the gonads not only differed between both groups but also changed within each group during progres- sive starvation. | HISTOLOGY OF STARVING DREISSENA 35 Mature Gonads: Initially and after 10 days starvation, the gonads of the 9 m specimens contained mature male and female gametes. Well-developed ovaries (Fig. 15) have a di- ameter of 200-400 um. The acinar epithe- lium consists of a layer of follicle cells and ova in different stages of development. Small, early oocytes are electron dense. This is due to the many free ribosomes and dark-staining mitochondria. During vitellogenesis, the oocytes enlarge and bulge into the acinar lu- men. The electron-lucent cytoplasm of the oocytes contain electron-dense lipid and pro- teinaceous yolk. The follicle cells contain mitochondria, sin- gle strands of rER, a well-developed lysoso- mal system, and as energy sources high amounts of glycogen rosettes and some lipid droplets. The testes are filled with mature spermato- zoa, many developing spermatids, spermato- cytes, and nutritive cells with glycogen and lipid. Mitotic figures were repeatedly observed. Partial Degeneration: Partial degeneration in the filled gonads was found in the males of the 9 m group after 30 days of starvation and after refeeding. In the 2 m group, there were two females (start and 30 days starvation) and one male (5 days refeeding) exhibiting partial degeneration. This means that the ac- inar epithelium of the ovaries and the oocytes shrink. Stored lipid and glycogen is dimin- ished. In some oocytes, the Golgi-vesicles fuse to vacuoles or they fuse with mitochon- dria (Fig. 17). Autolysosomes can be ob- served. The nuclear envelope of the devel- oped oocytes is dilated so that the resulting vacuoles protrude into the karyoplasm. Some testes are penetrated by haemocytes, which have a high content of glycogen and only few organelles and contain phagocytosed sperm cells in vacuoles (Fig. 18). Complete Degeneration of the Gonads: In 11 of 14 specimens of the 2 m group, which were taken from all phases of the experiment, the gonads were completely degenerated. In the 9 m group, only one female displayed a de- generated gonad after 30 days of starvation. In the degenerating ovaries of the 2 m spec- imens, in some acini all stages of partially or completely degenerated eggs (Fig. 16) were observed. The follicle cells contain autolyso- somes, glycogen, andlipid droplets. In the ova, the Golgi-stacks are reduced in number and their cisternae are distended but still prolifer- ating. With further degeneration, the vitelline coat, the outer membrane, the yolk, and the mitochondria become lytic. In other cases, where degeneration has probably progressed still further, some haemocytes have shed their granules onto the engulfed oocytes (Fig. 19). Other haemo- cytes contain heterophagosomes and lipid droplets. The mitochondria of the haemocytes exhibit a cristalline-like lattice (Fig. 20). There are also some acini with a small lumen and a thickened basement membrane. The epithe- lial cells of these acini contain large amounts of lipid and glycogen (Fig. 21). The lysosomal system metabolizes phagocytosed material. In the 2 m group, the degeneration of the testes can be compared with that of the ova- ries. Epithelial cells are filled with glycogen and lipid. Haemocytes, which phagocytose sperm cells, bear mitochondria with a lattice- like substructure and otherwise resemble those in the degenerating female gonads. Interstitial Haemocytes During the experiment, several forms of haemocytes (classification from Cheng, 1981) were observed (Fig. 18). Initially, mainly small (5 um) and large (5-10 рт) granulocytes occur. Lysosomes with hetero- geneous content (phagosomes) can be lo- cated in the periphery of the cytoplasm. In some haemocytes, the granules are lacking but the phagosomes predominate. All forms of granulocytes could be observed penetrat- ing the digestive gland epithelium. Further- more, hyalinocytes with an oval nucleus of up to 6.5 um diameter and a thin rim of electron- translucent cytoplasm containing only few or- ganelles were observed. During starvation, an increased number of haemocytes is found interstitially between the degenerating gonads (Fig. 16) and the diges- tive gland (Figs. 3, 4). The percentage of haemocytes containing phagocytosed mate- rial increased. Lipid droplets could be ob- served in all forms of those haemocytes in- vading the gland. DISCUSSION Digestive Gland Initial Observations: There is cytological evi- dence that in the 9 m group low environmental temperature and low oxygen supply might been the cause of the disturbed metabolism of the digestive gland. In the 9 m group, the scarce endocytic canals in the digestive cells indicate reduced resorptive activity. In the 36 BIELEFELD FIG. 5. Dreissena polymorpha, 2 m group, newly collected, digestive gland, electron micrograph of a digestive cell. Apically microvilli (MV), pinosomes (PS) and endocytic canals (EC), laterally desmosomes (DS) and septate junctions (SJ), vacuoles (V) with electrontransparent content. Small strands of the rough endoplasmic reticulum (RER) and mitochondria (M). 10,950 x , scale bar 1 pm. FIG. 6. Dreissena polymorpha, 2 m group, newly collected, digestive gland, electron micrograph of a basophilic pyramidal cell. Loosely packed strands of rough endoplasmic reticulum (RER), numerous Golgi stacks (GS) proliferating Golgi vesicles (arrows), which fuse to secretory granules (star), mitochondria (M). 9,500 x, scale bar 1 pm. FIG. 7. Dreissena polymorpha, 9 m group, newly collected, digestive gland, electron micrograph of a digestive gland duct. Duct cells with microvilli (MV), nucleus (N), mitochondria (M) and high amounts of lipid (LI), haemocytes (HC) invade the widened intercellular space. 4,750 x, scale bar 1 um. FIG. 8. Dreissena polymorpha, 9 m group, newly collected, digestive gland, electron micrograph of a basophilic pyramidal cell. Rough endoplasmic reticulum in short pieces (RER 1) or in dense stacks, laterally widened to vesicles (RER 2), nucleus (N), lipid (LI). 9,500 x, scale bar 1 pm. FIG. 9. Dreissena polymorpha, 9 m group, newly collected, digestive gland, electron micrograph of a basophilic pyramidal cell. (GS) Golgi stack proliferating vesicles (VE), which accumulate. 14,350 x , scale bar 1 um. 1 HISTOLOGY OF STARVING DREISSENA 37 basophilic pyramidal cells, protein synthesis of the rough endoplasmic reticulum is diminished and the accumulation of Golgi vesicles indi- cates reduced level of secretory activity. It is probable that both groups took up sim- ilar nutrients, because in both groups the di- gestive vacuoles have a similar translucent content and scarce residual bodies. There- fore, a nutritional factor probably had only a minor influence. Vacuoles in the basophilic py- ramidal cells, which also occurred in starved specimens, may also be interpreted as an in- dex of adverse environmental conditions. The high amounts of lipid in the 9 m group, confirmed by quantitative biochemical exam- ination (Sprung & Borcherding, submitted), might result from a reduced level of B-oxida- tion due to low oxygen supply. Lipid reserves in the digestive gland are described for Mytilus in late summer (Gabbot & Bayne, 1973), but they are accompanied by high levels of carbohydrates, which were neither histologically nor biochemically detected in the visceral sac of Dreissena (Sprung & Borcherding, submitted). Therefore, the ultrastructural observations suggest that the high lipid content and the greater weight of the 9 m group before star- vation (Sprung & Borcherding, submitted) must be very critically interpreted regarding physiological condition (Moore & Lowe, 1985). In this investigation, the percentage of tu- bules in a particular phase is taken as an in- dicator of the activity of the gland. Several phases have been documented in the gland of feeding specimens (cf. Morton, 1983), but the differentiation of a particular phase is dif- ficult (Robinson et al., 1981). Thus, | initially distinguished only the disintegrating phase (Langton, 1975), which is rather distinct from the other phases. In feeding Dreissena, the disintegrating phase has an excretory func- tion, and the tubules can regenerate after- wards (Morton, 1969). The percentage of the disintegrating phase is approximately 5% and thus similar to that found in Mytilus edulis (Langton, 1975). We refer to the other phases collectively as the integer phase, because the cells are resorbing and digesting. Starvation: Again, the ultrastructural altera- tions induced by starvation can be compared with those reported for Mytilus, but in Dreis- sena the adverse effects occur earlier. Following starvation, the height of the di- gestive cells and the number of late phagoly- sosomes is reduced (see Thompson et al., 1978: 13 days of starvation in Mytilus califor- nianus). Furthermore, the protein synthesizing apparatus of the basophilic pyramidal cells is disturbed, and later the vacuole membranes of the digestive cells disintegrate (cf. Thompson et al., 1974: 28 days of starvation in Mytilus edulis). True cell death is indicated by elec- tron-transparent cells (Thompson et al., 1974: 147 days starvation for Mytilus edulis). Further heterogeneity in the phases of the tubular cycle resulted from starvation in both groups. Loss of physiological condition in the 2 m group was more strongly marked after 10 days of starvation. The high percentage of tu- bules in the condensed holding phase indi- cates disturbed and reduced resorption and intracellular digestion. Although there was also some loss of cells, the tubules remained intact. Tissue reaction in the 9 m group was weaker, because only a minor increase in de- generating cells occurred in the transient phase. Probably two main factors are responsible for the more marked effects in the 2 m group: (1) the 9 m group could metabolize lipids, which had disappeared after 10 days starva- tion. This assumption is supported by the re- sults of Sprung & Borcherding (Submitted). In contrast, the digestive gland of the 2 m group predominantly catabolized tissue proteins. (2) In the 2 m group, the higher temperature en- hanced the autocatalytic process. Bayne et al. (1978) observed autolytic changes in the gland of Mytilus after spawn- ing, but because damage in the initial 2 m group was not observed, this might be of mi- nor influence. The histological results explain the initial decrease of the gland/body quotient, which was observed by morphometric analysis (Sprung & Borcherding, submitted): the tu- bules passed into the transient and con- densed holding phase. Thus, the initial de- crease of this quotient reflects adaptational rather than degenerative processes. In both groups, 30 days of starvation in- duced a further increase in the disintegrating and degenerating phases and more than 20% of the tissue was inactivated. Thirty days of starvation results in greater damage of gland tissue than 10 days. Refeeding: Ultrastructural changes in the gland cells are similar to those observed by Janssen (1981b) for Mytilus. The high per- centage of tubules in the degenerating and the disintegrating phase in both groups indi- 38 BIELEFELD FIG. 10. Dreissena polymorpha, 2 m group, 30 days starvation, digestive gland, electron micrograph of a digestive cell. Apically a reduced number of thin microvilli (MV), pinosomes (PS), vacuoles (V) with floculent content, late phagosomes (LPS), mitochondria (M). 11,550 x, scale bar 1 um. FIG. 11. Dreissena polymorpha, 2 m group, 30 days starvation, digestive gland, electron micrograph of the basal part of a basophilic pyramidal cells. (VA) vacuole with electron-dense vesicles. 20,900 x , scale bar 1 um. FIG. 12. Dreissena polymorpha, 2 m group, 30 days starvation, digestive gland, electron micrograph of a basophilic pyramidal cell. The Golgi stack (GS) is budding off vesicles, which fuse and contain electron lucent “bubbles.” 22,450 x, scale bar 1 рт. FIG. 13. Dreissena polymorpha, 2 m group, 30 days starvation, interstitium, electron micrograph of haemo- cytes. Large granulocytes (GC), small granulocytes (SGC), macrophage (MPH), (H) hyalinocytes, 3,050 x, scale bar 1 um. FIG. 14. Dreissena polymorpha, 2 m group, 30 days starvation, electron micrograph of a degenerating basophilic pyramidal cell. Swollen nucleus (N), the endoplasmic reticulum occuring in vesicles without ribosomes (arrows), Golgi stacks (GS) widened and reduced, but still proliferating, lipid (LI) and mitochondria (М). 8,500 x, scale bar 1 pm. HISTOLOGY OF STARVING DREISSENA 39 e” FIG. 15. Dreissena polymorpha, 9 m group, newly collected, ovaries, semithin section of a well-developed acinus. Mature oocytes with germinal vesicle (GV), nucleoli (NO), stacks of rough endoplasmic reticulum (RER) and cortical granules (CG). Young oocytes (YOC) are smaller and stain darker. Light microscopy 300x, scale bar 50 um. FIG. 16. Dreissena polymorpha, 2 m group, newly collected, ovaries, semithin section of acini, which have spawned. The remaining gametes are degenerating. In the acini, remainders of mature oocytes (OC), dark staining young oocytes (YOC) and blood cells (arrows). The basement lamina (BL) is thickened. In between the acini an increased number of haemocytes (HC). Light microscopy 300 x , scale bar 50 um. cate that reconstitution of the tissue is accom- panied by degeneration of damaged tubules. Because refeeding in both groups induced a shift of more than 70% of the tubules from the condensed holding to the transient and the integer phase, it is concluded that most tu- bules in the condensed holding phase can re- generate. Tubules in the degenerating and the majority of the tubules in the disintegrating phase are completely damaged. The occur- rence of tubules in the integer phase and the high percentage of transient phase tubules in- dicate that the 2 m group undergoes more regeneration than the 9 m group. Gonads Not all 2 m specimens had spawned, and spawning did not completely empty the go- nads (cf. for Mytilus, Thompson et al., 1978). However, for the majority of the specimens, the results are consistent with the quantitative experimental data. Complete degeneration in the 2 m group correlates well with the linear decrease of the gonad/body quotient (Sprung & Borcherding, submitted). In the 9 m group, only some loss in volume occurred, and the gonad/body quotient decreased only slightly (Sprung & Borcherding, submitted). The his- tological investigation clearly demonstrates that the increase of lipids in the gonads of the 2 m group (Sprung & Borcherding, submitted) indicates degeneration (cf. for Lymnaea, de Jong-Brink et al., 1983). “Degeneration lipid- ique” is also reported for spawned Dreissena after termal stress by Tourari et al. (1988). Starvation also affected the filled gonads so that in the 9 m group reproductive capacity 40 BIELEFELD FIG. 17. Dreissena polymorpha, 2 m group, newly collected, ovaries, electron micrograph of an oocyte showing partial lysis. Golgi stacks (GS) budding off vesicles, which fuse to what are probably lytic vacuoles (VA), lysis of mitochondria (M). 16,000 x , scale bar 1 pm. FIG. 18. Dreissena polymorpha, 2 m group, refeeding, testes, electron micrograph of a haemocyte with glycogen rosettes (GLY) and lysed spermatozoa (SZ). 12,750 x , scale bar 1 pm. FIG. 19. Dreissena polymorpha, 2 m group, starvation, ovaries, electron micrograph of a granulocyte engulfing an oocyte. Golgi stack (GS), granules (GR) are shed onto the surface (arrow) of the degenerating oocyte (OC). 17,400 x, scale bar 1 рт. FIG. 20. Dreissena polymorpha, 2 m group, starvation, ovaries, electron micrograph of a haemocyte. Granules (GR) phagosome (PHS), lipid (LI), mitochondrium (M) with cristalline-like lattice of a periodicity of 12-15 nm. 20,800 x , scale bar 1 рт. FIG. 21. Dreissena polymorpha, 2 m group, starvation, ovaries, electron micrograph of a degenerating acinus. Thickened basement lamina, follicle cell (FC) with lipid, haemocyte (HC) filled with lipid. 3,020 x, scale bar 1 pm. HISTOLOGY OF STARVING DREISSENA 41 becomes reduced. This can be considered as reversible. Haemocytes In Dreissena, follicle cells and haemocytes (see Bayne et al., 1978, for Mytilus; Tourari et al., 1988, for Dreissena) are involved in the resorption of degenerating gonadal tissue. The invasion of the digestive gland by haemo- cytes in the 2 m group indicates that haemo- cytes transport the resorbed and metabolized gonadal material to the digestive gland. His- tological observations of the haemocytes could clarify some of the mechanisms of de- generation and conservation of energy for the individual. The conspicious cristalline-like lattice in the matrix of the mitochondria was also found by Mix et al. (1979) in atypical cells of Mytilus edulis, but their function remains unclear. ACKNOWLEDGEMENTS The author received a grant from the state Baden-Wúrttemberg. | would like to thank Dr. M. Sprung, Dr. J. Borcherding (Cologne) and Prof. Dr. V. Storch (Heidelberg) for their help during the experiment. David Russell cor- rected the English manuscript. LITERATURE CITED BAYNE, B. L., 1976, Aspects of reproduction in bi- valve molluscs. Pp. 432-448, in: WILEY, M., ed., Estuarine processes. Vol. | Academic Press, New York. BAYNE, B. L., D. L. HOLLAND, M. N. MOORE, D. M. LOWE 4 J. WIDDOWS, 1978, Further studies on the effects of stress in the adult on the eggs of Mytilus edulis. Journal of the Marine Biological Association United Kingdom, 58: 825-841. CHENG, T. C., 1981, Bivalves. Pp. 233-300, in: RATCLIFFE, N. A. 4 A. F. ROWLEY, eds., Inver- tebrate blood cells. Vol. 1 Academic Press, Lon- don. de JONG-BRINK, M., H. H. BOER 4 J. JOOSSE, 1983, The Mollusca. Pp. 297-355, in: ADIYODI, A. & В. ADIYODI, eds., Invertebrate Reproduc- tion. Vol |. Oogenesis, Ovulation, Oosorption. Wiley Chichester, Brisbane, London, New York, Toronto. GABBOTT, P. A. & B. L. BAYNE, 1973, Biochemi- cal effects of temperature and nutritive stress on Mytilus edulis (L.). Journal of the Marine Biolog- ical Association of the United Kingdom, 53: 269— 286. JANSSEN, H. H., 1981a, Zur Histologie der Mittel- darmdruse von Mytilus edulis. |. Ultrastrukturelle Merkmale. Zoologisches Jahrbuch für Anatomie, 106: 289-322. JANSSEN, H. H., 1981b, Zur Histologie der Mittel- darmdrüse von Mytilus edulis. Il. Veränderungen in der Tubulus Feinstruktur. Zoologisches Jahr- buch für Anatomie, 106: 527-567. KARNOVSKY, M. J., 1971, Use of ferrocyanide- reduced osmium tetroxide in electron micros- copy. Journal of Cell Biology, 51: 284. LANGTON, R. W., 1975, Synchrony in the digestive diverticula of Mytilus edulis L. Journal of the Ma- rine Biological Association of the United King- dom, 55: 221-229. MAXWELL, W. L., 1983, Mollusca. Pp. 275-319. In: ADIYODI, K. G. & R. G. ADIYODI, eds., Re- productive biology of invertebrates, Vol Il. Sper- matogenesis and sperm function. Wiley, Chich- ester, New York, Brisbane, Toronto, Singapore. МХ, М. C., J.C. HAWKES 4 А. К. SPARKS, 1979, Observation on the ultrastructure of large cells associated with putative neoplastic disorders of mussels, Mytilus edulis, from Yaquina Bay, Ore- gon. Journal of invertebrate pathology, 34: 41- 56. MOORE, M. N. & D. M. LOWE, 1985, Cytological and cytochemical measurements. Pp. 46-80, in: BAYNE, B. L., D. A. BROWN, R. D. DIXON, A. IVANOVICI, R. D. LIVINGSTONE, D. M. LOWE, M. N. MOORE, A. R. D. STEBBING & J. WID- DOWS, The effects of stress and pollution on marine animals. Praeger, New York. MORTON, B. S., 1969, Studies on the biology of Dreissena polymorpha, Pall. Il. Correlation of the rhythms of adductor activity, feeding, digestion and excretion. Proceedings of the Malacological Society, London, 38: 412-414. MORTON, B. S., 1983, Feeding and digestion in Bivalvia. Pp. 65-147. т: SALEUDDIN, А. $. М. & K. M. WILBUR, eds., The Mollusca physiology II, Vol 5. Academic Press, New York, London. PIPE, В. K., 1987, Oogenesis in the marine mussel Mytilus edulis: an ultrastructural study. Marine Bi- ology, 95: 405-414. RICHARDSON, K. C., L. JARRETT 4 E. H. FINKE, 1960, Embedding in epoxy resins for ultrathin sectioning in electron microscopy. Stain Technol- ogy 35: 313-325. ROBINSON, W. E., M. R. PENNINGTON & R. W. LANGTON, 1981, Variety of tubule types within the digestive glands of Mercenaria mercenaria (L.), Ostrea edulis (L.) and Mytilus edulis (L.). Journal of Experimental Marine Biology and Ecology, 54: 265-276. SPRUNG, M., 1987, Ecological requirements of de- veloping Dreissena polymorpha eggs. Archiv fur Hydrobiologie/Supplementum, 79: 69-86. SPRUNG, M. & J. BORCHERDING (submitted), Physiological and morphometric changes in Dre- issena polymorpha (Mollusca; Bivalvia) during a starvation period. 42 BIELEFELD THOMPSON, R. J., N. A. RATCLIFFE & B. L. TOURARI, A. L., С. CROCHARD & J. С. PIHAN, BAYNE, 1974, Effects of starvation on structure 1988, Action de la température sur le cycle de and function in the digestive gland of the mussel reproduction de Dreissena polymorpha (Pallas). (Mytilus edulis L.). Journal of the Marine Biolog- Etude “in situ” et au laboratoire. Haliotis, 17: 85— ical Association of the United Kingdom, 54: 699— 98. 712. WIDDOWS, J. & B. L. BAYNE, 1971, Temperature THOMPSON, R. J., C. J. BAYNE, M. N. MOORE & acclimation of Mytilus edulis with reference to its T. H. CAREFOOT, 1978, Haemolymph volume, energy budget. Journal of the Marine Biological changes in the biochemical composition of the Association of the United Kingdom, 51: 827-843. blood, and cytological responses of the digestive cells in Mytilus californianus Conrad, induced by nutritional, thermal and exposure stress. Journal of Comparative Physiology, 127: 287-298. Revised Ms. accepted 17 July 1990 MALACOLOGIA, 1991, 33(1-2): 43-54 LIFE TABLES OF FRESHWATER SNAILS OF THE GENUS BIOMPHALARIA (B. GLABRATA, B. ALEXANDRINA, B. STRAMINEA) AND OF ONE OF ITS COMPETITORS MELANOIDES TUBERCULATA UNDER LABORATORY CONDITIONS J.P. Pointier, J.L. Toffart & M. Lefevre Centre de Biologie et d’Ecologie Tropicale et Méditérranéenne, Laboratoire de Biologie Marine et Malacologie, Ecole Pratique des Hautes Etudes, Université de Perpignan, avenue de Villeneuve, 66025, Perpignan cedex, France ABSTRACT Life tables of several strains of Biomphalaria glabrata, B. alexandrina and B. straminea have been established under laboratory conditions and compared with that of Melanoides tuberculata, a prosobranch snail used as a competitor of snail vectors of schistosomiasis in experiments of biological control. Results indicate a great range of intraspecific and interspecific variation within the Biomphalaria and a quite distinct demographic strategy of M. tuberculata. Species of Bi- omphalaria have a high intrinsic rate of natural increase (0.70 to 1.01) and a short mean generation time (5.5 to 9.3 fortnights). In contrast, the thiarid snail has a low intrinsic rate of natural increase (0.24) and a very long mean generation time (25.77 fortnights). These results fit well with several field observations showing that this snail is able to reach and maintain very high densities in permanent and stable habitats. In such habitats, the competition with pulmo- nate snails might be very strong and to the advantage of the thiarid species. Key words: life tables, laboratory, Biomphalaria glabrata, B. alexandrina, B. straminea, Mel- anoides tuberculata. INTRODUCTION Several authors have discussed the use of competitors to control pulmonate snail inter- mediate hosts of schistosomiasis (Michelson & Dubois, 1974; Malek & Malek, 1978; Frand- sen & Madsen, 1979; Barbosa et al., 1983; Madsen, 1984). Among the species cited by these authors, few have been the object of detailed field or laboratory investigations and rigorous research has been encouraged (Mc- Cullough, 1981; World Health Organization, 1984). In the Caribbean area, field observations of an introduced Oriental species belonging to the Thiaridae, Melanoides tuberculata, demonstrated its capacity massively to colo- nize many types of habitats, while at the same time limiting, even excluding, different Bi- omphalaria species (Prentice, 1983; Pointier, 1989; Pointier et al., 1989). Consequently, it is important to study the biology of this snail and of its target species. In this paper, life table parameters of two Caribbean and one African species of Bi- omphalaria (B. glabrata, B. straminea and B. alexandrina) have been established under 43 laboratory conditions and compared with those of M. tuberculata. MATERIAL AND METHODS Adult snails were collected in the field and transferred to laboratory aquaria: B. glabrata from Céligny pond and Dubelloy marsh, Guadeloupe; B. alexandrina from Kalyub and Kafr al Hamza canals, Egypt; B. straminea from Epinette and Madame Rivers, Marti- nique; M. tuberculata from Pointe-a-Pitre ca- nal, Guadeloupe. Viable eggs laid by these snails were used to start the experiments. In the case of the hermaphroditic В. glabrata, В. alexandrina and B. straminea, 81 eggs (Céligny pond), 97 eggs (Dubelloy marsh), 69 eggs (Kalyub canal), 71 eggs (Kafr al Hamza canal), 114 eggs (Epinette River) and 75 eggs (Madame River) were used, respectively. In the case of the parthenogenetic M. tuberculata, 40 newly liberated juveniles were taken. During the ex- periment, all the thiarid snails were tested separately for releasing juveniles in order to verify that they were all females. 44 POINTIER, TOFFART & LEFEVRE All newly hatched snails were transferred to small containers (200 cc) and then put in aquaria of one litre, four snails per aquarium to minimize crowding. Snails were reared in the presence of aquatic moss, Hygrohypnum eugyruum (Hypnacea) and fed with fresh and dried lettuce. Dechlorinated tap water was re- newed one or twice a week. The main chem- ical characteristics of the water were the fol- lowing: Ca = 90.0 mg/l; Mg= 4.4 mg/l; SO, = 16.9 mg/l; Cl= 23.8 mg/l; НСО. = 112 mg/l. Water temperature was maintained constant at 25°+ 1°C. Light was artificial and photope- riodically balanced (LD 12-12). Measures of growth (maximum size in mm), survivorship (|, = proportion of surviv- als from the original number of viable eggs or newly liberated juveniles) and fecundity (m, = mean viable eggs producted or mean newly liberated juveniles per female) were re- corded fortnightly. Cultures were stopped when numbers decreased to less than 20 to 30% of the original population. Below this value, the remaining snails were considered to be too different from the mean population. Life table parameters were calculated accord- ing to the methods of Birch (1948) and An- drewartha & Birch (1954). The following pa- rameters were calculated: Ro (net reproductive rate) = Ут, ; г (intrinsic rate of increase) from Sl,;m,-e ™ = 1; В (finite rate of increase) from В = e'; T (approximation of mean generation time in fort- nights) = Log R,/Log R. In order to compare growth of the different species, values of different parameters were calculated using the growth equation of Von Bertalanffy (1938): L= L, (1-е *“ 0) in which |, is the size of the snail at time t; LL, the value of L, when growth rate = О; К, a characteristic constant of growth; t, the age of the snail; tp, the hypothetical time at which the snail would have the O size. RESULTS Hatching At 25°C, the maturation time of eggs is 8-9 days for B. glabrata, B. alexandrina and B. straminea. The time at which eggs were laid has been taken as time=0 for calculations. For M. tuberculata, the maturation of eggs oc- curs in a brood pouch in which the different stages of development take place (Rama- moorthi, 1955). The maturation time at 25°C in the brood pouch is unknown and the time of juvenile release has been taken as time=0. Under laboratory conditions, 96.3% and 93.8% of the B. glabrata eggs from Celigny pond and Dubelloy marsh, 92.7% and 87.3% of the B. alexandrina eggs from Kalyub and Kafr al Hamza canals, and 89.5% and 90.7% of the B. straminea eggs from Epinette and Madame Rivers, respectively, hatched. Growth in Size The growth of the Biomphalaria and Mel- anoides populations is shown in Figure 1. Be- cause the results for the two B. glabrata, B. alexandrina and B. straminea populations are quite similar, the growth curve of only one population of each species is shown. The growth of B. glabrata is quite different from those of B. alexandrina and B. straminea and especially from that of M. tuberculata. B. glabrata reaches a size of 17.3mm (which corresponds to 95% of its maximal theoretical size, L..) in 21.9 fortnights (k= 0.13). B. alex- andrina has a similar growth rate (k= 0.15) but reaches a size of 13.6 mm (95% of L..) in 20.6 fortnights, and B. straminea reaches a size of 8.4 mm (95% of L.) in 15.9 fortnights (k= 0.19) (Table 1). The growth parameters of T. tuberculata are quite different: k= 0.04 and the species reaches a size of 30.2 mm (95% of L..) in 85.1 fortnights (Table 1). Survivorship and Fecundity The results, presented in Figures 2 and 3 and Tables 2—6, show marked differences between the pulmonates, B. glabrata, B. al- exandrina and B. straminea, and the proso- branch, M. tuberculata. Because survival curves of the two populations of each species of Biomphalaria are similar, only one curve and table are presented for each species. Marked differences can be noted among the three species, B. glabrata having a better sur- vival than B. alexandrina and B. straminea. The decline of populations of the species of Biomphalaria occurs between 6 and 9 months of culture. In comparison, decline of M. tuber- culata occurs much later: more than 50% of the individuals survived after five years of cul- ture (Fig. 3). Its fecundity is much lower than those of species of Biomphalaria however: a maximum of 18 newly liberated juveniles per female per fortnight against 150-400 fertile LIFE TABLES OF FRESHWATER SNAILS 45 mm 30 25 15 10 0 10 20 30 40 50 FORTNIGHTS FIG. 1. Growth curves for Biomphalaria glabrata (Bg), B. alexandrina (Ba), B. straminea (Bs) and Mel- anoides tuberculata (Mt) under laboratory conditions. Above: calculated growth curves using Von Bertalanffy equation. Below: observed growth curves and standard deviations. Arrows indicate age of first reproduction. 46 POINTIER, TOFFART & LEFEVRE TABLE 1. Values of parameters k and L.. from growth equation of Von Bertalanffy for several species of Biomphalaria and Melanoides tuberculata reared in laboratory. Cohorts Biomphalaria glabrata Celigny pond, Guadeloupe Biomphalaria glabrata Dubelloy marsh, Guadeloupe Biomphalaria glabrata St Lucia (data from Sturrock & Sturrock, 1972) Biomphalaria alexandrina Kalyub canal, Egypt Biomphalaria alexandrina Kafr al Hamza canal, Egypt Biomphalaria straminea Epinette River, Martinique Biomphalaria straminea Madame River, Martinique Biomphalaria pfeifferei Tanzania (data from Sturrock, 1966) Biomphalaria pfeifferi Zaire (data from Loreau & Baluku, 1987) Melanoides tuberculata Pointe-a-Pitre canal, Guadeloupe Water Temp. 25°C 25°C 25°C 256 2550 25°C 25°C 25°C 18°C-24°C 25°C K 0.134 0.121 0.123 0.15 0.131 0.187 0.236 0.192 0.05 0.035 Time to reach = 18.2 19.39 20.29 14.35 14.71 8.81 7.6 13.25 12.6 31.8 95% of [E in fortnights 21.9 24.9 24.5 20.6 229 15.9 12.8 16.1 58.1 85.1 TABLE 2. Life table for Biomphalaria glabrata, Céligny pond, Guadeloupe, reared in laboratory (25°C). Pivotal age in fortnights Survival Ix 1.0000 1.0000 0.9630 0.9383 0.9383 0.9383 0.9383 0.9383 0.9383 0.9383 0.9383 0.9383 0.9383 0.8888 0.8888 0.8395 0.7405 0.6914 0.5926 0.4938 0.4444 0.3456 Fecundity mx 0.0000 0.0000 0.0000 0.0000 0.0000 3.7600 23.3000 69.1700 130.1200 168.8300 222.8700 210.9500 97.3700 143.1700 155.7500 136.4500 97.4100 42.1000 37.1200 44.3000 35.0700 25.2500 Ro= Net reprod. rate Ro 3.5280 21.8624 64.9022 122.0916 158.4132 209.1189 197.9344 91.3623 127.2495 138.4306 114.5498 72.1321 29.1079 21.9973 21.8753 15.5851 8.7264 1418.8670 Finite Mean rate of generation increase time R 1 1.3234 4.4994 1.8369 5.3188 2.1060 6.0460 2.2345 6.6647 2.2881 7.1471 2.3141 7.5839 2.3238 7.8944 2.3257 8.0185 2.3268 8.1755 2.3273 8.3276 2.3275 8.4404 2.3275 8.5068 2.3275 8.5326 2.3275 8.5518 2.3275 8.5705 2.3275 8.5836 2.3275 8.5909 Intrinsic rate of increase rm rm(%) 0.2802 33.1676 0.6081 71.9816 0.7448 88.1629 0.8040 95.1705 0.8277 97.9759 0.8390 99.3134 0.8432 99.8106 0.8440 99.9053 0.8445 99.9645 0.8447 99.9882 0.8448 100.0000 0.8448 100.0000 0.8448 100.0000 0.8448 100.0000 0.8448 100.0000 0.8448 100.0000 0.8448 100.0000 Stable age distribution (%) 47.0429 30.8353 12.7580 5.3408 2.2946 0.9859 0.4236 0.1820 0.0782 0.0336 0.0144 0.0062 0.0027 0.0011 0.0005 0.0002 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 LIFE TABLES OF FRESHWATER SNAILS 47 Ix mx 1,0 be 400 28 300 0,6 200 0,4 mx 0,2 100 0,0 0 0 2 4 6 8 101214 1618 20 22 Biomphalaria glabrata age in fortnights Ix mx 1,0 400 0,8 Ix 300 0,6 200 0,4 0,2 100 0,0 0 0 2 4 6 8 1012 14 16 18 20 22 Biomphalaria straminea age in fortnights 1,0 400 QE 300 0,6 Ix 200 0,4 0,2 100 0,0 0 0 2 4 6 8 1012141618 20 22 Biomphalaria alexandrina age in fortnights FIG. 2. Graphical representation of life tables of three species of Biomphalaria reared in laboratory. |,: survivorship (proportion of survivals from the original number of viable eggs); m,: fecundity (mean viable egg production per snail per fortnight). 48 POINTIER, TOFFART & LEFEVRE Ix mx 1,0 25 0,8 20 0,6 к 15 0,4 10 mx 0,2 5 0,0 0 0 20 40 60 80 100 120 140 age in fortnights FIG. 3. Graphical representation of life tables of a population of Melanoides tuberculata reared in labora- tory. |,: survivorship (proportion of the original number of newly liberated juveniles); m,: fecundity (mean newly liberated juveniles per female per fortnight). TABLE 3. Life table for Biomphalaria alexandrina, Kafr al Hamza, Egypt, reared in laboratory (25°C). Intrinsic Pivotal Net Finite Mean ео! Stable age in reprod. rate of generation a age fortnights Survival Fecundity rate increase time __Inerease ____. ‘distribution x Ix mx Ro R ЛГ rm rm(%) (%) 0.0 1.0000 0.0000 44.1399 0.5 1.0000 0.0000 31.0707 les 0.8732 0.0000 13.4432 2:5 0.8591 0.0000 6.5535 3.5 0.7183 0.0000 27150 4.5 0.6197 4.3600 2.7019 1.2472 4.4996 0.2209 31.4583 1.1606 5.5 0.5352 20.6600 11.0572 1.6461 5.2602 0.4984 70.9769 0.4967 6.5 0.5070 42.1400 21.3650 1.8291 5.8941 0.6038 85.9869 0.2331 129 0.4507 68.7200 30.9721 1.9167 6.4419 0.6506 92.6516 0.1027 8.5 0.4084 1203400 49.1469 1.9699 7.0015 0.6780 96.5537 0.0461 9.5 0.3662 133.7700 48.9866 1.9921 7.4017 0.6892 98.1487 0.0205 10.5 0.3662 171.0800 62.6495 2.0051 TET A 0.6957 99.0743 0.0101 11.5 0.3099 204.2700 63.3033 2.0111 8.1158 0.6987 99.5016 0.0043 12:5 0.2958 272.6200 80.6410 2.0150 8.4438 0.7006 99.7721 0.0020 13.5 0.2817 287.2000 80.9042 2.0168 8.7143 0.7015 99.9003 0.0009 14.5 0.2535 338.2800 85.7540 2.0176 8.9570 0.7019 99.9573 0.0004 15:5 0.1690 391.0800 66.0925 2.0180 9.1196 0.7021 99.9858 0.0001 16.5 0.1549 333.8200 51.7087 2.0182 9.2354 0.7022 100.0000 0.0001 175 0.1127 399.2500 44.9955 2.0182 9.3299 0.7022 100.0000 0.0000 18.5 0.0986 110.0000 10.8460 2.0182 9.3518 0.7022 100.0000 0.0000 Ro= 711.1244 LIFE TABLES OF FRESHWATER SNAILS 49 TABLE 4. Life table for Biomphalaria straminea, Epinette River, Martinique, reared in laboratory (25°C). Pivotal Net Finite Mean ue Stable age in reprod. rate of generation age fortnights Survival Fecundity rate increase time New as? distribution x Ix mx Ro li rm rm(%) (%) 0.0 1.0000 0.0000 49.0990 0.5 1.0000 0.0000 31.5505 1.5 0.8947 0.0000 11.6560 2.5 0.8772 0.0000 4.7189 3.5 0.7982 2.0100 1.6044 1.1447 3.4992 0.1351 15.2742 1.7730 4.5 0.7895 24.9000 19.6586 2.0049 4.3947 0.6956 78.6433 0.7241 5:5 0.7632 48.3000 36.8626 2.2646 4.9702 0.8174 92.4138 0.2891 6.5 0.7193 65.6100 47.1933 2.3589 5.4265 0.8582 97.0266 0.1125 YES 0.7193 79.7300 57.3498 2.3984 5.8204 0.8748 98.9033 0.0464 8.5 0.7105 66.5400 47.2767 2.4107 6.0767 0.8799 99.4799 0.0189 9.5 0.6667 93.4200 62.2831 2.4172 6.3524 0.8826 99.7852 0.0073 10.5 0.6140 136.3400 83.7128 2.4206 6.6457 0.8840 99.9435 0.0028 1:5 0.5965 87.5600 52.2295 2.4215 6.7975 0.8844 99.9887 0.0011 12.5 0.3158 94.1400 29.7294 2.4218 6.8762 0.8845 100.0000 0.0002 13:5 0.1228 141.9300 17.4290 2.4218 6.9203 0.8845 100.0000 0.0000 14.5 0.1053 61.8300 6.5107 2.4218 6.9364 0.8845 100.0000 0.0000 Ro= 461.8398 eggs produced per species of Biomphalaria per fortnight (Figs. 2,3). Calculated Life Table Parameters The following life table parameters have been calculated: intrinsic rate of natural in- crease (r), mean generation time (T), finite rate of increase (R) and net reproductive rate (Ro) (Tables 2-6). B. glabrata, B. alexandrina, and B. stra- minea have a high intrinsic rate of natural in- crease (0.84 and 0.86, 0.78 and 0.70, 0.88 and 1.01, respectively) and a short mean gen- eration time (8.59 and 8.87, 8.60 and 9.35, 6.94 and 5.47 fortnights, respectively). M. tu- berculata, on the contrary, has a low intrinsic rate of natural increase (0.24), and a very long mean generation time (25.77 fortnights, calculated for 135 fortnights of culture). DISCUSSION AND CONCLUSIONS Several authors have studied growth and demography of tropical freshwater pulmo- nates in the laboratory; a review of the main results obtained under optimal conditions of temperature is presented in Tables 1 and 6. Precise comparison of these results is dif- ficult because of differences in laboratory rearing conditions (food, water volume, water quality, etc.). Several aspects of the results are noteworthy, however. Among the species of Biomphalaria, B. al- exandrina has the lowest intrinsic rate of nat- ural increase (0.70—0.78). This is due to mor- tality, which is important in all stages of development (Fig. 2). In contrast, the mortal- ity of B. glabrata is quite different and the in- trinsic rate of increase is higher (0.84—0.88). In regard to B. straminea, the intrinsic rate of increase is slightly higher (0.88-1.01) but the mean generation time is shorter. This species has a distinct demographic strategy and these results fit well with field data on the ecology of this snail: B. straminea is more re- sistant to drought and better adapted to tem- porary and fluctuating environments than is B. glabrata (Barbosa, 1973; Guyard & Point- ier, 1979). Results for the B. pfeifferi group show a great range of variation. According to the or- igin of the strain, the intrinsic rate of increase can vary greatly (0.24 to 2.37; Table 6). Care is needed in comparing these results, how- ever, owing to the different experimental con- ditions used by the authors: r=0.24 in Zaire, but with water temperature varying between 18.5°C and 24°C and different food (Loreau & Baluku, 1987); r= 0.48 in Rhodesia, but the study was disturbed by atypical migratory be- haviours (Shiff & Garnett, 1967); r= 0.86 in Tanzania (Sturrock, 1966); r= 2.37 in South Africa (De Kock & Van Eeden, 1981), but us- 50 POINTIER, TOFFART & LEFEVRE TABLE 5. Life table for Melanoides tuberculata, Pointe-a-Pitre canal, Guadeloupe, reared in laboratory (25°C). Pivotal Net Finite Mean ОВС Stable age in reprod. rate of generation Я аде fortnights Survival Fecundity rate increase time Incledse distrib. x Ix mx Ro R T rm rm(%) (%) 0.0 1.0000 0.0000 20.4327 0.5 0.9750 0.0000 17.6638 ES 0.9500 0.0000 13.5305 25 0.9375 0.0000 10.4971 3.5 0.9375 0.0000 8.2524 4.5 0.9375 0.0000 6.4876 5:5 0.9375 0.0000 5.1003 6.5 0.9375 0.0000 4.0096 7.5 0.9125 0.0000 3.0681 8.5 0.9000 0.0000 2.3790 9.5 0.8875 0.3040 0.2698 0.8712 9.5002 —0.1379° —57.3150 1.8443 10.5 0.8875 0.6377 0.5660 0.9826 10.1941 —0.0176 —73150 1.4499 11.5 0.8875 2.9710 2.6368 1.1180 11.1649 0.1115 46.3425 1.1398 12.5 0.8875 4.1159 3.6529 1.1810 11.8003 0.1664 69.1604 0.8961 13.5 0.8875 3.1014 2.7525 1.2061 12.2214 0.1874 77.8886 0.7045 14.5 0.8750 1.8971 1.6600 1.2164 12.4841 0.1959 81.4214 0.5460 15.5 0.8750 3.6176 3.1654 1.2300 12.9858 0.2070 86.0349 0.4293 16.5 0.8750 4.6176 4.0404 1.2416 13.5437 0.2164 89.9418 0.3375 175 0.8750 5.5882 4.8897 1.2513 14.1064 0.2242 93.1837 0.2653 18.5 0.8750 4.6176 4.0404 1.2570 14.5189 0.2287 95.0540 0.2086 19.5 0.8750 4.7500 4.1563 1.2611 14.9156 0.2320 96.4256 0.1640 20.5 0.8750 2.3382 2.0459 1.2629 15.0930 0.2334 97.0075 0.1289 21.5 0.8750 3.9853 3.4871 1.2647 15.4203 0.2348 97.5894 0.1013 22.5 0.8750 3.2500 2.8438 1.2659 15.6660 0.2358 98.0050 0.0797 23.5 0.8750 2.2353 1.9559 1.2667 15.8271 0.2364 98.2544 0.0626 24.5 0.8750 2.7941 2.4448 1.2673 16.0317 0.2369 98.4622 0.0492 25:5 0.8750 4.3235 3.7831 1.2681 16.3339 0.2375 98.7116 0.0387 26.5 0.8750 9.4412 8.2611 1.2693 16.9263 0.2385 99.1272 0.0304 27.5 0.8750 7.3636 6.4432 1.2701 17.3343 0.2391 99.3766 0.0239 28.5 0.8500 5.8235 4.9500 1.2706 17.6207 0.2395 99.5428 0.0183 29.5 0.8500 5.3235 4.5250 1.2709 17.8746 0.2397 99.6259 0.0144 30.5 0.8000 4.5625 3.6500 1.2711 18.0643 0.2399 99.7091 0.0106 31.5 0.7750 3.4839 2.7000 1.2712 18.2018 0.2400 99.7506 0.0081 32.5 0.7750 5.0645 3.9250 1.2714 18.3964 0.2401 99.7922 0.0064 33.5 0.7750 5.6333 4.3658 1.2715 18.6025 0.2402 99.8337 0.0050 34.5 0.7500 5.3333 4.0000 1.2716 18.7814 0.2403 99.9753 0.0038 35.5 0.7000 8.8214 6.1750 1.2716 19.0540 0.2403 99.8753 0.0028 36.5 0.7000 7.9286 5.5500 1.2718 19.2766 0.2404 99.9169 0.0022 37.5 0.6500 5.6154 3.6500 1.2718 19.4216 0.2404 99.9169 0.0016 38.5 0.6500 5.6538 3.6750 1.2719 19.5545 0.2405 99.9584 0.0013 39.5 0.6500 3.8846 2.5250 1.2719 19.6486 0.2405 99.9584 0.0010 40.5 0.6250 5.4800 3.4250 1.2719 19.7730 0.2405 99.9584 0.0007 41.5 0.6250 9.7200 6.0750 1.2719 19.9849 0.2405 99.9584 0.0006 42.5 0.6250 10.9200 6.8250 1.2719 20.2107 0.2405 99.9584 0.0005 43.5 0.6250 14.9200 9.3250 1.2719 20.5006 0.2405 99.9584 0.0004 44.5 0.6250 14.3600 8.9750 12719 20.7618 0.2405 99.9584 0.0003 45.5 0.6250 9.7200 6.0750 1.2720 20.9211 0.2406 100.0000 0.0002 46.5 0.6250 6.2800 3.9250 1.2720 21.0260 0.2406 100.0000 0.0002 47.5 0.6250 7.2000 4.5000 1.2720 21.1432 0.2406 100.0000 0.0001 48.5 0.6250 4.7500 2.9750 1.2720 21.2188 0.2406 100.0000 0.0001 49.5 0.6250 7.8333 4.8958 1.2720 21.3404 0.2406 100.0000 0.0001 50.5 0.6000 10.0833 6.0500 1.2720 21.4860 0.2406 100.0000 0.0001 51.5 0.6000 13.3750 8.0250 1.2720 21.6715 0.2406 100.0000 0.0001 52.5 0.6000 17.4166 10.4500 1.2720 21.9012 0.2406 100.0000 0.0000 53.5 0.6000 17.6250 10.5750 1.2720 22.1215 0.2406 100.0000 0.0000 LIFE TABLES OF FRESHWATER SNAILS 51 TABLE 5. (continued) Pivotal Net Finite Mean eee Stable age in reprod. rate of generation age fortnights Survival Fecundity rate increase time Сы distrib. x Ix mx Ro R 1 rm rm(%) (%) 54.5 0.6000 13.9166 8.3500 1.2720 22.2875 0.2406 100.0000 0.0000 55.5 0.6000 12.5416 7.5250 1.2720 22.4317 0.2406 100.0000 0.0000 56.5 0.5750 19.9130 11.4500 1.2720 22.6418 0.2406 100.0000 0.0000 57.5 0.5750 14.8695 8.5500 1.2720 22.7921 0.2406 100.0000 0.0000 58.5 0.5750 16.3478 9.4000 1.2720 22.9513 0.2406 100.0000 0.0000 59.5 0.5750 18.0000 10.3500 1.2720 23.1198 0.2406 100.0000 0.0000 60.5 0.5750 18.6521 10.7250 1.2720 23.2875 0.2406 100.0000 0.0000 61.5 0.5750 17.3913 10.0000 1.2720 23.4380 0.2406 100.0000 0.0000 62.5 0.5750 15.4347 8.8750 1.2720 23.5671 0.2406 100.0000 0.0000 63.5 0.5750 17.8695 10.2750 1.2720 23.7118 0.2406 100.0000 0.0000 64.5 0.5750 11.0869 6.3750 1.2720 23.7991 0.2406 100.0000 0.0000 65.5 0.5750 11.4782 6.6000 1.2720 23.8875 0.2406 100.0000 0.0000 66.5 0.5750 9.2600 5.3245 1.2720 23.9576 0.2406 100.0000 0.0000 67.5 0.5750 7.7826 4.4750 1.2720 24.0155 0.2406 100.0000 0.0000 68.5 0.5750 9.9130 5.7000 1.2720 24.0882 0.2406 100.0000 0.0000 69.5 0.5750 9.3043 5.3500 1.2720 24.1553 0.2406 100.0000 0.0000 70.5 0.5750 10.4783 6.0250 1.2720 24.2295 0.2406 100.0000 0.0000 ZAS 0.5750 9.3043 5.3500 1.2720 24.2944 0.2406 100.0000 0.0000 72:5 0.5750 8.7400 5.0255 1.2720 24.3544 0.2406 100.0000 0.0000 73.5 0.5750 6.6957 3.8500 1.2720 24.3998 0.2406 100.0000 0.0000 74.5 0.5750 6.7826 3.9000 1.2720 24.4452 0.2406 100.0000 0.0000 79:5 0.5750 9.1304 5.2500 1.2720 24.5057 0.2406 100.0000 0.0000 76.5 0.5750 10.1738 5.8499 1.2720 24.5720 0.2406 100.0000 0.0000 71-9 0.5750 8.9565 5.1500 1.2720 24.6296 0.2406 100.0000 0.0000 78.5 0.5750 10.2174 5.8750 1.2720 24.6942 0.2406 100.0000 0.0000 79.5 0.5750 9.2600 5.3245 1.2720 24.7520 0.2406 100.0000 0.0000 80.5 0.5750 12.2609 7.0500 1.2720 24.8273 0.2406 100.0000 0.0000 81.5 0.5750 95217 5.4750 1.2720 24.8848 0.2406 100.0000 0.0000 82.5 0.5750 6.9130 3.9750 1.2720 24.9261 0.2406 100.0000 0.0000 83.5 0.5750 7.7826 4.4750 1.2720 24.9720 0.2406 100.0000 0.0000 84.5 0.5750 3.9130 2.2500 1.2720 24.9950 0.2406 100.0000 0.0000 85.5 0.5750 5.0870 2.9250 1.2720 25.0246 0.2406 100.0000 0.0000 86.5 0.5750 4.0870 2.3500 1.2720 25.0482 0.2406 100.0000 0.0000 87.5 0.5750 6.0000 3.4500 1.2720 25.0827 0.2406 100.0000 0.0000 88.5 0.5750 7.0000 4.0250 1.2720 25.1225 0.2406 100.0000 0.0000 89.5 0.5750 3.2174 1.8500 1.2720 25.1407 0.2406 100.0000 0.0000 90.5 0.5750 5.6522 3.2500 1.2720 25.1725 0.2406 100.0000 0.0000 91.5 0.5500 3.1816 1.7499 1.2720 25.1895 0.2406 100.0000 0.0000 92.5 0.5500 4.5000 2.4750 1.2720 25.2134 0.2406 100.0000 0.0000 93.5 0.5500 4.2727 2.3500 1.2720 25.2360 0.2406 100.0000 0.0000 94.5 0.5500 7.2272 3.9750 1.2720 25.2740 0.2406 100.0000 0.0000 95.5 0.5500 4.8636 2.6750 1.2720 25.2993 0.2406 100.0000 0.0000 96.5 0.5500 4.6363 2.5500 1.2720 25.3233 0.2406 100.0000 0.0000 97.5 0.5250 3.2857 1.7250 1.2720 25.3395 0.2406 100.0000 0.0000 98.5 0.5250 3.0952 1.6250 1.2720 25.3547 0.2406 100.0000 0.0000 99.5 0.5250 1.6666 0.8750 1.2720 25.3628 0.2406 100.0000 0.0000 100.5 0.5250 2.9524 1.5500 1.2720 25.3772 0.2406 100.0000 0.0000 101.5 0.5250 4.3809 2.3000 1.2720 25.3985 0.2406 100.0000 0.0000 102.5 0.5250 4.3809 2.3000 1.2720 25.4196 0.2406 100.0000 0.0000 103.5 0.5250 3.4286 1.8000 1.2720 25.4361 0.2406 100.0000 0.0000 104.5 0.5250 6.1905 3.2500 1.2720 25.4657 0.2406 100.0000 0.0000 105.5 0.5250 5.4286 2.8500 1.2720 25.4915 0.2406 100.0000 0.0000 106.5 0.5250 6.6190 3.4750 1.2720 25.5227 0.2406 100.0000 0.0000 107.5 0.5250 6.3333 3.3250 1.2720 25.5523 0.2406 100.0000 0.0000 108.5 0.5250 6.0476 3.1750 1.2720 25.5804 0.2406 100.0000 0.0000 continued 52 POINTIER, TOFFART & LEFEVRE TABLE 5. (continued) Pivotal Net age in reprod. fortnights Survival Fecundity rate x Ix mx Ro 109.5 0.5250 5.2381 2.7500 110.5 0.5250 4.7143 2.4750 MAS 0.5250 3.9048 2.0500 1125 0.5250 3.2857 1.7250 113.5 0.5250 2.5714 1.3500 114.5 0.5250 1.7619 0.9250 1155 0.5250 2.1429 1.1250 116.5 0.5000 2.7000 1.3500 117-5 0.5000 1.9500 0.9750 118.5 0.5000 1.6500 0.8250 119.5 0.5000 1.7000 0.8500 120.5 0.5000 0.8500 0.4250 121.5 0.5000 1.2000 0.6000 122:5 0.5000 1.4000 0.7000 123.5 0.5000 1.8000 0.9000 124.5 0.5000 0.9000 0.4500 125:5 0.4750 0.7895 0.3750 126.5 0.4750 0.5785 0.2748 127.5 0.4750 0.4210 0.2000 128.5 0.4500 0.4737 0.2132 129.5 0.3500 0.6429 0.2250 130.5 0.3250 0.3846 0.1250 131.5 0.3250 0.3077 0.1000 132.5 0.2750 0.9091 0.2500 133.5 0.2500 1.5000 0.3750 Ro = 492.5179 ing artificial food and the authors do not indi- cate the time taken as time=O for their cal- culations (egg laying time or hatching time?). This omission is important because a slight difference in the pivotal age greatly alters the calculations of the intrinsic rate of increase; for example, r= 2.37 becomes r= 1.66 if a fortnight is added in the pivotal age used by De Kock & Van Eeden (1981). Some other studies carried out on other pul- monates such as Bulinus globosus (r= 0.66), B. tropicus (r= 3.37) and Lymnaea natalensis (r= 1.81) also show a great range of variation (Shiff, 1964; De Kock & Van Eeden, 1985). In comparison, calculated life table param- eters of M. tuberculata indicate a quite differ- ent demographic strategy. This species has a very low intrinsic rate of natural increase (r= 0.24) but a very long mean generation time (T= 25.77 fortnights). These results support numerous field ob- servations, which record that this snail can reach and maintain very high densities for a Finite rate of 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 1.2720 Mean Е Stable generation р аде increase time INGIE ASS, distrib. Г rm rm(%) (%) 25.6047 0.2406 100.0000 0.0000 25.6263 0.2406 100.0000 0.0000 25.6442 0.2406 100.0000 0.0000 25.6591 0.2406 100.0000 0.0000 25.6708 0.2406 100.0000 0.0000 25.6788 0.2406 100.0000 0.0000 25.6885 0.2406 100.0000 0.0000 25.7001 0.2406 100.0000 0.0000 25.7084 0.2406 100.0000 0.0000 25.7155 0.2406 100.0000 0.0000 25.7227 0.2406 100.0000 0.0000 25.7264 0.2406 100.0000 0.0000 257315 0.2406 100.0000 0.0000 25.7374 0.2406 100.0000 0.0000 25.7451 0.2406 100.0000 0.0000 25.7489 0.2406 100.0000 0.0000 25.7521 0.2406 100.0000 0.0000 25.7544 0.2406 100.0000 0.0000 25.7561 0.2406 100.0000 0.0000 25.7579 0.2406 100.0000 0.0000 25.7598 0.2406 100.0000 0.0000 25.7608 0.2406 100.0000 0.0000 25.7617 0.2406 100.0000 0.0000 25.7638 0.2406 100.0000 0.0000 25.7670 0.2406 100.0000 0.0000 long time in permanent and stable habitats (Murray & Wopschall, 1965, 4,798 snails/ft?; Roessler et al., 1977, about 7,000 to 37,000 snails/m*; Pointier et al., 1989, between 9,900 and 13,400 snails/m?). In such habitats, thiarid snails have a competitive advantage and in particular situations their advantage leads to the elimination of this group of snails (Pointier et al., 1989). The mechanisms of the competition remain to be studied more pre- cisely but it is established that competition oc- curs whenever thiarids reach and maintain high densities. In that case, there is probably competition for space or by interference. Thiarids and Biomphalaria also have a similar diet (detritivorous and microphagous, includ- ing microalgae, bacteria, organic materials, etc.) and it is therefore probable that compe- tition for food occurs in some cases. Consequently, it can be predicted that bio- logical control using thiarid snails is likely to be efficient in permanent and stable habitats, but not in temporary or unstable ones. LIFE TABLES OF FRESHWATER SNAILS 53 TABLE 6. Calculated parameters of life tables for some tropical freshwater pulmonates and for a population of Melanoides tuberculata reared in laboratory. r = intrinsic rate of increase, R = finite rate of increase, R, = net reproduction rate, T = mean generation time. Water Origin of Snails Temp. Ro r R T (fortnights) Biomphalaria glabrata 25°C 1418.87 0.84 2.33 8.59 Celigny pond, Guadeloupe Biomphalaria glabrata 25°C 2143.52 0.86 2.37 8.87 Dubelloy marsh, Guadeloupe Biomphalaria glabrata 25°C 2054.1 0.88 2.42 8.63 St Lucia (data from Sturrock, 1972) Biomphalaria alexandrina 25°C 806.83 0.78 ANA 8.6 Kalyub canal, Egypt Biomphalaria alexandrina 256 Hive O7 2.02 9.35 Kafr al Hamza canal, Egypt Biomphalaria straminea 25°C 461.84 0.88 2.42 6.94 Epinette River, Martinique Biomphalaria straminea 25°C 255.82 1.01 2.75 5.47 Madame River, Martinique Biomphalaria pfeifferi 25°C 111.43 0.48 1.62 9.79 Rhodesia (data from Shiff & Garnett, 1967) Biomphalaria pfeifferi 25°C 182.14 0.86 2.36 6.05 Tanzania (data from Sturrock, 1966) Biomphalaria pfeifferi 26°C 2696.19 2.37 10.72 5.89 South Africa (data from De Kock & Van Eeden, 1981) Biomphalaria pfeifferi 18.5°-24°C 216.6 0.24 1627 11.36 Zaire (data from Loreau & Baluku, 1987) Bulinus globosus 25°C 467.31 0.66 1.93 7 Rhodesia (data from Shiff, 1964) Bulinus tropicus 26°C 1655.64 3.37 29.15 3.36 South Africa (data from De Kock & Van Eeden, 1985) Lymnaea natalensis 26°C 1371.91 1.81 6.11 4.99 South Africa (data from De Kock & Van Eeden, 1985) Melanoides tuberculata 25°C 492.5 0.24 127, 25.77 Pointe-a-Pitre canal, Guadeloupe ACKNOWLEDGEMENTS REFERENCES ANDREWARTHA, H. G. 4 L. C. BIRCH, 1954, The distribution and abundance of animals. University of Chicago Press, Chicago, 782 pp. This research received financial support from the UNDP/World Bank/WHO special programme for Research and Training in BARBOSA, Е. S., 1973, Possible competitive dis- Tropical Diseases. We should like to thank Dr placement and evidence of hybridization be- Fergus McCullough, formerly Senior Scien- tween two Brazilian species of planorbid snails tist, WHO, VBC, for his critical review of the inhabiting northeastern Brazil. Memorias do In- manuscript. stituto Oswaldo Cruz, 76: 361-366. 54 POINTIER, TOFFART & LEFEVRE BARBOSA, F. S., D. P. PEREIRA DA COSTA & F. ARRUDA, 1983, Competitive interactions be- tween species of freshwater snails |. Laboratory: la. General methodology. Memorias do Instituto Oswaldo Cruz, 79: 163-167. BERTALANFFY, B. VON, 1938, A quantitative the- ory of organic growth. Human Biology, 10 (2): 181-213. BIRCH, L. C., 1948, The intrinsic rate of natural increase of an insect population. Journal of Ani- mal Ecology, 17 (1): 15-26. ОЕ КОСК, К. М. & J. A. VAN EEDEN, 1981, Life table studies on freshwater snails: The contribu- tion of the egg production of successive pivotal age groups of a cohort of freshwater snails to- wards the innate capacity of increase (rm). Wetenskaplike Bydraes van die Potchefstroom Universiteit vir Christelike Höer Onderwys. Reeks B: Natuurwetenskappe, 108: 1-4. DE KOCK, K. N. & J. A. VAN EEDEN, 1985, Effect of constant temperature on population dynamics of Bulinus tropicus (Krauss) and Lymnaea natal- ensis Krauss. Journal of the Limnological Society of Southern Africa, 11 (1): 27-31. FRANDSEN, F. & H. MADSEN, 1979, A review of Helisoma duryi in biological control. Acta Tropica, 36: 67-84. GUYARD, A. & J. P. POINTIER, 1979, Faune malacologique dulçaquicole et vecteurs de la schistosomose intestinale en Martinique (Antilles frangaises). Annales de Parasitologie, 54 (2): 193-205. LOREAU, M. & B. BALUKU, 1987, Population dy- namics of the freshwater snail Biomphalaria pfe- ifferi in eastern Zaire. Journal of Molluscan Stud- les, 53: 249-265. MADSEN, H., 1984, The effect of water conditioned by either Helisoma duryi or Biomphalaria camer- unensis on the growth and reproduction of juve- nile B. camerunensis (Pulmonata: Planorbidae). Journal of Applied Ecology, 21: 757-772. MALEK, E. A. & R. R. MALEK, 1978, Potential bi- ological control of schistosomiasis intermediate hosts by helisome snails. Nautilus 92: 15-18. MCCULLOUGH, F., 1981, Biological control of the snail intermediate hosts of human Schistosoma spp.: a review of its present status and future prospects. Acta Tropica, 38: 5-13. MICHELSON, E. H. & L. DUBOIS, 1974, Lymnaea emarginata, a possible agent for the control of the schistosome snail host, Biomphalaria gla- brata. Nautilus, 88: 101-108. MURRAY, H. D. & L. J. WOPSCHALL, 1965, Ecol- ogy of Melanoides tuberculata (Muller) and Tare- bra granifera (Lamarck) in south Texas. Bulletin of the American Malacological Union: 25-26. POINTIER, J. P., 1989, Comparison between two biological control trials of Biomphalaria glabrata in a pond in Guadeloupe, French West Indies. Journal of Medical and Applied Malacology, 1: 83-95. POINTIER, J. Р., А. GUYARD & А. MOSSER, 1989, Biological control of Biomphalaria glabrata and B. straminea by the competitor snail Thiara tuberculata in a transmission site of schistosomi- asis in Martinique, French West Indies. Annals of Tropical Medicine and Parasitology, 83: 263- 269. PRENTICE, M. A., 1983, Displacement of Bi- omphalaria glabrata by the snail Thiara granifera in field habitats in St Lucia, West Indies. Annals of Tropical Medicine and Parasitology, 77: 51- 59. RAMAMOORTHI, K., 1955, Studies in the embry- ology and development of some melanid snails. Journal of the Zoological Society of India, 7 (1): 25-34. ROESSLER, M. A., G. L. BEARDSLEY & D. C. TABB, 1977, New record of the introduced snail, Melanoides tuberculata (Mollusca: Thiaridae) in South Florida. Florida Sciences, 40: 87-94. SHIFF, C. J., 1964, Studies on Bulinus (Physopsis) globosus in Rhodesia |. The influence of temper- ature on the intrinsic rate of natural increase. An- nals of Tropical Medicine and Parasitology, 58: 98-105. SHIFF, C. J. & B. GARNETT, 1967, The influence of temperature on the intrinsic rate of natural in- crease of the freshwater snail Biomphalaria pfeifferi (Krauss) (Pulmonata: Planorbidae) Ar- chiv für Hydrobiologie, 62 (4): 429-438. STURROCK, R. F., 1966, The influence of temper- ature on the biology of Biomphalaria pfeifferi (Krauss), an intermediate host of Schistosoma mansoni. Annals of Tropical Medicine and Para- sitology, 60: 100-105. WORLD HEALTH ORGANIZATION, 1984, Report of an informal consultation on research on the biological control of snail intermediate hosts. UNDP World Bank/WHO Special Programme for Research and Training in Tropical Diseases, TDR/BCV-SCH/ 84.3: 1-41. Revised Ms. accepted 15 May 1990 MALACOLOGIA, 1991, 33(1-2): 55-62 FLUCTUATIONS AND IMMOBILITY IN AGE-STRUCTURED SNAIL POPULATIONS L. M. Cook Department of Environmental Biology, University of Manchester, Manchester, M13 9PL, United Kingdom ABSTRACT In helicid snails, the density of adults and large juveniles may affect output. The question whether this feedback effect could generate regular changes in population size is discussed. A bifurcation diagram is produced for a model population with life history characteristics like those of Cepaea species, and it is shown that cycles of several years could readily arise. In practice, however, these are not likely to be distinguishable from random fluctuations. In temperate regions, many helicids take one or more years to become mature. It is argued that this delay selects for adult longevity, so that generations become overlapping rather than allochronic. It is suggested that the very low mobility characteristic of terrestrial pulmonates both increases the chance of large population fluctuations and explains the fact that the group is hermaphroditic. INTRODUCTION Two studies of the dynamics of Cepaea ne- moralis populations that have recently been carried out have features in common (Cain & Cook, 1989; Cain et al., 1990). In both cases, population data are available for 20 years, and there are fluctuations in numbers from high to low over periods of 6—10 years. Numbers of first-year and second-year juveniles have also been estimated, and quite frequently an abun- dant year class is followed by a year in which recruitment is very low. This pattern could arise if high density of adults and large (year 2) juvenile individuals interfered with egg hatching, perhaps by disturbing the ground where eggs are laid, or reduced the chance of successful egg production or hatchling growth in some other way. For example, mucus pro- duction in dense populations may have a neg- ative effect on very young animals (e.g. Heller & Ittiel, 1990). If good recruitment in one year tends to reduce the output in a following year, this raises the question whether the interaction between year classes could generate cycles in the adult population. Study of cycles and other population fluctuations have come a long way since the early work on lynxes, hares and sun- spots (e.g. Elton, 1942; Cole, 1954), but snails have a characteristic that makes it worth look- ing further at the kind of fluctuations to which they are liable. They are relatively long-lived for small invertebrates and have both a high potential output and an extended juvenile pe- 55 riod. In addition, it is rarely possible to census both juveniles and adults, so that in practice the fluctuations to be explained refer to only part of the population. PATTERNS OF POPULATION CHANGE If the species is annual, such as members of the helicid sub-family Helicellinae in tem- perate regions, then adults of successive generations are non-overlapping. Although growth, mating and egg laying show a pattern over the season, the change in numbers (N) from one generation (n) to the next may rea- sonably be represented by simple recurrence relations of the form, N N„R(1-(N,/K)) (1) n+1 or Nae = N,exp(r(1-(N,/K))) (2) In these equations, R or r measures rate of increase and K the restricting effect of the en- vironment. When the population is very small, R is the net rate of increase of the population in generation п, andr = In A. In the first equa- tion, the final steady level which the popula- tion approaches when R is small is K(R-1)/R; in the second it is K. Both equations have been used in analyses of population fluctua- tions by May and others (e.g. May, 1975; May 8 Oster, 1976); | shall use the second, 56 COOK which avoids the possibility of negative num- bers. The essential feature of these relations is that numbers in one year are determined by numbers in the previous year, so that a con- siderable time lag occurs. When r is between O and 1, the population approaches K mono- tonically, but when it lies between 1 and 2, there are damped oscillations, and above 2 other things happen. When r is between 2 and 2.5, the equation predicts that numbers should fluctuate from a stable value above K to a stable value below it, taking one genera- tion to move from the upper to the lower value. Thus, for these values of r, there are two stable values, one above and one below K. As increasingly large values of r are imag- ined, the population moves from a two-point to a four-point and then to an eight-point limit cycle, and after that moves into a chaotic dis- tribution, mimicking the sort of patterns that might be expected if random mortality effects impinged on the population. May (1975) gives critical values at which these changes in be- haviour occur. The bifurcation diagram, in which these end points are plotted on the val- ues of r that generate them, has been repro- duced many times in books and articles on chaos (e.g. Gleick, 1988; Stewart, 1989). If the average rate of increase in such an an- nual species is greater than exp(2) or 7.4, cy- cles of two or more years or erratic fluctua- tions may be generated simply as a result of the interaction between the intrinsic rate of increase and the carrying capacity of the en- vironment. Such levels could be attained by snail species. This observation is now well known, but it leads on to consideration of spe- cies with more complicated life histories. Some snail species, such as Arianta arbus- torum living in favourable conditions (Baur, 1984), lay eggs one year that become adult the following year. If after reproduction they were then to die, adults would be present in alternate years, and the pattern of population change may be written in the form of the Les- lie matrix (Williamson, 1972) as, The top row represents output, so that in this case +. > is equivalent to В. The f,, term rep- resents survival from birth to adult, and in a larger matrix becomes the top term of a sub- diagonal sequence; x, are numbers in succes- sive age classes. May et al. (1974) describe the stability conditions for a situation of this kind. Density dependence could limit recruit- ment or survival to adulthood. In either case, it could be brought about by juvenile or adult crowding. In the Cepaea data, the implication was that output was influenced by adult num- bers or by numbers of adults and large juve- niles together, so an effect on output will be assumed. The stability criteria show that this system may be stable. On the other hand, if fecundity is density-independent but survival of juveniles depends on adult density, the model is unstable (May et al., 1974) so pos- sible outcomes are now more complicated. Density-dependence has been introduced by multiplying x>f,2 by exp(r(1-x./K)). The bifur- cation pattern when output is limited by adult density is essentially similar to that for the simple model. In this model, adults are present only in al- ternating years. Theoretically, a second pop- ulation could exist out of step by one year, although there would be competition between the two. Adult numbers in one population would limit output in the other without any feedback control, and it might be expected that one of the allochronic populations would be driven to extinction. Some moths, such as the oak eggar, Lasiocampa quercus, in Eu- rope have a one-year generation in southern parts of their range and a two-year generation further north or at higher altitudes. In these, adults do not occur in each year, probably because of inter-population effects. Mikkola (1976) discusses several species of Xestia that are biennial and have adult populations in alternate years in northern regions, and sug- gests that parasites supported by the larger of the two populations may tend to extinguish the smaller one. In these circumstances, snails, but proba- bly not moths, could live longer than one year as adults, so as to lay eggs in each year of adulthood and fill the alternating gaps. There would be selection for longevity and iteropar- ity, and the population becomes one with eggs produced in all years and overlapping generations. The basic matrix is, x, 0 fio fig x x = fo, 0 0 Xo X's 0 f2 0 Хз FLUCTUATIONS AND IMMOBILITY IN SNAILS 57 "^ log N F 2 1 er N ‘ ES DE Ds es 1 2 FIG. 1. r 3 4 Bifurcation diagram for a “Cepaea-like” snail species with two juvenile and five adult year classes. Numbers of adults only are shown. Values of output and survival have been chosen so that the population is in equilibrium and density-dependent feedback of adult and large juvenile numbers on output has been imposed using equation (2). If such a population is allowed to increase uninterruptedly, the pattern of increase even- tually becomes exponential and the ratio of numbers in successive year classes becomes constant (Williamson, 1972). A possible cause of fluctuation in numbers could be an extrinsic factor that increases or decreases one age class relative to the others (Leslie, 1948). The aberrant class then passes as a “pulse” through the population from year to year, damped or magnified by such intrinsic factors as survival or feedback control. This pulse is likely to be of one-year duration, so that numbers would fluctuate from one year to the next but not with long periods (Bernardelli, 1941, discussed by Williamson, 1974). The fluctuations brought about by feedback control are superimposed on these changes in age distribution, and the observed change in adult numbers can become complex. As- 58 COOK leak (u) UN Bo] FIG.2. Population change for r = 3. The trajectory shows numbers, N, plotted on year, п. The web diagram represents N, , , plotted on N... pects of models of this kind have been stud- ied by a number of authors (e.g. Beddington, 1974; De Angelis, 1975; Levin & Goodyear, 1980; Oster & Takahashi, 1974; Usher, 1972). In Cepaea nemoralis, juveniles take two years to become adult, so that the mini- mum matrix dimension to provide overlapping generations is four. We have estimated adult annual survival rate to be about 0.52 (Cain et al., 1990), so that a seven-year time span en- compasses most of the expected survival of a cohort. The model has then been set up as before, but in this case with numbers of adults and large juveniles exerting density-depen- dent control on output. The bifurcation dia- gram is shown in Figure 1. There are regions of regular but complex and of irregular fluctu- ation, and four regions (at around г = 2,r = 3, г = 3.4, andr = 3.7) which exhibit 4-point, 13-point, 9-point, and 14-point cycles respec- tively. Figure 2 illustrates the trajectory that occurs when r = 3. This shows that, in addi- tion to having a 13-step periodicity, the cycle can be resolved into three successive com- ponents of 4, 4 and 5 points. Williamson (1974) has previously shown that these sub- sets occur within population sequences in models with age classes. FLUCTUATIONS AND IMMOBILITY IN SNAILS 59 TABLE 1. Serial product-moment correlation coefficients for steps of length k and a range of values of the feedback coefficient r, using the Cepaea model illustrated in Figure 2 and sequences of 400 years. Step length k 1 1.5 2.0 1 1.00 —0.01 —0.01 2 1.00 — 0.88 — 0.92 3 1.00 0.22 —0.01 4 1.00 0.97 1.00 5 1.00 — 0.23 — 0.01 6 1.00 — 0.82 — 0.92 7; 1.00 0.44 —0.01 8 1.00 0.88 1.00 9 1.00 —0.44 —0.01 10 1.00 — 0871 —0.92 11 1.00 0.65 —0.01 12 1.00 0.75 1.00 13 1.00 — 0.62 —0:01 14 1.00 = 0:55 — 0.92 15 1.00 0.81 —0.01 This type of effect is general for populations with year classes living in an annually cycling environment. If the simulation model is changed so that the species is assumed to reproduce in its second year, as well as later, a pattern similar to Figure 1 is produced, but with cycles of 9, 14 and 5 years occurring at around г = 2.7, г = 3.3, andr = 4 respec- tively. In either case, the pattern changes di- rectly from a monotonic to a complex one as r increases from a low value, without the bi- furcation seen in simpler systems. Other as- sumptions lead to other patterns. The type of pattern is consequent upon use of difference equations to describe the population change, but it is not limited to models of this kind. De- layed differential models will also give rise to cycles. Time delays between the imposition of eco- logical constraints and response by the pop- ulation can therefore lead to a wide range of types of fluctuation in numbers. These oper- ate when the response time is relatively short in relation to the delay (May et al., 1974). In age-structured populations, deviations from the stable age distribution also cause popu- lation fluctuation. These two kinds of effect have been considered in relation to land snail populations, where individuals are relatively long-lived and have a long pre-reproductive stage, and it is assumed that high density has an inhibiting effect on output. The fluctuations are sometimes relatively short, being cycles of two, three or four years. If the system is sufficiently reactive, however (r is sufficiently Feedback coefficient 2.5 3.0 3.5 4.0 0.02 0.06 0.09 0.15 —0.83 — 077, — 0:71 —0'59 0:21 — 0.36 — 0.44 — 0.52 0.94 0.80 0.60 0.28 0.25 0.47 0.60 0.77 — 0.78 —0.63 —0.45 — 0.14 — 0.41 — 0.63 —0.71 —0.65 0.81 0.47 0.09 —0.33 0.46 0.80 1.00 0.59 —0.68 —0.36 0.09 0.59 = 0.57 — 077, — 0:71 —0.33 0.64 0.06 —0.44 —0.65 0.65 1.00 0.60 —0.14 = 0:56 0.06 0.60 0.77 —0.69 = 0:77 —0.45 0.28 large), cycles of 13 or 9 years could be gen- erated, such as are sometimes apparent in populations of Cepaea. In a field study, one of the methods available for detecting cycles is to calculate serial correlations between suc- cessive population values. This has been done for sequences of 400 values, equivalent to 400 years in reality, using the simulation that produced Figure 1. The results for a range of step lengths and values of the rate of increase r are shown in Table 1. It can be seen that as r increases, changing patterns of positive and negative correlation appear. They provide high positive correlations at gaps of 4, 8 and 12 years when r = 2, chang- ing to 13 years when г = 3 and 9 years when r = 3.5. It is difficult to know what average clutch sizes these values correspond to, be- cause we do not know the survival rate from egg to first year and therefore do not know the clutch size necessary just to maintain the population. A value of r = 3.5, or 33 times the maintenance level of output, could be possi- ble, however, and r = 2, or 7 times the main- tenance level, may be quite readily achieved. It may therefore be concluded that cycles of several years’ duration could arise as a result of the interplay of fecundity and reactivity to population constraints. In practice, a real population would also be subject to random extrinsic mortality factors. If a species is effectively established in a re- gion, these are likely to operate as random deviates about a survivable mean value. Their effect can be simulated by generating 60 COOK TABLE 2. Serial product-moment correlation coefficients for steps of length k, using the Cepaea model kept in equilibrium with r = 1 but with random deviation added. (a) Long runs. Three representative runs of sequences of 400 years. (b & c) Short runs. Three runs each for sequences of 30 years and 20 years. Step length k (a) 400 years 1 0.00 0.05 0.03 0.39 2 —0.05 —0.01 — 0.03 0.11 3 — 0.06 0515 01 — 021 4 0.05 —0.04 0.01 —0.06 5 0.06 0.09 0.01 0.06 6 —0.03 —0.04 —0.03 0.16 7 —0.07 —0.03 —0.06 0.16 8 0.05 0.00 0.02 —0.08 9 0.03 —0.02 0.12 0.11 10 —0.08 0516 0:12 0.00 11 0.02 — 0:07 —0.01 0.08 12 0.10 0.04 0.06 0.12 13 —0.04 0.08 —0.01 0.40 14 —0.02 —0.09 —0.06 0.49 15 0.00 0.00 0:05 0.51 random deviates about unity and multiplying the output by these factors in each genera- tion, while r is kept small to avoid intrinsic fluctuations. When this is done for sequences of 400 years, the correlation analysis pro- duces no evidence of cycles (Table 2a). No one is ever likely to census a Cepaea popu- lation for 400 years, however, and a realistic sample length would be at most 20 or 30 years. If sequences of this length are anal- ysed, then fluctuations of several years’ du- ration that look superficially like the intrinsic cycles in a serial correlation analysis could be produced at random (Table 2b and c). The Markovian behaviour of the population, cou- pled with the effect of movement of peaks through the age classes, has a tendency to produce relatively long cycles of numbers in some runs. In a long sequence, these do not repeat regularly and the effect disappears, but that is not always evident in the se- quences of a few years, which are all that the experimentalist can hope for. Intrinsic and random cycles are therefore impossible to distinguish in practice in species that are at all long-lived. DISCUSSION The study of Cain & Cook (1989) was car- ried out on eight artificial populations kept in enclosures which prevented migration from one to another. Numbers fluctuated greatly and most of the populations become extinct. (b) 30 years (c) 20 years 0.19 0.27 — 0.23 0.36 0.35 0.01 — 0.08 0.39 — 0.05 — 0.04 — 0.30 —0.13 — 0.08 = 0:12 0.21 —0.28 0.05 0.51 0.15 0.34 —0.07 0.52 0516 0.27 0.20 0.11 0.15 0.72 0.45 0.09 0.24 — 0.03 — 0.14 0.37 —0.18 0.31 OZ 0.58 0.08 0.07 0.08 0.02 0.00 0.10 0.55 0.15 0.07 0.62 0.29 0.32 — 0.08 0.00 0.03 0.65 0.10 — 0.08 0.11 0.58 0.66 0.61 —0.33 0.56 0.25 0.09 0.79 0.02 0.12 0.40 0.20 0.43 0.13 0313 0.08 0.70 0.64 Under these conditions, cycling in numbers could be generated by feedback response of output to population density. It is probable that if there had been continuity between the enclosures the centre of density of the popu- lation would have moved from one point to another as conditions changed and became unfavourable in particular places, so that, looked at on a larger scale, the fluctuations in numbers would have appeared less extreme. The natural population studied by Cain et al. (1990) was situated in an area of varied veg- etation. At the outset, there were two patches of nettles (Urtica dioica) in the centre, which harboured high densities of snails. As time went on, the nettles became less dense and the density of animals went down, probably in part as a result of migration. Nevertheless, the likelihood of large fluctuations in numbers in a given spot is increased by the very low mobility characteristic of land snails. Low mobility may also explain the normal breeding pattern. Some pulmonate molluscs are parthenogenetic whereas others self, but most are obligate cross-fertilizing hermaphro- dites, as in the examples considered here. There are three types of reasons why species should be cross-fertilizing hermaphrodites (Charnov, 1982; Charnov et al., 1976; May- nard Smith, 1978), and some probable costs associated with the system (Heath, 1977). Reasons for hermaphroditism are (1) that in- dividuals may have difficulty during their life times in finding individuals with which to mate, (2) that there may be an advantage in spread- FLUCTUATIONS AND IMMOBILITY IN SNAILS 61 ing reproductive effort over a long period of life and of body size, and (3) that if offspring live very close to their parents throughout their lives, there is an advantage in investing in both kinds of gametes, because excess ef- fort in either direction is likely to lead to dimin- ishing returns. As the number of male ga- metes produced goes up, there will come a point when no more female gametes are within reach with which they can unite. As the number of female gametes increases, so will the potential for sib-sib competition. A distri- bution of resource through both routes is therefore favoured. This non-obvious conse- quence of low mobility is explained in terms of generation of concave fitness sets by Char- nov et al. (1976). It is unlikely that explanation (1) is important in pulmonate snails, since al- though, or perhaps because, they are immo- bile they generally live in patches of relatively high density, and they are long-lived so that the chance of passing a lifetime without find- ing a suitable mate is small. The second ex- planation can be imagined in some circum- stances, for example when snails live in relatively severe Mediterranean climates. It may then be advantageous to exchange sperm at one time of the year and having done so, to use them to fertilize eggs later, perhaps after accumulation of substantially more energy during a period when more food is available. Most species probably live in cli- matic conditions that are not so restrictive. All species are relatively immobile, however, and have to lay eggs close to where they live as adults. Offspring-parent and sib-sib competi- tion are therefore possibilities, and hermaph- roditism may reduce the likelihood that they occur. Oligochaetes are another group with similar traits, to which the same argument may be applied. In this connection, pulmo- nates differ in an important respect from prosobranch molluscs. Whereas adult proso- branchs may be as restricted in their move- ment as are adult pulmonates, the group is basically marine, rather than terrestrial, and the reproductive strategy involves separate sexes and a planktonic larva. Many excep- tions occur, some of which can be explained as responses to particular conditions. For ex- ample, some slipper limpets (Crepidula spp.) are protandrous hermaphrodites with plank- tonic larvae, but because they settle upon each other it is easy to see that there is an advantage to the newly settling individual be- ing the right sex to to mate with one which is already established. Periwinkles (Littorina spp.) have separate sexes and some have planktonic larvae while others brood eggs. An explanation may be sought in terms of differ- ing need for local adaptation in different spe- cies. Such examples are exceptions, how- ever, rather than the rule. The existence of planktonic dispersal phase removes the direct competition effect and, arguably, the selective pressure to be hermaphroditic. Immobility and the interactions that generate population fluc- tuations may therefore have had profound ef- fect on the evolution of pulmonate molluscs. ACKNOWLEDGMENTS | am grateful to S. E. R. Bailey, A. J. Cain, R. A. D. Cameron, J. D. Currey, G. S. Mani and M. H. Williamson for helpful comments. LITERATURE CITED BAUR, B., 1984, Early maturity and breeding in Ari- anta arbustorum (L.) (Pulmonata: Helicidae). Journal of Molluscan Studies, 50: 241-242. BEDDINGTON, J. R., 1974, Age distribution and the stability of simple discrete time population models. Journal of Theoretical Biology, 47: 65- 74. BERNARDELLI, H., 1941, Population waves. Jour- nal of the Burma Research Society, 31: 1-18 [not seen]. CAIN, A. J. & L. M. COOK, 1989, Persistence and extinction in some Cepaea populations. Biologi- cal Journal of the Linnean Society, 38: 183-190. CAIN, A. J., L. M. COOK, & J. D. CURREY, 1990, Population size and morph frequency in a long- term study of Cepaea nemoralis. Proceedings of the Royal Society of London B, 240: 231-250. CHARNOV, E. L., 1982, The theory of sex alloca- tion. Princeton University Press, Princeton, N.J. 355 pp. CHARNOV, E. L., J. MAYNARD SMITH & J. J. BULL, 1976, Why be an hermaphrodite? Nature, London, 263: 125-126. COLE, L. C., 1954, Some features of random cy- cles. Journal of Wildlife Management, 18: 2-24. DE ANGELIS, D. L., 1975, Global asymptotic sta- bility criteria for models of density-dependent population growth. Journal of Theoretical Biol- ogy, 50: 35—43. ELTON, C., 1942, Voles, mice and lemmings: prob- lems in population dynamics. Oxford University Press, Oxford. GLEICK, J., 1988, Chaos. Making a new science. Heinemann, London. 352 pp. HEATH, D. J., 1977, Simultaneous hermaphrodit- ism; cost and benefit. Journal of Theoretical Bi- ology, 64: 363-373. 62 COOK HELLER, J. & H. ITTIEL, 1990, Natural history and population dynamics of the land snail Helix texta in Israel (Pulmonata:Helicidae). Journal of Mol- luscan Studies, 56: 189-204. LESLIE, P. H., 1948, Further notes on the use of matrices in population mathematics. Biometrika, 35: 213-245. LEVIN, S. A. & C. P. GOODYEAR, 1980, Analysis of an age-structured fishery model. Journal of Mathematical Biology, 9: 245-274. MAY, R. M., 1975, Biological populations obeying difference equations: stable points, stable cycles, and chaos. Journal of Theoretical Biology, 51: 511-524. MAY, R. M. 8 G. F. OSTER, 1976, Bifurcations and dynamic complexity in simple ecological models. American Naturalist, 110: 573-599. MAY, В. M., С. В. CONWAY, M. P. HASSELL 4 T. R. E. SOUTHWOOD, 1974, Time delays, den- sity-dependence and single-species oscillations. Journal of Animal Ecology, 43: 747—770. MAYNARD SMITH, J., 1978, The evolution of sex. Cambridge University Press, Cambridge. 222 pp. MIKKOLA, K., 1976, Alternate-year flight of north- ern Xestia species (Lep., Noctuidae) and its adaptive significance. Annales Entomologici Fennici, 42: 191-199. OSTER, G. F. & TAKAHASHI, Y., 1974, Models for age specific interactions in a periodic environ- ment. Ecological Monographs, 44: 483-501. STEWART, |., 1989, Does God play dice? The mathematics of chaos. Blackwell, Oxford. USHER, M. B., 1972, Developments in the Leslie matrix model. Pp. 29-60, in JEFFERS, J. М. R., ed., Mathematical models in ecology. Blackwell, Oxford. WILLIAMSON, M., 1972, The analysis of biological populations. Arnold, London, 180 pp. WILLIAMSON, M., 1974, The analysis of discrete time cycles. Pp. 17—33, in, USHER, M. B. 8 M. H. WILLIAMSON, eds., Ecological stability. Chap- man and Hall, London. Revised Ms. accepted 14 August 1990 MALACOLOGIA, 1991, 33(1-2): 63-70 COSTS OF REPRODUCTION: A STUDY ON METABOLIC REQUIREMENTS OF THE GONADS AND FECUNDITY OF THE BIVALVE DREISSENA POLYMORPHA Martin Sprung’ Lehrstuhl für Physiologische Ökologie, Zoologisches Institut, Universität Köln, Weyertal 119, D-5000 Köln 41, Germany ABSTRACT In two sets of experiments, the fecundity of the freshwater mussel Dreissena polymorpha and the metabolic requirements of mature eggs within the animal have been estimated. Female Dreissena polymorpha can spawn more than 10° eggs, and the males up to nearly 10'° sperms depending on size. Both release about the same gamete weight, which contributes in many cases more than 30% of the body weight prior to spawning. Oxygen consumption and ammonia excretion rates of individual females were estimated at 10°C and 15°C before and after spawning; data indicate a lower weight-specific respiration rate of the stored gametes compared with that of the whole animal; it was independent of temper- ature in this range. The eggs contribute to only a small and insignificant proportion of the total ammonia excretion rate. Respiration and excretion rate of the whole animal, in contrast, were highly dependent on temperature. Key words: Dreissena polymorpha, fecundity, oxygen consumption, ammonia excretion, O/N ratios. INTRODUCTION The freshwater mussel Dreissena polymor- pha has a mode of reproduction that is pecu- liar and common at the same time. It is com- mon, because many bivalves that live in the marine environment reproduce in the same way, by means of a free-swimming larva. It is on the other hand peculiar, because no other freshwater bivalve has this free-swimming larva, at least with respect to Europe. This unspecialized reproductive behaviour runs parallel with high losses of the unprotected larval stage (Sprung, 1989). Dreissena can only be competitive when it produces high numbers of gametes. And, in fact, it has been very successful when colonizing lakes and rivers in Europe (Stanczykowska, 1977) or re- cently in North American (Hebert et al., 1989)—or becoming a pipe-fouling pest and attaining by this means economic interest (e.g. Clarke, 1952). Gametes develop during autumn and spring months (Antheunisse, 1963), but also during winter at low water temperatures (Кабапома, 1961; Walz, 1978; Borcherding, 1989). As demonstrated in an earlier paper (Sprung, 1987), Dreissena eggs can only be fertilized when the water temperature is higher than 10°C. It is reasonable to assume that the animals will not release their eggs below this mark. The consequence is that many animals have morphologically ripe go- nads for a significant length of time before they actually spawn. The aim of this paper is to quantify these phenomena from two sides: (a) to obtain fig- ures for the size of the gonad and the number of gametes an animal can produce; (b) to es- timate the physiological costs to maintain a gonad of that size. MATERIAL AND METHODS Sampling and Processing Twice animals were collected by SCUBA diving from a lake north of Cologne: (1) In August 1985 from 9 m depth: these animals were used for estimating fecundity; they were stored at 4°C in distilled water with 1% seawater added and stimulated to spawn (see below) one week later. (2) In May 1987 from 4 m depth: mainte- nance costs of the gonads were estimated “Present address: U.E. Ciéncias e Tecnologias dos Recursos Aquáticos, Universidade do Algarve, Campus de Gambelas, Apartado 322, P-8004 Faro Codex, Portugal 64 SPRUNG with these animals; they were labelled individ- ually, stored at 10°C in a temperature- controlled room and fed every working day with an algal suspension (in most cases Chlamydomonas reinhardii) except the day prior to the following measurements: (a) For each animal individually, oxygen consumption and ammonia excretion rate were estimated after at least one week of ad- aptation at 10°C; the last animals were pro- cessed seven weeks after collection. (b) They were transferred to 15°C and after an adaptation of at least one week these pa- rameters were also recorded individually at that temperature; subsequently these animals were stimulated to spawn and the eggs counted, if spawning was successful. (c) A few days (typically one week) after the spawning attempt, oxygen consumption and excretion were estimated again at 15°C. (d) They were transferred to 10°C and mea- surements repeated after at least one week of adaptation. Subsequently the soft parts were excised for weight estimation and sex determined if still unknown. Fecundity Estimates and Spawning Spawning was induced by transferring the animals to water of room temperature (ap- proximately 18 to 20°C) with homogenized soft parts of Dreissena added. After two hours the animals were separated individually to Petri dishes; the gametes were collected in Erlenmeyer flasks. Gametes were counted with a Coulter Counter (Model TAll; 50 рт aperture tube for the sperm; 280 um tube for the eggs). The water for counting was diluted with a NaCl solution of 16%. to achieve an adequate conductivity. Weight Estimates Freshly spawned eggs were transferred to a beaker by means of a pasteur pipette to form a highly concentrated suspension. Egg density in this suspension was estimated us- ing a Coulter Counter. Egg weight was as- sessed in five replicates each from 300 ul of the egg suspension transferred to pre- weighed and preashed aluminium trays. Sperm weight was estimated from a sus- pension of Known density on preashed and preweighed glass fibre filters (1 ml containing about 10’ sperms on Whatman GF/C). The weight estimate was corrected by a blank. For the tissue weight, soft parts of the ani- mals were exised and transferred to alumin- ium trays. After one day (eggs and sperm) or two days (soft parts) at 105°C the trays were reweighed (dry weight) and subsequently ashed for three hours at 450°C (loss: ash free dry weight). Oxygen Consumption Oxygen consumption was recorded in a flow-through apparatus. Water from an aer- ated and temperature equilibrated container was sucked by means of a peristaltic pump through five chambers of about 30 ml content and a bypass at about 60 ml/h. The exact rate was checked twice at the beginning and at the end of the measurements by a graduated су|- inder. Electrical valves directed the water flow in such a way that every five minutes the out- flow of another chamber or the bypass passed an oxygen electrode (Radiometer Copenhagen PHM 72). Five repeated cycles were registered on a chart recorder after an adaptation time of at least 30 minutes. The system was calibrated by aerated and by deoxygenated water (5% Na,SO, solution). Oxygen consumption was calculated from the product of the flow rate, the difference in ox- ygen tension between the bypass and the chamber corrected by a blank run and ß, which is the absorption coefficient for oxygen in water calculated according to Benson & Krause (1980). Ammonia Excretion Animals were transferred individually to 50 ml Erlenmeyer flasks. Excretion was calcu- lated from the difference in ammonia concen- tration as determined after the animals had opened their shells and after 2 h. Estimation was performed with a test kit (Spectroquant, Merck), measuring the extinction in 5 cm cu- vettes in a photometer; a calibration curve was established with a NH,CI solution. Statistical Evaluation For the oxygen consumption and ammonia excretion rate (Fig. 3), a power curve was fit- ted versus tissue weight. Data refer to an an- imal of 50 mg ash free dry weight, because this was close to the mean of all animals ex- amined. Statistics (confidence intervals, stan- dard deviations, test for the significance of the COSTS OF REPRODUCTION OF DREISSENA 65 1.4 + | Dreissena polymorpha 8 A: 1.42 | = = (9): y=0.606 x * = = =< bom = =. 10 г= 0.971; п=12 2 =) Ne) = E o x = a U a Je E Е Е 24 4 2 i 2 06 зи“ 20 + on 3 © = 0% \ Ail eae | (4: y=250x10°x201 2 a 02 г= 059]: п-9 0 15 20 25 30 Shell length [mm] (x) FIG. 1. Dreissena polymorpha: Gamete output of females and males of different size; for each sex re- gressions are indicated; n: number of estimates; r: correlation coefficient. difference of the slope of a regression from zero) were calculated following Sachs (1978). RESULTS Fecundity Large Dreissena polymorpha females could release more than a million eggs and a male nearly 10 billion sperms. For one egg, an ash-free dry weight of 34.4 + 9.9 ng (mean + standard deviation; 30 estimates) has been determined and for one sperm 5.8 + 1.3 pg (8 estimates). Taking these weights into ac- count, it could be demonstrated that male and female Dreissena release roughly the same gamete weight (Fig. 1). Unfortunately, only the body weight of the last series of animals (maintenance costs of the gonad) has been recorded. The data reveal that from 18 fe- males spawned, six released more than 30% of the initial body weight as gametes and one even 45.1%. Maintenance Costs Ripe gametes in the female showed only a moderate oxygen demand. This could be cal- culated from the difference in respiration rate before and after spawning at both tempera- ture levels (difference from estimate (a) to (d) for 10°C and from (b) to (c) for 15°C respec- tively described under Material and Methods). Although there is a high standard deviation in the slope in Figure 2, it is roughly significantly different from zero for p = 0.05: the 10°C estimate is slightly below this level, the 15°C estimate slightly above. Data indicate a tem- perature-independent oxygen consumption о slope: (-9.624.5)x10"6mL0,h leg"! x r:-0.366 =n: 31 [e) 15:6 slope:(-10.0+ 391)x10" mmol NH 3h leg” r:-0.047; п: 31 + 66 SPRUNG о 15 р * slope: (-99+51)x10-ml0,h egg”! wc 10r « r:-0.358; n:28 — 5 oe . oe TE . . a OFS © + e £ T= -5 . e < = 0“ А -15 | ? E Ba ll A E 150, 7 100F* stope:t-50225 71x10 mmol NH; egg” mn 50 zu 5;-0.038;n:28 = . = о ei; Ll. = 1 e 5 : e E-=505* < — -100 A EPA ie Я: ee -150 08 1.0 Eggs released [x 106] FIG. 2. Dreissena polymorpha, females: Differences in oxygen consumption and ammonia excretion be- fore and after the application of a spawning stimulus versus egg number released; each data point presents the untransformed estimate for one animal; the slope of the regression with its standard deviation is indi- cated; n: number of estimates; r: correlation coefficient. rate of the eggs between 10 and 15°C, in con- trast to the overall oxygen consumption (Fig. 3). It is interesting to note that those females not in the state of spawning tended to show a higher respiration rate at 10°C than the ripe females and, furthermore, that these animals did not react in such a pronounced manner to an elevation of the temperature to 15°C. This is documented by the lower Q,, values. The ammonia excretion rate showed only an insignificant decrease with the egg number released (Fig. 2). This indicates that only a very small part, if any, of the proteins metab- olized were derived from the eggs. This makes sense, because it would otherwise mean that proteins from already intact egg structures would be respired. Here in turn those animals that did not show any spawning success react more dis- tinctly to the temperature elevation. The con- sequence is a clearly lower O/N ratio at the higher temperature. This decrease, which was not observed in ripe animals, must be interpreted as an indication of protein de- growth to meet the needs of reproduction (see Aldridge et al., 1986, for the snail Vivi- parus georgianus). The spawning stimulus it- self did not affect the oxygen consumption and excretion rates of the animals in this se- ries, because they showed nearly the same level after the stimulus had been applied. DISCUSSION Fecundity There are different approaches to estimate fecundity. Stimulating the animal and count- ing the gametes is only one, and it has one severe shortcoming. Walz (1978) showed that Dreissena can spawn repeatedly during a reproductive cycle. On average, he found that about half of the eggs produced are released at the first spawning event. The same phe- nomenon has been documented for Merce- COSTS OF REPRODUCTION OF DREISSENA 67 Oxygen consumption Ammonia excretion O/N-Ratio Ex Ee Spawning success 39 3.3 0.1 10.155’ 102 Temperature [°C] FIG. 3. Dreissena polymorpha, females: Oxygen consumption, ammonia excretion, and O/Nammonia Patio of two experimental groups processed as described in Materials and Methods: one reacting to a spawning stimulus (indicated by the arrow, 18 animals), the other not (9 animals); data and 95% confidence interval for an animal of 50 mg ash-free tissue dry weight are presented as calculated from a power curve; the numbers indicate Q,, values for the temperature range; rates for the eggs in the gonads are derived from Figure 2. naria mercenaria and Crassostrea virginica (Loosanoff & Davis, 1963). Ansell (1967), however, also stresses the fact that the pro- portion released by M. mercenaria during the first spawning event may vary greatly, from only a small part to nearly the whole stock of eggs. By this means the data obtained here may represent fecundity quite well, although only the good spawners were selected for the estimates. Kacanova (1961) indicates for Dreissena polymorpha fecundity numbers of 30,000 to 36,000 for animals of 18-20 mm shell length and 40,000 for 32 mm shell length. These val- ues should be more or less stable under dif- ferent environmental conditions. This is in contrast to my own and literature data in three ways: (1) Fecundity estimates are roughly 20 times higher in my study, in accordance with Walz (1978). (2) Egg and sperm number produced in- 68 SPRUNG crease several times in the size range men- tioned above; this is also indicated by the high exponents in Figure 1. (3) Fecundity of bivalves varies with the prevailing food and temperature conditions (Thompson, 1979). That is why to a certain degree differences in fecundity are natural, and no ideal fecundity curve of a species can be established as a function of size. But some comparisons may be valid. Dreissena is not an extreme case, com- pared with other bivalves from the marine en- vironment with the same mode of reproduc- tion. Many species produce more eggs because of their generally larger size: e.g., Crassostrea virginica, 23-86 x 10° eggs (Davis & Chanley, 1956); Mya arenaria, 1-5 x 10° eggs (Stickney, 1964); M. mercenaria, 0.4-30 x 10° eggs (Ansell, 1967); Mytilus edulis, 1-10 x 10° eggs (Sprung, 1983); Pla- copecten magellanicus, up to 270 x 10% eggs (Langton et al., 1987). Even adult Os- trea edulis, which release fully developed D- larvae, produce 1.5 x 10° propagules (Walne, 1979), as does the freshwater pearl mussel Margaritifera margaritifera, with 8 x 10° glochidia (Bauer, 1987). For an appropri- ate comparison, however, the relation of ga- mete output to body weight should be valid. The data obtained here fall well within the range reported in literature: e.g., Tellina tenuis, 22% (Trevallion, 1971); Nucula niti- dosa, 31% (Rachor, 1976); Macoma balthica, 25% (de Wilde & Berghuis, 1978); Tellina fab- ula, 23% (Salzwedel, 1979); Mytilus edulis, 20-30% (Thompson, 1979) or 35—40% (Fis- cher, 1983); and Mytilus chilensis, 24-35% (Navarro & Winter, 1982). This is not surprising because these ani- mals have the same mode of reproduction by means of a planktotrophic larva. Skirkyavichena (1970) estimated 46.6% of the body consisting of gonads. Walz (1978) gives a lower value: 6.5% to 16.5% of the total carbon consists of gametes prior to spawning. Comparing these data, it must also be taken into account that only a part of the gonads is formed of gametes and that the proportion of gonad to body weight varies greatly with body size. Physiological Costs of the Gonad Gonads consist of energetically rich tissue, because they contain the fuel for a certain time span of the larval life. However, animals developed ways to handle synthesis and maintenance in an efficient manner. Calow (1983) provides evidence for invertebrate species that gonad tissue is built up at a higher efficiency than somatic tissue. In this context also, the moderate metabolic mainte- nance costs found here must be seen. They agree with Kawaii (in Giese, 1969), who esti- mated only average oxygen demand in pieces of gonad tissue of Pinctada martensii when compared to other tissues. Temperature Response Temperature is the most important external factor controlling the reproductive cycle of many aquatic invertebrates (Giese, 1959). With respect to Dreissena polymorpha, a crit- ical temperature is 12°C. It is only at higher temperatures that the first larvae appear in the plankton (for literature, see Sprung, 1989). Morton (1969) registered an enhancement of growth and ripening of the gonads beyond 11°C. For this reason, the extremely high Q,, values in this range are not unexpected. This agrees with Bayne’s (1984) postulation of an increased Q,, value for the respiration rate during gametogenesis based on observations with Mytilus edulis (Bayne et al., 1976). The new detail added by this study is that a dis- tinction must be made between the oxygen demand caused by gametogenesis, which is high and temperature dependent, and that caused by ripe gametes present in the ani- mal, which is moderate and temperature in- dependent. This is a meaningful separation considering the size of the gonad. This may also interpret the apparently par- adox situation found for Dreissena by Lyash- enko & Kharchenko (1989): prior to spawning the weight exponent of the respiration rate was lower (0.66 instead of 0.75) despite an overall elevated respiration rate at this time of the year—although the largest animals should have the largest gonads. According to my laboratory observations, a temperature rise alone causes spawning of Dreissena only in exceptional cases. How- ever, the elevated respiration rate at higher temperatures may enhance susceptibility of the animals towards a spawning stimulus and may thus provide the background for a mode to establish the synchrony of gamete release. ACKNOWLEDGEMENTS My cordial thanks are due to Prof. Dr. D. Neumann, Dr. J. Borcherding (both Zoolo- COSTS OF REPRODUCTION OF DREISSENA 69 gisches Institut, Köln), Dr. D. W. Golding (Dove Marine Laboratory, Cullercoats), and two anonymous reviewers for their valuable comments on this paper and to Mrs. P. Horta e Costa for correcting the English. LITERATURE CITED ALDRIDGE, D. W., W. D. RUSSEL-HUNTER & D. E. BUCKLEY, 1986, Age-related differential ca- tabolism in the snail, Viviparus georgianus, and its significance in the bioenergetics of sexual di- morphism. Canadian Journal of Zoology, 64: 340-346. ANSELL, A. D., 1967, Egg production of Merce- naria mercenaria. Limnology and Oceanography, 12: 172-176. ANTHEUNISSE, L. J., 1963, Neurosecretory phe- nomena in the zebra mussel Dreissena polymor- pha Pallas. Archives Néerlandaises de Zoologie, 15: 237-314. BAUER, G., 1987, Reproductive strategy of the freshwater pearl mussel Margaritifera margaritif- era. Journal of Animal Ecology, 56: 691—704. BAYNE, B. L., 1984, Aspects of reproductive be- haviour within species of bivalve molluscs. In: ENGELS, W., et al., eds., Advances in inverte- brate reproduction, Vol. 3. Elsevier Publishers, Amsterdam, 357-366. BAYNE, В. L., В. J. THOMPSON & J. WIDDOWS, 1976, Physiology |. In: BAYNE, B. L., ed., Marine mussels: their ecology and physiology. Cam- bridge University Press, Cambridge, 121-206. BENSON, B. B. & D. KRAUSE, JR., 1980, The con- centration and isotopic fractionation of gases dissolved in freshwater in equilibrium with the at- mosphere. 1. Oxygen. Limnology and Ocean- ography, 25: 662-671. BORCHERDING, J., 1989, Die Reproduktionsleis- tung der Wandermuschel Dreissena polymorpha. Dissertation Univ. Köln, 168 pp. CALOW, P., 1983, Energetics of reproduction and its evolutionary implications. Biological Journal of the Linnean Society, 20: 153-163. CLARKE, K. B., 1952, The infestation of water- works by Dreissena polymorpha, a fresh water mussel. Journal of the Institute for Water Engi- neering, 6: 370-379. DAVIS, H. C. & P. E. CHANLEY, 1956, Spawning and egg production of oysters and clams. Biolog- ical Bulletin (Woods Hole), 110: 117-128. DE WILDE, P. A. W. J. & E. M. BERGHUIS, 1978, Laboratory experiments on the spawning of Ma- coma balthica; its implication for production re- search. In: MCLUSKY, D. S. & A. J. BERRY, eds., 12th European Symposium on Marine Biol- ogy, Stirling, Scotland. Pergamon Press, Oxford, 375-384. FISCHER, H., 1983, Shell weight as an indepen- dent variable in relation to cadmium content of molluscs. Marine Ecology Progress Series, 12: 59-75. GIESE, A. C., 1959, Comparative physiology. An- nual reproductive cycles of marine invertebrates. Annual Review of Physiology, 21: 547-576. GIESE, A. C., 1969, A new approach to the bio- chemical composition of the mollusc body. Oceanography and Marine Biology Annual Re- view, 7: 175-229. HEBERT, P. D. N., B. W. MUNCASTER £ G. L. MACKIE, 1989, Ecological and genetic studies on Dreissena polymorpha (Pallas): a new mol- lusc in the Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences, 46: 1587-1591. KACANOVA, A. A., 1961, Some data on the repro- duction of Dreissena polymorpha Pallas in the Uchinsk reservoir. Trudy Vsesoyuznogo Gidrobi- ologicheskogo Obshchestva, 11: 117-121. LANGTON, R. W., W. E. ROBINSON & D. SCHICK, 1987, Fecundity and reproductive effort of the sea scallops Placopecten magellanicus from the Gulf of Maine. Marine Ecology Progress Series, 37: 19-25. LOOSANOFF, V. L. & C. H. DAVIS, 1963, Rearing of bivalve mollusks. Advances in Marine Biology, 1: 1-136. LYASHENKO, A. V. & T. A. KHARCHENKO, 1989, Annual dynamics of energy metabolism in a freshwater clam. Hydrobiological Journal, 25 (3): 31-38. MORTON, B., 1969, Studies on the biology of Dreis- sena polymorpha Pall. Ill. Population dynamics. Proceedings of the Malacological Society of Lon- don, 38: 471-482. NAVARRO, J. M. & J. E. WINTER, 1982, Ingestion rate, assimilation efficiency and energy balance in Mytilus chilensis in relation to body size and different algal concentrations. Marine Biology, 67: 255-266. RACHOR, E., 1976, Structure, dynamics and pro- ductivity of a population of Nucula nitidosa (Bivalvia, Prosobranchia) in the German Bight. Berichte der Deutschen Wissenschaftlichen Kom- mission für Meeresforschung, 24: 296-331. SACHS, L., 1978, Angewandte Statistik—Planung und Auswertung, Methoden und Modelle. Springer, Berlin. SALZWEDEL, H., 1979, Reproduction, growth, mor- tality, and variations in abundance and biomass of Tellina fabula (Bivalvia) inthe German Bight 1975/ 76. Veröffentlichungen des Instituts für Meeres- forschung in Bremerhaven, 18: 111-202. SKIRKYAVICHENA, Z. YU., 1970, The regularities of weight ofthe Dreissena Il. The seasonal effect. Trudy Akademiia NAUK Litov, SSR, Ser. C, 2: 91-97. SPRUNG, M., 1983, Reproduction and fecundity of the mussel Mytilus edulis at Helgoland (North Sea). Helgoländer Meeresuntersuchungen, 36: 243-255. SPRUNG, M., 1987, Ecological requirements of de- veloping Dreissena polymorpha eggs. Archiv für Hydrobiologie, Supplement, 79: 69-86. SPRUNG, M., 1989, Field and laboratory observa- tions of Dreissena polymorpha larvae: abun- 70 SPRUNG dance, growth, mortality and food demands. Ar- chiv für Hydrobiologie, 115: 537-561. STANCZYKOWSKA, A., 1977, Ecology of Dreis- sena polymorpha (Pall.) (Bivalvia) in lakes. Pol- skie Archivum Hydrobiologii, 24: 461-530. STICKNEY, A. D., 1964, Feeding and growth of the juvenile soft- shell clams, Mya arenaria. U.S. Fish and Wildlife Service, Fishery Bulletin, 63: 635— 655. THOMPSON, R. J., 1979, Fecundity and reproduc- tive effort in the blue mussel (Mytilus edulis), the sea urchin (Strongylocentrotus droebachiensis), and the snow crab (Chionoectes opilio) from pop- ulations in Nova Scotia and Newfoundland. Jour- nal of the Fisheries Research Board of Canada, 36: 955-964. TREVALLION, A., 1971, Studies on Tellina tenius Da Costa Ill. Aspects of general biology and en- ergy flow. Journal of Experimental Marine Biol- ogy and Ecology, 7: 95-122. WALNE, P. R., 1979, Culture of Bivalve Molluscs. 50 Years of Experience at Conwy. Fishing News (Books), Surrey, 189 pp. WALZ, N., 1978, The energy balance of the fresh- water mussel Dreissena polymorpha Pallas in laboratory experiments and in Lake Constance Il. Reproduction. Archiv fur Hydrobiologie, Supple- ment, 55: 106-119. Revised Ms. accepted 16 July 1990 MALACOLOGIA, 1991, 33(1-2): 71-178 THE GENITALIC, ALLOZYMIC AND CONCHOLOGICAL EVOLUTION OF THE TRIBE MESODONTINI (PULMONATA: STYLOMMATOPHORA: POLYGYRIDAE) Kenneth C. Emberton Department of Malacology, Academy of Natural Sciences of Philadelphia, 19th and the Parkway, Philadelphia, Pennsylvania 19103, U.S.A. TABLE OF CONTENTS Abstract Conchological Variation Introduction Patterns of Genitalic Evolution Materials and Methods Patterns of Conchological Evolution Taxa Studied Discussion Collections Genitalic Analysis Dissections Allozymic Analysis Electrophoresis Phylogenetic Analysis Data Analysis Revised Classification Patterns of Genitalic Evolution Genitalic Evolution Patterns of Shell Evolution Shell Evolution Taxonomic History Recommendations for Future Research Results Acknowledgments Genitalic Analysis Literature Cited Variation Appendix 1. Systematic Review Descriptions Appendix 2. Alternative Phylogenetic Analy- Suggested Character-State Transforma- sis Using Successive Weighting tions Appendix 3. Alternative Phylogenetic Analy- Allozymic Analysis sis Treating Genitalic and Electrophoretic Phylogenetic Analysis Data Separately, Then Seeking a Con- Revised Classification sensus ABSTRACT The Mesodontini, Tryon, constitute a conchologically diverse radiation of 42 species of land snails in eastern temperate North America. The last monograph on the Mesondontini appeared 50 years ago, and was based primarily on shells. Dissection of the uneverted penial tubes of all 42 known species revealed a morphological diversity that was classified into five characters comprising 37 character-states. Starch-gel elec- trophoresis of pedal tissue detected 95 alleles among 16 loci. Maximum-parsimony cladistic analyses, with Ashmunella and Allogona as outgroups, were performed, assigning weights of 1, 2, 3, 4, 5 and 6 to the genitalic character-states relative to the allozymic character-states. Branch-by-branch visual comparison of all resulting trees resulted in a synthetic phylogenetic hypothesis. Two alternative approaches to phylogenetic analysis closely corroborated this hy- pothesis, and indicated a basic congruence of anatomical and biochemical data sets. Supraspecific revision based on this phylogenetic hypothesis divides the Mesodontini into ten genera and subgenera: Fumonelix gen. nov.; Inflectarius (Hubrichtius) subgen. nov.; I. (Inflec- tarius) Pilsbry, 1940; Mesodon (Akromesodon) subgen. nov.; M. (Aphalogona) Webb, 1954; M. (Appalachina) Pilsbry, 1940; M. (Mesodon) Férussac, 1821; Patera (Patera) Albers, 1850; P. (Ragsdaleorbis) Webb, 1954; and P. (Vesperpatera) subgen. nov. Genitalic and geographic comparisons between 29 pairs of sister taxa detected evolutionary trends similar to those previously found in the Triodopsini: sister taxa with peripatric geographical ranges generally showed little or no difference in penial sculpture; those with sympatric ranges showed no more than moderate differences; and all examples of great genitalic differences, as 71 72 EMBERTON well as most examples of moderate genitalic differences, between sister taxa occurred in those with parapatric or allopatric ranges. Population-level comparisons for 16 species failed to find any trace of reproductive character displacement with species of similar size and shell shape. These findings support and generalize the hypotheses made for the Triodopsini, that peripheral isolates generally do not differentiate, that differentiation due to reproductive character displace- ment is moderate at most, and that major differentiation is rare, rapid, and occurs in isolates. These findings do not support the hypothesis that vicariant isolates generally differentiate slowly. The pattern of shell evolution includes the relative conchological stasis of subgenera, with a few intriguing exceptions. A globose, toothless, imperforate, hairless shell-form occurs in all genera, and typifies the most plesiomorphic subgenera of the two most plesiomorphic genera. If this shell form is indeed plesiomorphic for the Mesodontini, then a parietal denticle evolved independently at least four times; and a basal lamella, an exposed umbilicus, and periostracal hairs each evolved at least three times. Hypotheses concerning the functions of these structures remain untested. The nature and limits of a species in the Mesodontini require further research. For example, there is extreme variation in penial length within and among nominal species, but the effect of this character on gene flow is unknown. The many sympatric convergences in shells between the Mesodontini (subfamily Polygyrinae) and the Triodopsini, Pilsbry (= Webbhelix, Neohelix, Xolotrema and Triodopsis) (subfamily Triodopsinae) provide naturally replicated experiments in evolutionary morphology. Key words: snails; evolution; genitalia; allozymes; shells; cladistics; character displacement; convergence. INTRODUCTION The family Polygyridae Pilsbry, 1894a, is autochthonous to North American and com- prises approximately 260 species currently classified into 17 genera in three subfamilies (Pilsbry, 1940; Webb, 1974; Hubricht, 1985; Richardson, 1986; Emberton, 1988). This pa- per deals with a presumed monophyletic ra- diation (see Emberton, 1986) in the subfamily Polygyrinae that has been known as Mes- odon Férussac, 1821, and is here elevated to the rank of tribe as Mesodontini, Tryon, comprising 42 species in the following ten subgenera and genera: Fumonelix gen. nov.; Inflectarius (Hubrichtius) subgen. nov.; I. (In- flectarius) Pilsbry, 1940; Mesodon (Akromes- odon) subgen. nov.; M. (Aphalogona) Webb, 1954b; M. (Appalachina) Pilsbry, 1940; M. (Mesodon) Ferussac, 1821; Patera (Patera) Albers, 1850; P. (Ragsdaleorbis) Webb, 1954b; and P. (Vesperpatera) subgen. nov. The Mesodontini are restricted to eastern temperate North America, east of the Great Plains. They are a common, large (shell di- ameter about 8-40 mm), and sometimes dominant and conspicuous element of the in- vertebrate faunas of leaf-litter and floodplain habitats. For example, the density of Mes- odon thyroidus on a floodplain in Illinois was 63,330 snail per hectare, with a standing tis- sue biomass of 26 kg per hectare, exceeding maximal fish biomass in the most productive river in the state (Foster, 1937). Dead plant material and living herbs and fungi seem to be their chief foods (Pilsbry, 1940), which are di- gested by presumably endogenous cellulases (Runham, 1975). Mesodontins are eaten by a wide variety of mammals, reptiles, amphibi- ans and insects (references in Emberton, 1986; personal observations), and thus are an important link in the food chain. Large me- sodontins are a potential North American source of anti-A agglutinin for typing human blood, as helicid snails now are in Europe (Miles, 1983). Some mesodontins are inter- mediate hosts to parasites of various game and non-game mammals, some of them lethal (Maze & Johnstone, 1986). Although human meningoencephalitis, which is carried by land snails and can cause paralysis and blindness in humans, is restricted to the tropics, it has been shown experimentally to be transmissi- ble by the (temperate-climate) mesodontin Mesodon thyroidus (Say, 1817) (see Schultz, 1982). Were this disease to invade the United States, phylogenetic knowledge of its poten- tial local carriers could be essential to its con- trol. As in the case of the tribe Triodopsini, Pils- bry (here erected to comprise the genera Webbhelix, Neohelix, Xolotrema and Triodop- sis sensu Emberton, 1988), the large size, high density, low vagility and easy markability of the Mesodontini make them favorable sub- jects for studies in ecology (e.g. Foster, 1937; Solem, 1955; Blinn, 1963; Randolph, 1973; Emberton, 1981, 1986; Asami, 1988a, 1988b) and population genetics [R. K. Selander, per- sonal communication concerning an unpub- MESODONTINI EVOLUTION 73 lished study of Patera roemeri (Pfeiffer, 1848); Hubricht (1985) and personal observa- tions concerning the six sympatric color-and- banding morphs of Mesodon altivagus on Clingman’s Dome, Great Smoky Mountains National Park, Tennessee-North Carolina]. Because of their diversity of courtship dis- plays and methods of external sperm ex- change, the Mesodontini are good subjects for studies on systems of mate recognition and speciation (Webb, 1947a, 1947b, 1954b, 1968a, 1968b, 1983). The Mesodontini also exhibit numerous cases of sympatric concho- logical convergences with the Triodopsini (which have brief courtship and internal sperm exchange), and thus offer superb nat- urally replicated experiments in evolutionary morphology (see Pilsbry, 1940; Solem, 1976; Emberton, 1986, 1988). The last monograph on the Mesodontini (Pilsbry, 1940) is now 50 years old, and was based primarily on shells. The purposes of this paper are (1) derivation of a phylogenetic hypothesis for the Mesodontini based on male genitalia and allozymes; (2) revision of the Mesodontini above the level of species based on this phylogenetic inference; and (3) analysis of phylogenetic patterns of variation in both genitalia and shell, in order to gener- ate testable hypotheses about the evolution- ary processes that produced these patterns. For this analysis, the species designations of Hubricht (1985) have been followed, ex- cept for elevating Mesodon altivagus and M. trossulus to the status of species, removing Inflectarius verus from synonomy with /. sub- palliatus and synonomizing Fumonelix cling- manicus with F. wheatleyi. According to this scheme there are 42 species in the tribe Me- sodontini. Genitalic data are presented here for all 42 species. Most of these data are new. The pe- nial morphologies of ten species of the Mes- odontini (Patera appressa, Fumonelix christyi, Mesodon clausus, M. elevatus, I. in- flectus, P. kiowaensis, P. pennsylvanica, P. sargentiana, M. thyroidus and M. zaletus) were previously studied by Webb (1954b; 1968a, 1968b, 1983). As noted for the Triod- opsini (Emberton, 1988), Webb’s figures tend to omit some structural and sculptural detail. In addition, the anatomical distortion some- times produced by Webb's technique of killing with boiling water and then often crushing the shell can be even more pronounced in mes- odontins than in triodopsins, owing to their lack of rigid sculptural elements. The dissec- tive method for studying penial morphology— slitting and pinning open the uneverted penial tube (Emberton, 1988: Fig. 1)—was recom- mended by Pilsbry (1940) for future revisers but, until this study, had been used for only eight species of mesodontins: four species (Mesodon elevatus, M. sayanus, M. thyroidus and M. zaletus) by Leidy (1851); one (Inflec- tarius subpalliatus) by Pilsbry (1940); and three (Patera binneyana, Mesodon clausus and P. clenchi) by Solem (1976). All of these studies included excellent, detailed illustra- tions, but only one per species, with little or no discussion of intraspecific variation. Thus, in genital morphology, 28 of the 42 species of mesodontins never had been ex- amined before, and of the 14 that had, none had been examined for individual variation and six had possible distortion or omission of details or both. Limited additional information on penial sculpture was available from Pilsbry's (1940) one or two sections of the penis for each of 17 species of mesodontins (Patera binneyana, P. clarki, Mesodon clausus, M. elevatus, Inflectarius ferrissi, |. in- flectus, Fumonelix jonesiana, P. laevior, I. magazinensis, P. pennsylvanica, P. peri- grapta, P. roemeri, P. sargentiana, M. say- anus, M. thyroidus, F. wheatleyi and M. zale- tus), as well as from the histological studies of the penes of two species by Cox (1979: Me- sodon zaletus) and Cookson (1982: Mesodon thyroidus). The value of combining morphological and molecular data for phylogenetic studies has been well reviewed by Hillis (1987). Starch- gel electrophoresis is now a traditional source of molecular data for phylogenetic analysis (Buth, 1984; Hillis, 1987). New electro- phoretic data are reported herein for 39 of the 42 species of the Mesodontini. For only one of these species (Mesodon normalis) has elec- trophoretic data previously been published (McCracken 8 Brussard, 1980, as reevalu- ated by Emberton, McCracken & Wooden, in preparation). MATERIALS AND METHODS Taxa Studied The following taxa were studied. Taxa are arranged according to this revision, but here alphabetically by genus and by species within each genus. 74 EMBERTON Fumonelix gen. nov. archeri (Pilsbry, 1940) christyi (Bland, 1860) jonesiana (Archer, 1938) orestes (Hubricht, 1975) wetherbyi (Bland, 1874) wheatleyi (Bland, 1860) Inflectarius Pilsbry, 1940 Inflectarius (Hubrichtius) subgen. nov. downieanus (Bland, 1861) kalmianus (Hubricht, 1965) Inflectarius (Inflectarius) approximans (Clapp, 1905) edentatus (Sampson, 1889) ferrissi (Pilsbry, 1897) inflectus (Say, 1821) magazinensis (Pilsbry & Ferriss, 1907) rugeli (Shuttleworth, 1852) smithi (Clapp, 1905) subpalliatus (Pilsbry, 1893) verus (Hubricht, 1954) Mesodon Férussac, 1821 Mesodon (Akromesodon) subgen. nov. altivagus (Pilsbry, 1900) andrewsae Binney, 1879 normalis (Pilsbry, 1900) Mesodon (Aphalogona) Webb, 1954b elevatus (Say, 1821) mitchellianus (Lea, 1838) zaletus (“Say” Binney, 1837) Mesodon (Appalachina) Pilsbry, 1940 chilhoweensis (Lewis, 1870) sayanus (Pilsbry, in Pilsbry & Ferriss, 1906) Mesodon (Mesodon) clausus (Say, 1821) sanus (Clench & Archer, 1933) thyroidus (Say, 1817) trossulus Hubricht, 1966 Patera Albers, 1850 Patera (Patera) appressa (Say, 1821) clarki (Lea, 1858) laevior (Pilsbry, 1940) panselena (Hubricht, 1976) perigrapta (Pilsbry, 1894b) sargentiana (Johnson & Pilsbry, 1892) Patera (Ragsdaleorbis) Webb, 1954b pennsylvanica (Green, 1827) Patera (Vesperpatera) subgen. nov. binneyana (Pilsbry, 1899) clenchi (Rehder, 1932) indianorum (Pilsbry, 1899) kiowaensis (Simpson, 1888) leatherwoodi (Pratt, 1971) roemeri (Pfeiffer, 1848) Collections Principal field work was conducted April— June, 1982, in the eastern United States (“GS” series), and was supplemented by col- lections from the lower Ohio River Valley in April, 1980, (“H” series) and from the south- ern Appalachian area in March—June, 1983, (“SC” series). All collections were deposited in the Field Museum of Natural History, Chi- cago (FMNH). County-level localities, field- station numbers, and FMNH catalog numbers of dissected and electrophoresed voucher material are listed under each species in the systematic review (Appendix 1). Detailed lo- cality data are available from the author upon request or from the FMNH catalog. Snails in each lot were marked individually on their shells: 1, 2, 3, etc. for snails from which tissue samples were taken; and A, B, C, etc. for snails not so sampled. Appendix 1 shows which individual specimens from each lot were dissected, electrophoresed, and illus- trated anatomically or conchologically or both. Additional anatomical materials (total 16 lots) were borrowed from FMNH, from the Academy of Natural Sciences of Philadelphia (ANSP), and from Mr. Leslie Hubricht's pri- vate collection (which has been willed to FMNH). Dissections The uneverted penial tubes of 203 snails were dissected, and the everted penes of 19 snails were examined, from a total of 96 pop- ulations comprising all 42 species of the Me- sodontini. In order to assess qualitatively the variation in penial-morphological characters at the in- dividual, populational and specific levels, mul- tiple dissections and illustrations were made of two populations each of Mesodon zaletus and M. elevatus wherever they were sym- patric. These two species are very similar in size and shape of shell, and Leidy’s (1851: plate X, figs. Ill, V) illustrations of their penial sculpture indicated an essential similarity. The first sympatric locality was a wooded hill- side above a creek, 1.5 miles north of the Sherwood post office, Franklin County, Ten- nessee (GS-104), collected 1 May 1982 by Ken and Ellen Emberton. The shells of M. za- letus and M. elevatus were so similar at this site that field identification of many individuals was in doubt; the hypothesis that this repre- sented hybridization, however, was rejected MESODONTINI EVOLUTION 19 based upon electrophoretic evidence—M. za- letus and M. elevatus were fixed at alternative alleles for six of the 16 loci examined (Mdh-1, lcd, Gd-1, Got-2, Lap, Mpi; see Table 2). Closer examination of the shells also re- vealed features reliably distinctive of the two species. Population densities were high at this site owing to the presence of a small trash dump. Of the 41 adults collected in two per- son-hours, six were M. zaletus and 35 were M. elevatus. The collecting area was small, on the order of 50 square meters; the two species were almost certainly capable of en- countering each other. Five individuals of each species were randomly chosen for dis- section, illustration and comparison of the male genitalia. The second locality studied with sympatric M. zaletus and M. elevatus was “in the vicinity of the shelter in the primitive camping area, McCormick’s Creek State Park, Spencer, Owen County, Indiana,” collected 1 June 1974 by Glenn Goodfriend. The collection consists of nine adult and four juvenile M. za- letus (FMNH 214785) and of nine adult and two juvenile M. elevatus (FMNH 214655), al- though several “adults” of both species were mature in shell only, with their genitalia imma- ture. Neither the collecting area nor the de- gree of sympatry was recorded for this collec- tion. The two species were much more readily separable by shell morphology at this site than at that in Tennessee. For interspecific systematic comparisons, at least three adults per species were dis- sected whenever possible. Because of the limitations of the material, however, three species (Patera clenchi, P. pennsylvanica and Mesodon trossulus) were represented by only a single dissection each, and nine spe- cies (/nflectarius approximans, Fumonelix archeri, Mesodon chilhoweensis, |. downiea- nus, P. indianorum, P. kiowaensis, P. leather- woodi, F. orestes and M. sanus) were repre- sented by only two dissections each. The remaining 30 species were represented by three to 16 dissections each, often with at least three from a single population. The number of populations dissected per species ranged from one to ten. A single population was dissected of each of the 22 mesodontin species; two populations were dissected of each of 11 of the species (M. altivagus, P. appressa, |. approximans, M. chilhoweensis, M. clausus, M. elevatus, |. fer- rissi, F. jonesiana, P. leatherwoodi, M. mitch- ellianus and I. rugeli); three populations of two species (P. laevior, M. thyroidus); four popu- lations of one species (I. inflectus); five pop- ulations of two species (P. binneyana, M. nor- malis); six populations of one species (P. clarki); seven populations of one species (F. wheatleyi); nine populations of one species (M. zaletus); and ten populations of one spe- cies (P. perigrapta). Not all of the populations of M. zaletus and P. perigrapta were actually dissected; in some the genital morphology was visible on specimens that had everted their penes in the drowning jar. Comparisons of genitalic anatomies of out- groups of the Mesodontini were made from published illustrations. According to a prelim- inary phylogenetic analysis of the Polygyridae (Emberton, 1986), the sister group to the Me- sodontini consists of a lineage that split into Polygyra sensu lato and Praticolella, and its next nearest outgroup is Stenotrema. Penial- morphological data on Polygyra sensu lato were available from Pilsbry (1940), Webb (1950, 1967), Pratt (1981) and Emberton (1986). Data on Praticolella were available from Pilsbry (1940), Webb (1967) and Ember- ton (1986); and data on Stenotrema were available from Pilsbry (1940), Archer (1948) and Webb (1948, 1950). More distant out- groups used for comparison were two genera of the subfamily Ashmunellinae, Ashumunella Pilsbry & Cockerell, 1899, and Allogona Pils- bry, 1939. Genitalic data for Ashmunella were taken from Pilsbry (1940) and Webb (1954a); and for Allogona from Leidy (1851) and Pils- bry (1940). For the Mesodontini, one representative dissection per species was illustrated in de- tail, using a camera lucida. These illustrations were compared for character-state differ- ences, which were grouped into presumably homologous characters (Patterson, 1982; Wagner, 1989). For each character, a phylog- eny of its character-states was proposed, based on the criterion of continuity of forms (Hennig, 1966; Wiley, 1981). These phyloge- nies were polarized by outgroup comparison (Hennig, 1966; Watrous & Wheeler, 1981). Electrophoresis Biochemical methods were identical to those used for the Triodopsini (Emberton, 1988). Posterior pedal tissues (“snail tails’) were excised from field-activated snails and stored in screw-top cryogenic vials in liquid nitrogen. Horizontal starch-gel electrophore- sis followed methods of Selander et al. (1971) 76 EMBERTON and Shaw & Prasad (1970), as modified by Davis et al. (1981) and by Emberton (1988). Twelve enzyme systems from a wide array of biochemical pathways (Richardson et al., 1986) yielded 16 variable loci: Sordh, Mdh- 1&2, Me, Icd, Pgd, Gd-1&2, Sod-1&2, Got- 1&2, Pgm, Lap, Mpi and Gpi. Methodological details are given in Emberton (1988: Appen- dix A). The electrophoresed material comprised 706 snails from 81 populations representing 38 or 39 species. The three species for which tissue samples were lacking were F. archeri, |. downieanus and Г. jonesiana. The tissue sample for P. clenchi was from a single juve- nile specimen, the identity of which was nec- essarily suspect. All electrophoresed species were repre- sented by at least one population with com- plete data for all loci. Sixteen species were represented by a single population each (M. andrewsae, |. approximans, Е. christyi, P. clenchi?, M. elevatus, P. indianorum, |. kalm- ianus, P. kiowaensis, P. leatherwoodi, |. mag- azinensis, M. mitchellianus, F. orestes, P. pennsylvanica, M. sanus, M. trossulus and F. wetherbyi); nine species were represented by two populations each (P. binneyana, M. chilhoweensis, |. edentatus, M. normalis, P. panselena, P. roemeri, P. sargentianus, M. sayanus and I. subpalliatus); 11 species were represented by three populations each (M. al- tivagus, P. appressa, P. clarki, M. clausus, I. ferrissi, |. inflectus, P. laevior, P. perigrapta, |. rugeli, |. smithi and F. wheatleyi); one spe- cies was represented by five populations (M. thyroidus); and one species was represented by eight populations (M. zaletus). Catalog numbers of the voucher specimens for all electrophoresed populations are given in Ta- ble 2 and in the systematic review (Appendix 1): Of the electrophoresed snails, 140 be- longed to a single population (Monte Sano, Alabama) of Mesodon zaletus that was used as the control on all gels for both the Mesod- ontini and the Triodopsini (Emberton, 1988). Eliminating this and five other populations of M. zaletus that were analyzed separately (Emberton, 1986; in preparation), the number of snails electrophoresed per population for the 75 populations listed in Table 2 ranged from one to 22, with a mean of 7.1 and a standard deviation of 5.2. The closest outgroups of the Mesodontini from which comparable electrophoretic mate- rial was available were Ashmunella danielsi dispar Pilsbry & Ferriss, 1915, of which one population with a sample size of two was electrophoresed, and Allogona profunda (Say, 1821), of which one population with a sample size of ten was included in the anal- ysis, using data from Emberton (1988: Table 2). Data Analysis Three methods of phylogenetic analysis were used: (1) maximum-parsimony analysis of combined data, with various weights as- signed to anatomical over allozymic data, and a consensus tree constructed from the result- ing cladograms; (2) maximum-parsimony analysis of combined data, with successive weighting (Farris, 1988); and (3) maximum- parsimony analyses of anatomical and of al- lozymic data, and Distance-Wagner analyses of allozymic data, with a consensus tree con- structed from the weighted resulting trees (Emberton, 1988). The first of these seems the most objective and informative of all avail- able methods (see Discussion), and is the only method presented in the body of the text. The second and third methods were used only for comparison, and are presented in Ap- pendices 2 and 3. For phylogenetic analysis (first and pre- ferred method), transformations of the geni- talic character-states and the allozymic alleles (see Michevich & Johnson, 1976; Buth, 1984; Hillis, 1987; Swofford & Olsen, 1990) were binary-coded as present (1) or absent (0) for each species. Ashmunella danielsi dispar and Allogona profunda, both of which scored 0 for all genitalic transformations, were used as out- groups. All autapomorphies were deleted, and the data matrix was analyzed using Hennig86 (Farris, 1988), a cladistics program that uses the criterion of Wagner unrestricted parsimony (Kluge & Farris, 1969; Farris, 1983) and that empirically has been found superior to all other currently available programs (Platnick, 1989). Separate analyses were performed using six different weighting schemes, with each geni- talic transformation assigned a weight of 1, 2, 3, 4, 5 and 6 times the weight of each allozymic allele. Each analysis used the mhennig* bb* options. The mhennig* option produces one or more preliminary trees by single passes through the data, then applies branch-swap- ping to them; the bb* option subjects the re- sulting trees to further branch-swapping, and outputs all trees of the maximal discovered parsimony. A Nelson (1979) strict consensus MESODONTINI EVOLUTION 77 tree was then calculated from this set of max- imum-parsimony trees, using the “nelsen” op- tion of Hennig86. The six analyses were com- pared visually to identify those clades most robust to assumptions about the relative reli- abilities of morphological and electrophoretic data sets (see Hillis, 1987). A general phylo- genetic hypothesis for the Mesodontini was constructed from these comparisons, for use in evaluating evolutionary patterns in genitalic and shell morphology. Patterns of Genitalic Evolution Patterns of evolution in penial morphology were analyzed by comparing sister taxa (spe- cies or terminal clades appearing dichoto- mously in the general phylogenetic hypothe- sis). For each of 29 pairs of sister taxa, their difference in penial morphology (judged from the illustrated dissections) was ranked as great, moderate, slight, or none; and the geo- graphical relationship of their ranges was classified as allopatric, sympatric, parapatric, or peripatric (in which one taxon is a small- ranged endemic peripheral to, or isolated within, the much broader range of the other taxon). Geographic ranges were taken from Hubricht (1985), with slight modification based upon collecting results. The importance of reproductive character displacement was assessed by first dissect- ing the genitalia of populations of 16 species that were sympatric with another mesodontin species of similar shell size and shape, then by comparing these dissections with addi- tional dissections of populations that were al- lopatric with the same species. Patterns of Shell Evolution To analyze conchological evolution at the generic, subgeneric and species-group lev- els, a representative shell was chosen for each species and was mounted in its proper position on the general phylogenetic hypoth- esis. Patterns of change through time were interpreted under the assumptions that the phylogenetic hypothesis was correct, and the shell morphology of each (unknown) ancestor was intermediate to the morphologies of its living descendents. To aid in analyzing shell evolution, and to assist users of this paper in making identifications, a representative shell of each species, whenever available, was il- lustrated in detail. Two views are shown—in the apertural plane and perpendicular to the apertural plane—that simultaneously display as many important shell features as possible. For most species, the illustrated shell was from the same population of which the geni- talia were illustrated. TAXONOMIC HISTORY The origin of the generic name Mesodon has been outlined by Rosenberg & Emberton (in press). The name Mesodon first appeared in Rafinesque (1819: 66) as Mesodon leuc- odon. Both the generic and specific names were nude, however. Férussac [1821: 33 (quarto) or 37 (folio)] listed Mesodon leuc- odon Rafinesque in the synonymy of Helix (Helicodonta) thyroidus Say, 1817, and listed Mesodon helicinum Rafinesque, a manuscript name, in the synonymy of his Helix (Helic- odonta) knoxvillina, then a nude name. The first use of Mesodon as an available name was that of Rafinesque (1831: 3), who briefly described the genus, thereby making the name available from the time of its appear- ance in synonymy in Férussac (1821), and included only Mesodon maculatum Rafi- nesque, 1831 (a nomen dubium). Thus by the provisions of the International Commission on Zoological Nomenclature (1985) Articles 11e, 50g and 671, the correct citation for the genus is Mesodon Ferussac, 1821, with type spe- cies Helix thyroidus Say, 1817, by monotypy. Martens (1860) later expanded Mesodon to include all eastern North American land snails with capacious, subglobose shells with a small parietal tooth or toothless, thereby dis- tinguishing it from Triodopsis Rafinesque, 1819, in which he included all eastern North American species with depressed, bidentate or tridentate shells (see Pilsbry 1940: 790). Martens’ (1860) diagnoses of Mesodon and Triodopsis were followed by Tryon (1867), Binney & Bland (1869), and some later au- thors, although many later authors (e.g. Baker, 1939) continued to synonymize the two genera under the genus Polygyra Say, 1818. The description of Polygyra seemingly had been overlooked by Martens and other European workers. Pilsbry’s (1930) dissections, leading to his monograph on Mesodon (Pilsbry, 1940: 702— 778), were the first clearly to characterize Me- sodon anatomically and to recognize it as oc- cupying the entire range of shell shapes also occupied by Triodopsis and Allogona Pilsbry, 78 EMBERTON 1939. Pilsbry divided Mesodon into four sub- genera, Mesodon s. str. Ferussac, 1821; Pa- tera Albers, 1850; Appalachina Pilsbry, 1940; and /nflectarius Pilsbry, 1940, based upon shell shape. Taxonomic changes in Mesodon following Pilsbry’s (1940) monograph have been sum- marized by Miller et al. (1984) and by Rich- ardson (1986). Additional species were de- scribed by Hubricht (1954, 1965, 1966, 1975, 1976) and Pratt (1971). Additional distribu- tional records and a few life-history notes were published by several workers, primarily in the journals Sterkiana and Nautilus, but scattered among many journals and regional faunal surveys. Webb (1954b, 1968a, 1968b, 1983) published a series of reports on the re- productive behavior and everted genitalia of selected species of the Mesodontini, and pointed out important variations in penial mor- phology, upon which he based his new sec- tion Aphalogona Webb, 1954b, and subgenus Ragsdaleorbis Webb, 1954b. Solem’s (1976) “Comments on Eastern North American Polygyridae” laid solid groundwork for future revision of the Meso- dontini. In that work, Solem compared the sympatric, conchologically similar Patera bin- neyana, Mesodon clausus and Patera clenchi with one another and with two sympatric, con- chologically similar species of the Triodopsini. He compared shell, radular structure, jaw structure, external aspect of the reproductive system and dissected penial morphology, he emphasized the need for comparisons of sympatric species to establish criteria for dis- tinguishing allopatric species, and he indi- cated penial morphology as potentially the best source of systematically useful charac- ters. Using SEM, Solem detected ectocones on the marginal and lateral radular teeth of Mesodon clausus, which had been missed by Pilsbry (1940: 703). Hubricht’s (1985) book of range maps for land snails of the eastern United States made species-level taxonomic changes within the Mesodontini, resulting in a total of 40 species, but did not mention supraspecific taxa. Richardson’s (1986) bibliographic cata- log of polygyrid species went to press too early to incorporate Hubricht's taxonomic changes. Concerning the name of the family to which the Mesodontini belong, the International Commission on Zoological Nomenclature has recently been petitioned (Emberton, 1989a) to grant precedence of Polygyridae Pilsbry, 1894a, over Mesodontidae Tryon, 1866, be- cause of its long and stable usage. RESULTS Genitalic Analysis Variation A mesodontin snail copulates by placing its sperm mass on the everted penis of its mate, usually with the two penes inter- twining (Fig. 1). The penis (Fig. 2) bears a terminal, ventrally opening, cup-like chalice (Webb, 1954b) for molding the sperm mass; lateral pilasters for containing the sperm mass and often—by differences in their lengths or rigidities—for coiling the penis; and sometimes dorsal ridges or other frictional sculpture for retaining the received sperm mass. The basin of the chalice usually has a lining of minute, parallel riblets that presum- ably increase the surface area for resorb- ing water to concentrate and solidify the autosperm mass. Although there is some de- gree of preservational distortion of the chalice walls in the uneverted penial tube, the struc- ture of the chalice can be studied effectively by dissection (Emberton, 1988: Fig. 1) [e.g., compare Figs. 2 and 11b; compare Figs. 17, 4a-e, 6a—c; and compare Webb (1954b: fig. 6) and Solem (1976: fig. 6b)]. The term chal- ice is used here in a broader sense than in Webb (1954b), in that all mesodontins are considered to have a Chalice, even in cases in which its walls are simply unmodified terminal extensions of the lateral pilasters, and even in those cases in which the chalice is everted during copulation, and thus resembles a glans rather than a basin. Figure 3, which illustrates the unopened, uneverted penial tubes of five randomly se- lected Mesodon zaletus and five randomly se- lected M. elevatus from a site of sympatry in Tennessee, shows that individual variation is greater than interspecific differences in penial length, penial shape, vas deferens length and retractor-muscle length: the species cannot be separated by these characters. Figure 4 shows these same ten penial tubes, in the same order, cut open to reveal their sculptural details. Both M. zaletus and M. elevatus have right and left lateral pilas- ters, dorsal, longitudinal ridges of variable number and distribution, and thick chalice walls forming a A-shaped cleft, with the right chalice wall thicker than the left. In all these characters individual variation exceeds inter- MESODONTINI EVOLUTION 79 FIG. 1. External sperm exchange in Mesodon normalis. a,b. Intertwining of fully everted penes. c. Deposition of sperm mass on mate’s penis. d. Inversion of penes and their attached sperm masses. specific variation. The only consistent differ- ences in penial sculpture between M. zaletus and M. elevatus of GS-104 that are readily detectable are: the right chalice wall of M. za- letus is more thickly and more abruptly swol- len than that of M. elevatus; the left chalice wall of M. zaletus is extended into a slight lateral flap that is lacking in M. elevatus; and the surface texture of the right pilaster differs, bearing tiny pock-marks in M. zaletus, but bearing angled folds in M. elevatus. Even these differences can be difficult to detect in individual cases; M. elevatus #39 (Fig. 4i), for example, could easily be mistaken for M. za- letus because of local distortion in its chalice walls; and M. zaletus #12 (Fig. 4c) is also confusing because its penial tube was inad- vertently opened from the ventral side rather than from the dorsal side as in all the other dissections. Figure 5 diagrammatically illustrates dis- sections of three randomly chosen specimens of M. zaletus and three randomly chosen specimens of M. elevatus from another site of sympatry in Indiana, with the unopened, un- everted penial tubes shown in situ. As in M. zelatus and M. elevatus from Tennessee, in- dividual variation in penial size and shape easily outweighs any difference between spe- cies (the same is true of the length of the vas deferens, not shown in Figure 5). Another possible difference between the species—the relative points of insertion of the penial retrac- tor muscle and what seems to be the cephalic aorta (also shown diagrammatically in Figure 5)—proved also to be invalidated by individ- ual variation. Although the muscle and artery generally occur closer together in M. zaletus than in M. elevatus, in some specimens of M. zaletus (e.g. #A, Fig. 5a) they are just as far apart as in M. elevatus. Penial sculptures of M. zaletus and M. ele- vatus sympatric at the Indiana site are shown in Figure 6, in which four of the six dissected penes (M. zaletus #B, D; M. elevatus #B, C) are those shown undissected in Figure 5. As in the case of specimens from the Tennessee site, the two species (Fig. 6) have similar right and left lateral pilasters, dorsal longitudinal ridges of variable pattern, and thick chalice walls forming a A-shaped cleft. Also as in the previous comparison, the two species dif- fer in that the right chalice wall of M. zaletus was more thickly and abruptly swollen than that of M. elevatus, and the left chalice wall of M. zaletus was extended into a lateral flap lacking in M. elevatus. The difference be- tween M. zaletus and M. elevatus in the sur- face texture of the right chalice wall and upper right pilaster detected in the Tennessee pop- ulations (Fig. 4) is lacking, however, in the Indiana population (Fig. 6). The radically dif- ferent appearance of M. elevatus #B from the 80 EMBERTON 5mm FIG. 2. Dorsal and ventral views of fully everted penis of Mesodon andrewsae (GS-11 #2), showing major structural features. BB = basal bulge, Cl = chalice inside, CO = chalice outside, CW = chalice wall, DR = dorsal ridge, LP Indiana site (Fig. 6e) is due to its having been opened from the ventral rather than the dorsal side; this difference emphasizes the impor- tance of using standardized dissecting meth- ods for making anatomical comparisons of this type. Comparison of M. zaletus from the Tennes- see with Indiana populations (Fig. 4, top; Fig. 6, top) showed no populational difference greater than individual variation, except for the difference between pock-marked and smooth right chalice walls, as mentioned above. Interpopulational differences in M. el- evatus, on the other hand, are more pro- nounced: the Indiana population has chalice walls more nearly equal in size, fewer and less variable dorsal ridges, and a less corded surface structure of the right chalice wall. Thus the result of these individual, popula- tional and specific comparisons of M. zaletus and M. elevatus (Figs. 3-6) was to eliminate much of the variation in penial morphology from phylogenetic analysis. The characters that seem to be relatively stable are the over- all structures of the lateral pilasters, of the left pilaster, P = pore = exit of vas deferens, RP = right pilaster. chalice and of the dorsal ridges and bulges. Similar conclusions were drawn from unillus- trated comparisons among many populations of other species, especially Patera binney- ana, P. clarki and Fumonelix wheatleyi. Even without individual and artificial varia- tion, the dissected mesodontin penis, the ma- jor features of which are shown in Figure 4 (compare with Figure 2), is moderately rich in sculptural variation useful for phylogenetic analysis. The lateral pilasters, although somewhat inconspicuous in the everted penis (Figs. 2, 17, 20, 23, 24), are easily visible in the dissected, uneverted penis (compare Fig- ure 2 with Figure 11b; the everted with the uneverted portions of the penis in Figure 23; Figure 20a with Figure 20b; Figure 17 with Figures 4a—e, ба—с). The relative sizes and shapes of the lateral pilasters seem to be fairly stable within any given species. These pilasters are usually continuous with, or in some way connected to, the walls of the chalice. For the purpose of description and analysis, however, the lateral pilasters and the chalice have been treated as MESODONTINI EVOLUTION 81 Mesodon zaletus Mesodon elevatus FIG. 3. Uneverted penial tubes of two Mesodon species sympatric at GS-104 (Tennessee). а-е. Mesodon zaletus #2, 6, 12, 21, 27 ||. Mesodon elevatus #16, 14, 29, 39, 11. independent characters, because, although there is some correlation between the two, they are usually discrete, and because treat- ing them as a single character would have caused unnecessary complication. In the de- scriptions below, the convention has been fol- lowed of labeling the pilasters and the chalice walls “right” and “left” according to their po- sitions in the illustrated dissections, even though these sides are reversed in the everted, functioning penis. In a few of the dis- sections, these directional conventions were reversed or confused by opening the penial tube from the ventral rather than the standard dorsal side. All such cases are noted in the labels or in the figure captions. Descriptions Measurements in the following genitalic descriptions are taken solely from the illustrations (Figs. 4-25) and do not in any way reflect natural variation. [See Emberton (1989b) for an analysis of sources of individ- ual variation in penial length.] Penial length was measured from the apex to the vaginal opening; very curved penes (e.g. I. ferrissi) were measured with a Minerva curvimeter from illustrations. Penial width was measured as the distance between the outside edges of the lateral pilasters in the middle region of the penis. Fumonelix archeri Pilsbry, 1940. (Fig. 7) Dissections: two from one population. Length 13.4 mm, width 2.0 mm. Left lateral pilaster thick (0.5 mm), solid, running full length of penis; at three-fourths of distance from penial apex merging dorsally with right lateral pi- laster, expanding to width of 0.9 mm just be- fore merging; at about mid-length of penis, branching ventrally to form left arm of ventral, 82 EMBERTON M. zaletus FIG. 4. Opened penial tubes of two Mesodon species sympatric at GS-104 (Tennessee). а-е. Mesodon zaletus #2, 6, 12, 21, 27. +]. Mesodon elevatus #16, 14, 29, 39, 11. Same specimens and in same order as in Figure 3. С = chalice, CW = chalice wall, DR = dorsal ridge, LP = left pilaster, RP = right pilaster. All were opened from dorsal side except for c, which was opened from ventral side. V-shaped pouch or notch. Right lateral pilas- ter solid, cord-like, variable in width but gen- erally narrow (0.2 mm) in its mid-region and expanding apically (to 0.7 mm) to its junction with right chalice wall and basally (to 0.6 mm) to its merger with left lateral pilaster. Basal pilaster (formed by fusion of lateral pilasters) situated on basal 3.9 mm of penis, massive, solid, 1.3 mm wide and with narrow notch api- cally, gradually tapering to 0.4 mm wide ba- sally. Ovate ventral bulge, 1.1 mm long and 0.6 mm wide, present on right side of ventral surface of penis at about its mid-length. On left side of ventral surface, adjacent to ventral bulge, is a V-shaped ventral notch composed of right and left, short (2.0 mm) cord-like or fold-like arms, originating from just above ventral pilaster and from branch of left lateral pilaster, respectively, and meeting in sharply acute angle. Ejaculatory pore flush with pe- nial wall. Chalice walls relatively thick (ca. 0.5 mm) and firm, apical wall high (2.0 mm) and basally pocket-shaped, forming pointed, sym- metric hood over chalice. Penial walls free of other structures. Fumonelix christyi (Bland, 1860). (Fig. 8a) Dissections: three from one population. Length 8.6 mm, width 1.9 mm. Left lateral pi- laster short (ca. 2.2 mm long) and thin (ca. 0.05 mm wide). Right lateral pilaster stretch- ing nearly entire length of penis (length 7.7 mm), narrow throughout (width 0.05 mm), ba- sally connected with dorsal bulge by oblique ridge. Right side of dorsal surface of penis MESODONTINI EVOLUTION M. zaletus M. elevatus 83 FIG. 5. Penial retractor muscle and cephalic aorta in two Mesodon species sympatric at McCormick Creek, Indiana. a—c. Mesodon zaletus +A, В, D. d-f. Mesodon elevatus #B, С, D. bearing pilaster-like Биде, thick (0.6 mm), massive, firm and long (4.2 mm), ending about 1.4 mm above base of penis, tapering to blunt point both apically and basally, in its middle third parallel and very close to right lateral pilaster, but angling slightly away from right lateral pilaster in both its apical and basal thirds, the basal third connected to lower right lateral pilaster by obliquely angled ridge 0.9 mm long and 0.1 mm wide. Ejacu- latory pore flush with penial wall. Chalice walls 0.1 mm thick at their edges, continuous with lateral pilasters, and forming basally pointing, moderately deep (0.6 mm), symmet- ric, rounded hood or cup, 1.1 mm in outside diameter. Penial wall otherwise lacking in structures (the minute ridges and patterns in Figure 8a are seemingly artifacts of preserva- tion). Fumonelix jonesiana (Archer, 1838). (Fig. 8b) Dissections: five from two populations (in- cluding topotypes). Length 14.7 mm, width 2.6 mm. Left lateral pilaster 0.6 mm wide, rounded, firm, extending about 3.5 mm from its juncture with left side of left chalice wall to its effacement in penial wall just above ventral bulge. Right lateral pilaster low and nearly ob- solete, 5.3 mm long and 0.6 mm wide, nar- rowing basally. Penis dominated by firm, massive ventral bulge, 8.5 mm long, 1.0 mm wide apically, swelling basally to width of 2.1 mm, occupying usual position of basal portion of left lateral pilaster, but seemingly neither continuous nor homologous with it. Ejacula- tory pore flush with wall of penis. Chalice cup- shaped, 1.8 mm in diameter, ca. 1.2 mm deep, and with edges ca. 0.3 mm thick, rim continuous for four-fifths of the circumfer- EMBERTON 84 и 5 — O > © © => FIG. 6. Opened penial tubes oftwo Mesodon species sympatric at McCormick Creek, Indiana. a-c. Mesodon zaletus #B, C, D. d-f. Mesodon elevatus #A, B, C. Nearly the same specimens and ordered as in Figure 5. All were opened from dorsal side except for e, which was opened from ventral side. MESODONTINI EVOLUTION 85 FIG. 7. Opened uneverted penial tube. Fumonelix archeri, SC-279 #A (also dissected #B). LP = left pilaster, RP = right pilaster, VB = ventral bulge. ence, lower or missing on basal side. Penial walls otherwise free of structures (the thin ventral connections between the lateral pilas- ters, as well as the dark rugosities on the left dorsal wall, in Figure 8b, are seemingly pres- ervational artifacts). Fumonelix orestes Hubricht, 1975. (Fig. 8d) Dissections: two from one population (topo- types). Length 15.0 mm, width 3.0 mm. Left lateral pilaster apically firm, cord-like, and widening from 0.2 mm at its inception on left side of chalice to 0.5 mm near midpoint of penis, below which, after possible hiatus, it is uniformly broad (0.5 mm) and flattened against penial wall. Right lateral pilaster also extending full length of penis, firm, cord-like, gradually widening, with a possible hiatus near the midpoint of the penis, from 0.3 mm at Chalice to 0.6 mm at base of penis, possibly turning dorsally to approach left lateral pilas- ter near its base. Ventro-basal penial wall bearing a massive, firm, fleshy bulge, some- what teardrop-shaped, length 2.4 mm, great- est width 1.4 mm. Ejaculatory pore flush with wall of penis. Chalice shaped like hood of a Cloak, its walls relatively thick (0.2 mm) and firm, ca. 1.8 mm across and 1.1 mm deep. Penial walls without other sculpture. Fumonelix wetherbyi (Bland, 1874). (Fig. 9b) Dissections: four from one population. Length 13.6 mm, width 3.0 mm. Left lateral pilaster a humped ridge, ca. 0.3 mm thick at its edge, 4.3 mm long from its point of merger with left chalice wall, ca. 0.8 mm maximal height near its midpoint. Right lateral pilaster extending full length of penis, rather thin (0.2 mm?), extremely high-standing, and in un- everted penis recurved to right to form tube 1.1 to 0.6 mm in diameter. Ejaculatory pore flush with wall of penis. Chalice walls thin and high-standing, forming deep, cylindrical cup; in uneverted penis, chalice walls are buckled and folded inward, falsely appearing thicker and more massive. Penial walls seemingly otherwise free of other structures (the fine basal ridges and basal nodular folds of the right lateral pilaster seem to be preservational artifacts). Fumonelix wheatleyi (Bland, 1860). (Fig. 8c) Dissections: 15 from seven populations. Length 6.0 mm, width 1.3 mm, expanded api- cally. Right and left lateral pilasters difficult to discern, but seemingly thickned bands (width ca. 0.2 mm) along right and left sides of mid- dle third of penis. Just dorsal to apices of lat- eral pilasters and just below and to sides of chalice, are two bulges, 1.1 mm long and 0.3 mm wide, longitudinally oriented. At base of penis, just above opening of vagina, is a me- dially placed, longitudinally oriented ventral bulge, 1.5 mm long and 0.6 mm wide. Ejacu- latory pore seemingly flush with penial wall. Chalice walls firm, fixed in shape, forming ba- sally-opening, spherical hood or cup 1.0 mm in diameter, its opening ca. 0.4 mm wide. Pe- nial walls otherwise smooth (I interpret the folds, or bands, between the ventral and apico-lateral bulges as artifacts). Inflectarius (Hubrichtiuss) | downieanus (Bland, 1861). (Fig. 10a) Dissections: two from one population; specimens had been placed live into isopropanol, so were highly retracted. Length 3.8 mm, width 1.3 mm. Left lateral pi- laster sausage-shaped, 1.8 x 0.5 mm, lower end bluntly rounded, upper and tapering rap- idly into chalice wall. Left pilaster somewhat 86 EMBERTON LP? FIG. 8. Opened uneverted penial tubes. a. Fumonelix christyi GS-161 #6. À sperm mass adhered to the pilaster. (Also dissected #2, 5.) b. Fumonelix jonesiana, SC-155 #5 (also dissected #4, 9). c. Fumonelix wheatleyi, GS-6 #5 [also dissected #1, 3; GS-10 #2, 9, 12; GS-153 #4, 7, 10; SC-144 (subspecies clingmanica) #1, 2; SC-192 #3; SC-202 #1, 2; SC-212 #2. Populations differ in length, in size and position of bulges and in whether main pilaster is divided.] d. Fumonelix orestes, GS-86 #4 (also dissected #2). DB = dorsal bulge, LP = left pilaster, RP = right pilaster, VB = ventral bulge. MESODONTINI EVOLUTION 87 A eo ht 2 AE E [ESE FIG. 9. Opened uneverted penial tubes. a. Inflectarius inflectus, SC-130 #2 (also dissected #3, 8; GS-16 #21, 27, 29; GS-95 #1; SC-26 #A). b. Fumonelix wetherbyi, GS-115 #20 (also dissected #1, 2, 7). Chalice actually a high-standing, thin-walled cup, with its base surrounding opening of vas deferens; in illustrated specimen, chalice walls are buckled and folded inward. Main pilaster very high and thin, and rolled over to left (ca. 1/4 turn here) in all specimens examined. c. Inflectarius smithi, GS-101 #4 (also dissected #1, 2). RP = right pilaster, 3P = third pilaster. rhomboidal with thin, rounded edge, slightly pendant below, 1.1 x 0.4 mm; trace of a low, simple, ridge-like right lateral pilaster, 0.2—0.1 mm wide, runs rest of length of penis. Ejacu- latory pore flush with penial wall. Upper chal- ice wall thick (0.2 mm) and high-standing, widening rapidly right and left into continuity with lateral pilasters. Penial wall free of other structures. Inflectarius (Hubrichtius) kalmianus Hu- bricht, 1965. (Fig. 11a) Dissections: four from one population. Length 6.9 mm, width 0.9 mm, slightly expanding apically. Left lateral pilaster extending along apical half of length of penis, width 0.2 mm basally, gradually wid- ening to 0.3 apically before flaring slightly as left wall of chalice. Right lateral pilaster seem- ingly extending full length of penis, thin (less than 0.1 mm) and inconspicuous basally, gradually widening apically to about 0.3 mm before continuously grading into right chalice wall. Ejaculatory pore flush with penial wall. Chalice walls soft, seemingly erectile, high- standing, right higher than left. Penial walls free of other structures (the small ridge- and pustule-like patterns shown in Figure 11a are seemingly preservational artifacts). Inflectarius (Inflectarius) approximans (Clapp, 1905). (Fig. 12a) Dissections: two from two populations. Length 5.7 mm, width 2.0 mm. Right and left lateral pilasters both extending entire length of penis, left one mas- sively larger than right. Upper two-thirds of left pilaster cylindrical, 0.8 mm wide; lower third tapering to 0.2 mm. Left pilaster ca. 0.1 mm wide. Ejaculatory pore flush with penial wall. Chalice undifferentiated, seemingly a contin- uation of right lateral pilaster. Penial wall free of other structures. Inflectarius (Inflectarius) edentatus (Samp- son, 1889). (Fig. 12c) Dissections: three from one population. Length 10.0 mm, width 2.1 mm, expanding at tip. Left lateral pilaster short (1.7 mm), broad (0.6 mm), merging into penial wall basally and continuous terminally with chalice wall. Right lateral pilaster running full length of the penis, rounded, gradually widening from 0.2 mm basally to 0.8 termi- nally. Ejaculatory pore flush with penial wall. Chalice walls thick (0.9 to 1.1 mm), soft, and 88 EMBERTON pam - FIG. 10. Opened uneverted penial tubes. a. /nflectarius downieanus, Hubricht 30825 #B (also dissected #A; both highly contracted specimens). b. Patera pennsylvanica, GS-206 #1. folded, seemingly composed of erectile tis- sue. Penial walls free of other adornment (the small ridges and reticulations shown in Figure 12c presumably are preservational artifacts). Inflectarius (Inflectarius) ferrissi (Pilsbry, 1897). (Fig. 13c) Dissections: six from two populations (including topotypes). Length 36.4 mm, width 4.0 mm, narrowing apically and basally. Left lateral pilaster running ap- proximately four-fifths length of penis, 0.7 mm wide apically, tapering gradually to less than 0.1 mm wide. Right lateral pilaster running full length of penis, 1.2 mm wide, tapering api- cally to 0.4 mm before joining chalice. A third pilaster occurs on dorsal surface of penis be- tween lateral pilasters from base of penis to about two-thirds of length of penis apically, tapering in width from 1.7 mm basally to 0.4 mm apically. Chalice uniquely spoon-shaped with firm walls approximately 0.3 mm thick, its “handle” formed by right lateral pilaster, ejac- ulatory pore in its basin, and seemingly sep- arate from left lateral pilaster. Penial walls free of other structures (the nodular structures and the folds in the third pilaster shown in Figure 13c are seemingly artifacts of preser- vation). Inflectarius (Inflectarius) inflectus (Say, 1821). (Fig. 9a) Dissections: eight from four populations. Length 9.6 mm, width 1.4 mm. Left lateral pilaster solid and massive, running about one-third length of penis (2.5 mm), api- cally merging with chalice wall, uniformly broad (0.6 mm) for most of length until taper- MESODONTINI EVOLUTION 89 FIG. 11. Opened uneverted penial tubes. a. Inflectarius kalmianus, GS-116 #13 [also dissected #A; GS-188 (=GS116) #2, A, the latter with perfectly smooth walls]. b. Mesodon andrewsae, Roan Mountain, GS-11 #8 [also dissected #4 and examined #2 (fully everted—see Figure 2) and #3 (3/4 everted)]. c. Mesodon normalis, SC-158 (=SC145) #4 (also dissected SC-204 #2, 3; SC-154 #8, 10, 13; SC-184 #5, 6,10: SC-149 #3, 7, 11). This species highly variable in size and shape of chalice. 90 EMBERTON FIG. 12. Opened uneverted penial tubes. a. /nflectarius approximans, GS-57 #1 (also dissected Hubricht 23497 HA). b. Inflectarius magazinensis, GS-95 #5 (also dissected #6, 14). с. Inflectarius edentatus, GS-90 #2 (also dissected #7, 9). ing basally (to 0.3 mm). Right lateral pilaster running along about middle three-fifths of pe- nis, narrow (0.1 mm) at its midpoint, but wid- ening (0.3 mm) at two positions near its base and its apex where it merges with third pilas- ter. Third pilaster dorsal to, parallel with, equal in length to, and twice merging with right lateral pilaster; broad (0.4 mm). Ejacula- tory pore flush with penial wall. Chalice wall undifferentiated from left lateral pilaster, but on right side broader (0.4 mm) than right lat- eral pilaster to which it seemingly tapers, and apparently composed of soft, erectile tissue. Remaining penial walls seemingly free of other structures (the small dorsal ridges shown in Figure 9a seem to be preservational artifacts). Inflectarius (Inflectarius) magazinensis (Pilsbry & Ferriss, 1907). (Fig. 12b) Dissec- tions: three from one population (topotypes). Length 12.8 mm, width 1.7 mm, apically ex- panded. Left lateral pilaster extending along upper half of penis, broad (0.9 mm) and mas- sive throughout length, ending abruptly both basally and apically. Right lateral pilaster short (2.6 mm) and narrow (0.2 mm). Ejacu- latory pore flush with penial wall. Chalice walls low and solid, about 0.1 mm thick throughout, continuous with, but seemingly well differentiated from, right and left lateral pilasters. Penial walls free of other structures (the basal patterns and the thin apical exten- sion of the right lateral pilaster shown in Fig- ure 12b are presumably artifacts of preserva- tion). Inflectarius (Inflectarius) rugeli (Shuttle- worth, 1852). (Fig. 13b) Dissections: six from two populations. Length 14.0 mm, width 2.0 mm. Left lateral pilaster extending from 1.6 mm above base of penis to its juncture with left chalice wall, 0.2 to 0.3 mm wide, and branching basally to form pattern of nested, broadly U-shaped ridges. Right lateral pilaster extending to about mid-length of penis from its merger with right chalice wall, 0.3 mm wide. A third pilaster running mid-dorsally from about level of ejaculatory pore into first dorsal U-shaped ridge (total length 7.9 mm), parallel to and separate from lateral pilasters. Ejaculatory pore flush with wall of penis. Chalice walls high-standing, thin-edged, flared to left, composed of soft, presumably erectile tissue. Penial walls free of other struc- tures (other small sculpture shown in Figure 13b is presumably an artifact). Inflectarius (Inflectarius) smithi (Clapp, 1905). (Fig. 9c) Dissections: three from one population (topotypes). Length 20.0 mm, width 2.8 mm Left lateral pilaster extending along approximately upper two-fifths of length MESODONTINI EVOLUTION 91 | Зтт FIG. 13. Opened uneverted penial tubes. a. /nflectarius subpalliatus, GS-153 #2 (also dissected #1, 3). b. Inflectarius rugeli, SC-130 #2 (also dissected #3, 4; GS-3 #5, 10, A). c. Inflectarius ferrissi, GS-5 #3 (also dissected #1, 4; SC-216 #2, 3, 6). LP = left pilaster, RP = right pilaster, 3P = third pilaster. of penis, 0.3 mm wide throughout, seemingly ending abruptly basally. Right lateral pilaster twice as wide as the left (0.6 mm), uniform in width from its merger with right chalice wall to merger with basal bulge. Basal bulge mas- sive, firm, humped, an extension of right lat- eral pilaster, 4.4 mm long, 1.4 mm wide, 1.9 mm high at its midpoint, and 1.1 mm high at its ends, ending basally 3.6 mm above base of penis. Ejaculatory pore flush with penial wall. Chalice walls continuous with and slightly higher-standing and thinner than api- _ cal lateral pilasters. Penial walls otherwise featureless (the folds and tiny bumps shown in Figure 9c are preservational artifacts). Inflectarius (Inflectarius) subpalliatus (Pils- bry, 1893). (Fig. 13a) Dissections: three from One population. Length 16.1 mm, width 1.7 mm, expanding apically. Left lateral pilaster extending entire length of penis, widening ba- sally from 0.2 to 0.5 mm, with local thicken- ings, rapidly expanding near base of penis on dorsal side to width of 0.9 mm. Right lateral pilaster extending to mid-length of the penis, thin, tapering from 0.2 mm. A third pilaster, of about same size, length (6.5 mm) and vertical position as right lateral pilaster, runs along dorsal surface of penis between and parallel to lateral pilasters, lacking anastomoses with either. Ejaculatory pore flush with penial wall. Chalice walls quite high-standing, thin-edged, composed of flexible, presumably erectile tis- sue, flared to left. Penial walls free of other structures (the small basal ridges shown in Figure 13a are preservational artifacts). Inflectarius (Inflectarius) verus (Hubricht, 1954). (Fig. 14) Dissections: three from one population (topotypes). Length 6.0 mm, width 1.3 mm. Left lateral pilaster merging with left chalice wall so as to be indistinguishable from it, short, tapering basally to point less than one-third length of penis from its apex. Right lateral pilaster extending entire length of pe- nis, uniformly broad (0.5 mm) until tapering in its basal 0.7 mm, with several short (0.3 mm), oblique cord-like buttresses to dorsal penial wall. Ejaculatory pore flush with penis wall. Chalice walls more expanded than lateral pi- lasters, with which they are continuous, wide on right (0.7 mm), much narrower at apex and 92 FIG. 14. Opened uneverted penial tube of /nflectar- ius verus, GS-10 #8 (also dissected #3, 4). on left (0.3 mm). Small, longitudinally oblique, possibly expansive folds occur in chalice walls and lateral pilasters. Penial walls free of other structures. Mesodon (Akromesodon) altivagus (Pils- bry, 1900). (Fig. 15c) Dissections: three from two populations. Length 25.4 mm, width 3.9 mm. Left lateral pilaster seemingly extending full length of penis, but not clearly defined ow- ing to its branching into, or proximity to, sys- tem of dorsal ridges. Right lateral pilaster ex- tending entire length of penis, solid and cord- like, narrow apically and basally (0.3 mm), irregularly expanded medially (maximal width 1.8 mm), merging branch-like with at least two of dorsal ridges. Upper four-fifths of dorsal pe- nial wall wholly covered by fairly regular pat- tern of dorsal ridges. Dorsal ridges smooth, solid, cord-like, parallel, longitudinally ob- lique, meeting right lateral pilaster at angle of EMBERTON about 20 degrees, variously merging with or tapering alongside lateral pilasters, each varying in width, usually broadest medially (0.9 mm) and tapering laterally (to 0.3 mm), but with expanded ridge or section of ridge occurring both apically and basally (maximal widths 1.5 mm). Ejaculatory pore flush with penial wall. Chalice walls thin (0.3 mm), flex- ible, forming deep (2.1 mm) cup with thick- ened (0.6 mm), less flexible, conspicuously flared extension of left wall (length 2.4 mm). Basal penial wall with broad, loose fold asso- ciated with bases of the dorsal ridges, but oth- erwise free of additional structures. Mesodon (Akromesodon) andrewsae Bin- ney, 1879. (Figs. 2, 11b) Dissections: four from one population (two dissected, two al- ready everted). Length 16.1 mm, width 3.7 mm, expanded basally and apically. Left lat- eral pilaster thick (0.8 mm), firm, cord-like, ex- tending nearly entire length of penis, apically seemingly distinct from left chalice wall, and in basal fourth thickening and splitting into cluster of three or four basal bulges. Right lateral pilaster solid, cord-like, 0.5 mm wide and 9.8 mm long from point of merger with right chalice wall, branching in lower third to give rise to narrow (0.2 mm) dorsal ridge. Dor- sal surface of penis covered with cord-like ridges 0.2 to 0.4 mm in diameter; these run longitudinally, roughly parallel to lateral pilas- ters and to each other, but in their basal third variously divide or merge both with one an- other and with basal bulges. Basal bulges each about 3.4 mm long, 1.2 mm wide and 1.0 mm high; one of them is on ventral sur- face of penis, and differs from dorsal bulges by having apically-directed groove, or pocket. Ejaculatory pore flush with wall of penis. Chalice walls soft, thin-edged (0.2 mm), pre- sumably erectile, high-standing (ca. 2.4 mm), flared to left side. Penial wall otherwise free of structures (the transverse wrinkles on the ventral surface of the everted penis shown in Figure 2b are presumably preservational arti- facts). Mesodon (Akromesodon) normalis (Pilsbry, 1900). (Figs. 1, 11c) Dissections: 12 from five populations. Length 38.3 mm, width 4.4 mm. Left lateral pilaster extending full length of pe- nis, firm, cord-like, 1.5 mm wide from juncture with left chalice wall to about midpoint of pe- nis, below which it narrows as it gives rise to four or five baso-dorsal bulges. Right lateral pilaster short to absent, indistinguishable from right chalice wall. Lower half to third of dorsal penial wall covered with system of MESODONTINI EVOLUTION 93 | A AA A FIG. 15. Opened uneverted penial tubes. a. Mesodon sanus, GS-103 +2 (also dissected +1). b. Mesodon thyroidus GS-63 #69 [also dissected #1, 8, 14, 18; GS-202 #3; GS-78 (subspecies bucculentus) #6, 8, 12]. c. Mesodon altivagus, SC-144 #7 (also dissected #9, 10; GS-145 #4). basal bulges, ranging in shape from ridges (width 0.7 mm, length 3.6 to 9.9 mm) to lumps (diameter 2.6 mm, length 6.2 mm), that vari- ously merge with one another and with base of left lateral pilaster. Ejaculatory pore flush with wall of the penis. Chalice wall relatively thin (ca. 0.3 mm) and flexible, low on right, rapidly expanding to flare highly on left, form- ing asymmetric scoop that is diagnostic but variable in size and shape. Penial wall other- wise free of structures (the pattern of trans- | verse ridges on the upper dorsal surface and _ the small longitudinal ridges at the base _ shown in Figure 11c seem to be preserva- | tional artifacts). Mesodon (Aphalogona) elevatus (Say, 1821). (Figs. 3f-j, 4f-, 5d-f, 6d-f) Dissec- tions: eight from two populations. Length: for Tennessee population (n = 5), mean = 12.4 mm, standard deviation = 2.0 mm; for Indi- ana population (n = 3), mean = 17.1 mm, standard deviation = 4.0 mm. Length: Ten- nessee mean = 2.2 mm, standard deviation = 0.4 mm; Indiana mean = 2.4 mm, stan- dard deviation = 0.5 mm. Left lateral pilaster extending one-third to two-thirds length of pe- nis, cord-like, ca. 0.3 mm wide at juncture with left chalice wall, tapering to obscurity basally. Right lateral pilaster extending two-thirds to full length of penis, broad (0.5 to 1.0 mm), tapering basally, with substructure of parallel cords. Dorsal penial wall covered with system of about five to ten cord-like ridges about 0.1 mm wide, predominantly longtiduinally paral- lel, Occasionally merging with one another and with lateral pilasters, and sometimes co- alescing to form false dorsal pilaster (Fig. 4f). Ejaculatory pore flush with penial wall. Right and left chalice walls meeting apically in V- shaped notch, both usually with substructure of parallel cords; left wall 3.3 mm long, 0.6 mm wide, rounded (Fig. 4h) or slightly flap- like (Fig. 4i), basally tapering gradually (Figs. 4f, j) or rapidly (Fig. 6f); right chalice wall ei- 94 EMBERTON FIG. 16. Opened uneverted penial tubes. a. Mesodon mitchellianus, Hubricht 19406 # B (also dissected #A, C; GS-154 #1, 5, 6). b. Mesodon clausus, GS-116 #11 (also dissected #6, 17; GS-28 #A, B, C, E). c. Mesodon trossulus, GS-53 #3. ther identical to the left (Fig. 6d—f) or slightly to conspicuously broader (Fig. 4f-j). Penial walls otherwise free of structures. The abnor- mal appearance of the sculpture shown in Figure 6e is due to the fact that this penial tube was opened from the ventral rather than the standard dorsal side; apical knobs and bulges are artifacts of preservation and fold- ing, the dorsal ridges appear centrally, and the small apical ridges are at the basin of the chalice. The deficiency of sculptural features shown in Figure 6d and, to a lesser extent, in Figure 6f could be due to local variation or could be an artifact due to tight contraction of the penis, as indicated by its thick walls. Mesodon (Aphalogona) mitchellianus (Lea, 1838). (Fig. 16a) Dissections: five from two populations. Length 9.0 mm, width 1.7 mm. Right and left lateral pilasters both extending full length of penis and merging ventrally at its base; left pilaster beginning as thin (0.1 mm) arc from lower left side of chalice, continuing as longitudinal cord 0.2 mm wide; right pilas- ter thick (4.0 mm in diameter) and cord-like in upper three-fifths, angling ventrally and taper- ing rapidly (to ca. 0.1 mm diameter) in lower two-fifths before joining left lateral pilaster. Dorsal penial wall covered with field of cord- like ridges less than 0.1 mm in diameter, ranging in orientation from longitudinal to a 30-degree slant to left, and variously anasto- mosing among themselves and with upper left lateral pilaster. Ejaculatory pore opening on summit of fleshy, solid, cylindrical pedestal ca. 0.3 mm high, the sides of which are con- tinuous with chalice. Chalice completely en- closed, deep (ca. 0.9 mm) cylinder with thin, continuous wall (ca. 0.1 mm), right edge of which is scalloped, and basal-most point of which is continuous with right lateral pilaster. Other structures are not evident, although the baso-ventral longitudinal ridges shown in Fig- ure 16a could be real instead of artificial. Mesodon (Aphalogona) zaletus (“Say” Bin- ney, 1837). (Figs. За-е, 4a—e, 5a-c, 6a—, 17) Dissections: ten from four populations. Length: for Tennessee population (n = 5), mean = 14.3 mm, standard deviation = 1.5 mm; for Indiana population (n = 3), mean = 16.3 mm, standard deviation = 1.7 mm. Width: Tennessee mean = 1.9 mm, standard deviation = 0.3 mm; Indiana mean = 2.5 mm, standard deviation = 0.4 mm. Left lat- eral pilaster extending full length of penis, firm, cord-like, 0.3 to 0.6 mm wide, continuous with left chalice wall. Right lateral pilaster ex- tending full length of penis, firm, cordlike, wid- est (ca. 0.6 mm) just below indistinct junction with right chalice wall, tapering basally to about 0.3 mm wide. Dorsal penial wall bear- MESODONTINI EVOLUTION 95 5mm FIG. 17. Mesodon zaletus, with fully everted penis, from SC-71. a. Ventral view of whole animal, with body twisted such that head is in dorsal view but apex of penis is in ventral view. b. Ventral view of penis, uncoiled and with chalice pinned open. c. Dorsal view of same. ing approximately four cord-like ridges, ca. 0.1 mm wide, approximately parallel, and ori- ented either longitudinally or at a slight angle so as to merge with either or both of lateral pilasters. Ejaculatory pore flush with penial wall. Right and left chalice walls distinct, meeting apically in V-shaped notch; left wall thick (ca. 0.4 mm), firm, 2.7 to 4.2 mm long, at midpoint drawn into conspicuous flap, 1.2 to 1.5 mm high; right wall thicker (ca. 0.7 mm), massive, usually bulging in central region or just below. Penial wall otherwise free of sculp- ture. The dissection shown in Figure 4c seems unusual because the penial tube was opened from the ventral side instead of the standard dorsal side, hence the dorsal ridges are in the center, and the sides are reversed. Mesodon (Appalachina) chilhoweensis (Lewis, 1870). (Fig. 18b) Dissections: two from two populations. Length extreme: 203.7 mm, of which apical, basal, and middle re- gions totalling 38.7 mm are shown in Figure 18b. Width 3.3 mm. Left lateral pilaster stretching entire length of penis, 0.7 mm wide in apical and basal regions, 0.4 mm wide in mid-regions, rounded, cord-like. Right lateral pilaster extremely short, perhaps absent, in- distinguishable from right chalice wall. Ejacu- latory pore flush with penial wall. Chalice walls rather thin (0.3 mm) and firm, higher- standing on left (2.0 mm) than on right (at most 1.0 mm), fairly even in height on each side. Penial walls seemingly free of other structures (the apical folds, basal nodules and the transverse ridges throughout all seem to be artifacts of preservation). Mesodon (Appalachina) sayanus (Pilsbry, in Pilsbry & Ferriss, 1906). (Fig. 18a) Dissec- tions: three from one population. Length 20.2 mm, width 2.0 mm. Left lateral pilaster short (ca. 2.0 mm) and narrow (ca. 0.2 mm), taper- ing rapidly from left chalice wall, with which it is continuous. Right lateral pilaster extending perhaps two-thirds of length of penis, 0.2 to 0.3 mm wide, becoming inconspicuous ba- sally. Ejaculatory pore flush with wall of the penis. Chalice wall evenly high on left (ca. 0.7 mm), apically diminishing to become low on EMBERTON 96 Tr Ys mm D , GS-130 #6: penial wall actually much FIG. 18. Opened uneverted penial tubes. a. Mesodon sayanus , Hubricht 30943 #A (also thinner than shown here (also dissected #1, 4). b. Mesodon chilhoweensis dissected SC-263 #A: subadult with greatly flared chalice wall). MESODONTINI EVOLUTION 9% FIG. 19. Opened uneverted penial tubes. a. Patera sargentiana, GS-101 #1 (also dissected #9, 12). b. Patera clarki, GS-2 #5 (also dissected #6, 7; and other populations). c. Patera appressa, GS-104 #1 (also dissected #5, 7; GS-141 #2, 4, 6, 8). There is much variation in size of inflated chalice wall of P. appressa; illustrated specimen was opened from ventral side, such that opening of vas deferens appears to lie outside chalice. right (ca. 0.3 mm), thence grading inconspic- uously into right lateral pilaster. Penial wall otherwise free of structures (the transverse folds shown in Figure 18a are artifacts of preservation). Mesodon (Mesodon) clausus (Say, 1821). (Fig. 16b) Dissections: seven from two popu- lations. Length 9.8 mm, width 1.6 mm, curving strongly to left in upper third. Left lateral pi- laster extending full length of penis, thick (0.4 mm), firm, cord-like. Right lateral pilaster ex- tending entire length of penis, thick (0.3 to 0.4 mm), firm, and cord-like, except apically, where it is low and indistinct. Dorsal surface of penis covered with longitudinally parallel ridges, cord-like, ca. 0.2 mm wide basally, apically dividing to become narrower and less distinct, lateral-most ridges merging with api- cal right lateral pilaster or anastomosing with mid-region of left lateral pilaster. Ejaculatory pore seemingly flush with wall of penis. Right chalice wall 0.2 mm thick at edge, firm, flaring outward to form ear-like flap; right chalice wall thicker (0.3 mm), firm, cord-like basally, flar- ing outward apically to meet left chalice wall in apical point; chalice thus resembling pointed left ear, 1.7 mm long and 1.2 mm wide, pinna of which is rolled inward in uneverted, inactive state. Penial wall otherwise free of structures. Mesodon (Mesodon) sanus (Clench & Ar- cher, 1933). (Fig. 15a) Dissections: two from one population. Length 11.6 mm, width 2.2 mm. Left lateral pilaster seemingly extending full length of penis, firm, cord-like, 0.2 to 0.5 mm wide, seemingly merging baso-ventrally with right lateral pilaster. Right lateral pilaster extending whole length of penis, firm, cord- like, 0.2 mm wide. Entire dorsal surface of penis covered with field of closely-adjacent, longitudinally-arrayed, roughly parallel cord- like ridges, about 12 of them apically, averag- ing about 0.1 mm wide, but only about five of them basally and averaging ca. 0.2 mm wide by fusion with one another and with both lat- eral pilasters, with which they are intimately connected. Ejaculatory pore flush with wall of penis. Chalice a rounded ear-like flap, flared to left, about 0.1 mm thick at edge, and stand- ing a maximum of about 1.1 mm high, rolled over toward right side in the uneverted penis. Penial wall without other sculptural features. Mesodon (Mesodon) thyroidus (Say, 1817). (Fig. 15b) Dissections: nine from three popu- lations. Length 12.5 mm, width 2.5 mm, grad- ually expanding apically, then constricting nar- rowly just below chalice to form shoulder, or shelf, before expanding again slightly to apex; penis thus of a three-dimensional shape diffi- 98 EMBERTON FIG. 20. Penial tubes of Patera laevior. a. H-22 #3: fully everted penis, and remaining internal reproductive system (compare with Figure 10). b. H-22 #A: opened uneverted penial tube [also dissected #1 (pilaster much more inflated and almost forming a chalice); GS-125 #1, 4, 5; and examined three everted penes from SC-217]. Cult to pin out for clear viewing in two dimen- sions, hence the folds and distortions visible in Figure 15b. Left lateral pilaster extending full length of penis, firm, cord-like, ca. 0.5 mm wide throughout. Right lateral pilaster seem- ingly extending full length of penis, cord-like, apically separated from right chalice wall, ex- panding to width of approximately 0.8 mm in upper half of penis, then narrowing to about 0.4 mm in basal half, variable in width throughout. Dorsal wall of penis covered with field of somewhat parallel, somewhat longitu- dinally arranged cord-like ridges, ca. 0.3 mm wide, variously merging with one another and with both lateral pilasters (in Figure 15b sev- eral are merged to form the analog of a third pilaster beside the left lateral pilaster). Ejacu- latory pore flush with wall of penis. Chalice in shape of left ear, rolled over to right in unev- erted penis, 1.3 mm high at point. Penial wall seemingly free of other structures. Mesodon (Mesodon) trossulus Hubricht, 1966. (Fig. 16c) Dissections: one from one population (topotype). Length 9.8 mm, width 1.4 mm, apex bent to left. Left lateral pilaster extending entire length of penis, its apex not continuous with left chalice wall, but to left of chalice, firm, cord-like, 0.3 mm wide apically, gradually tapering to ca. 0.1 mm basally. Right lateral pilaster firm, cord-like, 0.3 mm wide apically, tapering to 0.2 mm basally. Dorsal surface of penis covered with about 12 approximately parallel, thin (ca. 0.05 mm), cord-like ridges, variously branching and merging, most of them branching off at angle of about 15 degrees, step-like, from right lat- eral pilaster. Ejaculatory pore flush with wall of penis. Chalice shaped like left ear with api- cally pointed pinna, rolled to right in uneverted penis; unrolled length 2.0 mm, width 1.4 mm, edge thickness ca. 0.2 mm. Penial wall oth- erwise seemingly free of other structures (the wavering ventral gutter shown in Figure 16c is presumably an artifact). Patera (Patera) appressa (Say, 1821). (Fig. 8c; opened ventrally, dorsal ridges central, lateral pilasters reversed as labeled) Dissec- tions: seven from two populations. Length 7.2 MESODONTINI EVOLUTION 99 FIG. 21. Opened uneverted penial tubes. a. Patera panselena, GS-142 #9 (also dissected #2, 6). b. Patera perigrapta, GS-98 #8: whole length = 64 mm [also dissected #20, A, and examined #7, 9, and 18, each with partly everted penis (penis length varies greatly within this population); GS-3 #12; GS-57 #A; GS-90 #1; and examined one speci- men with everted penis from each of the following populations: GS-170; SC-61; SC-65; SC-66; SC- 67; SC-97]. mm, width 3.2 mm, apically expanded. Left lateral pilaster running entire length of penis, thin, cord-like, 0.2 mm wide at junction with left chalice wall, gradually tapering to minimal width of 0.1 mm at mid-penis. Right lateral pilaster running full length of penis, cord-like, uniformly wide in apical two-thirds (0.3 mm), whence tapering abruptly to 0.1 mm. Dorsal penial wall evenly covered with thin, cord-like, nearly parallel ridges, approximately alike in width, apically twice as wide (0.1 mm) as ba- sally, variously branching and anastomosing with one another and with the lateral pilasters (Figure 19c shows a very close anastomosis between one of these ridges and the right lat- eral pilaster). Ejaculatory pore (shown above chalice in Figure 19c) flush with penial wall. Chalice (inverted in Figure 19c) with rather thin (0.3 mm), flexible walls, forming seem- ingly symmetrical, broad (3.0 mm) hood with estimated central depth of 0.7 mm. Penial walls otherwise free of scultpure. Patera (Patera) clarki (Lea, 1858). (Fig. 19b) Dissections: three from one population. Length 9.1 mm, width 3.0 mm. Left lateral pi- laster seemingly running entire length of pe- nis, solid, thick, uniform in width (0.6 mm), abruptly but slightly tapering apically so as to become discrete from otherwise similar left wall of chalice. Right lateral pilaster running entire length of penis, solid, fairly uniform in width (0.6 mm), flattened in upper two-thirds so as to grade into penial wall, especially on dorsal side. Ejaculatory pore flush with penial wall. Chalice a symmetrical cowl, 2.7 mm long, 1.9 mm wide, distinct from lateral pilas- ter, its walls uniformly thick (0.3 mm). Penial walls free of sculpture. Patera (Patera) laevior (Pilsbry, 1940). (Fig. 20) Dissections: nine from three popula- tions (one population with penes already everted). Length 12.6 mm, width 2.0 mm. Left lateral pilaster extending entire length of pe- nis, wavy or bulbous in outline, apically grad- ing into left chalice wall, 0.8 mm wide in upper third, tapering gradually and irregularly to 0.3 mm in lower third. Right lateral pilaster absent or imperceptible. Ejaculatory pore flush with penial wall. Chalice inconspicuous, its left wall (0.5 mm wide) continuous with left lateral pi- laster, and apically tapering rapidly to mere low arching ridge. Penial walls free of sculp- ture (the patterns of transverse folds and wavy, elongated pits shown in Figure 20b are interpreted as preservational artifacts). Patera (Patera) panselena (Hubricht, 1976). (Fig. 21a) Dissections: three from one population. Length 15.1 mm, width 1.6 mm. Left and right lateral pilasters running full length of penis, equally narrow (0.2 mm) al- most throughout, grading apically into chalice walls. Ejaculatory pore flush with penial wall. Chalice a shallow, thin-walled hood (0.8 mm deep, wall 0.3 mm thick), slightly asymmetric, extending 5.0 mm on right and 3.5 mm on left, tapering smoothly into lateral pilasters. Penial walls free of sculpture (all the diagonal folds shown in Figure 21a presumably are preser- vational artifacts). Patera (Patera) perigrapta (Pilsbry, 1894b). (Fig. 21b) Dissections: 19 from ten popula- 100 EMBERTON FIG. 22. Opened uneverted penial tubes. a. Patera binneyana “short”, GS-95 #2 (also dissected Hubricht 31615 #A, B; Hubricht 33898 #A). b. Patera leatherwoodi, GS-67 #1 (also dissected GS-68 #1). c. Patera roemeri, GS-63 #21 (also dissected #4, 6, 7, 15). tions (including 13 everted penes from seven populations). Length 65.4 mm (extremely variable in other examined specimens), width 1.8 mm. Left lateral pilaster extending entire length of penis, cord-like, fairly uniform in width (0.5 mm), gradually tapering basally (to 0.2 mm wide), apically grading into left wall of chalice. Right lateral pilaster extending along upper half of penis, where it is similar in width and appearance to left lateral pilaster. Ejacu- latory pore flush with penial wall. Chalice a symmetric hood, 1.2 mm deep at apex, about 3.1 mm long, and with walls 0.2 mm thick that become lower basally to grade into lateral pi- lasters. Penial walls free of sculpture (the transverse grooves and diagonal folds shown in Figure 21b seem to be artifacts of preser- vation). Patera (Patera) sargentiana (Johnson & Pilsbry, 1892). (Fig. 19a) Dissections: three from one population. Length 7.7 mm, width 2.9 mm. Left lateral pilaster seemingly run- ning entire length of penis, solid, thick, broad (0.7 mm), seemingly discontinuous but uni- form in width in lower half. Right lateral pilas- ter short (ca. 1.4 mm), flat, tapering rapidly from right chalice wall to merge with penial wall. Ejaculatory pore flush with penial wall. Chalice a thick-walled, symmetric hood (2.6 mm long, 3.1 mm wide, about 1.2 mm deep, walls 0.6 mm thick), its walls grading into lat- eral pilasters. Penial walls free of sculpture (Figure 19a shows both large- and small- scale structural artifacts due to severe con- traction and folding of the specimen). Patera (Ragsdaleorbis) pennsylvanica (Green, 1827). (Fig. 10b) Dissections: two from two populations. Length 13.8 mm, width 2.2 mm. Left lateral pilaster inconspicuous, extending about one-third of length of penis, 0.3 mm wide, not highly elevated above wall of penis. Right lateral pilaster extending to about mid-length of penis, like a thick (ca. 0.5 mm), high-standing (ca. 1.0 mm) ridge imperceptibly continuous with right chalice wall and basally tapering in height and, to lesser extent, in width. Ejaculatory pore on summit of some- what barrel-shaped apical plug, or pedestal, 1.7 mm long, 1.7 mm wide at center. Chalice wall uniformly high (ca. 1.0 mm), about 0.2 mm MESODONTINI EVOLUTION 101 10 mm | ER, Q / 7 FIG. 23. Opened partly everted penial tube and remaining internal reproductive system of Patera binneyana “long”, FMNH 176008 #C (also dissected +A, В; FMNH 176195 #A). Compare wtih Figure 12a. thick at edge, well differentiated from left lat- eral pilaster to form flap on left, but undiffer- entiated from right lateral pilaster. Penial wall seemingly otherwise free of structures (the massive system of angular ridges shown in Figure 10b seems to consist of transient folds best considered preservational artifacts). Patera (Vesperpatera) binneyana (Pilsbry, 1899). Two forms were found (see Appendix 1) and are described separately here. P. binneyana “short.” (Fig. 22a) Dissec- tions: eight from eight populations. Length 30.6 mm, width 4.2 mm. Left lateral pilaster short, extending along only upper fifth to fourth of length of penis, 0.6 mm wide, cord- like at junction with left wall of chalice, flatten- ing basally to merge gradually with penial wall. Right lateral pilaster extending entire length of penis, solid, lumpy, varying in width from 0.6 to 1.4 mm, wide at merger with right chalice wall. Ejaculatory pore flush with penial wall. Chalice an asymmetric hood (left side 4.3 mm long, right side 6.5 mm long), about 4.0 mm wide, 1.2 mm deep, its walls contin- uous with lateral pilasters and about 0.5 mm thick. Penial walls free of sculpture (patterns of folds shown in Figure 22a are preserva- tional artifacts). 102 EMBERTON FIG. 24. Partly everted penial tube of Patera binneyana “long”, FMNH 176008 #C. a. Actual and extrap- olated everted penis (see Figure 10). b. Tip of actually everted penis showing lack of chalice or functional structure. P. binneyana “long.” (Figs. 23, 24) Dissec- tions: four from two populations. Length 91.0 mm, width 2.7 mm. Left and right lateral pi- lasters seemingly extending along entire length of penis, left one very thin (0.1 mm) and cord-like, increasing in height (but not width) apically to juncture with left chalice wall. Right lateral pilaster thick and sausage- like throughout, 0.9 mm wide at mid-penis, 1.2 mm wide apically before constricting slightly at junction with right chalice wall. Ejac- ulatory pore flush with penial wall. Chalice a simple hood, its walls about 1.6 mm high and 0.5 mm thick, asymmetric, right side 7.1 mm long and left side 4.9 mm long. Penial walls free of sculpture (apical pits and mid-penial grooves shown in Figure 23 are considered preservational artifacts). Patera (Vesperpatera) clenchi (Rehder, 1932). (Fig. 25a) Dissections: one from one population. Length 8.2 mm, width 1.1 mm. Left lateral pilaster obsolete, consisting of thickened region, 0.3 mm x 0.8 mm, at base of left chalice wall. Right lateral pilaster ex- tending entire length of penis, solid, cord-like, 0.4 mm wide for most of length, tapering to 0.3 mm in upper fifth. Ejaculatory pore flush with penial wall. Chalice a simple hood, 1.1 mm wide, 1.1 mm long, its wall about 0.2 mm thick, right wall higher (0.3 mm) than left wall (0.2 mm). Penial walls free of sculpture (diag- onal folds shown in Figure 25a seem to be due to contraction during preservation). Patera (Vesperpatera) indianorum (Pilsbry, 1899). (Fig. 25b) Dissections: two from one population. Length 28.1 mm, width 2.8 mm. Left lateral pilaster seemingly extending full length of penis, flat, 0.9 mm wide, in some places difficult to distinguish from wall of pe- nis, seemingly lessening considerably in width before joining left chalice wall. Right lat- eral pilaster extending along full length of pe- nis, solid, cord-like, variable in width (0.4 to 1.3 mm), wide at junction with right chalice wall. Ejaculatory pore flush with penial wall. Chalice a simple hood, 4.9 mm wide, 4.9 mm long, its wall about 0.6 mm thick, right wall higher (1.5 mm) than left wall (1.1 mm). Pe- nial walls free of sculpture (transverse folds shown in the lower half of Figure 25b are ar- tifacts). Patera (Vesperpatera) kiowaensis (Simp- son, 1888). (Fig. 25c) Dissections: two from one population. Length 15.9 mm, width 1.4 MESODONTINI EVOLUTION 103 FIG. 25. Opened uneverted penial tubes. a. Patera clenchi, Hubricht 25210 #A. b. Patera indianorum, GS-87 #1 (also dissected #5). c. Patera kiowaensis, GS-84 #12 (also dissected #18). mm. Left lateral pilaster seemingly absent, or at most a thickened streak at base of left chalice wall. Right lateral pilaster running full length of penis, solid, cord-like, variable in width (0.4 to 0.5 mm) for most of length, tapering basally to merge with wall of penis, apically grading imperceptibly into right chal- ice wall. Ejaculatory pore flush with penial wall. Chalice a simple hood, 1.4 mm wide, 2.0 mm long, wall about 0.3 mm thick, right wall higher (0.5 mm) than left wall (0.3 mm). Pe- nial walls free of sculpture (apical longitudinal ridges and transverse basal folds shown in Figure 25c seem to be artifacts of preserva- tion). Patera (Vesperpatera) leatherwoodi (Pratt, 1971). (Fig. 22b) Dissections: two from two populations. Length 9.1 mm, width 1.4 mm, expanding apically to about 3.0 mm. Left lat- eral pilaster stretching about three-fourths of length of penis, very thin (0.1 mm) and incon- spicuous. Right lateral pilaster reaching full length of penis, solid, cord-like, 0.3 mm wide at juncture with right chalice wall, gradually tapering basally to 1.0 mm wide. Ejaculatory pore flush with penial wall. Chalice like a spat- ula, right wall flared high (1.4 mm maximum), long (3.9 mm) and rounded; left wall lower (0.7 mm) and seemingly appressed to penial wall. Penial walls free of sculpture (the ob- lique folds and the large, short, База! thicken- ing shown in Figure 22b are interpreted as preservational artifacts). Patera (Vesperpatera) roemeri (Pfeiffer, 1848). (Fig. 22c) Dissections: five from one population. Length 16.6 mm, width 1.5 mm. Left lateral pilaster seemingly absent. Right lateral pilaster extending three-fourths of length of penis, solid, cord-like, and 0.4 mm wide below junction with right chalice wall, ta- pering and flattening in lower half to merge with penial wall. Ejaculatory pore flush with penial wall. Chalice like a spatula, right wall flared high (about 1.4 mm maximum), long (3.8 mm) and rounded; left wall vestigial or absent. Penial walls free of sculpture (left- hand, transverse grooves shown in Figure 22c are artifacts of preservation). 104 EMBERTON FIG. 26. Suggested character-state transformations in Mesodontini penial morphology: Character 1, lateral pilasters. Suggested Character-State Transformations. The total variation in penial morphology was classified into five characters comprising 37 character states. These are arranged into their suggested phylogenies in Figures 26— 28, in which the suggested character-state transformations are numbered 1-34. The lateral pilasters (Character 1) vary greatly in the Mesodontini. Twelve states were detected, none of which seemed to be convergent. Their suggested phylogeny (Fig. 26) contains Transformations 1-11. The shape of the chalice (Character 2) is the most variable feature of the mesodontin genitalia. Fourteen character states were de- tected, for which 14 transformations (Trans- formations 12-25) are suggested (Fig. 27). According to this hypothesis, there are three convergences: thick-walled, A-shaped chal- ices (Transformations 13 and 23); deep, cy- lindrical chalices (Transformations 16, 17-20, 25); and thin, high-standing chalice walls (Transformations 14, 22). Character 3, baso-ventral structures, com- prises three states, connected by suggested Transformations 26 and 27 (Fig. 28). Dorsal structures (Character 4) yielded six character states, none of them detectably convergent (Fig. 28) and connected by six suggested transformations (Transformations 28-33). Other dorsal features, which are not included under this particular character, but which could have been, and which probably serve the same function, are the third pilaster (Fig. 13; Transformations 3, 4) and the basal bulges branching from the left pilaster (Fig. 11b, c; Transformation 11). Other dorsal sculptural features that appear in the illus- trated dissections but were interpreted as preservational artifacts include the oblique folds in Patera panselena (Fig. 21a), P. peri- grapta (Fig. 21b), P. sargentiana (Fig. 19a), P. leatherwoodi (Fig. 22b), P. roemeri (Fig. 22c), P. laevior (Fig. 20b), P. clenchi (Fig. 25a), P. indianorum (Fig. 25b), P. kiowaensis (Fig. 25c), Mesodon sayanus (Fig. 18a), M. chil- howeensis (Fig. 18b), Inflectarius edentatus (Fig. 12c), I. inflectus (Fig. Эа), I. smithi (Fig. 9c), I. ferrissi (Fig. 13c), |. downieanus (Fig. 10a), Patera pennsylvanica (Fig. 10b) and P. kalmianus (Fig. 11a); the irregular-network pattern in Mesodon zaletus (Fig. 4f), Patera binneyana (Fig. 23), P. laevior (Fig. 20b), In- flectarius edentatus (Fig. 12c), Fumonelix jo- nesiana (Fig. 8b) and Inflectarius kalmianus (Fig. 11a); the beaded and/or cuneiform sculpture next to a pilaster in Mesodon chil- howeensis (Fig. 18b) and Inflectarius ferrissi (Fig. 13c); and the transverse waves in Mes- odon normalis (Fig. 11c). The fifth and final character, peripheral structures, has only two states, connected by Transformation 34 (Fig. 28). In presenting each of the 34 suggested character-state transformations below, the same format has been used throughout: (1) MESODONTINI EVOLUTION 105 FIG. 27. Suggested character-state transformations in Mesodontini penial morphology: Character 2, chalice. identification number as used in Figures 26— 28; (2) number(s) of transformation(s) sug- gested to have preceded it in evolution; (3) suggested plesiomorphic state; (4) outgroup taxa having the suggested plesiomorphic state; (5) suggested apomorphic state; (6) taxa whose ancestor(s) are suggested to have had the apomorphic state, although these taxa lack the state now; (7) taxa that now have the suggested apomorphic state; and (8) discussion of the suggested transfor- mation, including any further explanation, and the reasoning behind its suggestion. In defin- ing the transformations, the terms distal and apical are used interchangeably, as are prox- imal and basal. Transformation 1. Preceding transforma- tions: none. Plesiomorphic state: left pilaster distally higher than broad. Present in (outgroups): Polygyra, Praticolella, some Stenotrema, all Mesodontini except /. inflectus, |. approxi- mans, |. magazinensis and I. downieanus. Apomorphic state: left pilaster distally twice as broad as high, thick and fleshy. Formerly present in: I. inflectus (Fig. Эа), I. approxi- mans (Fig. 12a), I. magazinensis (Fig. 12b), I. downieanus (Fig. 10a). Now present in: I. in- flectus (Fig. 9a). Discussion. The homology of the broad, thick left pilaster of these taxa is uncertain. The most problematic of these is that of I. downieanus, in which contractile distortion due to immersion of the specimen live in iso- propynol renders interpretation difficult. The left pilaster of I. inflectus is attenuated, ex- tending only half the length of the penis, whereas the left pilasters of /. approximans and /. magazinensis (and possibly /. downiea- nus) are full-length. These differences in de- gree of pilastral attenuation were not used for phylogenetic analysis because of possible in- traspecific variation (compare Figs. 4h, 4i). Transformation 2. Preceding transforma- tions: 1. Plesiomorphic state: left pilaster distally twice as broad as high, thick, fleshy; right pi- laster pronounced. Present in (outgroup): /. inflectus (Fig. 19a). Apomorphic state: left pilaster extremely thick and rounded, right pilaster obsolete to absent. Formerly and now present in: /. ap- proximans (Fig. 12a), /. magazinensis (Fig. 125), |. downieanus (Fig. 10a). Discussion. As mentioned for Transforma- tion 1, interpretation of /. downieanus is diffi- cult and its homology for this character state is highly problematic, especially because its 106 EMBERTON | А a 30 34 FIG. 28. Suggested character-state transformations in Mesodontini penial morphology: Characters 3, 4 and 5, baso-ventral structures, dorsal structures and peripheral structures. right pilaster, although attenuated, is much more fully developed than in /. approximans and /. magazinensis. The latter two species, on the other hand, seem very similar and are much more likely to be homologous. Transformation 3. Preceding transforma- tions: none. Plesiomorphic state: two lateral pilasters only. Present in (outgroups): Polygyra, Prati- colella, most Stenotrema, all Mesodontini ex- cept I. inflectus, |. rugeli, I. subpalliatus and I. ferrissi. Apomorphic state: third pilaster present and partly attached to right pilaster. Formerly present in: /. subpalliatus (Fig. 13a), I. ferrissi (Fig. 13c), I. inflectus (Fig. 19a), I. rugeli (Fig. 13b). Now present in I. inflectus (Fig. 9a). Discussion. The longitudinally divided right pilaster of I. inflectus is unique. It is thought to represent an intermediate stage toward a complete longtiduinal division to produce a third pilaster (Transformation 4). Transformation 4. Preceding transforma- tions: 3. Plesiomorphic state: third pilaster partly at- tached to right pilaster. Present in (outgroup): |. inflectus (Fig. Эа). Apomorphic state: third pilaster separate from and parallel to right pilaster. Formerly and now present in: /. subpalliatus (Fig. 13a), |. ferrissi (Fig. 13c), I. rugeli (Fig. 13b). Discussion. The structure here called a third pilaster is approximately equal in size, length and general appearance to the right and left lateral pilasters and is entirely parallel to them. It therefore seems not to be homol- ogous with other dorsal structures such as long bulges (e.g. Fig. 8a), oblique branches from the pilasters (e.g. Fig. 7), and smaller and/or oblique basal ridges (e.g. Figs. 11b, 15c, 17b). Because it is near the right pilaster, the third pilaster is thought to be the result of longitudinal separation from the right pilaster, homologous to the condition in /. inflectus. Transformation 5. Preceding transforma- tions: none. Plesiomorphic state: right pilaster basally unmodified. Present in (outgroups): Polygyra, Praticolella, Stenotrema, all Mesodontini ex- cept /. smithi. : Apomorphic state: right pilaster with basal MESODONTINI EVOLUTION 107 swelling. Formerly and now present in: /. smithi (Fig. 9c). Discussion. This character state seems not to be homologous with either basal bulges that are isolated from the right pilaster (Fig. 8), or basal enlargement due to merger of the right and left pilasters (Fig. 7). Transformation 6. Preceding transforma- tions: none. Plesiomorphic state: neither lateral pilaster higher than four times its width. Present in (outgroups): Polygyra, Praticolella, Stenotre- ma, all Mesodontini except P. pennsylvanica. Apomorphic state: both right and left pilas- ters about five times higher than wide, the right half-length and the left a short distal flap. Formerly and now present in: P. pennsylvan- ica (Fig. 10b). Discussion. This unique character state is most similar to that of F. wetherbyi (Transfor- mation 7), from which it differs in the equally high-standing pilasters and the flap-like end of the left pilaster. Transformation 7. Preceding transforma- tions: none. Plesiomorphic state: Lateral pilasters no higher than five times their respective widths. Present in (outgroups): Polygyra, Praticolella, Stenotrema, all Mesodontini except F. weth- erbyi. Apomorphic state: right pilaster extremely high and thin (height more than six times width), full-length, and rolled over; left pilaster about three times higher than wide, and half- length. Formerly and now present in: F. weth- erbyi (Fig. 9b). Discussion. The unique right pilaster of F. wetherbyi is easily mistaken for a thick, broad pilaster whenever it is rolled over (Fig. 9b). Only when it is unrolled is its extreme height and thinness evident (see cross-section, Fig. 26). Its appearance in the everted penis is unknown, and its functional significance is un- clear. Transformation 8. Preceding transforma- tions: none. Plesiomorphic state: at least one lateral pi- laster pronounced and conspicuous. Present in (outgroups): Polygyra, Praticolella, Sten- otrema, all Mesodontini except F. christyi, F. wheatleyi and F. jonesiana. Apomorphic state: both lateral pilasters greatly reduced to absent. Formerly and now present in: F. christyi (Fig. 8a), F. wheatleyi (Fig. 8c), F. jonesiana (Fig. 8b). Discussion. This character state seems to be homologous in the three taxa, despite some differences in detail (Fig. 8). It is most clearly visible in F. christyi, in which the right and left pilasters are reduced to small but un- mistakable traces, with the distal left pilaster the strongest remnant. In F. wheatleyi, this character state is complicated by the pres- ence of seemingly innomologous distal and basal bulges, but structures tentatively inter- preted as remnant lateral pilasters occur as low ridges (Fig. 8c). In F. jonesiana, this char- acter state is complicated by the presence of a long basal bulge, which might be a hyper- trophied left pilaster, but which is interpreted here as a ventral basal bulge (Transformation 11), with the left pilaster fading proximally from a distally more pronounced region; the right pilaster appears fairly clearly as a slight trace (Fig. 8b). Transformation 9. Preceding transforma- tions: none. Plesiomorphic state: lateral pilasters broadly separated basally. Present in (out- groups): Polygyra, Praticolella, some Sten- otrema, all Mesodontini except F. orestes and F. archeri. Apomorphic state: lateral pilasters meeting at their basal termini. Formerly present in: F. orestes (Fig. 8d), F. archeri (Fig. 7). Now present in: F. orestes (Fig. 8d). Discussion. The right pilaster of F. orestes seems to be interrupted distally, whereas that of F. archeri is not; their basally-joining lateral pilasters might be independently derived, rather than homologous as here hypothe- sized. Transformation 10. Preceding transforma- tion: 9. Plesiomorphic state: lateral pilasters meet- ing at their basal ends. Present in (outgroup): F. orestes (Fig. 8d). Apomorphic state: lateral pilasters basally joined for about one-fourth their total lengths. Formerly and now present in: F. archeri (Fig. 7). Discussion. See discussion under Trans- formation 9. Transformation 11. Preceding transforma- tions: none. Plesiomorphic state: left pilaster basally a single ridge; right pilaster short to long. 108 EMBERTON Present in (outgroups): Polygyra, Praticolella, some Stenotrema, all Mesodontini except M. normalis and M. andrewsae. Apomorphic state: left pilaster thick and di- viding basally into network of bulges; right pi- laster short. Formerly and now present in: M. normalis (Fig. 11c), M. andrewsae (Figs. 2, 11b). Discussion. This clearly derived and pre- sumably homologous character state differs in M. andrewsae and M. normalis: in the former the branching basal bulges are shorter and more pronounced. Because there is no basis for identifying one of these configura- tions as primitive, however, | have combined them as a single character state. Transformation 12. Preceding transforma- tions: none. Plesiomorphic state: chalice walls undiffer- entiated from lateral pilasters. Present in (out- groups): Polygyra, Praticolella, some Sten- otrema, and the mesodontins P. panselena, P. perigrapta, P. sargentiana, P. clarki, P. ap- pressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. approximans, |. magazinensis, |. edentatus, |. verus, |. inflec- tus, |. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: right chalice wall abruptly flared above right pilaster; left chalice wall and pilaster undifferentiated and greatly reduced. Formerly and now present in: P. roemeri (Fig. 22c), P. leatherwoodi (Fig. 22b). Discussion. The flared right chalice wall is shown in side view in Figure 22b and in top view in Figure 22c. This change in elevation from the right pilaster is decidedly more abrupt than in P. panselena (Fig. 21a), P. perigrapta (Fig. 21b), P. clarki (Fig. 19b), or P. clenchi (Fig. 25), and these latter species also have more strongly developed left walls of the chal- ice, and thus their partial similarity is attributed to homoplasy or preservational artifact. Transformation 13. Preceding transforma- tions: none. Plesiomorphic state: chalice walls not dif- ferentiated from lateral pilasters. Present in (outgroups): Polygyra, Praticolella, some Stenotrema, and the mesodontins P. pansel- ena, P. perigrapta, P. sargentiana, P. clarki, P. appressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. approximans, |. magazinensis, |. edentatus, |. verus, 1. in- flectus, |. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: right and left chalice walls thick and evenly rounded, enlarging smoothly from the lateral pilasters, and form- ing a A-shaped cleft. Formerly and now pres- ent in: M. elevatus (Figs. 4+, 6d-f). Discussion. As discussed previously, the two dissected populations of M. elevatus dif- fer considerably in their manifestation of this character state. The similarity in shapes of chalices of M. elevatus and the occasionally sympatric M. zaletus (Transformation 23) is attributed to homoplasy owing to the differ- ences in structural detail previously dis- cussed. Transformation 14. Preceding transforma- tions: none. Plesiomorphic state: chalice walls not dif- ferentiated from lateral pilasters. Present in (outgroups): Polygyra, Praticolella, some Sten- otrema, and the mesodontins P. panselena, P. perigrapta, P. sargentiana, P. clarki, P. ap- pressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. approximans, |. magazinensis, |. edentatus, |. verus, |. inflec- tus, I. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: chalice walls abruptly higher-standing than lateral pilasters, thin and symmetric or expanding slightly to right. For- merly and now present in: /. subpalliatus (Fig. 13a), / rugeli (Fig. 13b). Discussion. This thin, high-standing chalice wall differs from that in M. andrewsae, M. nor- malis and M. altivagus (Transformation 22) in being symmetric or flared to the right rather than strongly flared to the left. Transformation 15. Preceding transforma- tions: none. Plesiomorphic state: chalice walls not dif- ferentiated from lateral pilasters. Present in (outgroups): Polygyra, Praticolella, some Sten- otrema, and the mesodontins P. panselena, P. perigrapta, P. sargentiana, P. clarki, P. ap- pressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. approximans, |. magazinensis, |. edentatus, |. verus, |. inflec- tus, |. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: chalice resembling thick wooden spoon, with right pilaster as its han- die. Formerly and now present in: I. ferrissi (Fig. 13c). Discussion. This type of chalice is both unique and very aberrant, without plausible connection with any other existing type. MESODONTINI EVOLUTION 109 Transformation 16. Preceding transforma- tions: none. Plesiomorphic state: chalice walls not dif- ferentiated from lateral pilasters. Present in (outgroups): Polygyra, Praticolella, some Stenotrema, and the mesodontins P. pansel- ena, P. perigrapta, P. sargentiana, P. clarki, P. appressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. approximans, |. magazinensis, |. edentatus, |. verus, I. in- flectus, |. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: chalice floor deeply re- cessed to form symmetric, cylindrical pit skirted by undifferentiated chalice walls. For- merly and now present in: P. pennsylvanica (Fig. 10b). Discussion. Despite a superficial resem- blance to the deep chalices of F. christyi, F. jonesiana, F. wheatleyi, F. orestes, F. archeri and M. mitchellianus (Transformations 17- 20, 25), the chalice of P. pennsylvanica is al- most certainly inhomologous owing to its unique relationship to the normal chalice walls, which are continuous with and undiffer- entiated from the lateral pilasters. Transformation 17. Preceding transforma- tions: none. Plesiomorphic state: chalice walls not dif- ferentiated from lateral pilasters. Present in (outgroups): Polygyra, Praticolella, some Stenotrema, and the mesodontins P. pansel- ena, P. perigrapta, P. sargentiana, P. clarki, P. appressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, I. approximans, |. magazinensis, |. edentatus, |. verus, |. in- flectus, |. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: apical chalice wall form- ing symmetrical hood with moderately thick rim. Formerly present in: F. christyi (Fig. 8a), F. wheatleyi (Fig. 8c), F. jonesiana (Fig. 8b), F. orestes (Fig. 8d), F. wetherbyi (Fig. 9b). Now present in: F. christyi (Fig. 8a). Discussion. Despite a superficial resem- blance to the symmetrical, cylindrical chalice of P. pennsylvanica (Transformation 16), dif- ference in structural detail suggests that this type of chalice arose by modification of the primitive walls of the chalice rather than by sinking of its floor. Transformation 18. Preceding transforma- tion: 17. Plesiomorphic state: floor of hooded chal- ice continuous with ventral penial wall; chalice walls straight-sided. Present in (outgroup): F. christyi (Fig. 8a). Apomorphic state: floor of hooded chalice separated from ventral penial wall by a con- tinuous, moderately thick circular rim around chalice; chalice walls straight-sided to weakly convex. Formerly present in: F. wheatlyi (Fig. 8c), F. jonesiana (Fig. 8b), F. orestes (Fig. 8d), F. archeri (Fig. 7), F. wetherbyi (Fig. 9b). Now present in: none. Discussion. It is most parsimonious, and seems developmentally most likely, that all symmetrical, circular-rimmed cup-like or cy- lindrical chalices formed by development of their walls as both homologous and derived from the simple hooded chalice of F. christyi (Fig. 8a) by way of this hypothetically inter- mediate stage. The ventral part of the contin- uous rim seems weak in F. jonesiana and F. archeri, stronger in F. wheatleyi, and very strong in F. orestes; these differences have been pooled because of possible individual variation. The ventral rim in F. wetherbyi is extreme. Transformation 19. Preceding transforma- tions: 17, 18. Plesiomorphic state: circular rim of chalice rather thick; walls of chalice straight-sided to weakly convex. Present in (outgroup): hypo- thetical ancestor. Apomorphic state: circular rim of chalice thin; chalice walls thin, very high-standing, and straight-sided to weakly convex. Formerly and now present in: F. wetherbyi (Fig. 9b). Discussion. This highly derived type of chalice is difficult to illustrate (Fig. 26), and the representation of it in Figure 9b is mis- leading owing to initial misinterpretation of its structure. In this figure, the high, thin walls are folded down into the mouth of the chalice, making the chalice seem shorter, thicker, and smaller-mouthed than it really is; in its natu- rally extended state, the chalice probably resembles a tall, symmetric cylinder. As dis- cussed previously, its affinities are problem- atic, but seem closest to the type of F. wheat- leyi (Figs. 7, 8b-d), hence the hypothesized homology. It differs from the homoplasic chal- ice of M. mitchellianus (Transformation 25) in its symmetry and smooth, unserrated rim. Transformation 20. Preceding transforma- tions: 17, 18. Plesiomorphic state: circular rim of chalice rather thick; chalice walls straight-sided to 110 EMBERTON weakly convex. Present in (outgroup): hypo- thetical ancestor. Apomorphic state: circular rim of chalice very thick; chalice walls convexly rounded. Formerly and now present in: F. wheatleyi (Fig. 8c), F. jonesiana (Fig. 8b), F. orestes (Fig. 8d, F. archeri (Fig. 7). Discussion. This character state often re- sembles a recumbent, thick-walled Chinese teacup, best seen in the illustration of F. wheatleyi (Fig. 8c). As mentioned in the dis- cussion of the preceding Transformation, the differences in this type of chalice among the four species having it have not been scored separately, because the relationship of these interspecific differences to individual variation is uncertain. Transformation 21. Preceding transforma- tions: none. Plesiomorphic state: chalice walls not dif- ferentiated from lateral pilasters. Present in (outgroups): Polygyra, Praticolella, some Stenotrema, and the mesodontins P. pansel- ena, P. perigrapta, P. sargentiana, P. clarki, P. appressa, P. binneyana, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. approximans, |. magazinensis, |. edentatus, |. verus, I. in- flectus, I. smithi and I. kalmianus (Figs. 9a,c, 11a, 14, 18-25). Apomorphic state: left wall of chalice mod- erately flared. Formerly present in: M. say- anus (Fig. 18a), M. chilhoweensis (Fig. 18b), M. normalis (Fig. 11c), M. altivagus (Fig. 15c), M. andrewsae (Figs. 2, 11b), M. zaletus (Figs. 4a—e, 6a—c, 16), M. mitchellianus (Fig. 16a), M. clausus (Fig. 16b), M. trossulus (Fig. 16c), M. thyroidus (Fig. 15b), M. sanus (Fig. 15a), /. downieanus? (Fig. 10a). Now present in: M. sayanus (Fig. 18a), M. chilhoweensis (Fig. 18b). Discussion. The chalices of all these taxa are flared to the left, usually with a left-hand flap. It is hypothesized, therefore, that they are homologous and derived from a common ancestor with the sort of moderate left-hand flare that occurs in M. sayanus and is more developed in M. chilhoweensis. Transformation 22. Preceding transforma- tion: 21. Plesiomorphic state: left wall of chalice rather flared. Present in (outgroup): M. say- anus (Fig. 18a), M. chilhoweensis (Fig. 18b). Apomorphic state: chalice asymmetrically scoop-shaped, left wall flared, high-standing, and thin-walled. Formerly and now present in: M. normalis (Fig. 11c), M. altivagus (Fig. 15c), M. andrewsae (Figs. 2, 11b). Discussion. This kind of chalice differs from that of I. subpalliatus and I. rugeli (Transfor- mation 14), which it superficially resembles in its strong left-hand asymmetry, whereas the other is symmetrical or flared to the right. The chalice of the illustrated specimen of M. alti- vagus (Fig. 15c) has a puckered rim; relaxed, undistorted examples resemble those of M. andrewsae and M. normalis in shape. This character state seems to be derivable from the kind occurring in M. chilhoweensis (Fig. 18b) by uniformly greater growth of the chal- ice walls. Transformation 23. Preceding transforma- tion: 21. Plesiomorphic state: left wall of chalice rather flared. Present in (outgroup): M. say- anus (Fig. 18a), M. chilhoweensis (Fig. 18b). Apomorphic state: right and left walls of chalice thick and rounded; enlarging rapidly from lateral pilasters, the right larger than the left, the left bearing a flap; and forming a A- shaped cleft. Formerly and now present in M. zaletus (Figs. 4a-e, 6a—, 17). Discussion. This unique sort of chalice seems to be convergent with that of M. ele- vatus (Figs. 4f-j, 6d-f); see discussion under Transformation 13. Its left-hand flap suggests a possible derivation from the sort in M. chilhoweensis (Fig. 18b) by thickening of both walls and restriction of the left-hand flare. Transformation 24. Preceding transforma- tion: 21. Plesiomorphic state: left wall of chalice rather flared. Present in (outgroup): M. say- anus (Fig. 18a), M. chilhoweensis (Fig. 18b). Apomorphic state: chalice narrowly triangu- lar, inclined to the left, rolled over, thick- edged, with pore on right side. Formerly and now present in: M. clausus (Fig. 16b), M. trossulus (Fig. 16c), M. thyroidus (Fig. 15b), M. sanus (Fig. 15a). Discussion. This sort of chalice is very dis- tinctive and is almost certainly homologous in the four taxa. It conceivably could have de- rived from the kind in M. normalis (Transfor- mation 22), rather than directly from that in M. chilhoweensis (Transformation 21) as hypoth- esized here. Transformation 25. Preceding transforma- tion: 21. Plesiomorphic state: left wall of chalice MESODONTINI EVOLUTION 111 rather flared. Present in (outgroup): M. say- anus (Fig. 18a), M. chilhoweensis (Fig. 18b). Apomorphic state: chalice an asymmetric cylinder, taller on left side; wall and rim of chalice thin, the right rim serrated. Formerly and now present in: M. mitchellianus (Fig. 16a). Discussion. Affinities of this unique kind of chalice are problematic. Despite a superficial resemblance to the deep chalices repre- sented by Transformations 17-20 and 16, it differs in its asymmetry, its serrated right rim, and its continuity with the right pilaster but not the left pilaster (in which it is reminiscent of that in M. ferrissi, Fig. 13c). Its taller left wall suggests derivation from a homologue of the sort in M. chilhoweensis by way of unknown intermediates. Transformation 26. Preceding transforma- tions: none. Plesiomorphic state: ventral penial wall smooth and featureless. Present in (out- groups): some Polygyra, some Praticolella, some Stenotrema, all Mesodontini except F. wheatleyi, F. jonesiana, F. orestes, F. archeri. Apomorphic state: ventral penial wall bear- ing thick basal bulge. Formerly and now present in: F. wheatleyi (Fig. 8c), F. jonesiana (Fig. 8b), F. orestes (Fig. 8d), F. archeri (Fig. 7). Discussion. This baso-ventral bulge varies somewhat in size, shape and position among F. wheatleyi, F. orestes and F. archeri, but it is assumed to be homologous. In F. jonesiana it is displaced to the left and is very large and elongate; thus it might not be homologous with that of the other three species. It does seem to be homologous with the bulge in F. wheatleyi, however. Even if the mid-ventral and left-ventral bulges are inhomologous, that of F. jonesiana still falls within the range of variation evident in F. wheatleyi. Transformation 27. Preceding transforma- tions: none. Plesiomorphic state: ventral penial wall smooth and featureless. Present in (out- groups): some Polygyra, some Praticolella, some Stenotrema, all Mesodontini except F. wheatleyi, F. jonesiana, F. orestes, F. archeri. Apomorphic state: ventral penial wall bear- ing basal pocket that opens toward apex. For- merly and now present in: M. mitchellianus (Fig. 16a). Discussion. This basal pocket, shown in Figure 16a as a notch to the left of and slightly above the vaginal opening, is formed by a fold looping between the bases of the right and left lateral pilasters. This fold seems independent of the pilasters themselves, and therefore is not homologous with the basally joined pilas- ters of F. orestes (Fig. 8d) and F. archeri (Fig. 7), despite a superficial resemblance. This pocket, or pouch, occurred in all six dissected specimens (from two populations) of M. mitchellianus; it certainly is not an artifact of preservation. Transformation 28. Preceding transforma- tions: none. Plesiomorphic state: dorsal wall of penis smooth and featureless. Present in (out- groups): many Polygyra, many Praticolella, some Stenotrema, and the mesodontins P. panselena, P. perigrapta, P. sargentiana, P. clarki, P. binneyana, P. leatherwoodi, P. roe- meri, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. magazinensis, |. edentatus, |. verus, 1. inflectus, Е. wetherbyi, |. smithi, 1. subpalliatus, |. rugeli, |. ferrissi, I. downiea- nus, P. pennsylvanica, Е. jonesiana, Е. or- estes, F. archeri, I. kalmianus and M. norma- lis (Figs. 7, 8b,d, 9, 10, 11a,b, 12-14, 18, 19a,b, 20-25). Apomorphic state: dorsal wall of penis with elongate bulge on left side. Formerly and now present in: F. christyi (Fig. 8a), some speci- mens of F. wheatleyi. Discussion. This dorsal bulge parallels the left lateral pilaster and might join it basally. In this respect it is homoplasic with the third pi- laster (Transformation 4), from which it differs in being much thicker and more rounded, and is therefore more like the baso-ventral bulge (Transformation 26). Transformation 29. Preceding transforma- tions: none. Plesiomorphic state: dorsal wall of penis smooth and featureless. Present in (out- groups): many Polygyra, many Praticolella, some Stenotrema, and the mesodontins P. panselena, P. perigrapta, P. sargentiana, P. clarki, P. binneyana, P. leatherwoodi, P. roe- meri, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. magazinensis, |. edentatus, |. verus, I. inflectus, Е. wetherbyi, I. smithi, 1. subpalliatus, |. rugeli, I. ferrissi, I. downieanus, P. pennsylvanica, F. jonesiana, F. orestes, F. archeri, I. kalmianus and M. normalis (Figs. 7, 8b,d, 9, 10, 11a,b, 12-14, 18, 19a,b, 20-25). Apomorphic state: dorsal wall of penis bearing isolated apico-lateral bulges, one on 112 EMBERTON each side. Formerly and now present in: some specimens of F. wheatleyi (Fig. 8c). Discussion. This unique character state is clearly derived. It seems not to be associated with the lateral pilasters. Transformation 30. Preceding transforma- tions: none. Plesiomorphic state: dorsal wall of penis smooth and featureless. Present in (out- groups): many Polygyra, many Praticolella, some Stenotrema, and the mesodontins P. panselena, P. perigrapta, P. sargentiana, P. clarki, P. binneyana, P. leatherwoodi, P. roe- meri, P. laevior, P. clenchi, P. indianorum, P. kiowaensis, |. magazinensis, |. edentatus, |. verus, I. inflectus, Е. wetherbyi, I. smithi, 1. subpalliatus, |. rugeli, I. ferrissi, I. downiea- nus, P. pennsylvanica, F. jonesiana, F. or- estes, F. archeri, I. kalmianus and M. norma- lis (Figs. 7, 8b,d, 9, 10, 11a,b, 12-14, 18, 19a,b, 20-25). Apomorphic state: dorsal wall of penis bearing multiple, semi-parallel ridges. For- merly present in: P. appressa (Fig. 19c), M. zaletus (Figs. 4а-е, 6a—c, 17), M. elevatus (Figs. 4f-, 6d-f), М. andrewsae (Figs. 2, 11b), M. mitchellianus (Fig. 16a), M. clausus (Fig. 16b), M. trossulus (Fig. 16c), M. sanus (Fig. 15a), M. thyroidus (Fig. 15b), M. altiva- gus (Fig. 15c). Now present in: none. Discussion. Because there is considerable and incompletely understood variation within each of the three types of dorsal ridges here tentatively recognized (Fig. 28; Transforma- tions 31-33), it is hypothesized that all three evolved from a common ancestral type of un- Known appearance. More careful investiga- tion of the dorsal ridges, including studies of ontogenetic and individual variation, should be very valuable and might well alter this sug- gestion of general homology. Transformation 31. Preceding transforma- tion: 30. Plesiomorphic state: dorsal wall of penis bearing several semi-parallel ridges. Present in (outgroup): hypothetical ancestor. Apomorphic state: dorsal ridges thick, ir- regular in width, and semi-parallel. Formerly and now present in: M. andrewsae (Figs. 2, 11b). Discussion. These dorsal ridges are usually more than 0.5 mm wide and longitudinal. Per- haps they evolved from thin dorsal ridges (Transformation 32), or vice versa; as a com- promise it is suggested that Transformations 31 and 32 arose independently from a com- mon ancestor. Transformation 32. Preceding transforma- tion: 30. Plesiomorphic state: dorsal wall of penis bearing multiple, semi-parallel ridges. Pres- ent in (outgroup): hypothetical ancestor. Apomorphic state: dorsal ridges thin, irreg- ular in width, and semi-parallel. Formerly and now present in: P. appressa (Fig. 19c), M. zaletus (Figs. 4а-е, 6a—c, 11), M. elevatus (Figs. 4f-j, 6d-f), M. mitchellianus (Fig. 16a). Discussion. These dorsal ridges are usually less than 0.25 mm wide and longitudinal, but they can be slightly oblique (e.g. Figs. 4c, 6b, 16a). Whether the dorsal ridges in all four species are homologous is very problematic; because the variation is so great, and occa- sionally overlapping, in both M. zaletus and M. elevatus (Figs. 4, 6), and because their collective variation seems to overlap the con- ditions in both P. appressa and M. mitch- ellianus, homology is hypothesized. Transformation 33. Preceding transforma- tion: 30. Plesiomorphic state: dorsal wall of penis bearing several semi-parallel ridges. Present in (outgroup): hypothetical ancestor. Apomorphic state: dorsal ridges thin to thick, uniform in width, and parallel. Formerly and now present in: M. clausus (Fig. 16b), M. trossulus (Fig. 16c), M. sanus (Fig. 15a), M. thyroidus (Fig. 15b), M. altivagus (Fig. 15c). Discussion. These dorsal ridges are usually 0.25—0.5 mm wide (but ca. 1 mm in М. altiv- agus) and usually oblique. They are much more even and corrugated than the other two kinds of dorsal ridges (Transformations 31, 32). The dorsal ridges of M. altivagus, al- though similar in general appearance, are proportionately larger than those of the other species, and might not be homologous. Transformation 34. Preceding transforma- tions: none. Plesiomorphic state: both peripheries of penis smoothly straight or curving. Present in (outgroups): Polygyra, Praticolella, Sten- otrema, all Mesodontini except M. thyroidus. Apomorphic state: right periphery of penis abruptly stepped near apex to form shoulder. Formerly and now present in: M. thyroidus (Fig. 15b). Discussion. Only in the naturally everted 113 MESODONTINI EVOLUTION TABLE 1. Distribution of 34 suggested genitalic character-state transformations (Figs. 26-28) among 42 species of Mesodontini. For correct gender endings of species, see Table 6 and text, excluding Appendix 3. Transformation Number 123 45 6 7 8 91011 121314 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 OQUIGROUPS=0 00 OOO DDOO0 ODO ODO 0 0 0,0 00-000. 00000000 0 0 Species ORO MO s ORONO OO SOKO) 0 HOMO 000. 07100) (0) 0008 0 07 0) 0) 0) 0 (0M 0) 070) 0 perigraptus 02023072070207020Z0202.07:020707.07.0707.07102.0707.020707.0707.0707.0.:07.0702070 panselenus 02029202.079207.020702.02.07.07.070702.07.0707207 0110 о ооо 0.0) 0:02.00. 070 sargentianus laevior clarki 00050’ 0.0 0000000’ 0 10) ооо ооо 0) 017000 01 07 070. 07.0 0202020202092.0202:0502. 07100. 0720207.02070707.070202.0702.02020.02.02.07.02070 02020202020920200202.0707020207.0702:07. 002070702070 707.020 07070-020270 binneyanus indianorum clenchi 029702020207 07020202. 070720770502 020707.0207.0710702.07.0207.07. 0 © 020707020 0202070=.02107.07.070707:0707070707 050) 0" O00 0:0) 0 00 оо 0 0: 00.070 0100’ 0 10) OF 0) 0) 0,0) 0 0 0 0 ооо O OO 0 0 0 0 0 0 0700 02.070700 080% 10) 05020 050/2010 0080) 0) 10 0 0): 0 0 00) 0 020) 0702 0 0 0) 0, 0 0) 0 0 kiowaensis edentatus verus OF ONO) 10) 010) 005010501050 010) 0110: 10) 01040: 710) 97 © 01 0) 0 0) 0 0) 0202070, OF ONO) ON OO ONON O0 0; 10 0) 20510" 0) 0) 100) 0) 0) 1050 0 100) 0:0) 0) 108 00) 070 1 1 1 kalmianus | 10) 10)50) 0) 10.0) 10) 0) оо 10 10) 10) 0/0" (0) о 0) (0) 0 10,0 1050 1050/0010 0) 0) 0 approximans 1,100) 010! 0 ооо 0) 0) 0 0 0 0 0 0 00 00 00 OO 07.00 OF 0 0) 0 0 magazinensis 120207020207.0710702.070207.0707.02.07.070707070707:070 1000). 0) 0) 0 10) 020 80 1 1000) 00100101020’ 0710’ 07070702020707. 0) 0: ооо 0) 10) 0) 0) 0) 0) 0 downieanus inflectus rugeli 1100) 0501070, 0100 1 0,0) 0) 0 10 0" 0) 07070707 072070 10) 00) 0) 0) 0 01 4 4 ONO MOMO 1 0) OF 00/0) 10) 10) 40/50/ 100/0010 0) 00) 10) OF 0! 0) 10" 0) 50) 0) 010) 10) 00 0 0 0 0 0 1=0702020707.02070207 1 0 0) 0 0 0 0) ооо 0) 0: ооо 0, 0) 01 0) 020 12 0207202070207.0202.020 1 subpalliatus ferrissi smithi 00) 10) 0 0) 0) O05 02.0207. 0207.0202.0207070 ооо о ооо 1) 0500) о ооо оо 020202020507.02:020 0700009 ооо 070707070! 0771 ONOMO 0 1040) 01) (0.0070 6 07070771 ORONO" 100500) 1) (0°00) 0 0) O00 04 07100710100) 0) i 0801051010) 00 O41 ONO} 00050) 0 00) 110.0 20) 000 | pennsylvanicus 0 0 0 0 O 1 wetherbyi christyi 120202020720202070702.07107.0702070 1 07070707 10) 0 0 00) O 1710702072070. 0 От Оооо IO 1 0 0) 0 0 0 wheatleyi 1707170707 07070717 0) 0) 0 1050/0070 10) 1 20) OSORNO) OST ONO; 0720202.02070 170212020205020517.02072020207.02070 jonesianus archeri 020,20707 0707,07.071707 070700001 orestes 02020205097070205.02.071070 120207020202. OF 0-1 020202070207. 020707.0710702.0207.0707.07.02.05,02.070 1 0202020202070202.02.0251 7020207207 0050) 0) 071 011010000 00.001 normalis 1707070 00000 00001 andrewsae roemeri 02070705070707.020707 071 ооо OMC 0 00 0 1 0 0 0 0 0 0 оо 0 0 070070707070 070707070 0 000 0'0 0 0 00 00 1 00000 00 0 0 0 0 0 0 00.0 1 leatherwoodi elevatus sayanus ооо 0400100’ 0.10) 0) 00) 10; 010: 1050/50 JO 0) оо 1, 0 о ооо ооо 0) 07 OF 00 02020507. 01°00 0) 010) 0 0! 0 0) 0) 00) 0 0) 0 717.0 07.070707. 0720) 0 070) 01 00 O10 0 0 DO 0 0 0 O10 0 0010-10 0 0 0 0 0 0 chilhoweensis zaletus 0712070707070 7. 071707 177070 020707177071 1 ооо OO OMONO OM ON о 02020207202 0202070207207202 0771 mitchellianus clausus 05020207.070710770210707107.07:07070107070707.0717.07.0717 02070720707 17.070 1770 02020207020707.07070707.02.07.071070707 0 0) 0 1! 010 10.0) 0) 0) 0 7 00000 010 0, O00 0 O 0 OO 00001 07071209 trossulus sanus 0010 0:10:11 10°10; 0170) 0) 1 1 07071 0202020702070 712029710 ORONO ORONO! OF OOF 0 100) 05070) O90 0202021070217 02070207071 010) 00 10) 0) 00) 00) 10 0 0) 0707070707070 1 thyroidus 1 altivagus 0202020707020207.07020202070720207.0207070720=0702:07072070202:0712.0212.020 appressus Table 2 summarizes the complete electro- Allozymic Analysis penis is this penial shoulder obvious (Webb 1954b: figs. 3, 4). In the opened, uneverted penis (Fig. 15b), the shoulder appears as a kink in the wall that easily might be mistaken for a preservational artifact were it not for its consistent presence in all dissections. In this table, each allele (electromorph) is represented by its migration phoretic results. distance relative to the control, Mesodon za- letus from Monte Sano, Alabama (FMNH 214772, 214773), the migration distance of Presence or absence of each of these 34 in each species of the Mesodontini is presented in Table 1. transformations suggested anatomical which was arbitrarily set at 100 mm. For the 75 populations presented in Table 2, 95 EMBERTON 114 001 96 16 S6 16 001 001 001 86 £OL 001 501 001 001 66 00! + 969712 = sisuauizeBew W (os')004 sol 001 LOL 26 16 oll 00: (0S')SOL 66 £OL 001 501 001 001 001 00! 2 €-269p 12 Ipoomiayjeaj W (e9)16 (88°)56 (21)46 (Sz')86 (sz’)ooı (88`)001 (21:)001 001 (88)86 (ZOOL 16 001 00+ (SZ’)Z01 66 505 ool (zt )EoL 001 001 001 505 + 069712 JOIA8E) 'W 001 001 16 86 16 06 001 001 001 50 OO! 501 001 001 001 00! + p89r1z SISUBBMOIY W (05')56 (06`)56 (09:)00+ (06)56 (sz’)ooL LOL 96 (0S )26 (01')00+ 16 501 001 001 001 505 001 (or)eoL 001 001 (01)96 (sz)zot $5 289712 snueiwjey ‘И (05')96 (oz')£6 (9t')86 (s2')86 (02 ’)ooL 96 (08')66 (ps')001 16 16 001 001 86 vol 001 50: (S2)ooL 001 66 ool << 999pLz 5трауи! ‘И (6/`)66 001 001 96 (sz’)eoı 16 001 101 001 66 50: O0! 501 001 001 96 00 + 599712 шплоив!ри Ww (05`)5`56 001 16 16 (09°)26 16 16 oll 001 66 505 OO! 901 vol 001 66 00 + 659712 15$ша} W (26)001 (82 )26 (£o')cot 96 16 (zz:)001 16 001 001 001 001 zOL 001 501 001 001 96 00! 81 959712 тела! ‘И (55')86 (21 )€6 (05`)001 (ZL )26 (05`)001 (21)56 001 96 (€8°)S6 (ZL)EOL 16 (€8°)00L 001 001 001 50: 00: (0s)eol 001 001 (€8)66 00 € €S9riz snjejuapa ‘И (os')001 00! 201 96 16 16 001 001 001 66 201 001 (05’)501 001 66 001 zo 1 Zs9bLz &142иарр ‘И (16)26 (20)66 001 96 001 (£0')004 001 501 001 001 001 Е O0! 501 70: (£6)001 001 ool 9 €r9riz snsnejo W 001 001 £6 86 16 001 001 001 66 Е O0! oll 001 001 96 204 € ee9bLz DEJO И (21)56 (90)46 (50')46 001 96 (v6')66 (8266 16 LOL 001 001 001 201 001 16 001 001 001 505 6 159715 IAISUYO WN (50`)56 (90')S6 (06)001 (26')96 (2v)86 (66°) LOL (zo)soL (€0')001 16 (Zr ’)ooL 16 501 001 001 001 201 001 £01 (20')bvoL 001 001 ool St 129$%12 SISUSAMOYNLYI ‘И (05')56 00! 001 16 (0/`)66 16 001 101 001 66 505 OO! 501 001 001 96 ook $5 Szyrız snueÄauuig W 001 96 86 501 16 16 001 001 66 505 O0! 501 001 001 66 008 1 ez9bLz sueuxoldde ‘W (12)16 (98`)96 (9b')S6 (61')96 (26)06 (z1 )ooı 00: (pv5)86 (85')26 16 001 001 (80°)001 86 €0! 001 501 001 001 001 z0ı €l 989712 snssesdde W (11)96 (s8')001 (LE )z0L (96 )001 (Si )soL 96 (85 )poı 501 16 (o')eol 011 001 66 LOL O0! 501 001 001 001 001 61 8L9pLz S9ESMAJPUE W (8S°)001 001 001 96 (Zp €or 16 001 001 001 001 00 OO! 001 001 001 001 00! 61 vL9bLz зпбелие W ido id de шба 2.109 1109 Z-poS L-PoS PD PD Pb pol aw z-UPW L-UPW ypıos N 19QUINN $90995 эпбо!е1е HNW4 ‘9 a[qe 1 995 sjuawubis -Se 91eu9B рэзила! 10} :(0y61) A1qSIId JO 9Sou] эле (uoposayy = ‘W) Suewuubisse sueuen “WW 001 Jo eoueysip e paubisse Ашелуиюие ‘(1681г HNWA ‘snjajez uoposayy) uoneindod |одиоэ au} jo ajajje UOWIWIO9 JSOW ay, о} anejas ww ul $|э6 eu} uo seoueJsIp лэц} AG рацциар! эле 190) 91 Jo yoea 10} Saja||ÿ ‘в/эипшуз\у dnolbino эц} jo uoneindod auo pue 'IUNUOPOSaY jo saisads GE бизидиоэ suonendod 18 10} eyep awAzolly “2 FIGVL 1115 MESODONTINI EVOLUTION (рапициоэ) v6 v6 (81:56 (28`)00+ 001 (68)001 (tt')001 (11 )eoı (9s')zo!L 001 96 001 96 001 96 001 001 (11 )96 (€8')00L 96 (€L')S6 (90)26 (18`)001 96 (64`)00+ (sz)voL 001 (55')001 (29°)SOL 001 (06)96 001 (OL')ooL sol 001 001 20+ 001 001 96 201 (88')001 (Z1)s0L 001 001 201 (oz')001 (08`)501 96 66 (60')86 (16)001 001 (55')66 (7) 16 (s2')66 (09)86 (or )ooı 16 96 16 (v6)26 (90')z01 (05)96 (09°)26 (os')s6 (os')001 66 LOL (9t')56 (75`)86 (05')96 (05')66 16 (88°)26 (z1’)86 66 86 5'56 (21)86 (88°)OOL 001 (ZL)L6 (Sz')OOL (85')501 (04)56 (05')86 (zv’)s’96 (8S")001 (08`)06 (02°)s6 (55`)001 (19°) 501 (61°)26 (61:)66 (z9')e01 (88 )eoL (z1)901 001 (0t')26 (09')96 (06 )68 (o1')26 (85')66 (Zp) LOL 16 (09')86 (07')001 (p2)56 (85')001 (85')501 96 (06)66 (0, )eor 16 001 16 16 001 001 16 16 (b6')86 (90')vo! 46 16 16 16 16 16 (01)v6 (0626 16 16 001 501 (12°) 46 (6Z)00L (82°) LOL (zz Jeo LOL 501 501 16 (e€)26 (29001 501 16 501 16 501 001 501 001 (61) от (sz’)eoı 001 (06`)001 (or )eor 001 001 (84001 (zz")v01 LOL 001 001 LOL 001 (90`)68 (76`)001 00+ 001 00! 001 001 001 001 (z9')001 (8e')ZoL 00! ZOL 06 001 001 00+ 001 00+ 001 00L 00+ 001 001 00+ col 001 001 (08')001 (02)201 06 001 00+ 86 001 00+ 001 001 001 86 66 66 66 001 66 66 66 001 86 001 001 00+ v6 001 501 LOL LOL LOL £ot 501 LOL 501 LOL 50+ 501 cot 501 LoL LOL cot LOL 001 001 001 00! 00! 001 001 001 001 001 001 00+ 001 001 00+ 001 00+ 001 00+ 96 00+ 501 50+ 501 50+ 901 001 (sz’)ooL (S2’JeoL 001 (11) 501 (€8")SOL 501 501 501 001 501 50+ 001 001 LOL 00! 00! 00! 00+ LOL 00+ 00+ (18')86 (61)001 00+ 00+ 001 001 00! 00+ 86 oot (02')86 (08")001 00+ LOL 00! 66 00+ 00! 00! 00+ 001 00! 00! 00+ 001 00! 66 00+ 66 00+ 001 00+ 16 00+ 96 00+ 001 06 001 66 96 001 00! 66 00! (29`)96 (85`)001 96 00+ 96 001 00! col 00! 001 00! 00! 00! 00! 00! (88`)001 (ZL) LOL col 001 00+ 00! col 00+ 501 001 001 00! Zt ce 168715 [ya 274 19112 АУЛ АХ 72474 147447 (0) 74474 854716 65491 [0774 204 elote ‘yeoun 6LZbLZ $04716 $04915 004715 86971 6671 96971< Isja1uep Bjlaunwysy snjajez W ıAaneaym ‘И ¡ÁQUIUIOM * snınssoN ' W W snpıosAyy WN smejedqns ‘И и Is | SNUBÁES : зпиециабиез | snues | W W W 1¡96n1 'W пашао/ 'W W snıdesßbued | snuajasued ‘pw S8]S810 W SIJBWJOU W SNUBI/AYIIW ‘И EMBERTON 116 (96')001 (29`)56 (8€)z6 (pO')pOL z0ı (ge)86 (29°66 16 001 001 001 66 20: 001 50 001 66 96 501 21 1049712 г—тавибиаа ‘и’ (05')56 (0S")OOL zol 16 86 16 001 001 001 86 LOL 00+ £OL 86 66 001 501 ! ‘yeoun a—snusjssued ‘И (S2’)26 001 г0 (5/0 86 16 001 001 001 00 20; 00: 00! 001 001 001 00! y 996712 г—ившшои ‘И (ZL )L6 (cf) 16 96 001 86 (e8)86 16 001 001 00! 66 £OL 001 00! O0! 00: (29‘)001 501 € 689712 £—JOIA8E] ‘И (60`)16 (06)56 (06)56 (01')001 00! 00: (s0')86 16 001 = 001 = — 001 00: 001 00! 001 501 o! {89712 z—Jorae] ‘И (se )86 001 96 16 (s2')oo1 16 001 001 00! 86 50. 001 00! O0! 001 66 001 Zz 899512 в—траци! ‘И (2t')96 (80°)96 (os)s6 (EE)BE (80`)06 (z6')001 96 (os)6 (sz)o0L 16 16 001 001 005 €0! 00: £OL 001 (26‘)001 66 00! 9 199712 г—тоаци! ‘И (12:)56 (£2)001 16 16 $'96 16 16 oll 001 — 650, 001 90. Hol 00! 66 00! LE 859712 6—4$5$шща; W 00! 16 16 16 16 16 = 001 = — OO! got = 001 66 001 01 1599515 &—455$шв; W (LL )€6 (05`)00+ (ZL)004 00! 96 (€8°)S6 (0S )eoL 16 001 001 001 66 £OL 001 (£8)£OL 001 001 66 001 € 799512 г—зтешере W (os)ooı (0S)96 (0S)26 (os)z6 — (09`)001 (05')66 (sz’)ooL (05`)66 — (05)96 (0S°)001 00: (os)zoı (os’)ooı (0S)86 (os’)ooı (09`)50+ 00! 00: (os)ooı 20, 00, (s2)eoı 00, (os’)ooı (os)001 (0s')z0!L г 059712 £—SNSNEJO ‘И (27) 16 (9b")96 001 00! 00: (85‘)001 001 501 001 001 66 20: 001 £OL 001 001 (S")OOL 16 21 vrobLz Z—SNSNEJO ‘И 001 001 96 96 16 001 001 001 = — 001 501 — 001 96 20: el ve9bLz E—PHEJO И (ss)e6 (81)26 (LF)EOL 001 001 (17)96 (28`)86 001 501 = 001 = — 60, (6S°)901 = 001 96 zo! LE Ze9bLz е—чжер ‘И (s8')001 (04)86 (s1:)s01 96 16 (0£)001 16 501 001 001 = — 001 50 LOL 001 001 001 o! 8z9b1Lz Z—SISUSIMOYNJO ‘И 00! 00! 16 66 16 001 101 00! oO! 50, 001 50 001 001 96 001 L gzgrız z@—snuefauuiq W (0€)56 (08')S6 (09')66 (0S')06 (oz')00! 001 56 (or )eol 16 001 00: (05’)00+ 86 50 001 50 001 001 001 201 5 Lz9P 12 €—snsseaidde ‘И (Or) 16 (or )e6 (0z)56 (65')96 (s2)96 — (05°)/6 (Sp')OOL 96 (s9)86e (01')66 16 001 — 06 — — 001 501 — 001 001 201 Ol 619515 z—snsseaidde W 00! 96 201 501 16 001 oll 001 66 r0L 001 £OL 001 001 001 001 L $19512 5—пбелще W (09')001 001 001 96 — (Or EOL 16 00! — 001 — — 00 00! 001 001 001 001 01 eL9pLz г—эпбелще W id) IdW de w6d 2-109 1109 Z-poS L-poS 2-PD кро pbd pol 2 WN &-UPW L-UPW yp10S N JaqunNn $90245 enboee9 HNW4 Ъ—— eee eee EEE (рэпициоэ) ‘г 319V1 17 MESODONTINI EVOLUTION 8 (68)001 (LL)EOL 00! 001 00+ 00+ 00+ 001 (os')001 (os')£ot 00! 00+ 001 001 (09')001 (or )roL (01:)96 (06')001 001 sol (o€')S6 (02Z')SOL y (vr )ooL (95')z01 96 00+ 96 96 96 96 96 001 96 96 96 001 96 96 00+ cot 001 (09)96 (or )66 96 S6 16 (05)26 (05`)00+ 86 16 26 (21)e6 (99')56 (21)26 16 16 (s1)96 (58`)/6 (01)68 (06)26 96 LOL (09°)S6 (0p)86 91! 001 (or )26 (09)001L 16 (29')96 (e2:)86 (SL )ooL (05')26 (os')o01 (62)S6 (sz')eoL 96 (e€)26 (82')86 (6e')001 (05')s6 (05')26 (55')86 (Z9')OOL (05`)86 (Gt')001 (so JeoL (sz')86 (gz')001 (0£')001 (02 )eoL (01)96 (06)001L 001 68 (06)46 (01)66 26 v6 16 16 00+ 00+ 001 00+ 16 16 16 86 16 16 16 16 16 9 (82°) LOL (ez')£oL 20+ LOL (v0')p6 (96)26 £0L 501 50+ 501 16 00! 001 001 16 001 16 501 00+ 5 (8/`)001 (z2’)vol 101 00+ 001 001 001 001 00+ 001 001 001 (os')001 (05')Z01 00! 00! 001 00! 001 001 (0t')06 (09')001 06 00! 001 00! 00! 00! 001 001 001 00+ 001 001 001 cot 001 00+ 66 66 00+ 501 501 001 001 001 00+ 001 001 00! 001 501 501 (08`)501 (05`)901 501 501 501 901 50+ 50+ 50+ 00! (07')56 (09')z01 501 501 $01 00! 001 86 001 LOL (01)86 (06)001L 001 001 001 00! 001 001 00! 001 001 00+ 00+ 001 001 001 00! 00+ 00+ 00+ 66 001 96 (8s')66 (zv")001 66 00! 06 00+ 96 66 66 66 96 001 66 66 00! 00! 00+ 00+ 001 00+ 001 001 001 001 00! 001 001 co! 001 001 001 (09)26 (ot')zoL (LOL =<) (9=x) (S6=<) 0+ el el OL 01 LE [eyo] A\uo в/эипшцзу uoposey 484915 854716 659971 954915 194915 054915 ‘yeoun eririe [51374274 464915 984915 147440 854716 774 474 [0774 474 8llrle (72474 ‚sajally 1101 z—snjajez 5—чЛазеацм г—Иазеацм | SNIBA | S—SnpiosAy} * у—прю^иц; * €—snpiosAy} * Z—SnpiosAy} | z—snjejedgns | E YyuIs | е—чцише | 2— 5ПИРАЕЗ | z—snuenuabies | 5—/абп | z—yebns | —118W801 ' €—snidesbued W W == === и 118 EMBERTON alleles were detected in the Mesodontini, and an additional six alleles were detected in the outgroup, Ashmunella. The most variable loci were Pgm and Lap, with 17 and 12 alleles. Icd and Gpi each had nine alleles; Mdh-1, Gd-1, and Got-1 each had six alleles; Sordh, Sod-2, Got-2 and Mpi each had five alleles; Mdh-2 and Me each had four alleles; Gd-2 and Sod- 1 each had three alleles; and Pgd had two alleles. The allozymic data for the second out- group, Allogona profunda, have been pub- lished in Emberton (1988: Table 2). Heterozygosity within populations was ex- tremely low. Most populations were mono- morphic for all but two or three loci, with a maximum of four alleles per locus (Table 2). Phylogenetic Analysis The results of the first and preferred method of phylogenetic analysis only (see Discussion) are presented here. Results of the two alternative methods are presented in Appendices 2 and 3. Removal of all autapomorphies from the data sets resulted in 21 informative genitalic transformations and 68 informative allozymic alleles. These 89 informative character states are identified in Table 3; their distributions among species are presented in Table 4. The six Nelson consensus trees produced from these data are shown in Figures 29-34; their statistics are presented in Table 5. These trees differ in the degree to which the anatomical data are more heavily weighted than the biochemical data, these weights ranging from one to six. Because informative allozymic alleles were approximately three times as numerous as genitalic character states, Figure 31 (weight = three) approxi- mately equates the sets of anatomical and biochemical data, and Figure 34 (weight = six) accords the set of anatomical data ap- proximately twice the weight of the set of bio- chemical data. The level of phylogenetic resolution of each of the six consensus trees (Figs. 29-34) de- pends at least partly on the number of trees from which it is constructed. Thus Figure 31, the least informative tree, is the consensus of 33 trees, the greatest number among the six consensuses (Table 4). One reason for this negative correlation between the number of trees and the phylogenetic resolution of their consensus is that the Nelson strict consensus method collapses clades that differ only in the placement of a single taxon (Hillis, 1987: TABLE 3. Numbering of character states in Table 4, as used for phylogenetic analysis. All autapo- morphic character states have been removed. For allozymic data (Table 2), all populations of each species have been pooled. Character No. Character State No. State O Transformation 1 (Fig. 26) 45 Gd2-98 1 Transformation 2 (Fig. 26) 46 Sod1-105 2 Transformation 3 (Fig. 26) 47 Sod1-100 3 Transformation 4 (Fig. 26) 48 Sod1-90 4 Transformation 8 (Fig. 26) 49 Sod2-110 5 Transformation 9 (Fig. 26) 50 Sod2-107 6 Transformation 11 (Fig. 26) 51 Sod2-101 7 Transformation 12 (Fig. 27) 52 Sod2-100 8 Transformation 14 (Fig. 27) 53 Got1-103 9 Transformation 16 (Fig. 27) 54 Got1-101 10 Transformation 17 (Fig. 27) 55 Got1-100 11 Transformation 19 (Fig. 27) 56 Got1-97 12 Transformation 20 (Fig. 27) 57 Got2-100 13 Transformation 21 (Fig. 27) 58 Got2-97 14 Transformation 24 (Fig. 27) 59 Pgm-103 15 Transformation 25 (Fig. 27) 60 Pgm-100 16 Transformation 26 (Fig. 28) 61 Pgm-99 17 Transformation 28 (Fig. 28) 62 Pgm-98 18 Transformation 30 (Fig. 28) 63 Pgm-97 19 Transformation 32 (Fig. 28) 64 Pgm-96.5 20 Transformation 33 (Fig. 28) 65 Pgm-96 21 Sordh-103 66 Pgm-95 22 Sordh-102 67 Pgm-92 23 Sordh-100 68 Pgm-89 24 Mdh1-100 69 Lap-104 25 Mdh1-99 70 Lap-102 26 Mdh1-96 71 Lap-101 27 Mdh1-95 72 Lap-100 28 Mdh1-91 73 Lap-99 29 Mdh2-100 74 Lap-98 30 Mdh2-99 75 Lap-97 31 Me-102 76 Lap-96 32 Me-100 77 Lap-95 33 Me-99 78 Lap-93 34 Me-97 79 Mpi-102 35 lcd-106 80 Mpi-100 36 Icd-103 81 Mpi-96 37 Icd-100 82 Mpi-94 38 Gd1-104 83 Gpi-105 39 Gd1-103 84 Gpi-104 40 Gd1-102 85 Gpi-103 41 091-101 86 Gpi-100 42 Gd1-100 87 Gpi-96 43 Gd2-100 88 Gpi-95 44 Gd2-99 fig. 2). In comparisons of Figures 29—34, the unfigured trees from which the Nelson con- sensuses were built were therefore consulted as well; these trees are available from the au- thor upon request. | The most robust clade is Fumonelix, which MESODONTINI EVOLUTION 19 FIG. 29. Nelson consensus tree of 18 maximum-parsimony cladograms generated from data matrix in Table 4, with genitalic character states weighted the same as allozymic character states. See Table 5 for statistics, text for discussion. appears consistently, regardless of the rela- tive weights accorded anatomical and bio- chemical data (Figs. 29-34). All other clades vary with different weightings. The second most robust clade is Mesodon (Mesodon). This subgenus shows the same phylogenetic structure in Figures 30-34. Of the 18 trees of which Figure 30 is the consen- sus, nine contain M. (Mesodon) intact, and nine differ in adding M. zaletus and M. mitch- ellianus at its terminus. The third most robust clade is /nflectarius (Inflectarius). Placement of I. downieanus (see below) within this subgenus in the trees is considered an artifact of the lack of electro- phoretic data for this species. I. (Inflectarius) is cohesive when anatomical characters are assigned weights of 1, 2, 3 and 4 [Figs. 29— 32; Figure 31 divides the subgenus only be- cause the alternative trees of which it is the consensus differ in placing /. kalmianus within (Inflectarius)]. Assignment of weights of 5 and 6 to anatomical states causes removal of /. rugeli from the subgenus, because of the ple- siomorphic penial morphology of this species. Mesodon (Akromesodon) is more strongly supported by the anatomical than by the elec- trophoretic data. The taxon occurs coherently EMBERTON 120 0010 0000! 0000! 00000 00000 00100 01000 00100 00100 01000 10010 01000 10000 11000 00000 00000 00000 00000 SISUSEMOIY EJ8JEJ EMOIN 0000 00010 OLOOL 00000 00010 00001 01000 01100 00100 01000 10110 01000 LOLLO OLLOO 00000 00000 00000 00000 $пивищехя snuejoajju; UWEX 6666 ¿ibid ¿ibid ¿ibid ¿ibid ¿ibid LLLLE ¿ibid ¿ibid ¿ibid ¿ibid LELLE LELLE 66660 00010 OOLLL 00000 10000 Puersauof хивиошп- sauol 0110 00010 OLLOL 01000 00001 OOLOL OLOLL 00100 OOLOL 01000 LLLLO 11000 10001 01000 00000 00000 00000 00,01 $пррауи SNHEJ98/}U] Sul 0010 0000! 00100 00000 00000 00010 11001 00001 00100 10000 10010 01000 10010 01000 00000 00000 00000 00000 WNJOUPIPU! Baje g Jupul 1010 00000 0000! 00000 00000 01000 01010 00000 10100 10000 1000! 00010 10001 01000 00000 00000 00000 01100 15$ ла; зпивеуи| $119} 0011 00010 0000! 00000 00000 01001 01001 00100 00100 01001 00010 01000 10010 01000 11000 00000 00000 00000 SNJEA9/9 UOPOSSWN 199 0010 00010 OLLOO 00000 00000 OOLOL LLOLL 00100 00100 11000 LOLLO 01000 10101 01000 00000 00000 00000 00000 SEJUS8pe snuejoajju; yuapa 5пиваимор 4666 66666 66666 66666 66666 66666 66666 66666 iiièè 66666 66666 66666 66666 66660 00000 00000 00000 OOOLL пивреуи IUMOP 0010 00000 10010 00000 00000 01000 01001 00100 00100 10001 00110 OLOOL 00000 10100 00000 00000 00000 00000 Iy2uap9 E18]Ed4 youjo 0010 00111: 00000 00100 00000 OOLOL 00100 01100 00100 LLOLL 00010 OLOLL 10010 LLOOL OLOOL 01000 00000 00000 snsnejo UOPOSaW snejo 0010 00001 01110 00000 00000 01100 OLLOL OLLOO 00100 10010 00011 01000 LOLLO 00100 00000 00000 00000 00000 MER E18JEd Я1е5 0010 00010 0000! 01000 00010 OLOLO 01000 10100 00100 OLOOL 00000 01000 10000 10010 00100 00001 00000 10000 IASUUO хивиошпу 15145 SISU88MOU/IY9 LOLO OLOLL O000! 00000 00010 OOLOL 01000 01100 00100 OLOOL 00010 00110 10000 11000 00000 01000 00000 00000 UOPOSSN ЧИ 0010 0000! 0000! 00000 00010 00010 01001 00001 00100 11000 10010 01000 10010 01000 00000 00000 00000 00000 BUPABUUIG в.эеа Auuiq 6666, 66666 66666 COUUU 66666 66666 66666 66666 66666 66666 ¿ebcl 66666 66666 668607 OOOLOPOOLLE 0000 00000 1194918 хувиошп- 1451 suewxodde 0010 00010 00000 10000 00000 00000 OLOLO 00100 00100 10000 10010 01000 10001 01000 00000 00000 00000 00011 зпивоауи Xidde LOLO 00011 OLLLO 10000 01010 01010 11001 00100 OLLOL 00000 10010 01000 10000 10100 11000 00000 00000 00000 2559/00 e18jE4 Sidde 0010 OLOLO 00010 00001 10000 00000 11001 01000 10100 10010 00010 01000 10000 11000 01000 11000 00010 00000 S9ESMS/PUE UOPOSSN MIPUE 0010 0000! 00010 00000 00000 0000! 11001 00100 00100 01100 00100 01000 10000 11001 01000 11000 00000 00000 зпбелще uoposayy ae 0010 1000! 00000 10001 01010 00100 OLOOL 00100 00000 10000 00000 00010 10000 00000 00000 00000 00000 00000 ерипуола виобоу бое 0000 00100 00000 01000 00000 00000 01000 01100 OLOLO 00000 00000 00100 01000 00100 00000 00000 00000 00000 /S/S/UEP Ejjaunwysy ипшзе 68-68 8-08 62-SL 7/-04 69-69 79-09 65-56 75-05 6b-Sb bb-0p 685-56 75-06 62-53 bc-0c 61-5, Pl-Ol 6-9 7-0 эшем $90995 ‘191qqY J9qUINN Jspeleyg ‘(pepos-Aieuiq эле salas цоцешио}зиед OeyuaB) | = juesald ‘0 = juasqy “(8861 ‘зшез) эвбшиэн Aq PEMOIIO} UOUSAUOD ay, цим P1099E ш 0 Je зе} (e SjqeL ul Pajsi)) seje]s лэрелецо 68 ay) jo Bunequiny ‘sisAjeue oneusbojÁyd 10; рэзп se ‘рэлошел saiydiowodeyne |je цим ‘вер опиАтоне pue oeyuan ‘у 378v1 125 MESODONTINI EVOLUTION LOLL 0010 0010 0010 0010 0011 001,0 OLLO LOLO 0010 0010 OLLO 0000 LOLO 0010 LOLO 0010 0010 0011 0010 0000 LOLO 00001 00011 00010 00010 00010 00010 00001 00010 00010 10001 01001 00011 01001 01000 00001 00000 01001 00000 00010 00010 01001 00001 10010 00010 00101 00100 00000 00001 00100 OLLOL 00001 00011 00100 00101 00000 10100 00010 10001 00001 1000! 00000 00001 00000 00001 10100 01000 01000 00000 10100 10100 00000 00000 00001 00000 00100 01000 00010 10000 01000 00000 10000 OLOLO 10000 00000 00010 10100 LLOLO 00000 00000 00001 00010 00011 00010 00000 00000 00000 00000 00101 00100 00000 00000 00000 00010 0000! 00000 00010 00100 OLOLO LLLOL OLOLL 01001 00101 00100 LLLOL 01000 00101 OLLEL 00001 00001 00001 00000 OLOLO 01000 00101 00001 00100 00010 00000 00000 00101 LOL LI 11000 11000 01010 00100 LOLOO 0.010 LLOLL LOOOL LLOLO 01000 OLOLL 01000 01001 01000 01001 11000 01001 LOLOL 01001 01000 01001 01100 LOLOL LOOLO 00100 01100 01100 00110 00100 OLLOL 00100 OLLOO 00100 OLLOO 00100 01100 00100 LLLOL 00100 01001 00100 00100 00100 00100 01100 00100 00100 00100 00100 00101 00100 00100 00100 00100 00101 00000 00100 00100 00101 01000 00100 00100 00101 00110 00100 01100 01000 0100 01000 01001 01010 00000 10000 10010 10000 OLOLO 10000 10000 LOOLL 01000 00010 OLOLO 01001 OLOLO 00000 10000 10000 00100 LOLLO 00010 10010 00010 LOOLL 10001 LOLLO LOLLO 10100 00010 LLOLO LOOLO 00010 00100 00010 00010 00100 00100 10010 10010 LOLLO 11000 01001 01000 10000 01000 01100 01000 01000 LL LOO 01000 01000 11000 01000 01001 01000 +0001 01000 11000 01000 01000 01000 01000 10000 LOOLL 10000 LOOOL 10000 +0010 10001 10001 10010 10000 10000 10001 10000 00010 10010 00000 1000! 10000 10000 1000! 10000 11000 11000 11000 11000 01000 LLOOL LLOOL LLOOO 01000 01000 10100 LLOOL 01000 11000 LOLLO 01000 10010 01000 11000 11000 01000 11000 10010 11000 00110 00000 00000 01001 01001 00000 00000 00000 00000 01001 00000 00000 00000 00000 00000 00010 00000 11000 00000 00000 00000 01000 OOLLL 00011 00000 01000 01000 00000 00000 01000 00000 01000 00000 00000 00000 00000 00000 00111 11000 01000 00000 00000 00000 00000 00000 00000 00000 00000 00000 01000 00000 00000 00000 00000 01000 00100 00000 10000 00000 00001 00010 00000 00000 00100 00000 00000 10000 00000 00000 00000 00000 OLLOO 00000 00000 00000 00000 OLLOO 00000 00000 00000 00000 00000 00000 00000 00011 00000 00000 snjajez uoposayy KaneayMm xijauouny Мдлэщам XılauoWnZ SNJ8A зпиваци $п/п55од UOPOSEN snpiosAy] uoposayy snjenjedgqns sn!/8]99//U/ цишз зпиврауи snuefÄes иорозаи вивциаб!ез в/эе4 snues uoposayy 1¡96n1 зпивоаци! 11911901 BIBJCg eidesBuad e1eje4 eaiuer/Äsuuad eJaje y Puajasued elajey se]seJo хувиоишпу 5/ешлои UOPOSEN зпив/ациш uoposayy sısuauızebew 5пивоауи! ‚роомлэщеа| 219]Ed JO/A98] 219784 Jejez Jeoum JU}OM SNJOA $$0Д рлАц} ¡dqns yyuus uefes 161es snues |эбпл JWa0l Bued suuad ¡sued ]saJo lou yon zeBbew yea] Inge] 122 EMBERTON FIG. 30. Nelson consensus tree of 15 maximum-parsimony cladograms generated from data matrix in Table 4, with genitalic character states assigned twice the weight of allozymic character states. See Table 5 for statistics, text for discussion. in Figures 32—34 (except for the incursion of Patera appressa, discussed below). In Figure 30, the taxon is intact except for the removal of M. elevatus; of the 33 trees of which Figure 31 is the consensus, M. (Akromesodon) oc- curs with M. elevatus in three, and without M. elevatus in 30. Only in Figure 29, in which anatomical and electrophoretic characters have equal weight, does M. (Akromesodon) break down, the only consistent elements be- ing the pairing of M. zaletus with M. mitch- ellianus, although other elements of the sub- genus recur among the 18 trees of which Figure 29 is the consensus. Patera (Patera) appears intact and mono- phyletic (except for the incursion of Allogona and Patera clenchi, discussed below) only in Figure 29, in which anatomical and allozymic character states receive equal weight. It ap- pears intact and paraphyletic in Figures 30 and 31 (because rotation of branches on any given node does not change the topologies of these trees), in the latter of which anatomical data as a whole are given the same weight as allozymic data as a whole. With increased weighting of anatomy (Figs. 32-34), how- ever, P. appressa moves from P. (Patera) to Mesodon (Akromesodon). This move is inter- MESODONTINI EVOLUTION 123 FIG. 31. Neison consensus tree of 33 maximum- parsimony cladograms generated from data matrix in Table 4, with genitalic character states assigned three times the weight of allozymic character states. See Table 5 for statistics, text for discussion. preted as the result of anatomical conver- gence between P. appressa (Fig. 19c) and M. elevatus (Fig. 4f-j). Less drastic breakdown of P. (Patera) upon heavy weighting of ana- tomical data—the isolation of Р. sargentiana in Figure 33 and of P. clarki in Figure 34—are also assumed to reflect genitalic conver- gence. In all of the 18 trees of which Figure 29 is the consensus, P. (Vesperpatera) (with the exception of P. clenchi, discussed below) ap- pears paraphyletically at the base of a clade leading to P. (Patera). The only reason this basal position does not occur in Figure 29 is that there is an incursion in nine of the 18 trees of M. sayanus and M. andrewsae as a sister group to P. indianorum; owing to the strict-consensus algorithm, this migrating unit causes the entire Patera clade to collapse into a polychotomy that belies much of the phylogenetic information. Similarly, (Vesperpatera) occurs paraphyl- etically in all 15 trees from which Figure 30 is derived, but is introgressed by P. sargentiana and P. laevior in 11 of the trees. The 33 trees comprising Figure 31 have P. (Vesperpatera) paraphyletic and intact in six, interrupted by either P. sargentiana and P. laevior or by the group P. sargentiana, Inflectarius subpalliatus and I. ferrissi. Patera (Vesperpatera) occurs phylogenetically intact and paraphyletic in Figure 32, but interrupted by the group P. sar- gentiana, Inflectarius subpalliatus, |. ferrissi and /. rugeli in Figure 33. In the 30 trees con- stituting Figure 34, the majority have Patera (Vesperpatera) paraphyletic and intact. The close affiliation of P. “clenchi” with P. peri- grapta throughout the consensus trees strongly supports the initial thought that the single electrophoresed juvenile was really a misidentified P. perigrapta, which was ex- tremely common at that site (Calico Rock, Ar- kansas). P. clenchi is tentatively placed in P. (Vesperpatera), with which it agrees not only in genital morphology but also in shape of shell and geographical distribution. The rela- tive isolation of P. kiowaensis within P. (Ves- perpatera) is due in large part to its sharing two allozymic alleles with many species of Mesodon and Fumonelix, as well as with other species (see especially Figure 30). Be- cause these bands are adjacent on the elec- trophoretic zymograms, P. kiowaensis might have only one of them, and a misinterpreta- tion of one or more gels might have assigned it two. The Inflectarius subpalliatus-I. ferrissi clade is a consistent feature of Figures 29-32. Only in Figures 33 and 34 is it divided by /. rugeli, which has a plesiomorphic genital morphol- ogy similar to that of /. subpalliatus, presum- ably by atavistic parallelism. Mesodon (Appalachina) is supported more by anatomy than by allozymes, for it occurs in Figures 32-34, but not in Figures 29-31. Of the 33 trees from which Figure 31 is derived, M. (Appalachina) appears in 15; in two other configurations its two species are separated by Mesodon normalis and by a clade consist- ing of M. elevatus, Patera pennsylvanica and 124 EMBERTON pues! [[ferrs subpl FIG. 32. Nelson consensus tree of 6 maximum-parsimony cladograms generated from data matrix in Table 4, with genitalic character states assigned four times the weight of allozymic character states. See Table 5 for statistics, text for discussion. Inflectarius kalmianus. The latter configura- tion obtains in Figure 30. In the 18 trees from which Figure 29 was computed, Mesodon sayanus is consistently split, with M. chilhow- eensis always paired with M. normalis, and with M. sayanus paired with either M. andrew- sae or Inflectarius kalmianus. Inflectarius (Hubrichtius) is the most tenta- tive taxon in this study. In the absence of elec- trophoretic data the position of /. downieanus in all consensus trees (Figs. 29-34) within I. (Inflectarius) serves only as evidence of its generic placement. Pairing this species with /. kalmianus is based solely upon the great con- chological similarity of these two species. Patera (Ragsdaleorbis) is the only mono- specific subgenus in this revision. Р. pennsyl- vanica appears isolated as the most plesio- morphic species of the Mesodontini in Figures 29, 32 and 33. In other censensus trees, it appears either between Mesodon elevatus and Inflectarius kalmianus (Fig. 30), or paired with Patera clarki. Because it is not divided into subgenera, further discussion the integrity of the genus Fumonelix in the cladograms is unnecessary. Inclusion of the /. subpalliatus-I. ferrissi clade in the genus Inflectarius is required by Fig- ures 29, 30 and 32; by 30 of the 33 trees from which Figure 31 was generated; and by the MESODONTINI EVOLUTION 125 FIG. 33. Nelson consensus tree of 6 maximum-parsimony cladograms generated from data matrix in Table 4, with genitalic character states assigned five times the weight of allozymic character states. See Table 5 for statistics, text for discussion. affiliation of this subgenus with I. rugeli in Fig- ures 33 and 34. Two facts support the inclu- sion of the problematic /. (Hubrichtius) in In- flectarius. One is the consistent association of |. downieanus (based upon genital morphol- ogy only, in the absence of pertinent allo- zymic data) with /. magazinensis and I. ap- proximans (Figs. 29-34). The other is the occurrence of I. kalmianus at the base of I. (Inflectarius) in Figures 32-34 and in all of the 18 trees from which Figure 29 was generated (not apparent in Figure 29 because in six of the 18 trees, Mesodon sayanus is paired with Inflectarius kalmianus), as well as within /. (Inflectarius) in three of the 33 trees of which Figure 31 is the consensus. The alternative positions of I. kalmianus in Figure 30 and in the majority of the trees comprising Figure 31 are tentatively deemed spurious and due to homoplasies. The type species of the genus Mesodon is M. thyroidus; M. (Mesodon) therefore belongs within the genus Mesodon a priori. M. (Ak- romesodon) also clearly belongs in Mesodon because it or the majority of its species form the sister group of (or rarely are intermixed with) M. (Mesodon) in all the trees from which Figures 29 and 31 were computed, and in Fig- 126 EMBERTON FIG. 34. Nelson consensus tree of 30 maximum-parsimony cladograms generated from data matrix in Table 4, with genitalic character states assigned six times the weight of allozymic character states. See Table 5 for statistics, text for discussion. ures 30 and 32-34. M. (Appalachina) ap- pears at the base of the clade M. (Mesodon) M. (Akromesodon) in Figures 32-34 and in 18 of the 33 trees comprising Figure 31. One member of the subgenus, M. chilhoweensis, retains this position in Figure 30 and in the additional trees comprising Figure 31; al- though it assumes another cladistic position in Figure 29, this species is still paired with Mesodon normalis. In 12 of the 18 trees of which Figure 29 is the consensus, M. say- anus is paired with M. andrewsae. M. (Appa- lachina) therefore is placed in Mesodon as its most plesiomorphic member. The association of M. sayanus with I. kalmianus in Figure 30 is expected, inasmuch as these two species are the plesiomorphic members of their respec- tive clades, /nflectarius and Mesodon. Inclusion of P. (Vesperpatera) in Patera (of which P. appressa is the type species) is re- quired by its close association with P. (Patera) in all trees, including those comprising Fig- ures 29 and 31 (in which resolution is lost owing to the strict consensus algorithm). The monospecific P. (Ragsdaleorbis) is placed tentatively in this genus as well, owing to its being plesiomorphic with respect to the rest of Patera in Figures 32 and 33, and in three of MESODONTINI EVOLUTION 127 TABLE 5. Statistics on the six cladograms in Figures 29-34, produced by Hennig86 (Farris, 1988) from 89 informative characters (21 genitalic, 68 allozymic) of the Mesodontini. Column headings refer to sequence of commands in this software program: “ccode /1” to “ccode /6” assign weights to genitalic character states (numbers “0” to “20”) of one to six times the weights of allozymic alleles; “mhenning*” trees are produced by single passes through data, and branch-swapping is then applied to them; “bb*” subjects these same trees to further branch-swapping, and outputs all trees of greatest discovered parsimony; “nelsen” produces a Nelson (1979) strict consensus tree from the set of most parsimonious trees. C.l. = consistency index, R.l. = retention index (Farris, 1989). ccode mhennig bb* nelsen /# 0.20 #Trees Length Cl. R.l #Trees Length Cl. R.l. Length Fig. # 1 1 374 23 47 18 374 23 47 450 29 2 1 409 26 50 15 409 26 50 417 30 3 2 443 29 53 33 443 29 53 554 31 4 4 473 32 56 6 472 32 56 473 32 5 3 500 34 59 6 499 34 59 500 33 6 1 528 36 61 30 527 36 61 533 34 the 33 trees from which Figure 31 is con- structed; and owing to its affiliation with P. clarki in Figure 34. The association of P. pennsylvanica with Inflectarius kalmianus in Figure 30 and in 30 of the 33 trees comprising Figure 31 is enigmatic. With respect to phylogenetic relationships among the four genera, Mesodon and Fu- monelix appear as sister groups in Figure 30, in 30 of the 33 trees comprising Figure 31, in half of each of the sets of six trees comprising Figures 32 and 33, and in half of the 30 trees from which Figure 34 was derived. In Figure 29 as well, all three species at the base of Fumonelix are members of Mesodon. Patera seems to be the most plesiomorphic genus in the Mesodontini: it occupies this position in Figures 30, 31-34, in 31 of the 33 trees con- stituting the consensus of Figure 31, and in seven of the 18 trees from which Figure 29 was constructed. Phylogenetically, then, /n- flectarius falls between Patera and Fu- monelix-Mesodon. The result of these comparisons of the cla- distic analyses is the following classification. Presentation of this phylogenetic hypothesis in the form of a tree is deferred to the consid- eration of conchological evolution. Revised Classification The complete systematic review is given in Appendix 1. Here the classification is simply outlined, with the genera, subgenera, and groups of species arranged phylogenetically, from most plesiomorphic to most apomor- phic. Species are listed alphabetically within groups. An asterisk denotes the type species of each genus and subgenus. Patera Albers, 1850 Patera (Ragsdaleorbis) Webb, 1954b pennsylvanica (Green, 1827)* Patera (Vesperpatera) subgen. nov. binneyana group binneyana (Pilsbry, 1899) clenchi (Rehder, 1932) indianorum (Pilsbry, 1899) kiowaensis (Simpson, 1888) roemeri group leatherwoodi (Pratt, 1971) roemeri (Pfeiffer, 1848)* Patera (Patera) perigrapta group penselena (Hubricht, 1976) perigrapta (Pilsbry, 1894b) appressa group appressa (Say, 1821)* laevior (Pilsbry, 1940) sargentiana (Johnson & Pilsbry, 1892) clarki group clarki (Lea, 1858) Inflectarius Pilsbry, 1940 Inflectarius (Hubrichtius) subgen. nov. downieanus (Bland, 1861) kalmianus (Hubricht, 1965)* Inflectarius (Inflectarius) edentatus group edentatus (Sampson, 1889) magazinensis (Pilsbry & Ferriss, 1907) smithi group smithi (Clapp, 1905) inflectus group approximans (Clapp, 1905) inflectus (Say, 1821)* 128 EMBERTON rugeli (Shuttleworth, 1852) verus (Hubricht, 1954) ferrissi group ferrissi (Pilsbry, 1897) subpalliatus (Pilsbry, 1893) Fumonelix gen. nov. christyi group christyi (Bland, 1860) wetherbyi group wetherbyi (Bland, 1874) wheatleyi group archeri (Pilsbry, 1940) jonesiana (Archer, 1938) orestes (Hubricht, 1975) wheatleyi (Bland, 1860)* Mesodon Férussac, 1821 Mesodon (Appalachina) Pilsbry, 1940 chilhoweensis (Lewis, 1870) sayanus (Pilsbry, in Pilsbry & Ferriss, 1906)* Mesodon (Aphalogona) Webb, 1954b elevatus (Say, 1821)* mitchellianus (Lea, 1838) zaletus (“Say” Binney, 1837) Mesodon (Akromesodon) subgen. nov. altivagus (Pilsbry, 1900) andrewsae (Binney, 1879) normalis (Pilsbry, 1900)* Mesodon (Mesodon) clausus (Say, 1821) sanus (Clench & Archer, 1933) thyroidus (Say, 1817)* trossulus (Hubricht, 1966) Table 6 permits comparison of this classifi- cation with those of Pilsbry (1940), based on shell morphology; and of Webb (1954b, 1968a, 1968b, 1983), based on genital anat- omy and behavior. Conchological Variation Conchological illustrations of 39 of the 42 species of Mesodontini are presented in Fig- ures 35-50. Of the three species not illus- trated, Fumonelix archeri and Inflectarius verus have extremely restricted ranges, and the shell of Patera indianorum resembles that of P. binneyana, except for its umbilicus, which is closed to narrowly chinked. Both F. archeri and P. indianorum are illustrated in Pilsbry (1940), whose monograph should be con- sulted for an understanding of variation within species in shell morphology. An illustrated key to most of the species is contained in Burch (1962). Range maps are available both in this paper (Fig. 51) and in greater detail and with ecological notes in Hubricht (1985). It is important to remember in identifying any shell of the Mesodontini that many species of the Triodopsini have closely convergent shells (see Emberton, 1988). Patterns of Genitalic Evolution The ranges of the 42 species of the Mes- odontini are presented in Fig. 51. These maps were compiled from Hubricht (1985), with a correction for Mesodon mitchellianus. The ranges of Inflectarius verus and Mesodon trossulus are from Hubricht (unpublished), and the range of M. altivagus is from data in Pilsbry (1940). These maps were used to assess the rela- tionships in geographic ranges among pairs of sister taxa, which were also compared in genital morphology (Figs. 2-25). The results, based on 29 pairs of sister taxa, are pre- sented in Table 7. Of the four pairs showing a great difference in penial morphology, three had parapatric ranges and one had an allo- patric range. Thirteen pairs of sister taxa were deemed moderately different in genital mor- phology, and of these, four were allopatric, two were peripatric, four were parapatric and three were sympatric. Of eight pairs of sister taxa whose genitalia differ only slightly four or five were mostly peripatric, one or two showed some allopatry and two were sympat- ric. Peripatry predominated over sympatry (three to one) among the four pairs of sister taxa judged not to differ in their penial mor- phologies. The tests for character displacement in re- productive organs at the level of population are summarized in Table 8. In none of these 16 comparisons was there any detectable dif- ference in penial morphology between allo- patric and sympatric populations. Patterns of Shell Evolution Figure 52 shows the phylogenetic pattern of shell morphology among all known species of the Mesodontini (these are the only illus- trations in this paper of the shells of Patera indianorum, Inflectarius verus and Fumonelix archeri). There is an evolutionary pattern of relative conchological stasis within subgen- era. In general, each subgenus is character- ized by a distinct shell form: shells of Mes- odon (Akromesodon) (Figs. 35c-f, 36e-f) are large, globose, imperforate, toothless and hairless; those of Mesodon (Mesodon) (Figs. MESODONTINI EVOLUTION 129 TABLE 6. Revised supraspecific classiciation compared with classifications of Pilsbry (1940) and of Webb (1954b, 1968a, 1968b, 1983). This Classification Genus, Subgenus, Species Fumonelix archeri christyi jonesiana orestes wetherbyi wheatleyi Inflectarius Hubrichtius downieanus kalmianus Inflectarius approximans edentatus ferrissi inflectus magazinensis rugeli smithi Inflectarius Inflectarius subpalliatus verus Mesodon Akromesodon altivagus andrewsae normalis Aphalogona elevatus mitchellianus zaletus Appalachina chilhoweensis sayanus Mesodon clausus sanus thyroidus trossulus Patera Patera appressa clarki laevior panselena perigrapta sargentiana Ragsdaleorbis pennsylvanica Vesperpatera binneyana clenchi indianorum kiowaensis leatherwoodi roemeri Previous Classifications Subgeneric Placement within Mesodon Pilsbry Mesodon Mesodon Patera Patera Mesodon Mesodon Inflectarius Inflectarius Mesodon Inflectarius Inflectarius Inflectarius Inflectarius Patera Mesodon Mesodon Mesodon Mesodon Mesodon Mesodon Appalachina Appalachina Mesodon Mesodon Mesodon Patera Mesodon Patera Patera Patera Mesodon Mesodon Mesodon Mesodon Mesodon Mesodon Webb Ragsdaleorbis Inflectarius Aphalogona Aphalogona Mesodon Mesodon Patera Patera Ragsdaleorbis Patera 130 EMBERTON FIG. 35. Shells. a,b. Inflectarius kalmianus, GS-116 #18. c,d. Mesodon andrewsae, Roan Mountain, GS-11 #1.e,f. Mesodon normalis, SC-158 #2. 36a-d, 37c-f) are generally medium-sized, globose, with a creviced umbilicus, toothless and hairless; those of Mesodon (Aphalogona) (Figs. 37a,b, 38) are generally large, globose, imperforate, bearing a single pronounced pa- rietal tooth and hairless; shells of Mesodon (Appalachina) (Fig. 39) are large, sub- globose, broadly umbilicate, with a baso-col- umellar tooth or node and hairless; those of Fumonelix (Figs. 40c-d, 41, 42) are generally small, subglobose, imperforate, with a single parietal tooth and hairless; shells of Inflectar- ius (Inflectarius) (Figs. 40a,b,e,f, 43, 44) are generally small, subglobose, imperforate, tri- dentate (with parietal, palatal, and basal teeth) and hairy; those of Inflectarius (Hu- brichtius) (Figs. 35a,b, 45a,b) are small, glo- bose, imperforate, toothless and hairless; shells of Patera (Patera) (Figs. 46, 47, 48c—d) are generally medium-sized, depressed, im- perforate, with a blade-like parietal tooth and a long basal lamella and hairless; those of Patera (Vesperpatera) (Figs. 48a,b, 49, 50) are generally medium-sized, depressed, with a creviced to open umbilicus, toothless and hairless; and shells of Patera (Ragsdaleorbis) MESODONTINI EVOLUTION 131 5mm FIG. 36. Shells. a,b. Mesodon sanus, GS-103 #8. c,d. Mesodon thyroidus, GS-63 #11. e,f. Mesodon altivagus, SC-144 #7. (Figs. 45c—d) are medium-sized, globose, im- perforate, toothless and hairless. Most of the shapes of shells of the various subgenera are unique and distinctive within the Mesodontini (but converge on those of other groups; see below). The resemblance is rather close, however, among Patera (Rags- daleorbis), Inflectarius (Hubrichitus), Mes- odon (Akromesodon) and Mesodon (Mes- odon). Some of these convergences or par- allelisms are confusing in the field. For exam- ple, specimens of I. kalmianus collected at GS-116, Knox County, Kentucky, initially were mistaken for M. clausus, with which they were microsympatric. In addition, there are several convergent deviations from the general pattern of con- chological stasis within subgenera. The most 132 EMBERTON Se = ITA Des N \ N A | \ , \ \ 1 \ FIG. 37. Shells. a,b. Mesodon mitchellianus, Hubricht 19406 +A. c,d. Mesodon clausus, GS-116 #3. e;f. Mesodon trossulus, GS-53 #A. striking of these is between Patera (Patera) clarki (Fig. 47c,d), which is an aberrantly high- spired member of its subgenus, and Mesodon (Aphalogona) elevatus (Fig. 38c,d), which dif- fers from other members of its subgenus in having a long, basal lamella. P. clarki looks superficially like a miniature M. elevatus, but close examination reveals important differ- ences in the relative size, shape and angle of the aperture. Other shell convergences among subgen- era are not so close. Patera (Vesperpatera) approaches Mesodon (Mesodon), but has more depressed spires and apertures (Fig. 52). Mesodon (Appalachina) has somewhat depressed spires, but has rounded apertures, very broad umbilici and coarse sculpture compared to the superficially similar Patera (Vesperpatera). The obviously unidentate Fumonelix (Figs. 40c,d, 41, 42c,d) somewhat resemble those species of /nflectarius (In- flectarius) in which the teeth on the apertural lip are effaced [/. magazinensis (Fig. 43c,d), I. verus (not individually figured) and /. sub- palliatus (Fig. 44a,b)]. Another convergence of members of different subgenera, also not very close, is that of Inflectarius smithi (Fig. 40e,f), which in its parietal tooth and basal lamella (Fig. 40e) resembles members of Patera (Patera) (Figs. 46, 47, 48c,d), which lack the periostracal hairs of /. smithi, how- ever. MESODONTINI EVOLUTION 133 cS \ ae d FIG. 38. Shells. a,b. Mesodon zaletus, GS-104 #21. c,d. Mesodon elevatus, GS-104 #33. Further minor convergences occur among subgenera within a genus, such as the ab- sence of periostracal hairs in both /. (Hubrich- tius) (Figs. 35a,b, 45a,b) and I. (Inflectarius) (I. ferrissi, Fig. 44e,f). Within Mesodon, pari- etal teeth occur, seemingly independently, in M. (Appalachina) (M. sayanus, Fig. 39c), M. (Aphalogona) (M. zaletus, Fig. 38a; M. eleva- tus, Fig. 38c), and M. (Mesodon) (M. thyroi- dus, Fig. 36c); a reddish-brown color band occurs convergently in M. (Aphalogona) (some specimens of M. elevatus: Pilsbry, 1940: Fig. 440b), M. (Akromesodon) (some specimens of M. altivagus: Pilsbry, 1940: Fig. 437e), and M. (Mesodon) (M. trossulus, Fig. 37f); and an exposed umbilicus occurs in both M. (Appalachina) (Fig. 39a,c) and M. (Mes- odon) (M. clausus, Fig. 36c; M. thyroidus, Fig. 37c). Within Patera, parietal teeth occur in both P. (Vesperpatera) (P. roemeri: Pilsbry, 1940: Fig. 449b,c; P. leatherwoodi, Fig. 49c,d) and P. (Patera) (Figs. 46, 47, 48c,d); and elevated spires occur in both P. (Rags- daleorbis) (P. pennsylvanica, Fig. 45d) and P. (Patera) (P. clarki, Fig. 47d). Within a subgenus, the pattern of distribu- tion of a few minor shell characters among species is mosaic, with cases of convergence or parallelism. Within Fumonelix, periostracal hairs appear independently in F. wetherbyi and F. wheatleyi (subspecies F. wheatleyi clingmanica Pilsbry, 1904; Pilsbry, 1940: 736); and an enlarged shell with a reduced parietal tooth occurs in both F. wheatleyi (Fig. 42a,b) and F. orestes (Fig. 42e,f). Within /. Inflectarius, reduction or loss of teeth on the outer lip occurs in /. magazinensis (Fig. 43c), |. verus (not figured), /. subpalliatus and I. fer- rissi (Fig. 44a,e). DISCUSSION Genitalic Analysis As Pilsbry predicted (1940), penial sculp- ture proved a useful source of systematic characters in the Mesodontini. That it yielded fewer characters and character states than in 134 EMBERTON | 10 N || 4 a — N а TT] FIG. 39. Shells. a,b. Mesodon chilhoweensis, Hubricht 30943 +A. c,d. Mesodon sayanus, GS-130 #7. the Triodopsini (Emberton, 1988) is hardly surprising. Because it is inserted during cop- ulation, the triodopsin penis is subject in its morphology to many more forces of natural selection, including sexual selection, than is the mesodontin penis, which merely touches or intertwines with the mate’s penis during copulation, but is never inserted (Fig. 1). Owing to their plasticity and probable erec- tility, many of the penial structures in the Me- sodontini sometimes are only tentatively inter- pretable. Investigation of variations among individuals and populations was essential to distinguishing real structures from preserva- tional artifacts. The suggested character-state transforma- tions (Figs. 26-28) vary in plausibility. Char- acters might have been differently delineated, for example, the chalice might have been combined with the lateral pilasters, or each lateral pilaster treated entirely separately. The effect of such alternate approaches upon the outcome of phylogenetic analysis is unknown. Thorough documentation of the suggested character-state transformations provides ob- jective, falsifiable hypotheses that might facil- itate future, more enlightened revisions. Allozymic Analysis Starch-gel electrophoresis produces es- sentially one-dimensional characters: migra- tion distances of stained bands on gels. De- tection of convergences, whenever possible, depends on running the doubted alleles side- by-side, and preferably staggered, on the same gel. In this study, most suspected con- vergences were run on the same gel, but the magnitude of the project prevented side-by- side comparisons in many cases. Further, the large number of taxa and compared popula- tions increased the probability of undetect- able homoplasies. Additional error is introduced by undetected alleles that indeed are possessed by a given species (Swofford & Olsen, 1990). Compari- son of eight populations of Mesodon zaletus (Emberton, 1986), for example, showed two potential sources of undetected alleles in a species: low frequency and uneven geo- graphical distribution. Both sources undoubt- edly introduced a substantial number of spu- rious homoplasies into the allozymic data, resulting in low consistency indices for the cladograms (Table 5). The predictably high MESODONTINI EVOLUTION 135 FIG. 40. Shells. a,b. Inflectarius inflectus, SC-130 #1. c,d. Fumonelix wetherbyi, GS-115 #4. e,f. Inflectarius smithi, GS-20 #1. FIG. 41. Shell. a,b. Fumonelix christyi, GS-161 #10. 136 EMBERTON FIG. 42. Shells. a,b. Fumonelix wheatleyi, GS-6 #3. c,d. Fumonelix jonesiana, SC-155 #3. e,f. Fumonelix orestes, Hubricht 40465 #A. incidence of real and false homoplasy in the allozymic data is offset, however, by the rel- atively large number of alleles that were de- tected. Thus, if all sources of error are ran- dom, the phylogenetic “signal” will be detectable through the “noise.” Phylogenetic Analysis With the widespread use and acceptance of the maximum-parsimony method, and of the method of combining morphological with biochemical data for phylogenetic analysis, justification probably is not needed. In view of the controversy and flux in phylogenetic methodology, however, it seems worthwhile to document the reasoning behind the choices made in this study. As systematists, we sample from a distri- bution of characters that have changed during the true phylogeny. Estimating the true phy- logeny, then, is ideally a statistical problem, in which confidence limits can be placed upon each phylogenetic hypothesis based upon the size and distribution of samples, and upon in- herent error of sampling (Felsenstein, 1983a, 1983b; Kim & Burgman, 1988). Realizing this ideal, however, depends upon having a prob- abilistic model of the evolutionary process (Felsenstein, 1982; Farris, 1983). The model commonly used for the statistical (maximum likelihood) approach to phylogenetic recon- MESODONTINI EVOLUTION 137 FIG. 43. Shells. a,b. Inflectarius approximans, Hubricht 23497 #A. c,d. Inflectarius magazinensis, GS-95 #17. ef. Inflectarius edentatus, GS-90 XA. struction based upon gene frequencies, intro- duced by Edwards & Cavalli-Sforza (1964), assumes that the arc-transformed frequen- cies are in Brownian motion on an infinite scale (Felsenstein, 1982; Rohlf & Wooten, 1988; Kim & Burgman, 1988). Although re- lated random models of morphological change effectively mimic robust phylogenetic hypotheses from real data (D. M. Raup, pers. comm.), correlation does not imply causation. Such random models might be sufficiently ac- curate depictions, however, of evolutionary adaptations to effectively random climatic and biotic changes (L. Van Valen, pers. comm.). If such models are accepted, then the most accurate and reliable method of phylogenetic estimation is maximum likelihood, using Fels- enstein’s (1986) CONTML program, part of his PHYLIP package (see Kim & Burgman, 1988). From the little that is known of the pop- ulation biology and population genetics of po- lygyrid snails (McCracken, 1976; McCracken & Brussard, 1980; Emberton, 1986), there is not any reason to reject this model of evolu- tion. The goal of this research, however, is to establish a phylogenetic framework for under- standing the mechanisms of evolution in this group of snails. Until these mechanisms are understood, therefore, phylogenetic recon- struction should use methods devoid of as- sumptions about these mechanisms (Farris, 1983). 138 EMBERTON 5 mm FIG. 44. Shells. a,b. Inflectarius subpalliatus, GS-153 #8. c,d. Inflectarius rugeli, GS-130 #8. e,f. Inflectarius ferrissi, SC-144 HA. In the absence of probabilistic assumptions about the evolutionary process, the array of methods for phylogenetic reconstruction can be divided into those using the character data directly, and those that begin by reducing the data to distances among examined taxa. Not only is much critical information lost by con- densing structured data into a single index, but all methods that use distance data (whether clustering or pairwise; Felsenstein, 1982) carry the sometimes, perhaps often, false assumption of equal rates of evolution (e.g. Farris, 1983). The effect of unequal rates of evolution on the ability of a distance method to detect a “true” simulated phylog- eny can be alarmingly severe (Kim & Burg- man, 1988). For these reasons, distance methods of phylogenetic reconstruction were not used in this study. Non-distance, non-probabilistic methods all converge on Hennig’s method (1966) when- ever the data are completely free of unde- tected homoplasies (i.e., parallelisms, con- vergences and reversals). This traditional method, seemingly first codified by Mitchell (1901, cited in Nelson & Platnick, 1981) finds the shortest branching tree (cladogram) along which the taxa are connected by unique tran- sitions between states (variants) of charac- ters (homologous structures). Rarely is such a data set encountered, and characters are usually found to be incongruent. Whenever MESODONTINI EVOLUTION 139 FIG. 45. Shells. a,b. Inflectarius downieanus, Hubricht 30825 #A. c,d. Patera pennsylvanica, SC-246 #1. the characters are incongruent, and when- ever careful reassessment fails to detect all homoplasies, then various algorithms having different biological assumptions can serve to reach a compromise solution to the conflicting data (Felsenstein, 1982). None of these algorithms is free of risk, and all are most prone to error (as defined by in- consistency in the maximum-likelihood model; see Farris, 1983) whenever homoplasy is common and whenever evolutionary rates are markedly variable (e.g. Felsenstein, 1982; Rohlf & Wooten, 1988; Kim & Burgman, 1988). These algorithms comprise two cate- gories: parsimony, which minimizes the num- ber of homoplasies among all characters; and compatibility, which minimizes the number of characters having homoplasies (Felsenstein, 1982). Compatibility, or clique, analysis (Mea- chum, 1981; Le Quesne, 1982) suffers from the covering assumption that if a character shows some homoplasy, then all points of similarity in that character are homoplasies— otherwise phylogenetic information is dis- carded whenever a character is discarded be- cause it shows homoplasy (Farris, 1983). Methods of parsimony currently are divisi- ble into those banning reversals (Camin- Sokal parsimony), those banning conver- gence and parallelism (Dollo parsimony), those lacking restrictions (Wagner parsi- mony), and those banning convergence and parallelism but permitting polymorphism (polymorphism parsimony) (Felsenstein, 1982). Given that specialists on stylommato- phoran land snails universally accept the prevalence of morphological convergences (e.g. Solem, 1978; Emberton, 1988; Tillier, 1989), and that rather drastic morphological reversals are postulated for various groups of land snails (W. B. Miller, pers. comm.; Nords- ieck, 1987; Emberton, 1986, 1988, in prep.), unrestricted parsimony is preferred. Of the plethora of available methods for phyloge- netic reconstruction, therefore, Wagner unre- stricted parsimony (Kluge & Farris, 1969; Far- ris, 1970) is the one used in this study. An exact maximum-parsimony solution was technically not feasible for the large data ma- trix generated (Table 4). The heuristic method used (mhennig* bb* options of Hennig86: Farris, 1988) has been found in recent, exten- sive, empirical tests to arrive at cladograms as short as, or shorter than, those produced by any other algorithm (Platnick, 1989). The importance of combining all available 140 EMBERTON a = PES >) \ ее > E | 0 e В : E AS | K | ES as DA x =. de: Ab FIG. 46. Shells. a,b. Patera panselana, GS-142 #A c,d. Patera perigrapta, GS-98 #3. data, both morphological and molecular, for phylogenetic analysis has been ably de- fended by Hillis (1987) and Kluge (1989), among others. Although one-dimensional al- lozymic data obviously should receive a lower weight than three-dimensional anatomical data, there is not any truly objective way to assign weights (see Emberton, 1988, and Ap- pendix 3 for one attempt to solve this prob- lem). The choice of assigning six different weights pursued in this study led to some very interesting results. This method has the ad- vantage of refining detection of discrepancies between allozymic and genitalic data, thus al- lowing detection of misleading convergences in both data sets (for example, the convergent dorsal penial structures in Patera appressa). A disadvantage of comparing the results of different and greater weightings of anatomical data with respect to allozymic data, however, is that the comparisons are tedious and sub- jective. An alternative approach is Farris’s method (1988) of successive weighting. This method first produces a set of maximum-par- simony trees, assigns each character a weight (0-10) according to its degree of fit to these trees, reruns the analysis with the as- signed weights, reweights each character in accordance with the resulting trees, and iter- ates this process until the character-weights stabilize. Although successive weighting “has the advantage of providing a means of basing groupings on more reliable characters without making prior decisions on weighting” (Farris, 1988), it has been rightly criticized for its in- herent circularity (Swofford & Olsen, 1990: 499). Application of successive weighting to the data set (Appendix 2) produces the strict con- sensus tree shown in Figure 53. This tree wholly supports that shown in Figure 52 in its phylogenies of Mesodon and Fumonelix, which therefore can be considered robust with respect to methodology. The resolution of the tree in Figure 53 is less for the other two genera, however, and there are some differ- ences from Figure 52. The most important dif- ference is in placing the pair /. subpalliatus and /. ferrissi as sister group to the pair I. inflectus and I. rugeli, rather than at the base of a larger clade including these two species. This solution is exactly that obtained in alter- native analyses of the some of the same data (Emberton, 1986, in press; Appendix 3), and MESODONTINI EVOLUTION 141 FIG. 47. Shells. a,b. Patera sargentiana, GS-101 #2. c,d. Patera clarki, GS-1 #3. e,f. Patera appressa, GS-104 #2. thus is a more robust hypothesis for evolution within Inflectarius. Separate analyses of genitalic and allozy- mic data, with construction of a consensus from weighted trees (Appendix 3), results in a topology (Fig. 60) very similar to that pro- duced by the preferred method (Fig. 52, Ap- pendix 1). This alternative and rather compli- cated analysis (procedure outlined in Figure 54) shows a high degree of congruence be- tween anatomical and electrophoretic data sets. Its resulting consensus differs most sig- nificantly in the placements of the Inflectarius ferrissi-subpalliatus clade (mentioned above) and of Mesodon (Appalachina), which admit- tedly is considered tentative in Figure 60. In sum, two alternative phylogenetic analy- ses, one of them quite different in approach and even in its outgroup, support to a high degree that shown in Figure 52. Revised Classification The revised classification (Appendix 1) fol- lows the phylogenetic hypothesis depicted in Figure 52. Within the limits of accuracy of the data, the phylogenetic hypotheses concern- ing Fumonelix and Mesodon can be consid- ered fairly robust, as noted above. Thus the classifications and inferred phylogenies of Fu- monelix and Mesodon are based entirely upon genitalic and allozymic data. For the more plesiomorphic genera, Patera and In- flectarius, these data are less conclusive, and 142 EMBERTON N : > | vA 5 mm Ne ye O ee 20 b SU d 5 mm FIG. 48. Shells. a,b. Patera binneyana “long”, FMNH 176008 #B. c,d. Patera laevior, H-22 #3. hence more reliance is placed on shell mor- phology for the final classification. Genitalic Evolution Genitalic and geographic comparisons be- tween 29 pairs of sister taxa (Table 7) de- tected evolutionary trends similar to those previously found in the Triodopsini (Ember- ton, 1988): sister taxa with peripatric geo- graphical ranges usually show little or no difference in penial sculpture; those with sympatric ranges show only moderate differ- ences; and all examples of great genitalic dif- ferences, and most examples of moderate genitalic differences, occur between sister taxa with parapatric or allopatric ranges. The caveats about interpreting these results were discussed by Emberton (1988: 236). In addition, population-level comparisons for 16 species (Table 7) failed to find any trace of character displacement in penial morphol- ogy. These findings support and generalize the hypotheses made for the Triodopsini (Ember- ton, 1988) that (1) peripheral isolates gener- ally do not differentiate, (2) differentiation due to reproductive character displacement is moderate at most, and (3) major differentia- tion is rare, rapid, and occurs in isolates; they do not support the hypothesis that vicariant isolates generally differentiate slowly. Full dis- cussion of these hypotheses and their impli- cations appear in Emberton (1988). It is sur- prising to find so much similarity in patterns of genitalic evolution between the Mesodontini and the Triodopsini, despite the very different ways the penis functions during copulation in these two tribes (Webb, 1961, 1974; Ember- ton, 1986). An important aspect of genitalic evolution in the Mesodontini is that a plesiomorphic or near-plesiomorphic morphology (lateral pilas- ters simple, chalice a simple continuation of the lateral pilasters, other sculpture absent) occurs in three of the four genera, including the most apomorphic genus (Fig. 52). Thus, although it is characteristic of Patera, this morphology also persists in /nflectarius (I. kalmianus, Fig. 11a; /. edentatus, Fig. 12c; I. verus, Fig. 14) and, surprisingly, with little modification in Mesodon (Appalachina) (Fig. 18). Such conservatism—or atavism, by an- other interpretation—was not found in the genitalic evolution of the Triodopsini (Ember- ton, 1988). MESODONTINI EVOLUTION 143 FIG. 49. Shells. a,b. Patera binneyana “short”, Hubricht 31615 #A. c,d. Patera leatherwoodi, GS-67 #1. e,f. Patera roemeri, GS-63 #4. Shell Evolution With important and striking exceptions, each genus has features characteristic of its shell morphology. Patera shells are generally medium-sized, smooth and depressed; /n- flectarius shells are typically small, hirsute and tridentate; those of Fumonelix tend to be small, smooth and unidentate; and Mesodon shells are usually large, smooth and globose. Each subgenus of the Mesodontini is rela- tively fixed, or evolutionarily static, in its shell morphology. Exceptions are rare and either unique [the high spire of Patera (Patera) clarki or parallel [reduction or loss of aper- tural tooth in /nflectarius (Inflectarius) ferrissi, and, to a lesser extent, in some other mem- bers of this subgenus, and in Fumonelix wheatleyi, F. orestes and Mesodon (Aphalog- ona) mitchellianus]. Some of the subgenera with static shell morphologies are conchologically similar. The most conspicuous example of this similarity is the globose, toothless, imperforate, hairless shell morphology that occurs in at least one subgenus of all four genera (Fig. 52). Accord- ing to the phylogenetic hypothesis, this sort of shell is plesiomorphic in the two most plesio- morphic genera (Patera and Inflectarius), and also occurs among the proposed close out- groups of the Mesodontini (Fig. 52). It is rea- sonable to hypothesize, therefore, that this 144 EMBERTON e FIG. 50. Shells. a,b. Patera clenchi, Hubricht 25210 #A. c,d. Patera roemeri. e,f. Patera kiowaensis, GS-84 #7. shell form is plesiomorphic within the Mesod- ontini. If this hypothesis—that the common ances- tor of the Mesodontini had a globose, tooth- less, imperforate, hairless shell—is correct, then several parallelisms among genera fol- low from an acceptance of the phylogenetic hypothesis (Fig. 52). First, a parietal tooth has evolved separately at least once within each of the four genera. Second, a basal lamella has evolved at least three times [in P. (Pa- tera), in I. (Inflectarius) smithi and in Mesodon (Aphalogona) elevatus]. Third, an exposed umbilicus has evolved three times [in Patera (Vesperpatera), in Mesodon (Appalachina) and in M. (Mesodon) ]. Fourth, periostracal hairs have evolved at least twice [in /. (Inflec- tarius) and in Fumonelix], and more probably three times, because the hairs of F. wetherbyi and F. wheatleyi clingmanica seem to be in- homologous. One can only speculate about the pre- sumed adaptive values of these parallel struc- tures. Apertural teeth, or denticles, have been interpreted as barriers against insect preda- tors (review in Goodfriend, 1986; Emberton, 1988), but without substantial experimental evidence. Some of the parietal teeth in the Mesodontini are rather small and only slightly obstruct the entire aperture, although they might substantially obstruct the open pneu- mostome. An alternative, or auxiliary, hypoth- esis is that a parietal tooth directs body wastes away from the body of the animal during crawling. The basal lamina must strengthen the lower lip of the aperture, a fea- MESODONTINI EVOLUTION \ CE IN ed Bee ко à ee i DE er \ NG \) Patera (Vesperpatera) : Patera (Ragsdaleorbis) & binneyana kiowaensis P. (Patera): clenchi leatherwoodi pennsylvanica panseleng indianorum roemeri perigrapta ON M VFS Y > Pe ( \ ) () LAN 4 у ( / м Ny nie ak ie ¡QA 9 4 À Inflectarius .(Inflectarius): approximans rugeli inflectus ferrissi subpalliatus verus EB J obey ee on „al / UN == ATA AN = SE ~ \ \ XS) и J a Mesodon (Akromesodon) : Mesodon (Aphal : ogona): altivagus РЕНИ z E andrewsae mitchellianus normalis zaletus Inflectarius (Inflectarius): an, ee I ЛЕ — Zo. - Zr, \ ( TF > \ 1 NE 5 N UE ¢ И у ¿ Ar A = =~ у € — y 1 o > \ rn ù y E y a y O VA DÍ as £ ) ses tee Rel oo N = A Patera (Patera): Inflectarius (In ( ) f1 : appressa clarki edentatus : ae laevior A magazinensis sargentian ый J ug у Inflectarius (Hubrichtius): ee therb Ownleanus ee on evi kalmianus ee Zn jonesiana Mesodon (Mesodon): Mesodon (Appalachina): chilhoweensis sayanus clausus sanus thyroidus trossulus FIG. 51. Range maps of 42 species of Mesodontini, arranged by revised subgenera. Adapted from Hubricht (1985). ture of benefit to cliff-dwelling species such as P. (Patera) laevior as they drag their shells into narrow crevices in the rock; the majority of mesodontin species with basal laminae are not cliff-dwellers, or even talus-dwellers, however (Emberton, 1986). The benefit, if any, of an exposed umbilicus is unknown. No mesodontin is known to brood its eggs in its umbilicus, nor does the umbilicus of litter- dwelling species seem to have the disadvan- tage of accumulating debris, although known burrowing species such as Fumonelix archeri (W. & A. Van Devender, pers. comm.) are im- perforate. Periostracal hairs presumably ei- ther protect the shell from contact with acidic, decaying leaves, or accumulate soil and de- bris that camouflage the shell from visual or tactile predators, or both. Although these con- chological characters seem to be adaptive because of their parallel derivations, their functions are unknown; hypotheses are at least available for testing. Unfortunately, not enough is yet known about the detailed ecol- ogy of these animals to test the hypotheses (see Emberton, in press). Recommendations for Future Research Several systematic problems in the Mesod- ontini remain unresolved. First, the mono- phyly of this tribe is still in question; the ple- siomorphic penial morphology occurs in various outgroups (Fig. 52), and no unequiv- ocal synapomorphy phylogenetically unites the Mesodontini. Second, the true phyloge- netic position of the aberrant Patera pennsyl- 146 EMBERTON TABLE 7. Comparison of difference in penial morphology with relationship between geographic ranges for 29 pairs of sister taxa of the Mesodontini according to the phylogeny in Figure 52. Taxa are designated by abbreviations used in Table 4. * = substantial difference in length only. Phylogenetically Adjacent Taxa rugel vs. apprx subpl vs. ferrs wethr vs. wheat group norml. vs. altiv penns vs. rest of Patera perig vs. pansl apprs vs. laevr apprs vs. sargt downi vs. kalmn (?) edent vs. magaz edent vs. smith magaz vs. smith inflc vs. rugel inflc vs. apprx chrst vs. rest of Fumonelix—wethr sayan vs. chilh zalet vs. mitch roemr vs. leath laevr vs. sargt binny group vs. apprs group—apprs wheat group (4 spp. inter se) elevt vs. zalet andrw vs. norml thyrd vs. claus thyrd vs. tross binny subgroup vs. clench subgroup perig subgroup vs. clark sanus vs. rest of thyrd group claus vs. tross vanica is unclear, whether is it a member of a monophyletic Patera clade, or a relict or unique member of an isolated, plesiomorphic genus within or without the Mesodontini. Third, the status of P. leatherwoodi as a sep- arate species, rather than as a small relict population of P. roemeri, needs to be deter- mined. Fourth, it is unclear whether the two genitalic forms of P. binneyana are separate species, or extremes of a continuum between populations with long and with short penes. In either case, the evolutionary mechanism by which extremely long penes evolve in land snails (Figs. 23, 24; Solem, 1974: fig. 12) re- mains to be determined, although presumably it involves runaway sexual selection (Fisher, 1930), a phenomenon poorly understood in hermaphrodites. Fifth, it is not clear whether the morphologically similar and geographi- cally parapatric P. perigrapta and P. pansel- ena constitute one or two true species. Sixth, the systematic status of the several geo- graphic forms of P. clarki remain incompletely Penial Geographical Shift Relationship great parapatric great parapatric great allopatric great parapatric (?) moderate parapatric moderate” allopatric moderate parapatric (?) moderate allopatric moderate parapatric moderate peripatric moderate allopatric moderate allopatric moderate sympatric moderate peripatric moderate sympatric (7?) moderate” parapatric moderate sympatric slight peripatric slight peripatric (?) slight allopatric slight allo- or peripatric slight sympatric slight peripatric slight sympatric slight peripatric none sympatric none peripatric none peripatric none peripatric (7?) assessed; for example, it is unknown whether the endangered P. clarki nantahalae is a spe- cies, a subspecies, or an ecophenotypic vari- ant. Seventh, the phylogenetic placement of Inflectarius downieanus needs to be tested with both relaxed anatomical material and with allozymic analysis. Because of their im- portant evolutionary implications, the plesio- morphic positions of /. kalmianus and I. down- jeanus within Inflectarius (Fig. 52) should be tested using additional data sets. Eighth, the species status of /. verus needs testing, in part to determine whether Hubricht (1985) was correct in synonymizing it with /. subpal- liatus. Ninth, more data are needed to allow better zoogeographic analyses of Patera and /nflectarius. The phylogenetic relation- ships among the Ozarkian and Appalachian members of both of these genera (Fig. 51) need more robust hypotheses (compare Fig- ures 53, 54). Tenth, relationships within the terminal taxa of Fumonelix require further in- vestigation, using more characters and more MESODONTINI EVOLUTION 147 TABLE 8. Localities (state:county) of populations dissected in searches for reproductive character displacement between pairs of conchologically similar species of the Mesodontini. Number of specimens dissected from each population in parentheses. TX = Texas, AL = Alabama, AR = Arkansas, TN = Tennessee, KY = Kentucky, NC = North Carolina, IN = Indiana. Species A Allopatry Sympatry Allopatry Species B roemeri TX:Travis (1) TX:Bastrop (5,5) TX:Cherokee (3) thyroidus bucculentus inflectus AL:Madison (1) AR:Logan (1,3) — magazinensis TN:Blount (3) KY:Henderson (3) inflectus AL:Madison (1) TN:Blount (3,3) NC:Swain (3) rugeli KY:Henderson (3) AR:Logan (1) wheatleyi NC: Avery (3) NC:Haywood (3,1) NC:Swain (3) clarki zaletus TN:Blount (1) TN:Franklin (5) — elevatus AR:Crawford (1) IN:Owen (3) NC:Swain (1) zaletus TN:Blount (1) NC:Swain (1,1) NC:Macon (3) normalis AR:Crawford (1) TN:Cocke (0,3) NC:Macon (2) TN:Franklin (5) IN:Owen (3) zaletus TN:Blount (1) NC:Swain (1,1) NC:Swain (2) altivagus AR:Crawford (1) TN:Franklin (5) IN:Owen (3) normalis NC:Macon (3) NC:Swain (1,1) NC:Swain (2) altivagus NC:Macon (2) normalis NC:Swain (1) NC:Macon (3) — thyroidus NC:Macon (2) TN:Blount (3) clausus TN:Blount (4) KY:Knox (3) — kalmianus populations. Preliminary studies (F. G. Thompson, unpublished; Emberton, unpub- lished) indicate complex and sometimes inter- grading variations in both shell and genital morphologies in the species of the Smoky Mountain and those of adjacent areas. This radiation is fascinating and poorly under- stood. Further, because of their highly aber- rant shell form and plesiomorphic genitalia, Mesodon sayanus and M. chilhoweensis should be more thoroughly investigated to test their current phylogenetic placement. Again, the position of M. mitchellianus within M. (Aphalogona) is likewise suspect. Further, the limits and relationships of species within M. (Akromesodon) remain problematic. For example, it is unclear whether M. altivagus is truly separate from M. andrewsae. This entire complex, like Fumonelix, is a variable and in- completely understood component of the land-snail fauna of the Southern Appala- chians. Finally, whether M. trossulus is truly separate from M. clausus or part of a local polymorphism is unclear. The Mesodontini, because of their species diversity, their phylogenetic hypothesis, their mapped ranges of species, and their broad conchological, genitalic and allozymic varia- tion, are an excellent system for further evo- lutionary studies. For example, the three widespread clades (Patera, Inflectarius and Mesodon) could be compared as to their modes of speciation; their covariations among the evolutionary rates of anatomy, shell and allozymes; their phylogenetic changes in ontogeny of the shell, as mea- sured from sections or X-radiographs of adult shells (Raup, 1966); their rates of spread from Pleistocene refugia as determined from geo- graphic variation of allozymes; and their evo- lutionary and phenotypic plasticities of shell shape. A most promising aspect of the Mesodon- tini for the study of evolution is the fact that 148 EMBERTON _ Fumonelix Mesodon FIG. 52. Evolution of shell morphology and upper penial sculpture in the Mesodontini. The ten clades designated as subgenera are (from left to right) Patera (Ragsdaleorbis), P. (Vesperpatera), P. (Patera), Inflectarius (Hubrichtius), |. (Inflectarius), Fumonelix, Mesodon (Appalachina), M. (Aphalogona), M. (Ak- romesodon), and M. (Mesodon). their conchological radiation has been iter- ated by the distantly related, confamilial tribe Triodopsini (Pilsbry, 1940; Emberton, 1988). These two tribes have very nearly the same geography, ecology, conchology and species richness (Emberton, 1986, 1988). This mono- graph on the Mesodontini complements that on the Triodopsini (Emberton, 1988) in laying the phylogenetic basis for using these conver- gent, sympatric radiations to address general questions about the evolutionary morphology of gastropod shells. ACKNOWLEDGMENTS This paper is adapted from part of a doc- toral dissertation approved by the Committee on Evolutionary Biology, University of Chi- cago. | am grateful to my advisor, the late Alan Solem, to whose memory this paper is dedicated, and to the members of my pro- posal and defense committees: David Raup, Michael Wade, Bradley Shaffer, Russell Lande, Lynn Throckmorton, James Teeri and Harold Voris. For physical, moral, and secre- tarial support throughout this project, | am most grateful to Ellen Emberton. This paper is a contribution of the Molecular Genetics Laboratory of the Department of Malacology, Academy of Natural Sciences of Philadelphia (ANSP). George Davis, Thomas Uzzell, Caryl Hesterman, John Hendrickson, Andrea Garback and Arthur Bogan were ex- tremely helpful and encouraging hosts during my electrophoretic stints at the Academy. For the loan of specimens under their care | thank the late Alan Solem, Field Museum of Natural History; George Davis, ANSP; and Leslie Hubricht, of Meridian, Mississippi. | also wish to thank Mr. Hubricht for generously providing then-unpublished range maps (Hu- bricht, 1985), precise locality data for several rare species, identifications of difficult mate- rial and advice for collecting. For carrying much of the burden of labelling and cataloging my collections into FMNH, | am indebted to Margaret Baker, Patricia Johnson and Lucy Lyon. For assistance in the field, | thank Ellen Emberton, Lucia Emberton, Ned Walker, Gene Bryant, Tony Bryant, Eugene Keferl, Leslie Hubricht, John Ahrens, John Petranka, Betsey Kirkpatrick, Glenn Webb, Wayne Van Devender, Amy Van Devender, Martha Van Devender, Wayne Evans, Arthur Bogan, Rob- ert Lawton, John Pinkerton, Mark Souther- land, Dennis Herman, Greg Mueller, Kisa Nishikawa, Phil Service, Joe Bernardo, Ken Baker, Alan Lo, David Kasmer and Brad Fos- ter. Thanks also go to the many park rangers and property owners who permitted me to col- MESODONTINI EVOLUTION 149 lect on their land, as well as to the many peo- ple who provided camping sites and other hospitality. Glenn Webb generously shared his vast knowledge of the Polygyridae and allowed me to study his slide-mounted voucher speci- mens. He also graciously provided the live specimens of Ashmunella. John Hendrickson (ANSP) has my very great gratitude for running most of the com- puter analyses of Appendix 3. This paper benefitted greatly from the com- ments of George Davis, Carol C. Jones, and two anonymous reviewers. All errors are my own. This work was funded by the following grants to the author: NSF Postdoctoral Fel- lowship BSR-87-00198; NIH Genetics Train- ing Grant GM07197-07; Jessup Fellowship, ANSP; Bequaert Award, American Malaco- logical Union; Hinds Fund, University of Chi- cago (UC); Research Grant, Highlands Bio- logical Station, Highlands, North Carolina; Louer Fund, FMNH; Field-Collection Grant, Division of Invertebrates, FMNH; and Student Computation Fund, UC. Additional funding was also provided by an NIH grant to George M. Davis, and a USDA grant to Michael J. Wade. LITERATURE CITED ALBERS, J. C., 1850, Die Heliceen, nach natürlicher Verwandschaft Systematisch Geord- net. Berlin, 262 pp. ARCHER, A. F., 1938, A new species of Polygyra from the Great Smoky Mountains, North Caro- lina. Nautilus, 51: 135-137. ARCHER, A. F., 1948, Land snails of the genus Stenotrema in the Alabama region. University of Alabama Museum Paper 28, 85 pp. ASAMI, T., 1988a, Competition and character dis- placement in the land snails Mesodon normalis and Triodopsis albolabris. Doctoral dissertation, University of Virginia, 206 pp. ASAMI, T., 1988b, Temporal segregation of two sympatric species of land snails. Venus, 47: 153-172. BAKER, F. C., 1939, Fieldbook of Illinois land snails. Illinois Natural History Survey, Manual 2, Urbana, 166 pp. BINNEY, A., 1837, A monograph of the helices in- habiting the United States. Boston Journal of Natural History, 1: 466-495, plates 12-21. BINNEY, W. G., 1879, On certain North American species of Zonites. Annals of the New York Acad- emy of Sciences, 1: 355-362, plates XIV, XV. BINNEY, W. G. 8 T. BLAND, 1869, Land and fresh water shells of North America. Part I. Pulmonata Geophila. Smithsonian Miscellaneous Collec- tions, 194: 1-316. BLAND, T., 1858-1861, Remarks on certain spe- cies of North American Helicidae. Annals of the Lyceum of Natural History of New York, 6: 280— 302 (1858); 7: 26-39 (1859), 115-142 (1860), 420-448 (1861). BLAND, T., 1874, Description of a new species of Helix, and note on H. mobiliana, Lea. Annals of the Lyceum of Natural History of New York, 10: 361-363. BLINN, W. C., 1963, Ecology of Mesodon thyroides [sic] and Allogona profunda. Ecology, 44: 498— 505. BURCH, J. B., 1962, How to know the eastern land snails. William C. Brown Company, Dubuque, lowa, 214 pp. BUTH, D. G., 1984, The application of electro- phoretic data in systematic studies. Annual Re- view of Ecology and Systematics, 15: 501-522. CLAPP, G. H., 1905, Descriptions of two new forms of Polygyra. Nautilus, 19: 73-74. CLENCH, W. J. 8 A. F. ARCHER, 1933, Two new land shells from the southern Appalachians. Nautilus, 46: 86-91. CLENCH, W. J. & R. D. TURNER, 1962, New names introduced by H. A. Pilsbry in the Mol- lusca and Crustacea. Academy of Natural Sci- ences of Philadelphia, Special Publication No. 4: 1-218. COOKSON, E. A., 1982, The comparative morphol- ogy and histology of the penial apparatus in three pulmonate gastropod species. Bios, 53: 73-84. COX, J. S., JR., 1979, The histology of the repro- ductive system of Mesodon zaletus (Pulmonata: Polygyridae). Master’s thesis, University of Tulsa, Oklahoma, 72 pp. DAVIS, G. M., W. H. HEARD, S.L.H. FULLER & C. HESTERMAN, 1981, Molecular genetics and speciation in Elliptio and its relationships to other taxa of North American Unionidae (Bivalvia). Bi- ological Journal of the Linnean Society, 15: 131— 150. EDWARDS, A. W. F. & L. L. CAVALLI-SFORZA, 1964, The reconstruction of evolutionary trees. Pp. 67-76 in: HEYWOOD, V. H. & J. MC NEILL, eds., Phenetic and phylogenetic classification. Systematics Association, London. EMBERTON, K. C., 1981, Ecological notes on two sympatric, conchologically convergent polygyrid snails in Ohio. Bulletin of the American Malaco- logical Union, 1980: 27-30. EMBERTON, K. C., 1986, The evolution of multiple sympatric homeomorphy among three genera of land snails. Doctoral dissertation, University of Chicago, 780 pp. EMBERTON, K. C., 1988, The genitalic, allozymic, and conchological evolution of the eastern North American Triodopsinae (Gastropoda: Pulmo- nata: Polygyridae). Malacologia, 28: 159—273. EMBERTON, K. C., 1989a, Case 2642: Polygy- ridae Pilsbry, 1894 (Mollusca, Gastropoda): pro- 150 EMBERTON posed precedence over Mesodontidae Tryon, 1866. Bulletin of Zoological Nomenclature, 46: 94-96. EMBERTON, K. C., 1989b, Retraction/extension and measurement error in a land snail: effects on systematic characters. Malacologia, 31: 157- 173. EMBERTON, K. C., in press, Ecology of a shell convergence between subfamilies of polygyrid land snails. Biological Journal of the Linnean So- ciety. FARRIS, J. S., 1970, Methods for computing Wag- ner trees. Systematic Zoology, 19: 83-92. FARRIS, J. S., 1983, The logical basis of phyloge- netic analysis. Pp. 7-36 in: PLATNICK, N. |. & V. A. FUNK, eds., Advances in Cladistics, Vol. 2. Columbia University Press, New York. FARRIS, J. S., 1988, HENNIG86 reference, version 1.5. Published by the author. FARRIS, J. S., 1989, The retention index and the rescaled consistency index. Cladistics, 5: 417- 419. FELSENSTEIN, J., 1982, Numerical methods for inferring evolutionary trees. The Quarterly Re- view of Biology, 57: 379-404. FELSENSTEIN, J., 1983a, Methods for inferring phylogenies: a statistical view. Pp. 315-334 in: FELSENSTEIN, J., ed., Numerical taxonomy: Proceedings of a NATO Advanced Study Insti- tute. Springer-Verlag, Berlin. FELSENSTEIN, J., 1983b, Inferring evolutionary trees from DNA sequences. Pp. 133-150 in: WEIR, B. S., ed., Statistical analysis of DNA se- quence data. Marcel Dekker, New York. FELSENSTEIN, J., 1986, PHYLIP, Version 2.9. Phylogeny inference package. University of Washington, Seattle. FERUSSAC, A.E.J.P.J.F., 1821, Tableau systema- tique de la famille des limagons, Cochleae. Pp. 1-90 in: Tableaux systématiques des animaux mollusques classés en familles naturelles. Ber- trand, Paris. FISHER, R. A., 1930, The genetical theory of nat- ural selection. Clarendon Press, Oxford, 291 pp. FOSTER, T. D., 1937, Productivity of a land snail, Polygyra thyroides [sic] (Say). Ecology, 18: 545— 546. GOODFRIEND, G. A., 1986, Variation in land-snail shell form and size and its causes: a review. Sys- tematic Zoology, 35: 204—223. GREEN, J., 1827, Contributions of the Maclurian Lyceum to the Arts and Sciences, 1: 8. HENNIG, W., 1966, Phylogenetic systematics. Uni- versity of Illinois Press, Urbana, 263 pp. HILLIS, D. M., 1987, Molecular versus morpholog- ical approaches to systematics. Annual Review of Ecology and Systematics, 18: 23-42. HUBRICHT, L., 1954, A new species of Mesodon from the Great Smoky Mountains National Park. Nautilus, 68: 65-66. HUBRICHT, L., 1965, Four new land snails from the southeastern United States. Nautilus, 79: 4— 7: HUBRICHT, L., 1966, Four new land snails. Nau- tilus, 80: 53—56. HUBRICHT, L., 1975, Four new land snails from the eastern United States. Nautilus, 89: 1-4. HUBRICHT, L., 1976, Five new species of land snails from the eastern United States. Malaco- logical Review, 9: 126-130. HUBRICHT, L., 1985, The distributions of the na- tive land mollusks of the eastern United States. Fieldiana, Zoology, New Series, No. 24: 1-191. INTERNATIONAL COMMISSION ON ZOOLOGI- CAL NOMENCLATURE, 1985, International code of zoological nomenclature, third edition. University of California Press, Berkeley, 338 pp. JOHNSON, C. W. & H. A. PILSBRY, 1892, A new American Helix. Nautilus, 6: 8-9. KIM, J. & M. A. BURGMAN, 1988, Accuracy of phy- logenetic-estimation methods under unequal ev- olutionary rates. Evolution, 42: 596-602. KLUGE, A. G., 1989, A concern for evidence and a phylogenetic hypothesis of relationships among Epicrates (Boidae, Serpentes). Systematic Zool- ogy, 38: 7-25. KLUGE, A. G. & G. S. FARRIS, 1969, Quantitative phyletics and the evolution of anurans. System- atic Zoology, 18: 1-32. LEA, 1., 1838, Description of new freshwater and land shells. Transactions of the American Philo- sophical Society, 6: 1-154, plates I-XXIV. LEA, |., 1858, Descriptions of a new Helix and two new Planorbes. Proceedings of the Academy of Natural Sciences of Philadelphia, 10: 41. LEIDY, J., 1851, Special anatomy of the terrestrial Gasteropoda of the United States. Pp. 198-366 in: BINNEY, A., The terrestrial air-breathing mol- lusks of the United States, and the adjacent ter- ritories of North America, Volume I. Charles C. Little & James Brown, Boston. Le QUESNE, W. J., 1982, Compatibility analysis and its applications. Zoological Journal of the Linnean Society, 74: 267-275. LEWIS, J., 1870, Notes on the land shells of east Tennessee. American Journal of Conchology, 6: 188-191. MARTENS, E. VON, 1860, Uber Martyn, the Uni- versal Conchologist, 1789, Bemerkungen zur Nomenklatur. Malakozoologische Blatter, 7: 141-148. MAZE, R. J. & C. JOHNSTONE, 1986, Gastropod intermediate hosts of the meningeal worm Pare- laphostrogylus tenuis in Pennsylvania: observa- tions on their ecology. Canadian Journal of Zo- ology, 64: 185-188. MC CRACKEN, G. F., 1976, The population biology of the white-lipped snail, Triodopsis albolabris. Doctoral dissertation, Cornell University, 136 pp. MC CRACKEN, G. F. & P. F. BRUSSARD, 1980, The population biology of the white-lipped land snail, Triodopsis albolabris: genetic variability. Evolution, 34: 92-104. MEACHUM, C. A., 1981, A manual method for character compatibility analysis. Taxon, 30: 591— 600. MESODONTINI EVOLUTION 151 MICHEVICH, E. F. & M. S. JOHNSON, 1976, Con- gruence between morphological and allozyme data in evolutionary inference and character ev- olution. Systematic Zoology, 25: 260-270. MILES, C. D., 1983, Land snails (Polygyridae) as a source of anti-A agglutinin for typing human blood. American Malacological Bulletin, 1: 97- 98. MILLER, W. B., R. L. REEDER, N. BABRAKZAI & H. L. FAIRBANKS, 1984, List of new and revised recent taxa in the North American terrestrial Mol- lusca since 19 March 1948. Part 1. Tryonia, 11: 1-14. NELSON, G., 1979, Cladistic analysis and synthe- sis: principles and definitions, with a historical note on Adanson's Familles des Plantes (1763-— 1764). Systematic Zoology, 28: 1-21. NELSON, G. & N. I. PLATNICK, 1981, Systematics and biogeography: Cladistics and vicariance. Co- lumbia University Press, New York, 567 pp. NORDSIECK, H., 1987, Revision des Systems der Helicoidea (Gastropoda: Stylommatophora). Ar- chiv für Molluskenkunde, 118: 9-50. PATTERSON, C., 1982, Morphological characters and homology. Pp. 21-74 in JOYSEY, K. A. & A. E. FRIDAY, eds., Problems of phylogenetic reconstruction. Academic Press, New York. PFEIFFER, L., 1848, Nachtrage zu L. Pfeiffer Monographia Helicearum viventium. Zeitschrift für Malakozoologie, 5: 113-123. PILSBRY, H. A., 1893, Polygyra subpalliata, new species. Nautilus, 7: 5-7. PILSBRY, H. A., 1894a, Manual of conchology. Sec- ond series: Pulmonata. Volume IX (Helicidae, Volume 7). Guide to the study of helices. Acad- emy of Natural Sciences of Philadelphia, 366 pp., 71 plates. PILSBRY, H. A., 1894b, Descriptive notes on cer- tain forms of Polygyra. Nautilus, 7: 139-141. PILSBRY, H. A., 1897, Polygyra ferrissi n. sp. Nau- tilus, 11: 92. PILSBRY, H. A., 1899, New southwestern forms of Polygyra. Nautilus, 13: 37-41. PILSBRY, H. A., 1900, Mollusca of the Great Smoky Mountains. Proceedings of the Academy of Natural Sciences of Philadelphia, 52: 110- 150. PILSBRY, H. A., 1930, Anatomy and relationships of some American Helicidae and Polygyridae. Proceedings of the Academy of Natural Sciences of Philadelphia, 82: 303-327. PILSBRY, H. A., 1939, Land Mollusca of North America (north of Mexico), Volume 1, Part 1. Academy of Natural Sciences of Philadelphia, Monographs, Number 3: 1-573. PILSBRY, H. A., 1940, Land Mollusca of North America (north of Mexico), Volume 1, Part 2. Academy of Natural Sciences of Philadelphia, Monographs, Number 3: 575-994. PILSBRY, H. A. & T. D. A. COCKERELL, 1899, A new genus of helices. Nautilus, 12: 107. PILSBRY, H. A. & J. A. FERRISS, 1906, Mollusca of the southwestern states. Il. Proceedings of the Academy of Natural Sciences of Philadelphia, 58: 123-175. PILSBRY, H. A. & J. A. FERRISS, 1907 [as 1906], Mollusca of the Ozarkian fauna. Proceedings of the Academy of Natural Sciences of Philadel- рта, 58: 529-567, plates ХХ-ХХИ. PILSBRY, H. A. & J. A. FERRISS, 1915, The New Mexican expedition of 1914 —Ashmunella. Nau- tilus, 29: 13-16, 29-35, 41-43. PLATNICK, N. 1., 1989, An empirical comparison of microcomputer parsimony programs, Il. Cladis- tics. The International Journal of the Willi Hennig Society, 5: 145-161. PRATT, W.L., 1971, Mesodon leatherwoodi, a new land snail from central Texas. Veliger, 13: 342— 343. PRATT, W. L., 1981, A revision of the land snail genus Polygyra in Texas. Doctoral dissertation, University of Arizona, 136 pp. RAFINESQUE, С. S., 1819, Conchology. Pp. 65- 66 in: MC MURTRIE, H., Sketches of Louisville and its environs. Louisville, Kentucky. [Rafin- esque's authorship of this part is noted on p. 62.] RAFINESQUE, C. S., 1831, Enumeration and ac- count of some remarkable natural objects in the cabinet of Prof. Rafinesque, in Philadelphia; be- ing animals, shells, plants, and fossils, collected by him in North America, between 1816 and 1831. Published by the author, Philadelphia, 8 Pp. RANDOLPH, P. A., 1973, Influence of environmen- tal variability on land snail population properties. Ecology, 54: 933-955. RAUP, D. M., 1966, Geometric analysis of shell coiling: general problems. Journal of Paleontol- ogy, 40: 1178-1190. REHDER, H. A., 1932, Two new Polygyras from northern Arkansas. Nautilus, 45: 129-131. RICHARDSON, B. J., P. R. BAVERSTOCK & M. ADAMS, 1986, Allozyme electrophoresis. A handbook for animal systematics and population studies. Academic Press, Sydney, 410 pp. RICHARDSON, L., 1986, Polygyracea: catalog of species. Part 1, Polygyridae. Tryonia, 13: 1-139. ROHLF, R. J., J. KISPAUGH & D. KIRK, 1972, NT- SYS: Numerical taxonomy system of multivariate statistical programs. Technical Report, State Uni- versity of New York, Stony Brook, New York (June 1974 version). ROHLF, F. J. & M. C. WOOTEN, 1988, Evaluation of the restricted maximum-likelihood method for estimating phylogenetic trees using simulated al- lele-frequency data. Evolution, 42: 581—595. ROSENBERG, G. & K. C. EMBERTON, in press, Comment on the proposed precedence of Po- lygyridae Pilsbry, 1894 over Mesodontidae Tryon, 1866 (Mollusca, Gastropoda). Bulletin of Zoological Nomenclature. RUNHAM, N. W., 1975, Alimentary canal. Pp. 53— 104 in: FRETTER, V. & J. PEAKE, eds., Pulmo- nates, Volume 1, Functional anatomy and phys- iology. Academic Press, London. 152 EMBERTON SAMPSON, F. A., 1889, Description of a new American Helix. Nautilus, 3: 85-86. SAY, T., 1817, Description of new species of land and fresh water shells ofthe United States. Jour- nal of the Academy of Natural Sciences of Phil- adelphia, 1: 123-134. SAY, T., 1818, Account of two new genera, and several new species, of fresh water and land shells. Journal of the Academy of Natural Sci- ences of Philadelphia, 1: 276-284. SAY, T., 1821, Descriptions of univalve shells ofthe United States. Journal of the Academy of Natural Sciences of Philadelphia, 2: 149-160. SCHULTZ, M. G., 1982, Parasitic zoonoses. CRC handbook in zoonoses; Section C. CRC Press, Boca Raton, Florida. SELANDER, R.K., M. H. SMITH, S. Y. YONG, W. E. JOHNSON & J. B. GENTRY, 1971, Biochem- ical polymorphism and systematics in the genus Peromyscus. |. Variation in the old-field mouse (Peromyscus polionotus). University of Texas Studies in Genetics, 4: 49-90. SHAW, C.R. & R. PRASAD, 1970, Starch-gel elec- trophoresis of enzymes—a compilation of reci- pes. Biochemical Genetics, 4: 297-320. SHUTTLEWORTH, R. J., 1852, Diagnosen neuer Mollusken. Mittheilungen der Naturforschenden Gesellschaft in Bern, 1852: 193-207. SIMPSON, C. T., 1888, Notes on some Indian Ter- ritory land and fresh-water shells. Proceedings of the United States National Museum, 11: 449— 454. SOLEM, A., 1955, Studies on Mesodon ferrissi (Gastropoda, Pulmonata) I. General ecology and biometric analysis. Ecology, 36: 83-89. SOLEM, A., 1974, The shell makers: Introducing mollusks. John Wiley & Sons, New York, 289 pp. SOLEM, A., 1976, Comments on eastern North American Polygyridae. Nautilus, 90: 25-36. SOLEM, A., 1978, Classification of the land Mol- lusca. Pp. 49-98, in: FRETTER, V. 8 J. PEAKE, eds., Pulmonates. Volume 2A. Systematics, ev- olution and ecology. Academic Press, London. SWOFFORD, D. L., 1983, PAUP. Phylogenetic analysis using parsimony, Version 2.1. Illinois Natural History Survey, Champaign. SWOFFORD, D. L. & G. J. OLSEN, 1990, Phylog- eny reconstruction. Pp. 411-501, in: HILLIS, О. M. & С. MORITZ, eds., Molecular systematics. Sinauer Associates Inc., Sunderland, Massachu- setts. SWOFFORD, D. L. & R. B. SELANDER, 1981, BIOSYS-1: A computer program for the analysis of allelic variation in genetics. University of Illi- nois, Urbana, 65 pp. TILLIER, S., 1989, Comparative morphology, phy- logeny and classification of the Stylommatophora (Mollusca: Gastropoda: Pulmonata). Malacolo- gia, 30: 1-303. TRYON, G. W., JR., 1866-1867, Monograph of the terrestrial Mollusca of the United States. Ameri- can Journal of Conchology, 2: 218-277 (1866), 306-327; 3: 34-80, 155-181, 298-324 (1867). WAGNER, С. P., 1989, The origin of morphological characters and the biological basis of homology. Evolution, 43: 1157-1171. WATROUS, L. E. & Q. D. WHEELER, 1981, On the out-group method of character analysis. System- atic Zoology, 30: 1-11. WEBB, G. R., 1947a, Studies of the sex-organs of mating polygyrid landsnails. /llinois Academy of Science, Transactions, 40: 218-227. WEBB, G. R., 1947b, The mating-anatomy tech- nique as applied to polygyrid landsnails. Ameri- can Naturalist, 81: 134-147. WEBB, G. R., 1948, The mating of Stenotrema fraternum. Nautilus, 62: 8-12. WEBB, G. R., 1950, The sexology of Polygyra sep- temvolva Say, life history notes, possible utility, and data on Stenotrema (Mollusca, Gastropoda, Pulmonata, Polygyridae). Transactions of the American Microscopical Society, 69: 387-393. WEBB, G. R., 1954a, The life-history and sexual anatomy data on Ashmunella with a revision of the triodopsin snails. Gastropodia, 1: 13-18. WEBB, G. R., 1954b, Pulmonata, Polygyridae, Po- lygyrinae: the sexology and taxonomy of seven species of land snails of the genus Mesodon. Gastropodia, 1: 19-20. WEBB, G. R., 1961, The phylogeny of American land snails with emphasis on the Polygyridae, Arionidae, and Ammonitellidae. Gastropodia, 1: 31-44. WEBB, G. R., 1967, Eratology of three species of Praticolella, and of Polygyra pustula. Nautilus, 80: 133-140. WEBB, G. R., 1968a, Notes on the sexology of the polygyrid landsnail, Mesodon sargentianus (Johnson & Pilsbry). Gastropodia, 1: 62-65. WEBB, G. R., 1968b, Sexological notes on Mes- odon (Ragsdaleorbis) christyi (Bland). Gastropo- dia, 1: 66. WEBB, G. R., 1974, The sexual evolution of the polygyrid snails. Gastropodia, 1: 85-90. WEBB, G. R., 1983, On the sexology of Mesodon kiowaensis (Simpson) Pulmonata, Polygyridae, Polygyrinae. Gastropodia, 2: 19-20. WILEY, E. O., 1981, Phylogenetics. The theory and practice of phylogenetic systematics. John Wiley & Sons, New York, 439 pp. Revised Ms. accepted 26 December 1990 APPENDIX 1. SYSTEMATIC REVIEW FAMILY POLYGYRIDAE PILSBRY, 1894a SUBFAMILY POLYGYRINAE PILSBRY, 1894a Tribe Mesodontini, Tryon (Figs. 1-25, 35-50, 51; Table 2; Fig. 52) Description Genitalia: upper penis (apical, sculptured region of penis in the Triodopsinae and the MESODONTINI EVOLUTION 153 Ashmunellinae) entirely absent; sperm ex- changed externally by deposition on mate’s everted basal penis; basal penis (i.e., the en- tire penis) with two longitudinal, lateral pi- lasters; penis short to extremely long; lateral pilasters thin and varying from simple and uni- form in length to unilaterally absent or thick- ened or extremely high and thin and to en- tirely absent; dorsal surface sometimes bearing an accessory pilaster, a system of thin to thick ridges, or one or more large bulges; apical penis bearing a basin-like chalice, varying from thin- to thick-walled, flexible to rigid, symmetric to asymmetric, con- tinuous to discontinuous with lateral pilasters, and uneverted (a basin or scoop) to everted (a glans) whenever functional; no penial sheath; penial retractor muscle attached to apex of penis; wall of spermathecal duct thin; no appendages, diverticulae or glands associated with genitalia. Shell: diameter 8 to 40 mm; widely umbili- cate to imperforate; high globose-conic to len- ticular; sculpture smooth to matte or coarsely ribbed or hirsute; parietal barrier absent, or present as a simple straight to slightly curved tooth; basal barrier absent, a simple tooth, a long lamella, or a baso-columellar peg; pala- tal tooth absent or a simple tooth. Genus Patera Albers, 1850 (Figs. 10, 19-25, 45-50; Table 2; Fig. 52) Type species: Helix appressa Say, 1821 Etymology: Latin “patera” (saucer). Gender: feminine. Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; chalice generally a simple hood, sometimes spatulate (right wall high and rounded, left wall inconspicuous) or seated atop barrel-shaped pedestal; dorsal structures usually absent (only type spe- cies has dorsal structures: thin, subparallel anastomosing cords); ventral structures ab- sent; peripheral structures absent. Shell: size medium (diameter 13-26 mm); shape usually depressed (height/diameter, 0.4-0.6), rarely globose (height/diameter, 0.7- 0.8); umbilicus narrow and broadly covered, chinked or open; parietal denticle absent, slight or pronounced and blade-like; basal denticle absent or present as long, thin lamella truncated palatally; palatal denticle absent; periostracal hairs or scales absent. Comparisons Shell: Patera comprises all the Mesodontini with flat, hairless shells. West of the Missis- sippi, there are no other Mesodontini with which they could be confused, although they closely resemble several species of Neohelix (Polygyridae: Triodopsinae), one species of which, N. lioderma Pilsbry, 1902, was origi- nally described as a subspecies of Patera in- dianorum. East of the Mississippi, all Patera shells have a characteristic blade-like parietal tooth and a long basal lamella; these fea- tures, together with a flat shape, make them confusingly convergent on the triodopsine ge- nus Xolotrema Rafinesque, 1819 (see Em- berton, 1988), with which they are sometimes sympatric. These shell characters easily dis- tinguish from other Mesodontini all species of Patera with the exceptions of P. perigrapta, which sometimes has a fairly elevated spire, and P. clarki, which is aberrantly domical. These two species can resemble some hair- less, large-toothed members of Fumonelix, particularly F. archeri and some specimens of F. wheatleyi; Patera clarki is also convergent on Mesodon elevatus, which is always at least twice as large and much more heavily calcified. Patera shells are never hirsute— i.e., they lack hair-like or scale-like periostra- cal processes—a feature that readily sepa- rates them from several species with somewhat convergent shell shapes: Inflectar- ¡us smithi, |. verus, |. subpalliatus, Fumonelix wetherbyi and F. jonesiana. Subgenus Ragsdaleorbis Webb, 1954b (Figs. 10b, 45c,d; Table 2; Fig. 52) Type species: Helix pennsylvanicus Green, 1827, by original designation. Etymology: Ragsdale (John P. Ragsdale, Jr., of Indianapolis, Indiana, Webb’s boyhood friend who died in the service of his country) + Latin “orbis” (disk). Gender: masculine. Description Genitalia: left lateral pilaster obsolete; right lateral pilaster thick, height twice the width; chalice moderately deep and seated atop a barrel-shaped, solid pedestal; dorsal, ven- tral and peripheral structures absent. Shell: size medium (diameter 15-20 mm); shape globose (height/diameter, 0.7-0.8); umbilicus imperforate; parietal denticle ab- sent; basal denticle absent; palatal denticle absent; periostracal hairs or scales absent; basal lip of aperture straight and pointing downward to make aperture somewhat trian- gular. 154 EMBERTON Patera pennsylvanica (Green, 1827) (Figs. 10b, 45c,d; Table 2; Fig. 52) (1) Ohio: Pike County (GS-206; FMNH 214703): one live adult, one tissue sampled— dissected (illustrated); electrophoresed. (2) Ohio: Pike County (SC-246; FMNH 214704): one live adult, one tissue sample—illustrated shell. Subgenus Vesperpatera, subgen. nov. (Figs. 22-25, 48a,b, 49, 50; Table 2; Fig. 52) Type species: Polygyra binneyana Pilsbry, 1899. Etymology: Latin “vesper” (west) + “patera” (saucer), because all known members occur west of the Mississippi River. Gender: feminine. Description Genitalia: left lateral pilaster variable to in- conspicuous or absent; right lateral pilaster long, cord-like; chalice a simple hood or spat- ula-shaped, right wall high and rounded; dor- sal, ventral and peripheral structures absent. Shell: size medium (diameter 14-26 mm); shape depressed-globose (height/diameter, 0.5-0.6); umbilicus imperforate to narrow and open; parietal denticle small to absent; basal denticle absent; palatal denticle absent; peri- ostracal hairs or scales absent; aperture a smooth oval. Species Group Patera binneyana (Pilsbry, 1899) Description Genitalia: left lateral pilaster variable to in- conspicuous or absent; right lateral pilaster long, cord-like, usually variable in width; chal- ice a simple hood; dorsal, ventral and periph- eral structures absent. Shell: size medium (diameter 14-26 mm); shape depressed-globose (height/diameter, 0.5-0.6); umbilicus imperforate to narrow and open; parietal denticle absent to weakly present; basal denticle absent; palatal denti- cle absent; periostracal hairs or scales ab- sent; aperture a smooth oval, upper margin of apertural lip abruptly reflected so that the lip is uniform in width. Included species Patera binneyana (Pilsbry, 1899) (Figs. 22a, 23, 24, 48a,b, 49a,b; Table 2; Fig. 52) (1) Oklahoma: LeFlore County (GS-89; FMNH 214625): no live adults, 20 tissue sam- ples—electrophoresed #1, 3, 4, 18, 19. (2) Arkansas: Yell County (GS-95; FMNH 214626): one live adult, ?two tissue sam- ples—dissected #2 (illustrated). (3) (Indian Territory: Sugarloaf Mountain) (ANSP-A2278- F): six live adults—dissected one. (4) Arkan- sas: (Petit Jean) (ANSP A2285): six live adults—dissected one. (5) Arkansas: Yell County (ANSP-A2299-C): four live adults— dissected one. (6) Arkansas: Polk County (FMNH 176008): three live adults— dissected #C (illustrated, both dissected and undis- sected); examined #A, previously dissected by Solem (1976); removed and examined genitalia of #B; illustrated shell #B. (7) Ar- kansas: Polk County (FMNH 176018): one live adult—dissected to examine penial length. (8) Arkansas: Polk County (FMNH 176169): two live adults—dissected one to examine penial length. (9) Arkansas: Scott County (FMNH 176195): one live adult—dis- sected to examine penial length. (10) Arkan- sas: Polk County (Hubricht 31615): two live adults—dissected both to examine penial length; illustrated shell (Hubricht 31621) #A. (11) Arkansas: Polk County (Hubricht 33898) three live adults—dissected one to examine penial length. Variation: There are two distinct penial lengths in P. binneyana. A relatively short penis (Fig. 22a, length 32.4 mm; Pilsbry 1940: Fig. 445D, length 33 mm) occurs in sampled populations #2, 3, 5, 7, 10 and 11, and a slightly longer penis occurs in populations #4 and 8. In populations #6 and 9, however, the penis is extremely long (Figs. 23, 24, length 91.0 mm; Solem 1976: Fig. 7, length >100 mm). W. L. Pratt reported (т litt. to Alan Solem, 8 November 1982) that the short- penis form has radular teeth with “extremely elongate, slender and bladelike [mesocones on the radular teeth], very different from [the mesocones of the long-penis form as illus- trated in Solem 1976, Figs. 18-21].” Thus, P. binneyana is almost certainly two separate species, which for the time being will be re- ferred to as binneyana “short” and bin- neyana “long.” Patera indianorum (Pilsbry, 1899) (Fig. 25; Table 2; Fig. 52) (1) Oklahoma: Atoka County (GS-87; FMNH 214665): two live adults, 14 tissue samples—dissected #1, 5 (illustrated #1); electrophoresed #1, 5, 6, 8. | MESODONTINI EVOLUTION 155 Patera clenchi (Rehder, 1932) (Figs. 25a, 50a,b; Table 2; Fig. 52) (1)? Arkansas: Izard County (GS-97; FMNH 214652): no live adults, one tissue sample—electrophoresed. (2) Arkansas: Izard County (Hubricht 25210): one live adult—dissected (illustrated); illustrated shell. Patera kiowaensis (Simpson, 1888) (Figs. 25c, 50c,d; Table 2; Fig. 52) (1) Oklahoma: Atoka County (GS-84; FMNH 214684): ca. nine live adults, 19 tissue samples—dissected #12, 18 (illustrated #12); electrophoresed #1, 6, 9, 16; illustrated shell #7. Species Group Patera roemeri (Pfeiffer, 1848) Description Genitalia: left lateral pilaster inconspicuous or absent; right lateral pilaster long, cord-like, usually uniform in width; chalice spatula- shaped, right wall high and rounded, left wall reduced; dorsal, ventral and peripheral structures absent. Shell: size medium (diameter 15-24 mm); shape depressed-globose (height/diameter, 0.5-0.6); umbilicus imperforate to narrow and creviced; parietal denticle absent to weakly present; basal denticle absent; palatal denti- cle absent; periostracal hairs or scales ab- sent; aperture a smooth oval, upper margin of apertural lip straight and unreflected. Included species Patera roemeri (Pfeiffer, 1848) (Figs. 22c, 49e,f; Table 2; Fig. 52) (1) Texas: Bastrop County (GS-63; FMNH 214718): ca. ten live adults, 23 tissue sam- ples—dissected #4, 6, 7, 15, 21 (illustrated #21); electrophoresed #2, 3; illustrated shell #4. (2) Texas: Travis County (GS-69; FMNH 214719): ca. ten live adults, 25 tissue sam- ples—electrophoresed #1, 8, 10, 12, 15. Patera leatherwoodi (Pratt, 1971) (Figs. 22b, 49c,d; Table 2; Fig. 52) (1) Texas: Travis County (GS-67; FMNH 214692): one live adult, one tissue sample— dissected (illustrated); electrophoresed; illus- trated shell. (2) Texas: Travis County (GS-68; FMNH 214693): one live adult, one tissue sample—dissected; electrophoresed. Subgenus Patera s. str. (Figs. 19-21, 25, 47, 50c,d; Table 2; Fig. 52) Description Genitalia: left lateral pilaster cord-like, run- ning entire length of penis; right lateral pilas- ter variable; chalice a simple hood; dorsal structures generally absent, rarely present as thin, parallel, anastomosing dorsal cords; ventral and peripheral structures absent. Shell: size medium (diameter 13-27 mm); shape generally depressed (height/diameter, 0.4-0.6), rarely globose (height/diameter, 0.7 mm); umbilicus imperforate, broadly covered; parietal denticle pronounced, blade-like; basal denticle present as long, thin lamella, palatally truncated; palatal denticle absent; periostracal hairs or scales absent. Species Group Patera perigrapta (Pilsbry, 1894) (Figs. 21, 46; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending full length of penis; right lateral pilaster extending half to full length of penis; chalice a simple hood; dorsal, ventral and peripheral struc- tures absent. Shell: size medium (16-23 mm); shape de- pressed (height/diameter, 0.4-0.5); umbilicus imperforate; parietal denticle pronounced, blade-like; basal denticle present as long, thin lamella; palatal denticle absent; periostracal hairs or scales absent; inter-strial microsculp- ture of conspicuous incised spiral lines. Included species Patera perigrapta (Pilsbry, 1894) (Figs. 21b, 46c,d; Table 2; Fig. 52) (1) Tennessee: Blount County (GS-3; FMNH 214705): ca. one live adult, 15 tissue samples—dissected #12; electrophoresed #1, 28,4, 556,78, 9) 10) Wiel 2 42) Tennessee: Blount County (GS-9; FMNH 214707): ca. five live adults, 13 tissue sam- ples— electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13. (3) Alabama: Perry County (GS- 57; FMNH 214708): one live adult, one tissue sample—dissected #A. (4) Arkansas: Craw- ford County (GS-90; 214709): 13 live adults, 13 tissue samples—dissected #1. (5) Arkan- 156 EMBERTON sas: Izard County (GS-98; FMNH 214710) 21 live adults, 21 tissue samples—dissected #8, 20, A (illustrated #8); examined partly everted penes of #7, 9, 18; electrophoresed #5, 7, 8, 9, 21; illustrated shell #3. (6) South Carolina: Lee County (GS-170; FMNH 214712): one live adult—examined everted penis. (7) Alabama: Cleburne County (SC-61; FMNH 214713): two live adults—examined everted penis of one specimen. (8) Alabama: Cleburne County (SC-65; FMNH 214714): one live adult—examined everted penis. (9) Alabama: Cleburne County (SC-66; FMNH 214715): one live adult—examined everted penis. (10) Alabama: DeKalb County (SC-67; ЕММН 214716): four live adults— examined everted penes of four specimens. (11) Ten- nessee: Marion County (SC-97; FMNH 214717): two live adults— examined everted penes of two specimens. Patera panselena Hubricht, 1976 (Figs. 21a, 46a,b; Table 2; Fig. 52) (1) West Virginia: Boone County (GS-142; FMNH 214700): 11 live adults, 11 tissue sam- ples—dissected #2, 6, 9 (illustrated #9); electrophoresed #1, 4, 5, 10, 11; illustrated shell #A. (2) West Virginia: Kanawha County (GS-204; FMNH uncataloged): one live adult, one tissue sample— electrophoresed. Species Group Patera appressa (Say, 1821) (Figs. 19a,c, 20, 47a,b,e,f, 49c,d; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending full length of penis; right lateral pilaster variable; chalice a simple hood; dorsal structures ab- sent or present as parallel, sometimes anas- tomosing cords; ventral and peripheral struc- tures absent. Shell: size medium (15-27 mm); shape de- pressed (height/diameter, 0.4-0.5); umbilicus imperforate; parietal denticle pronounced, blade-like; basal denticle present as long, thin lamella; palatal denticle absent; periostracal hairs or scales absent; inter-strial microsculp- ture generally smooth or pustular. Included species Patera appressa (Say, 1821) (Figs. 19c, 47e,f; Table 2; Fig. 52) (1) Kentucky: McCreary County (GS-12; FMNH 214619): 12 live adults, 12 tissue sam- ples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10. (2) Kentucky: Pulaski County (GS-13; FMNH uncat.): unknown number of live adults, 30 tissue samples—electrophoresed #11, 12, 13, 14; 15, 16, 17, 18) 1920722423 25. (3) Tennessee: Franklin County (GS-104; FMNH 214620): six live adults, eight tissue samples—dissected #1, 5, 7 (illustrated #1); illustrated shell #2. (4) Tennessee: Overton County (GS-111; FMNH 214621); ca. three live adults, nine tissue samples—elec- trophoresed #2, 3, 4, 6, 7. (5) West Virginia: Summers County (GS-141; FMNH 214622): ten live adults, ten tissue samples—dis- sected #2, 4, 6, 8. Patera laevior (Pilsbry, 1940) (Figs. 20, 48c,d; Table 2; Fig. 52) (1) Kentucky: Hancock County (H-22; FMNH 214685): three live adults—dissected #1, 3, A (illustrated #1, 3); illustrated shell #3. (2) Indiana: Jefferson County (GS-14; FMNH 214687): no live adults, 16 tissue sam- ples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10. (3) Kentucky: Fayette County (GS-112; FMNH 214689): no live adults, 11 tissue sam- ples—electrophoresed #1, 4, 6. (4) Ken- tucky: Edmonson County (GS-125; FMNH 214690): five live adults, five tissue sam- ples—dissected #1, 4, 5; electrophoresed #1, 2, 3, 4. (5) Illinois: Hardin County (SC- 217; FMNH 214691): two live adults—exam- ined everted penes of three specimens. Patera sargentiana (Johnson & Pilsbry, 1892) (Figs. 19a, 47a,b; Table 2; Fig. 52) (1) Alabama: Madison County (GS-20; FMNH 214728): no live adults, 20 tissue sam- ples— electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10. (2) Alabama: Madison County [GS-101 (= GS-20); FMNH 214729]: 12 live adults, 12 tissue samples—dissected #1, 9, 12 (illus- trated #1); electrophoresed #2, 5, 8, 9; illus- trated shell #2. Species Group Patera clarki (Lea, 1858) (Figs. 19b, 47c,d; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending full length of penis; right lateral pilaster extending entire length of penis; chalice a simple hood; dorsal, ventral and peripheral structures ab- sent. MESODONTINI EVOLUTION 157 Shell: size medium (13-18 mm); shape glo- bose (height/diameter, 0.7); umbilicus imper- forate; parietal denticle pronounced, blade- like; basal denticle present as long, thin, truncate lamella; palatal denticle absent; pe- riostracal hairs or scales absent; inter-strial microsculpture generally smooth, base sometimes malleate. Included species Patera clarki (Lea, 1858) (Figs. 19b, 47c,d; Table 2; Fig. 52) (1) Tennessee: Sevier County (GS-6; FMNH 214633): no live adults, three tissue samples—electrophoresed #1, 2, 3. (2) North Carolina: Haywood County (GS-10; FMNH 214634): five live adults, 14 tissue samples—dissected #1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14. (3) North Carolina: Swain County (GS-1; FMNH 214632): ca. seven live adults, 11 tissue samples—dissected #5, 6, 7; electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, ПО ТТ, Genus /nflectarius Pilsbry, 1940 (Figs. 9a,c, 10a, 11a, 12-14, 35a,b, 40a,b,e,f, 43,44c,d, 45a,b; Table 2; Fig. 52) Type species: Helix inflecta Say, 1821, by original designation. Etymology: Latin “inflecta” (name of type species) + “-arius” (Latinizing suffix). Gender: masculine. Definition Genitalia: left lateral pilaster variable; right lateral pilaster variable; third pilaster present or absent; chalice generally a simple hood, rarely (only in I. ferrissi) shaped like a thick spoon; dorsal structures absent (except for third pilaster, when present); ventral and pe- ripheral structures absent. Shell: size generally small, rarely medium (diameter 8-25 mm); shape subglobose to globose (height/diameter, 0.5-0.7); umbilicus narrow, broadly or narrowly covered, rarely creviced; parietal denticle generally pro- nounced, small or absent; basal denticle present or absent, rarely as a trace lamella; palatal denticle pronounced, reduced or ab- sent, flush with the aperture or, rarely, re- cessed; periostracal scales present (three types) or absent; body whorl with or without crest before preapertural deflection. Comparisons Shell. Inflectarius is the only genus in the Mesodontini with a palatal apertural tooth and the only member of the Mesodontini with scale-like periostracal hairs (the hirsute mem- bers of Fumonelix have thin, rounded hairs that are not scale-like, except for F. weth- erbyi, which species can be confusing). Only three species of Inflectarius lack hairs: /. fer- rissi, I. downieanus and I. kalmianus. Of these, the shell of /. ferrissi is distinctively un- like that of any of the other Mesodontini. On the other hand, the shells of /. downieanus and /. kalmianus are very much convergent on those of Mesodon (Akromesodon) and on those of Fumonelix wheatleyi; they can be distinguished by their smaller size and by the unique structure of the umbilicus, which is ei- ther barely creviced (/. kalmianus) or deeply imperforate with the apertural lip plunging deeply into the umbilical pit (/. downieanus). The shells of many species of the genus Tri- odopsis Rafinesque, 1819 (Emberton, 1988) resemble those of the tridentate /nflectarius, but shells of Triodopsis are always umbilicate and never hirsute, whereas shells of the tri- dentate species of Inflectarius are always im- perforate and always hairy. The shell of /. fer- rissi is very similar to that of the triodopsine Neohelix dentifera (Binney, 1837) (Emberton, in press). Subgenus Hubrichtius, subgen. nov. (Figs. 10a, 11a, 35a,b, 45a,b; Table 2; Fig. 52) Type species: Mesodon kalmianus Hubricht, 1965. Etymology: Hubricht (Mr. Leslie Hubricht of Meridian, Mississippi, an expert on the land snails of eastern North America) + Latin “-ius” (Latinizing suffix). Gender: masculine. Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; chalice a simple hood; dorsal, ventral and peripheral struc- tures absent. Shell: size small (diameter 9-15 mm); shape globose (height/diameter, 0.7); umbili- Cus narrow, narrowly covered, sometimes creviced; parietal denticle absent; basal den- ticle absent; palatal denticle absent; peri- ostracal hairs or scales absent; body whorl lacking crest before preapertural deflection. Included species Inflectarius downieanus (Bland, 1861) (Figs. 10a, 45a,b; Fig. 52) (1) Alabama: DeKalb County (Hubricht 30825): unknown number of live adults—dis- 158 EMBERTON sected #A, B (illustrated #B); illustrated shell #A. Inflectarius kalmianus Hubricht, 1965 (Figs. 11a, 35a,b; Table 2; Fig. 52) (1) Kentucky: Knox County [GS-116; FMNH 214683 (the specimens in this lot are unusually numbered from having been con- fused in the field with microsympatric Mes- odon clausus)]: three live adults, 3 tissue samples—dissected #13, A (illustrated #13); electrophoresed #18; illustrated shell #18. (2) Kentucky: Knox County [GS-188 (= GS- 116); FMNH 214682]: five live adults, five tis- sue samples—dissected #2, A; electro- phoresed #1, 2, 3, 4. Subgenus /nflectarius s. str. (Figs. 9a,c, 12, 13b, 14, 40a,b,e,f, 43, 44c,d; Table 2; Fig. 52) Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; third pilaster present or absent; chalice generally a simple hood, rarely (only in /. ferrissi) shaped like a thick spoon; dorsal structures absent (except for third pilaster, when present); ventral and pe- ripheral structures absent. Shell: size generally small, rarely medium (diameter 8-25 mm); shape subglobose (height/diameter, 0.5-0.6); umbilicus narrow, broadly covered; parietal denticle generally long and pronounced, rarely (only in /. ferrissi) short and small; basal denticle generally a tooth, rarely absent or as a trace lamella; pal- atal denticle pronounced, reduced or absent, flush with the aperture or, rarely, recessed; periostracal hairs or scales present (three types) or, rarely (only in I. ferrissi) absent; body whorl with or without crest before preap- ertural deflection. Remarks. The highly derived shell of /. fer- rissi is correlated with its extremely special- ized ecology (Emberton, in press). Species Group /nflectarius edentatus (Sampson, 1889) (Figs. 12b,c, 43c,f; Table 2; Fig. 52) Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; chalice a simple hood; dorsal, ventral and peripheral struc- tures absent. Shell: size small (diameter 13-14 mm); shape subglobose (height/diameter, 0.5-0.6); umbilicus narrow, broadly covered; parietal denticle long and pronounced; basal denticle a slight bump; palatal denticle a slight bump, flush with aperture; periostracal hairs or scales present, low, rounded, and blunt; body whorl with pronounced crest before preaper- tural deflection. Included species Inflectarius edentatus (Sampson, 1889) (Figs. 12c, 43e,f; Table 2; Fig. 52) (1) Arkansas: Crawford County (GS-90; FMNH 214653): 12 live adults, 12 tissue sam- ples—dissected #2, 7, 9 (illustrated #2); electrophoresed #1, 6, 10; illustrated shell #A. (2) Arkansas: Crawford County (GS-91; FMNH 214654): ca. three live adults, eight tissue samples— electrophoresed #2, 7, 9(?) Inflectarius magazinensis (Pilsbry & Ferriss, 1907) (Figs. 12b, 43c,d; Table 2; Fig. 52) (1) Arkansas: Yell County (GS-95; FMNH 214695): ca. 20 live adults, 26 tissue sam- ples—dissected #5, 6, 14 (illustrated #5); electrophoresed #2, 4, 12, 18; illustrated shell #17. Remarks. Clench & Turner (1962) gave the date of publication for this species as 1907. Their judgement has been followed here. Species Group /nflectarius smithi (Clapp, 1905) (Figs. 9c, 40e,f; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending less -than half length of penis; right lateral pilaster extending full length of penis, expanding ba- sally as massive bulge; chalice a simple hood; dorsal, ventral and peripheral struc- tures absent. Shell: size medium (diameter 14-17 mm); shape subglobose (height/diameter, 0.6); um- bilicus narrow, broadly covered; parietal den- ticle long and pronounced; basal denticle present as long, thin lamella truncated at pal- atal end; palatal denticle pronounced, broad, flush with apertural lip; periostracal hairs or scales present, very dense, long and sharply pointed; body whorl lacking crest before preapertural deflection. Included species MESODONTINI EVOLUTION 159 Inflectarius smithi (Clapp, 1905) (Figs. 9c, 40e,f; Table 2; Fig. 52) (1) Alabama: Madison County (GS-20; FMNH 214736): two live adults, 11 tissue samples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10; illustrated shell #1. (2) Alabama: Madison County [GS-101 (= GS- 20); FMNH 214737]: five live adults, five tissue sam- ples—dissected #1, 2, 4 (illustrated #4); electrophoresed #2, 3, 4. (3) Tennessee: Franklin County (GS-104; FMNH 214738): seven live adults, seven tissue samples— electrophoresed #1, 2, 7. Species Group /nflectarius inflectus (Say, 1821) (Figs. 9a, 12a, 13b, 14, 40a,b, 43a,b, 44c,d; Table 2; Fig. 52) Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; third pilaster some- times present; chalice a simple hood; dorsal, ventrai and peripheral structures absent. Shell: size small (diameter 8-16 mm); shape subglobose (height/diameter, 0.5-0.6); umbilicus narrow, broadly covered; parietal denticle long and pronounced; basal denticle a pronounced tooth or rarely absent; palatal denticle a pronounced tooth, either flush with aperture or moderately recessed, or rarely absent; periostracal hairs or scales present, low, bearing central points; body whorl bear- ing slight crest before preapertural deflection. Included species Inflectarius verus Hubricht, 1954 (Fig. 14; Table 2; Fig. 52) (1) North Carolina: Haywood County (GS- 10; FMNH 214756): 14 live adults, 14 tissue samples—dissected #3, 4, 8; electro- phoresed #1, 2; 3,.4, 5, 6, 7, 8, 9, 10, 11, 12, 13: Discussion. The status and name of this species are puzzling. Hubricht (1954b) did not illustrate the holotype. The holotype (USNM 607137) has not been examined, but three of the paratypes (ANSP 191211) are definitely Inflectarius subpalliatus, under which Hu- bricht (1985) eventually synonomized |. verus. Recently-collected specimens from the type locality of /. verus (FMNH 214756), the shells of which have not been illustrated, dif- fer substantially in penial morphology (Fig. 14) from /. subpalliatus (Fig. 13a). These pro- visionally have been called /. verus. Inflectarius inflectus (Say, 1821) (Figs. 9a, 40a,b; Table 2; Fig. 52) (1) Kentucky: Henderson County (GS-16; FMNH 214666): ca. 20 live adults, 30 tissue samples—dissected #21, 27, 29; electro- phoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15; 16, 17. 18. 19 20. 25.530" (2) Alabama: Clarke County (GS-53; FMNH 214667): 16 live adults, 16 tissue samples— electrophoresed #1, 2, 3, 4, 11, 16. (3) Okla- homa: LeFlore County (GS-89; FMNH 214668): 30 live adults, 30 tissue samples— electrophoresed #9, 11. (4) Arkansas: Yell County (GS-95; FMNH 214670): one live adult, one tissue sample—dissected. (5) Ala- bama: Madison County (SC-26; FMNH 214674): one live adult—dissected. (6) Ten- nessee: Blount County (SC-130; FMNH 214672): 12 live adults, 12 tissue samples— dissected #2, 3, 8 (illustrated #2); illustrated shell #1. Inflectarius rugeli (Shuttleworth, 1852) (Figs. 13b, 44c,d; Table 2; Fig. 52) (1) Tennessee: Swain County (GS-3; FMNH 214720): 11 live adults, 11 tissue sam- ples—dissected #5, 10, A; electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11. (2) Tennes- see: Polk County (GS-106; FMNH uncat.): unknown number of live adults, nine tissue samples— electrophoresed #1, 3, 4, 5, 6. (3) Alabama: Cleburne County (GS-180; FMNH 214725): ten live adults, ten tissue samples— electrophoresed #1, 2, 3, 6, 8. (4) Tennes- see: Blount County (SC-130; FMNH 214726): eight live adults, eight tissue samples—dis- sected #2, 3, 4 (illustrated #2); illustrated shell #8. Inflectarius approximans (Clapp, 1905) (Figs. 12a, 43a,b; Table 2; Fig. 52) (1) Alabama: Perry County (GS-57; FMNH 214623): one live adult, one tissue sample— dissected (illustrated); electrophoresed. (2) Alabama: Perry County (Hubricht 23497): un- known number of live adults—dissected #A; illustrated shell #A. Species Group Inflectarius ferrissi (Pilsbry, 1897) (Figs. 13a,c, 44a,b,e,f; Table 2; Fig. 52) Description Genitalia: left lateral pilaster variable; right 160 EMBERTON lateral pilaster variable; third pilaster present; chalice variable; dorsal ventral and peripheral structures absent. Shell: size medium (diameter 13-25 mm); shape subglobose (height/diameter, 0.5-0.6); umbilicus narrow, broadly covered; parietal denticle small and short to pronounced and long; basal denticle absent or present as an inconspicuous long, thin lamella, truncated palatally; palatal denticle absent; periostracal hairs or scales present, low and bearing cen- tral points, to entirely absent; body whorl lack- ing crest before preapertural deflection. Included species Inflectarius subpalliatus (Pilsbry, 1893) (Figs. 13a, 44a,b; Table 2; Fig. 52) (1) Tennessee: Carter County (GS-11; FMNH 214739): one live adult, one tissue sample—electrophoresed. (2) North Caro- lina: Avery County (GS-153; FMNH 214740): six live adults, 11 tissue samples—dissected #1, 2, 3 (illustrated #2); electrophoresed #2, 3, 4, 5, 8; illustrated shell #8. Inflectarius ferrissi (Pilsbry, 1897) (Figs. 13c, 44e,f; Table 2; Fig. 52) (1) North Carolina: Swain County (GS-1; FMNH 214657): ca. seven live adults, ten tis- sue samples— electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10. (2) Tennessee-North Carolina: Blount-Swain Counties (GS-2; FMNH 214658): ca. two live adults, 11 tissue sam- ples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11. (3) Tennessee-North Carolina: Blount-Swain Counties (GS-5; FMNH 214659): three live adults, four tissue sam- ples—dissected #1, 3, 4 (illustrated #3); electrophoresed #1, 2, 3, 4. (4) Tennessee- North Carolina: Blount-Swain Counties [SC- 144 (= GS-5); FMNH 214661]: ca. six live adults—illustrated shell #A. (5) Tennessee: Sevier County (SC-216; FMNH 214664): 13 live adults, 13 tissue samples—dissected #2, 3, 6. Genus Fumonelix gen. nov. (Figs. 7, 8, 9b, 40c,d, 41, 42; Table 2; Fig. 52) Type species: Helix wheatleyi Bland, 1860. Etymology: Latin “fumosa” (Smoky) + “mon- tana” (of mountains) + “helix” (snail), hence “snail of the Smoky Mountains.” Gender: feminine. Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; chalice a thick- walled, hooded cup; dorsal surface with or without a single or at most a pair of bulges; ventral surface with or without a bulge in mid- line; peripheral structures absent. Shell: size very small to medium (diameter 8-23 mm); shape subglobose to globose (height/diameter, 0.6-0.7); umbilicus narrow, broadly to narrowly covered; parietal denticle pronounced to absent; basal denticle a faint trace of long, thin lamella, or entirely absent; palatal denticle absent; periostracum with or without thin, rounded hairs or broad, obtusely pointed scales. Comparisons. See the Comparisons under Patera and In- flectarius. Species Group Fumonelix wetherbyi (Bland, 1874) (Figs. 9b, 40c,d; Table 2; Fig. 52) Description Genitalia: left lateral pilaster a humped ridge about half length of penis; right lat- eral pilaster full-length, extremely high- standing, and rolled over in inverted pe- nis; chalice a deep, thin-walled, cylindrical cup; dorsal, ventral and peripheral structures absent. Shell: size medium (diameter 17-18 mm); shape subglobose (height/diameter, 0.6); um- bilicus narrow, broadly covered; parietal den- ticle pronounced; basal denticle a faint trace of long, thin lamella; palatal denticle absent; periostracal scales present, small, low, broad and obtusely pointed. Included species Fumonelix wetherbyi (Bland, 1874) (Figs. 9b, 40c,d; Table 2; Fig. 52) (1) Kentucky: McCreary County (GS-115; FMNH 214757): eight live adults, 20 tissue samples— dissected #1, 2, 7, 20 (illustrated #20); electrophoresed #2, 3, 4, 5, 8, 11; il- lustrated shell #4. Species Group Fumonelix christyi (Bland, 1860) (Figs. 8a, 41; Table 2; Fig. 52) Description : Genitalia: left lateral pilaster short and thin; MESODONTINI EVOLUTION 161 right lateral pilaster long and thin; chalice a thick-walled, hooded cup; dorsal surface with thick, long bulge on right side; ventral and pe- ripheral structures absent. Shell: size very small (diameter 8-9 mm); shape subglobose to globose (height/diame- ter, 0.6-0.7); umbilicus narrow, broadly cov- ered; parietal denticle pronounced; basal denticle a faint trace of long, thin lamella; pal- atal denticle absent; periostracal hairs or scales absent. Included species Fumonelix christyi (Bland, 1860) (Figs. 8a, 41; Table 2; Fig. 52) (1) North Carolina: Burke County (GS-161; FMNH 214631): ten live adults, 11 tissue samples—dissected #2, 5, 6 (illustrated #6); electrophoresed #1, 2, 3, 4, 6, 8, 9, 10, 11; illustrated shell #10. Species Group Fumonelix wheatleyi (Bland, 1860) (Figs. 8b-d, 42; Table 2; Fig. 52) Description Genitalia: left lateral pilaster variable; right lateral pilaster variable; chalice a thick-walled, hooded cup; dorsal surface plain, with periph- eral-apical bulges, or with basal bulge formed by fusion of left and right lateral pilasters; mid-ventral surface with conspicuous bulge; peripheral structures absent. Shell: size small to medium (diameter 13- 23 mm); shape subglobose to globose (height/diameter, 0.6-0.7); umbilicus narrow, broadly to narrowly covered; parietal denticle pronounced to absent; basal denticle a faint trace of long, thin lamella, or entirely absent; palatal denticle absent; periostracal hairs sometimes present, thin, and rounded, or absent. Discussion. This group requires revision as soon as possible, because it includes three described species of potentially endangered status—F. jonesiana, F. archeri and F. orestes—and be- cause F. wheatleyi (q.v.) seemingly includes at least one other cryptic species, which also might be rare. F. G. Thompson’s un- published study assembled much alcohol- preserved material now housed in the Florida State Museum. Included species Fumonelix wheatleyi (Bland, 1860) (Figs. 8c, 42a,b, 51; Table 2; Fig. 52) (1) Tennessee: Sevier County (GS-6; FMNH uncat.): unknown number of live adults, unknown number of tissue samples— dissected #1, 3, 5 (illustrated #5); electro- phoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10; illus- trated shell #3. (2) North Carolina: Haywood County (GS-10; FMNH uncat.): unknown number of live adults, 21 tissue samples— dissected #2, 9, 12; electrophoresed #1, 2, 3, 4, 5:6; 7. 8, 92.107 11, 18,°212(8) (North Carolina: Avery County (GS-153; ЕММН 214767): ten live adults, ten tissue samples— dissected #4, 7, 10; electrophoresed #1, 3, 4, 5, 6, 8. (4) Tennessee-North Carolina: Sevier-Swain Counties (SC-144; FMNH 214762): two live adults, two tissue sam- ples—dissected #1, 2. (5) North Carolina: Swain County (SC-192; FMNH 214763): one live adult, three tissue samples—dissected #3. (6) North Carolina: Macon County (SC- 202; FMNH 214766): two live adults, two tis- sue samples—dissected #1, 2. (7) North Carolina: Macon County (SC-212; FMNH 214768): two live adults, two tissue sam- ples—dissected #2. Fumonelix jonesiana (Archer, 1938) (Figs. 8b, 42c,d; Table 2; Fig. 52) (1) North Carolina: Swain County (GS-1; FMNH 214678): two live adults, ten tissue samples—dissected #2, 10. (2) Tennessee: Sevier County (SC-155; FMNH 214679): ten live adults, ten tissue samples—dissected #4, 5, 9 (illustrated #5); illustrated shell #3. Fumonelix orestes (Hubricht, 1975) (Figs. 8d, 42e,f; Table 2; Fig. 52) (1) North Carolina: Haywood County (GS- 86; FMNH 214698): two live adults, nine tis- sue samples—dissected #2, 4 (illustrated #4); electrophoresed #1, 3, 4, 5. (2) North Carolina: Haywood County (Hubricht 40465): unknown number of live adults—illustrated shell #A. Fumonelix archeri (Pilsbry, 1940) (Fig. 7; Fig. 52) (1) Tennessee: Polk County (SC-279; FMNH uncat.): three live adults, two tissue samples—dissected #A, B. 162 EMBERTON Genus Mesodon Férussac, 1821 (Figs. 1-6, 11b,c, 15-18, 35c-f, 36-39; Table 2: Fig: 52) Type species: Helix thyroidus [sic] Say, 1817, by monotypy (see Taxonomic History). Etymology: Greek “mesos” (middie) + “odon” (tooth). Gender: masculine. Description Genitalia: left lateral pilaster rounded or cord-like, variable in length; right lateral pilas- ter rounded or cord-like, variable in length or absent; chalice variable, left wall higher than right; dorsal cords or ridges present, with or without enlarging into basal bulges, or absent; ventral structures generally absent, rarely present as a mid-ventral pouch; periph- eral structures generally absent, rarely present as a shoulder. Comparisons. East of the Mississippi River, Mesodon contains all of the globose Mesodontini ex- cept for Patera pennsylvanica, P. clarki, In- flectarius downieanus, |. kalmianus, Fu- monelix wheatleyi and F. orestes, and it contains all of the broadly umbilicate Mesod- ontini. Only four species of Mesodon occur west of the Mississippi (M. zaletus, M. eleva- tus, M. clausus and M. thyroidus; (Fig. 51). Of these, only M. thyroidus might be confused with any other western member of the Mes- odontini, primarily Patera roemeri, with which it is sometimes sympatric; but Mesodon thy- roidus is usually easily distinguished by its higher spire and duller surface. Several of the large, globose, toothless species of Mes- odon—especially M. normalis, the toothless morph of M. zaletus and the imperforate and toothless morph of M. thyroidus (subspecies bucculentus)—are very readily confused with species of the triodopsine genus Neohelix (Pilsbry, 1940; Solem, 1976; Emberton, 1988). Subgenus Appalachina Pilsbry, 1940 (Figs. 18a,b, 39; Table 2; Fig. 52) Type species: Polygyra sayana Pilsbry, in Pilsbry & Ferriss, 1906, by original designa- tion. Etymology: Appalachia (the major mountain- ous region of eastern North America) + Latin “ina” (Latinizing suffix). Gender: feminine. Description Genitalia: left lateral pilaster rounded, vari- able in length; right lateral pilaster present or absent; chalice somewhat spatulate, the left wall high and even; dorsal, ventral and pe- ripheral structures absent. Shell: size large (diameter 19-40 mm); shape subglobose (height/diameter, 0.6); umbilicus wide, open; parietal denticle small or absent; basal denticle a baso- columellar peg, or absent; palatal denticle ab- sent; periostracal hairs or scales absent. Included species Mesodon sayanus (Pilsbry, in Pilsbry & Ferriss, 1906) (Figs. 18a, 39c,d; Table 2; Fig. 52) (1) Kentucky: Harlan County (GS-122; FMNH 214732): no live adults, ten tissue samples—electrophoresed #2, 3, 4, 5, 7, 8, 9, 10. (2) West Virginia: Preston County (GS- 130; FMNH 214734): three live adults, seven tissue samples—dissected #1, 4, 6 (illus- trated #6); electrophoresed #3, 5; illustrated shell #7. Mesodon chilhoweensis (Lewis, 1870) (Figs. 18b, 39a,b; Table 2; Fig. 52) (1) Tennessee: Blount County (GS-3; FMNH 214627): ca. two live adults, 20 tissue samples— electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 20. (2) Tennessee: Blount County (GS-9; FMNH 214628): ca. one live adult, ten tissue samples —electro- phoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10. (3) North Carolina: Graham County (SC-263; FMNH 214630): two live adults—dissected #A. (4) Tennessee: Sevier County (Hubricht 30943): unknown number of live adults—dis- sected #A (illustrated); illustrated shell #A. Subgenus Aphalogona Webb, 1954b (Figs. 1-6, 16a, 17, 37a,b, 38; Table 2; Fig. 52) Type species: Helix elevata Say, 1821, by original designation. Etymology: Greek “a-” (without) + “phalos” (shining) [error for “lophos” (crest)] + “gona” [error for “gone” (that which produces seed), incorrectly translated as “sex organ”], thus, by original intent, “penis without a chalice” (Webb, 1954b: 19). Gender: feminine. Description Genitalia: left lateral pilaster cord-like, ex- tending one-third to full length of penis; right MESODONTINI EVOLUTION 163 lateral pilaster cord-like, extending two-thirds to full length of penis; chalice either a V- shaped notch defined by massive walls, and which everts whenever penis everts (Webb, 1954b: plate 12, figs. 8, 14-16), or a thin- walled deep, scalloped-edged cylinder atop a solid, cylindrical pedestral; dorsal wall bearing four to ten cord-like, subparallel, anastomos- ing ridges, running longitudinally to 30 de- grees obliquely; mid-ventral pouch present or absent; peripheral structures absent. Shell: size medium to large (diameter 15-31 mm); shape globose (height/diameter, 0.7- 0.8); umbilicus narrow, broadly covered; pari- etal denticle pronounced to absent; basal denticle a long, thin lamella truncated pala- tally; palatal denticle absent; periostracal hairs or scales absent. Remarks. Webb originally described Aphalogona as a section, which, according to the ICZN (1985: article 10e), is nomenclaturally equivalent to a subgenus. Included species Mesodon elevatus (Say, 1821) (Figs. 3f-j, 4f-j, 5d-f, 6d-f, 38c,d; Table 2; Fig. 52) (1) Tennessee: Franklin County (GS-104; FMNH 214655): 35 live adults, 35 tissue sam- ples—dissected and illustrated #11, 14, 16, 29, 39; electrophoresed #1, 3, 4, 5, 7, 8, 10, 13516. 17. 98. 19, 23, 26, 29. 33, 35, 38; illustrated shell #33. (2) Indiana: Owen County (FMNH 214656): ca. ten live adults— dissected and illustrated #A, B, C. Variation. Variation in penial morphology (Figs. 3f-j, 4f-j, 5d-f, 6d-f) was discussed above. Mesodon zaletus (Binney, 1837) (Figs. 3a-e, 4a-e, 5a-c, 6a-c, 17, 38a,b; Table 2; Fig. 52) (1) Tennessee: Blount County (GS-9; FMNH 214771): 17 live adults, 17 tissue sam- ples—dissected #2; Electrophoresed #1, 2, 314.15: 6, 7. 8,9, 10:11, 12,139; 14. 15, 16, 17. (2) Arkansas: Crawford County (GS-90; FMNH 214787): eight live adults, nine tissue samples (#1, 2, 5-11)—dissected #8; elec- trophoresed #1, 2, 5, 6, 7, 8, 9, 10, 11. (3) Tennessee: Franklin County (GS-104; FMNH 214774): six live adults, six tissue samples (#2, 6, 11, 12, 22, 27)—dissected and illus- trated #2, 6, 12, 22, 27; illustrated undis- sected penial tubes of #2, 6, 12, 22, 27; illus- trated shell #27. (4) Indiana: Owen County (FMNH 214785): ca. ten live adults— dissected and illustrated #A, B, D; illustrated undissected penes of #A, B, D. Variation. Variation in penial morphology was illustrated (Figs. 3a-e, 4a-e, 5a-c, 6a-c) and discussed previously. Because M. zale- tus was used as the control for all electro- phoresis, it has been tested for the ontoge- netic, temporal and geographic stability of alleles (Emberton, 1986; in preparation). Mesodon mitchellianus (Lea, 1838) (Figs. 16a, 37a,b; Table 2; Fig. 52) (1) North Carolina: Henderson County (GS- 154; FMNH 214696): eight live adults, eight tissue samples—dissected #1, 5, 6; electro- phoresed #1, 2, 4, 5, 6. (2) Ohio: Brown County (Hubricht 19406): unknown number of live adults—dissected #A, B, C (illustrated #B); illustrated shell #A. Remarks. The North Carolina population represents a significant southward extension of the range of this species presented in Hu- bricht (1985). This change is incorporated into the range map (Fig. 51). Subgenus Akromesodon, subgen. nov. (Figs..1;,.25 11b;c; 15¢35e-f; 36e;f; Table 2; Fig. 52) Type species: Polygyra andrewsae normalis Pilsbry, 1900. Etymology: Greek “akron” (summit, peak) + Mesodon (the generic name), because this group not only occupies some of the highest mountain peaks (e.g., Roan Mountain, Vir- ginia, and Clingman’s Dome, North Carolina- Tennessee) but also attains the greatest shell size of the entire tribe. Gender: masculine. Description Genitalia: left lateral pilaster cord-like, ex- tending full length of penis; right lateral pilas- ter present or absent; chalice a deep, thin- walled scoop, with left wall much higher than right; dorsal wall bearing eight to 12 cord-like, subparallel, anastomosing ridges, running longitudinally to 30 degrees obliquely, many of which are contiguous with one or both lateral pilasters, and many of which en- large basally to form a network of large basal bulges; ventral and peripheral struc- tures absent. Shell: size large (diameter 21-40 mm); 164 EMBERTON shape globose (height/diameter, 0.7-0.8); umbilicus very narrow, broadly covered; pari- etal denticle generally absent, rarely present as a trace; basal denticle absent; palatal den- ticle absent; periostracal hairs or scales ab- sent. Included species Mesodon andrewsae W. G. Binney, 1879 (Figs. 2, 11b, 35c,d; Table 2; Fig. 52) (1) Tennessee: Carter County (GS-11; FMNH 214618): 12 live adults, 22 tissue sam- ples—dissected #4, 8 (illustrated #8); exam- ined everted penes of #2, 3 (illustrated #2); illustrated shell #1. Mesodon normalis (Pilsbry, 1900) (Figs. 1, 11c, 35e,f; Table 2; Fig. 52) (1) Tennessee: Blount County (GS-3; FMNH 214979): ca. 25 live adults, 22 tissue samples—electrophoresed #2, 3, 4, 5, 8, 9, 11,12; 13. 142 15. 16 1/7 18; 19, 20. 21.23, 26, 30, 32, 34. (2) North Carolina: Watauga County (GS-200; FMNH 214966): two live adults, four tissue samples—electrophor- esed #1, 2, 3, 4. (3) Tennessee: Cocke County (SC-149; FMNH 214970) 12 live adults, 12 tissue samples—dissected #3, 7, 11. (4) Tennessee: Blount County (SC-154; FMNH 214980): 14 live adults, 14 tissue sam- ples—dissected #8, 10, 13. (5) North Caro- lina: Swain County (SC-158; FMNH 214977): four live adults, four tissue samples— dissected #4 (illustrated); illustrated shell #2. (6) North Carolina: Macon County (SC-184; FMNH 214987): ten live adults, ten tissue samples—dissected #5, 6, 10. (7) North Carolina: Macon County (SC-204; FMNH 214984): nine live adults, nine tissue sam- ples—dissected #2, 3. Variation. The dissected populations differ considerably in the degree to which the chal- ice walls are flared. Mesodon altivagus (Pilsbry, 1900) (Figs. 15c, 36e,f; Table 2; Fig. 52) (1) Tennessee: Blount County (GS-2; FMNH 214613): ten live adults, ten tissue samples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10. (2) Tennessee: Blount County (GS- 5; FMNH 214614): 20 live adults, 20 tissue samples—electrophoresed #1, 2, 3, 4, 5, 6, 758: 9, 10511 12,19, 14 190510 1741920: (3) North Carolina: Avery County (GS-205; FMNH 214615): one live adult; one tissue sample—electrophoresed. (4) North Caro- lina: Swain County (SC-144; FMNH 214616): 12 live adults, 12 tissue samples— dissected #7, 9 (illustrated #7); illustrated shell #7. (5) North Carolina: Swain County (SC-145; FMNH 214617): one live adult, four tissue samples—dissected #4. Remarks. The penial morphology of M. al- tivagus seems quite different from that of typ- ical M. andrewsae, indicating that these are separate species. Subgenus Mesodon s. str. (Figs. 15a,b, 16b,c, 36a-d, 37c-f; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending full length of penis, cord-like; right lateral pilaster extending full length of penis, cord-like; chal- ice a thick-walled, rounded or pointed ear- like flap, flared to the left, rolled over to right in uneverted penis; dorsal wall bearing about eight to 12 thin parallel ridges, equal in diameter, which is constant or gradually increases basally; ventral structures absent; peripheral step-like shoulder present or ab- sent. Shell: size medium to large (14-31 mm); shape subglobose to globose (height/diame- ter, 0.6-0.7); umbilicus narrow, partly to fully and broadly covered; parietal denticle present and small, or absent; basal denticle absent; palatal denticle absent; periostracal hairs or scales absent. Species Group Mesodon sanus (Clench & Archer, 1933) (Figs. 15a, 36a,b; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending full length of penis, cord-like; right lateral pilaster extending full length of penis, cord-like; chal- ice a thick-walled, rounded ear-like flap, flared to left, rolled over to right in uneverted penis; dorsal wall bearing about 12 thin par- allel ridges, alike in diameter, which gradually increases basally; ventral and peripheral structures absent. Shell: size medium (19-20 mm); shape subglobose (height/diameter, 0.6); umbilicus narrow, partly covered; parietal denticle ab- sent; basal denticle absent; palatal denticle absent; periostracal hairs or scales absent. Included species MESODONTINI EVOLUTION 165 Mesodon sanus (Clench & Archer, 1933) (Figs. 15a, 36a,b; Table 2; Fig. 52) (1) Tennessee: Franklin County (GS-103; FMNH 214727): two live adults, three tissue samples—dissected #1, 2 (illustrated #2); electrophoresed #1, 2, 3; illustrated shell #8. Species Group Mesodon thyroidus (Say, 1817) (Figs. 15a,b, 16b,c, 36a-d, 37c-f; Table 2; Fig. 52) Description Genitalia: left lateral pilaster extending full length of penis, cord-like; right lateral pilaster extending full length of penis, cord-like; chal- ice a thick-walled, rounded or pointed ear-like flap, flared to left, rolled over to right in unev- erted penis; dorsal wall bearing about eight to 12 thin parallel ridges, alike in diameter, which is constant or gradually increases ba- sally; ventral structures absent; peripheral step-like shoulder present or absent. Shell: size medium to large (14-31 mm); shape globose (height/diameter, 0.6-0.7); umbilicus narrow, partly to fully and broadly covered; parietal denticle present and small, or absent; basal denticle absent; palatal den- ticle absent; periostracal hairs or scales ab- sent. Included species Mesodon clausus (Say, 1821) (Figs. 16b, 37c,d; Table 2; Fig. 52) (1) Tennessee: Blount County (GS-9; FMNH 214643): 30 live adults, 30 tissue sam- ples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 29. (2) Illinois: Carroll County (GS-19; FMNH 214644): ca. ten live adults, 16 tissue samples—electrophoresed #1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 13, 15. (3) Ten- nessee: Blount County [GS-28 (= GS-9); FMNH 214645]: one live adult-dissected #A, B, C, E. (4) Kentucky: Knox County (GS-116; FMNH 214650): 21 live adults, 21 tissue sam- ples—dissected #6, 11, 17 (illustrated #11); illustrated shell #3. (5) Kentucky: Knox County [GS-188 (= GS-116); FMNH uncat.]: un- known number of live adults, unknown number of tissue samples—electrophoresed #1, 13. Mesodon trossulus Hubricht, 1966 (Figs. 16c,37e,f; Table 2; Fig. 52) (1) Alabama: Clarke County (GS-53; FMNH 214741); three live adults, five tissue sam- ples—dissected #3 (illustrated); electro- phoresed #1, 2, 3, 4, 5; illustrated shell #A. Remarks. At its type locality M. trossulus was found sympatric with M. clausus—the shell of which not only lacked the color band, but also had a greener background color and a slightly different shape—without concholog- ical intergradation. This variation in color might represent only a polymorphism; unfortunately live specimens of M. clausus were not found at the site for anatomical and electrophoretic tests. Mesodon thyroidus (Say, 1817) (Figs. 15b, 36c,d; Table 2; Fig. 52) (1) Kentucky: Pulaski County (GS-13; FMNH 214743): ca. two live adults, ten tissue samples— electrophoresed #1, 2, 3, 4, 5, 6, 8, 9, 10. (2) Texas: Bastrop County (GS-63; FMNH 214744): ca. 12 live adults, 18 tissue samples—dissected #1, 6, 8, 14, 18 (illus- trated #6); electrophoresed #1, 5, 7, 9, 10, 14; illustrated shell #11. (3) Texas: Trinity County (GS-74; FMNH 214750): ca. 40 live adults, 40 tissue samples— electrophoresed #2, 10. (4) Texas: Cherokee County (GS-78; FMNH 214751): ca. 20 live adults, 22 tissue sam- ples—dissected #6, 8, 12; electrophoresed #12. (5) Illinois: Kane-Cook Counties (GS- 207; FMNH uncat.): unknown number of live adults, unknown number of tissue samples— dissected #3; electrophoresed #3, 4. APPENDIX 2. ALTERNATIVE PHYLOGENETIC ANALYSIS USING SUCCESSIVE WEIGHTING. An alternative approach is Farris’s (1988) method of successive weighting. This method first produces a set of maximum-parsimony trees, assigns each character a weight (0-10) according to its fit to these trees, reruns the analysis using the assigned weights, re- weights each character according to the re- sulting trees, and iterates this process until the character-weights stabilize. “Successive weighting has the advantage of providing a means of basing groupings on more reliable characters without making prior decisions on weighting” (Farris, 1988). Application of successive weighting to the data set (Table 4, but with a single change: 166 EMBERTON removal of the convergence in the dorsal ridge from Patera appressa) resulted in 1077 equally parsimonious trees, the strict consen- sus of which is presented in Figure 53. ashmn pele eee clark clinch edent kalmn kiowa laevr ansl enns erig sargt smith erus f[Pinny indnr 1 joieath roemr 2 Саррех downi 3 agaz rugel ee 5 ferrs 4 т E Erres® m archr 10 jones heat nape dad DE FIG. 53. Nelson consensus tree of 1077 maximum- parsimony cladograms generated from data matrix in Table 4, using successive-weighting algorithm of Hennig86. Character-state changes (numbers refer to Table 3) at each number in the cladogram are: 1. 50; 2. 7, 67, 71, 86(loss); 3. 0, 1, 25; 4. 2, 25; 5. 38, 87.6. 3:7. 10, 54:8: 11:9. 12: 16:25:50; 10:5; 11. 4; 12. 13; 13: 18; 14. 14; 15. 37; 16. 19, 85; 17: 37, 57, 58(loss); 18. 15, 20; 19. 57, 58(loss), 72. This tree is very similar to, but not identical with, the preferred consensus tree (Fig. 52), which was used for taxonomic revision (Appendix 1). APPENDIX 3. ALTERNATIVE PHYLOGENETIC ANALYSIS TREATING GENITALIC AND ELECTROPHORETIC DATA SEPARATELY, THEN SEEKING A CONSENSUS. Methods Throughout this appendix, all subgeneric and generic assignments are those of Pilsbry (1940), and do not conform with the revised taxonomy arrived at in this paper (see Table 6). The analytical procedure used in this ap- pendix was the same as that developed for the Triodopsini (Emberton, 1988). The proce- dure is summarized in Figure 54, and is de- scribed below. Data for genitalic character states (Table 1) were analyzed cladistically by the Wagner cri- terion of unrestricted parsimony (Kluge & Far- ris, 1969; Farris, 1970), using the global branch-swapping algorithm in Swofford’s (1983) PAUP program. The resulting single most parsimonious cladogram was desig- nated the Anatomy Tree. Allozyme data (Table 2) were subjected to both cladistic and phenetic analysis. The Tri- odopsini (Emberton, 1988) were used as the outgroup. Alleles not shared with the Triod- opsini were considered apomorphic for max- imum-parsimony analysis using the indepen- dent alleles model (Michevich & Johnson, 1976). The first 50 trees generated by PAUP (Swofford, 1983), using global branch-swap- ping, were compared branch by branch to de- termine the most frequent configuration of each branch. The resulting consensus cla- dogram was designated the Alleles Tree. For phenetic analysis, the electrophoretic data were divided into two subsets. The first (Table 2, M. altivagus to Ashmunella danielsi) consisted of 38 species, one population each, with complete data for all 16 loci (total 88 al- leles). The second subset (Table 2, M. altiv- agus—2 to M. zaletus—2) consisted of 37 populations representing 23 species (all but one of these species, /nflectarius verus, were the same as those of the first subset); in this second subset, all loci with partly missing data (Me, Gd-1, Gd-2, Sod-2) were omitted, leaving 12 loci (total 78 alleles) for the analysis. In each of the two subsets, the same population of Ashmunella danielsi dispar was included as outgroup. Separate distance-Wagner trees (Farris, 1970), with branch-length optimiza- tion, were computed from the Prevosti genetic distance matrix of each data subset, using NT- MESODONTINI EVOLUTION 167 ALLOZYMES (reliab. index =1.0) 39 spp., 81 pops, 706 snails 16 loci, 95 alleles aa 38 spp., 38 pops. 23 spp., 37 pops. 16 loci, 88 alleles 12 loci, 78 alleles | | D-W Tree 2 (C.C.=.87) 39 spp. 44 apomorph. alleles GENITALIA (reliab. index =2.3) 42 spp., 96 pops., 222 snails 37 char. states 42 Spp. 5 c-s phylogs., 34 transfs. | Max.-Pars. Cladogram Max-Pars. Cladogram (C.1.=.73) (C.I.=.88) u Consensus Tree FIG. 54 Diagram of alternative procedure for phylogenetic analysis applied in Appendix 3. D-W Tree 1 = Wagner-1 Tree (Fig. 58); D-W Tree 2 = Wagner-2 Tree (Fig. 59); Max.-Pars. Cladogram (from allozymes) = Alleles Tree (Figs. 56, 57); Max.-Pars. Cladogram (from genitalia) = Anatomy Tree (Fig. 55); and Consensus Tree = Consensus Tree (Fig. 60). See Appendix 3 for explanation. SYS (Rohlf et al., 1972). These trees were designated the Wagner-1 Tree and the Wag- ner-2 Tree, respectively, for purposes of com- parison with the other two trees. The four trees (Anatomy, Alleles, Wagner- 1 and Wagner-2) were compared, branch by branch, to arrive at a Consensus Tree. Dis- crepancies among trees were resolved by in- voking their weights. The weight of each tree (Fig. 54) was calculated as the number of data units (alleles or transformations) used in its construction, times the reliability index of its data units. The reliability of anatomical with respect to electrophoretic data was estimated by dividing the number of homoplasies and reversals in the Anatomy Tree by the number of homoplasies and reversals in the Alleles Tree. This method, although arbitrary, is at least objectively calculated. Results Anatomy Tree. To simplify the analysis, spe- cies were pooled that were identical in their distributions of the 34 transformations of the genitalic character states (Table 1). This pro- cedure reduced the number of operational taxa from 42 species to 24. Eighteen of these groups consisted of a single species (inflec- tus, ferrissi, smithi, pennsylvanicus, weth- erbyi, christyi, wheatleyi, jonesianus, archeri, orestes, normalis, andrewsae, elevatus, Zal- etus, mitchellianus, thyroidus, altivagus and appressus). Each of the six remaining spe- cies groups was temporarily named for one of its better-known species, regardless of prior generic and subgeneric placement. By far the largest of these groups was the perigraptus group, containing 12 species (perigraptus, panselenus, sargentianus, laevior, clarki, in- dianorum, binneyanus, clenchi, kiowaensis, edentatus, verus and kalmianus), as well as the outgroups. The approximans group had three species (approximans, magazinensis and downieanus); the rugeli group had two species (rugeli and subpalliatus); the roemeri group had two species (roemeri and leather- woodi); the sayanus group had two species (sayanus and chilhoweensis); and the clau- sus group had three species (clausus, trossu- lus and sanus). Maximum-parsimony analysis of these 24 taxonomic units, using the global branch- swapping option of PAUP (Swofford, 1983), resulted in a single most parsimonious tree (Fig. 55). This tree, hereafter called the Anat- и 168 — Patera > ak \ = o ER Az N ; 5 ON. \ а : 7 \ = or > 3 Dm mi \ a | \ 29 a; | : \ = N r \ = , в — Inflectarius \ < dl =: NE k а ; FR = р те y O ae 7 N DEN Га 1 0) 1 SH SES ISO rev a N о | Lo \ a 18 =) = , = \ | > | 5 | © | п | № | | ES =.“ | 2 0 | | u = — o : = | 9 vi 1 > Ed = | Me ‚ Е Vi 7 | 8 (D) a ~ \ 1x vi 1 „ 212% [> \\ 419 28 a — 4 195, & Sos с = € 5 \ “ = a E! Dry ae! 4 = =; a С Е ! 2 ол Co” \ 17 а 1 18 = 12 \ 3 2 5) O \ Г jonesianus EMBERTON = 5 и AE a 3 © Mesodon s. str SAO 2 10 oa > o = ~~ n а = 34 о =. 2 un YN © 5 3 9 Е X 5 емо SS rev22 © 2 5 rev28 ES 5 Чу Bl = 24 а = = & Si © R = 11 3 nme >) © TD 25 22 10%) N = Ce с ro 23 27 rev32 Ws > 5 ES = pá Y la 13 21с 2 no! Se 5 n | 30 | 21C) 32 FIG. 55. Anatomy Tree: a phylogenetic hypothesis for the Mesodontini based on penial morphology (34 character-state transformations shown in Figs. 26-28). This is the single most parsimonious tree generated by PAUP, with a consistency index of 0.882. omy Tree, had two homoplasies (Transforma- tions 1 and 21) and six reversals (Transfor- mations 8, 18, 22, 28, 30 and 32), giving an overall consistency index of 0.882. None of these homoplasies and reversals seemed bi- ologically impossible. The two homoplasies and four of the reversals (Transformations 18, 22, 30 and 32) were seemingly robust. The reversals in Transformations 8 and 28, how- ever, could be obviated fairly parsimoniously by moving the jonesianus-orestes-archeri branch from its dichotomy with wheatleyi to a trichotomy with wetherbyi and a christyi- wheatleyi branch. This alternative substituted reversals in Transformations 8 and 28 for ho- moplasies in Transformations 8, 20 and 26, thereby slightly reducing the overall consis- tency index from 0.882 to 0.868. Because ho- moplasies in Transformations 20 and 26 seemed biologically unlikely, there was no good reason for choosing this less parsimo- nious alternative. Thus, Figure 55 shows the best cladogram to fit the suggested transfor- mations (Figs. 26-28). The branch lengths of this Anatomy Tree were scaled to the number of transformations they contain and are there- fore approximate indicators of the degrees of evolutionary change. Alleles Tree. Forty-four, or 46%, of the 95 al- leles detected in the Mesodontini (Table 2) were not detected in the Triodopsini and were therefore presumed to be apomorphic. These alleles are listed in Table 9, along with their distributions among the species of the Mes- odontini. Twenty-seven of these alleles were restricted to a single species, and each of the remaining 17 alleles was present in two to 14 species. Maximum-parsimony analysis of the data in Table 9 produced the cladogram shown in Figure 56. This tree, henceforth called the Al- MESODONTINI EVOLUTION 169 TABLE 9. Allozyme alleles presumed apomorphic in the Mesodontini (i.e., undetected in their outgroup, the Triodopsini) and the species in which they were detected. 1. Sordho, clausus, perigraptus 2. Mdh-1 9, approximans, edentatus, ferrissi, inflectus, magazinensis, rugeli, smithi, subpalliatus, verus, wheatleyi 3. Mdh-1,, edentatus, kalmianus 4. Mdh-1,, thyroidus 5. Meio; chilhoweensis, sayanus, thyroidus 6. Icd:10 clarki 7. lcd. clarki, ferrissi, subpalliatus, thyroidus 8. lcdios sanus 9. Icd,, christyi 10. Icdgs rugeli WAP clarki 12. Gd-1,04 inflectus, rugeli 13. Gd-1,00 altivagus, zaletus 14. Sod-1,,, laevior 15. Sod-1,,, leatherwoodi 16. Sod-1,,, panselenus 17. Sod-2,,, andrewsae, ferrissi 18. Sod-2,,, zaletus 19. Sod-2,,, subpalliatus, wetherbyi 20. Sod-2,, sayanus 21. Got-1,,, leatherwoodi 22. Got-1,,; andrewsae, chilhoweensis, clarki, clausus, kalmianus, mithcellianus, orestes, pennsylvanicus, roemeri, sanus, sayanus, thyroidus, trossulus, zaletus 23. Got-1,,, christyi, orestes, wetherbyi, wheatleyi, zaletus 24. Got-1,, verus 25. Got-14, kiowaensis 26. Got-2,,, panselenus 27. Got-25, зауапиз 28. Got-2,, panselenus, wheatleyi 29. PgmMiog Sargentianus 30. Pgm, 9, approximans 31. Pgm,.ı perigraptus 32. Pgmgss thyroidus 33. Pgmoss ferrissi 34. Рдто> leatherwoodi, roemeri, rugeli 35. Pgmso subpalliatus 36. Pgmgss roemeri 37. ap andrewsae 38. Гароз appressus, clarki, edentatus, inflectus, kalmianus, smithi 39. Mpio, ferrissi 40. GPisos andrewsae, chilhoweensis, leatherwoodi, orestes, perigraptus, roemeri, sanus 41. GPijo3 elevatus, mitchellianus, thyroidus, zaletus 42. GPiso: kalmianus 43. GPioz sayanus 44. GPige inflectus, rugeli, smithi leles Tree, represents the plurality consensus (see below) of the first 50 of an unknown number of equally parsimonious cladograms generated by the global branch-swapping op- tion of PAUP. In the Alleles Tree, homoplasies occurred in 14 of the 44 alleles [numbers 1, 2 (все); 3, 5, 7, 13, 17, 19; 22:23, 28, 34,40 (twice), and 41]; reversals occur т five о the 44 alleles [numbers 7, 22 (twice), 23, 38 (thrice), and 40]; and 28 of the 44 alleles oc- curred without homoplasy or reversal. To aid discussion, various branches of the Alleles Tree are labeled A through K in Figure 56. Branches A, B, C, E and F were stable in all of the 50 trees examined, and branch D occurred in the Alleles Tree in 49, or 98%, of the 50 trees. The remaining branches (G-K) occurred in the Alleles Tree more commonly than any alternative, and were represented in 32% to 94% of the 50 trees (Fig. 57). Figure 57 lists the alternative configurations of branches G-K, introduces the additional mi- nority branch L and lists the alternative topol- ogies of the Alleles Tree itself. Distance-Wagner Trees. The two subsets of allozyme data were analyzed using BIOSYS (Swofford & Selander, 1981). Prevosti dis- tance matrices (Emberton, 1986: Appendices C-1, C-2, available from the author upon re- quest) were calculated, then subjected to the distance-Wagner procedure with branch- length optimization, producing the two trees shown in Figures 58 and 59. The 38-species, complete-data tree (Wag- ner-1 Tree, Fig. 58) had a cophenetic corre- lation of 0.866, indicating only mild distortion of the original genetic distance matrix. The 23-species, 37-population, reduced-data tree (Wagner-2 Tree, Fig. 59) had a similarly high cophenetic correlation of 0.872. Consensus Tree. In the Anatomy Tree (Fig. 55), eight homoplasies and reversals oc- Patera и Y wa u 2 Oo o - > Inflectarıus £ EN 76 EN / Sek / / 2 \ 5 р \ = A \ o | 10 | Е | | 2 ex | Zac | _ a 1 ex = = 3c E | | © & ke р 2 trev38i © 32 = LA x | a = | u / \ ' =e | ES lees ı 16 rev38 ¡2 a \ 13 A r | 2 1% & Dee ei m (IC ES = + € № u lo a \ |e Е, / oO ¡a a Ч: {7} Г 2 a! \' 38 / ze 2 м \\ x ART k ' a 2 \ 2! he 2 ¡mi (a) 1 INS) ol 1c o yey | \ 2 ре oO, =) = is! \ NA \ ©, 6 o < 19! gn 2c \ 9, 0 | “ О u | y? ANO \ (В) / a \ ai SEE: \ y if E | sf pe ~~ 2c о | 0 29, ng cl EA “|, 32 Q So ME) EMBERTON Appalachına Г AN TERRE = o Be) a Mesodon s. str. Se 5 | | =a р E | | == | ) 201 + | © | — | œ = Г = 15) 7715 nN ı3 \ о ic ¡9 13c 15 x 2 reli iS A = 218 «3 u 2 м 18 \ 36 £ | = o 1185 a \ = 7c € à € 12 \ 15 | a < = | \ | = ! u EX 34с | Sc ; 137 E Y N ne 3 о — | EN (0) KO 7 © 22 E 3 u | FI ` Е - $ 7 = Mg 0 a Lic a u \ 5 о 33 fon 216 O a E) = | i = is Е 2 e 23 \ De 2 i=! Y © = £ 15 | 2c 5 а Lil ‘bec o \ k 2 \ fis! 2c 5 D - 5 2% с Y) FIG. 56. Alleles Tree: a phylogenetic hypothesis for the Mesodontini based on allozymes, with Triodopsini as outgroup. The 44 presumed apomorphic alleles are listed in Table 9. This tree is the plurality consensus of 50 trees of equal and maximal parsimony generated by PAUP. Gender endings conform to Pilsbry's usage (1940); revised endings shown in Table 6. curred among the 34 transformations (ratio 0.24, rounded to two decimal places), and in the Alleles Tree (Fig. 56), 24 homoplasies and reversals occurred among the 44 transforma- tions (ratio 0.55, rounded to two decimal places); thus the reliability index for anatom- ical with respect to electrophoretic data is 2.32 (rounded to two decimal places). The Wagner-1 Tree (Fig. 58) was based upon 88 electromorphs, and the Wagner-2 Tree (Fig. 59) was based upon 78 electromorphs. Thus the Anatomy, Alleles, Wagner-1 and Wagner- 2 Trees were constructed from the following numbers of roughly equivalent data units, re- spectively: 78.9 (i.e., 34 x 2.32), 44, 88 and 62. After multiplying the values for the two Wagner Trees by their respective cophenetic correlations to adjust for their distortion of the original distance data (this is an additional calculation not performed in the Triodopsini analysis of Emberton, 1988), the results were divided by 78.9 and rounded to get the follow- ing relative weights: Anatomy Tree 1.0 Alleles Tree 0.6 Wagner-1 Tree 1.0 Wagner-2 Tree 0.7 These weights are shown within the bottom arrows in Figure 54, which summarizes the procedure used for this analysis. The trees then were compared visually, branch by branch, to arrive at a consensus tree, using the weights to resolve discrepancies. To aid comparisons among Trees, three of the four subgenera of the Mesodontini recog- nized by Pilsbry (1940), Patera, Inflectarius and Appalachina, were delimited by dashed lines; all species and species groups were members of Pilsbry’s fourth subgenus, Mes- odon s. str. (Figs. 55-59). It is apparent in the four Trees (Figs. 29-33) that not all of these four nominal subgenera are discrete and co- herent. Appalachina is the only one that seems not to require modification from Pils- bry’s concept. This subgenus is equivalent to MESODONTINI EVOLUTION 174 (other) orest penns ©) ©) p x © (M) claus tross 14% claus u a o © г - al 4% est 05 se в = ul o с x or 18% 2 > ¿ ®OO@ & others) others) 12% 8% 2% (others) | ] o = o a others) 2% perig 6% FIG. 57. Alternative topologies of Alleles Tree (Fig. 56), of equal parsimony and among the first 50 trees generated by PAUP. the sayanus group; its two species (sayanus and chilhoweensis) are tightly linked and iso- lated in the Anatomy and Alleles Trees (total weight, 1.6) and are closely linked, in combi- nation with clausus, trossulus and thyroidus, in the Wagner-1 Tree (weight, 1.0). These linkages easily outweigh the separation of sayanus and chilhoweensis in the Wagner-2 Tree (weight, 0.7). Pilsbry’s (1940) subgenus /nflectarius is strongly supported by the Wagner-1 and Wagner-2 Trees (Figs. 32, 33; total weight, 1.7); less strongly supported by the Alleles Tree (Fig. 30; weight, 0.6), and weakly sup- ported, with some members combined with Patera and some Mesodon s. str. in the peri- graptus anatomical group, in the Anatomy Tree (Fig. 29; weight, 1.0). The combined weight of the evidence supports two conclu- sions: that /nflectarius is a coherent mono- phyletic group, and that it must be expanded to include M. (Mesodon) ferrissi and M. (Pa- tera) subpalliatus. The pairing of ferrissi and subpalliatus as a lineage within Inflectarius occurs in the Anatomy Tree (with subpalliatus a member of the rugeli group), in the Alleles Tree and in the Wagner-2 Tree; only ferrissi occurs in this position in the Wagner-1 Tree, 172 EMBERTON M-ALTIVAGUS M-ZALETUS M-NORMALIS Mesodon s. str. M-SANUS M-PENNSYLVANICUS = M-APPROXIMANS ~~ Y : | M-INFLECTUS ı |nflectarıus / M-RUGELI 7 M-FERRISSI 7 M-EDENTATUS e ] 7 ia a M-AALMIANUS ES M-SMITHI IR M-MAGAZINENSIS ) = M-CHILHOMEENSIS »Appalach ina x aa] N - | ate Ba ae = = M-SAYANUS ) M-THYROIDUS M-TROSSULUS M-CHRISTYI M-WETHERBY y a EN M-MITCHELLIANUS M-ELEVATUS a M-BINNEYANUS M-INDIANORUM M ANDREUSAE M-ORESTES M-WHEATLEYI Patera M-KIOWAENSIS — M-SUBPALLIATUS < \ M-SARGENTIANUS Pa M-APPRESSUS M-LAEVIOR _7 Fa M-LEATHERWOODI ! / M-ROEMERI M-CLARKI M-CLENCHI M-PANSELENUS 7 ze 7 ASHMUNELLA-DANL———2— OUt group $a a mn - 2 42 mme me mm mm 0.0 0.06 0.12 017 o 23 0.29 0.35 0 40 0.46 0.52 0. 58 FIG. 58. Wagner-1 Tree: distance-Wagner tree for 38 species of Mesodontini, with Ashmunella as outgroup. Computed from Prevosti distance matrix based on 16 allozyme loci (Table 2, upper half). Cophenetic correlation is 0.866; branch lengths are optimized. Gender endings conform to Pilsbry's usage (1940); revised endings shown in Table 6. in which subpalliatus is well separated, but atomical group), but also it clusters closest to this separation is outweighed (2.3 to 1.0) by edentatus in the Wagner-1 Tree. In the Alleles the position of subpalliatus in the other three Tree kalmianus also occurs within /nflectar- trees. M. (Mesodon) kalmianus also seems to ius; it was not included in the Wagner-2 Tree. belong in Inflectarius: not only does it have One other member of Pilsbry’s Mesodon s. penial morphology indistinguishable from that str., downieanus, seems to belong to Inflec- of edentatus (both are in the perigraptus an- tarius because of its apparent membership in MESODONTINI EVOLUTION M-ALTIVAGUS-2 173 M-LAEVIOR-2 M-LAEVIOR-3 } M-SARGENTIANUS-2 ” M- ZALETUS-2 M-NORMALIS-2 M-WHEATLEYI-2 M-WHEATLEYI-3 M-ALTIVAGUS-3 M-SAYANUS-2 M- pro M-APPRESSUS-2 \, Patera M-PANSELENUS- 2 San N y M-PERIGRAPTUS-3 ) у M-PERIGRAPTUS-2 Y BINNEYANUS-2 == — A M-SMITHI-2 _ | Е + M-EDENTATUS-2 M-SMI M=I M-INFLECTUS-2 RS | SS M-SUBPALLIATUS-2 ‚ Inflectarius / \ / 7 M-RUGELI-2 EN Appalachına / THI-3 a a = rsa M-FERRISSI-2 M-FERRISSI-3 M-SUBPALLIATUS-3 e / K ) / NFLECT M-RUGELI-3 = M-CHILHOWEENS-2 | M-ROEMERI-2 M-CLARKI-2. M-CLAUSUS-3 M-CLAUSUS-2 M-THYROIDUS-4 Mesodon s. str. M-THYROIDUS-3 M-THYROIDUS-5 M-THYROIDUS-2 ASHMUNELLA-DANL ——>— OUtgroup ES pa 04 02-4004 44444 444 - = nes + 0.0 0.06 0.13 0.19 0.26 0.32 о зв 0. 45 O. 51 0.58 0. 54 FIG. 59. Wagner-2 Tree: distance-Wagner tree for one additional species and 29 additional populations of 22 of the species of the Mesodontini represented in Wagner-1 Tree (Fig. 58), with Ashmunella as outgroup. Computed from Prevosti distance matrix based on 12 allozyme loci for which all populations had complete data (Table 2, lower half). Cophenetic correlation is 0.872; branch lengths are optimized. Gender endings conform to Pilsbry's usage (1940); revised endings shown in Table 6. the approximans anatomical group. Unfortu- nately, because electrophoretic data were not available, the position of downieanus is highly problematic; nevertheless it is tenatively transferred to Inflectarius on the basis of its penial morphology. Pilsbry's (1940) subgenus Patera seems to be a coherent group, both anatomically and electrophoretically, requiring expansion to in- clude M. (Mesodon) clarki and Pilsbry's bin- neyanus group (binneyanus, indianorum, clenchi, kiowaensis, roemeri and leather- woodi), also of Mesodon s. str.; and requiring the removal of two species, subpalliatus and wetherbyi. The removal of subpalliatus to In- flectarius was justified above. Tentative re moval of wetherbyifrom Paterato Mesodon s. str. is based upon its clear and consistent iso- lation from that subgenus in the Anatomy, Al- leles and Wagner-1 Trees (wetherbyi was not included in the Wagner-2 Tree). Removal of clarkito Patera is indicated by its membership in the perigraptus anatomical group, by its clustering within Patera in the Wagner-2 Tree 174 EMBERTON and by the clustering of one of its two popu- lations within Patera in the Wagner-2 Tree. The combined weight of these positions in trees [1.0 + 1.0 + (0.5)(0.7) = 2.35] strongly outweights the combined weight of the Alleles Tree and of one of the two populations in the Wagner-2 Tree [0.6 + (0.5)(0.7) = 0.95], in which clarki appears outside Patera. Trans- ferral of Pilsbry’s binneyanus group to Patera is indicated by the membership of binney- anus, indianorum, clenchi and kiowaensis in the perigraptus anatomical group, (which also includes all of Patera except appressus) by the separation of roemeri and leatherwoodi from this anatomical group by a single trans- formation (Transformation 12; Fig. 29); by the clustering of roemeri, leatherwoodi and clen- chi within Patera and of kiowaensis, binney- anus and indianorum adjacent to or very close to Patera in the Wagner-1 Tree (Fig. 32); and by the clustering of binneyanus within Patera, and of roemeri very close to Patera in the Wagner-2 Tree, in which these two species are the sole representatives of the binneyanus group. This conclusion is also supported by the Alleles Tree, in which kiowaensis is the only member of Pilsbry’s binneyanus group to appear because all of its other members lack derived alleles (Table 9); kiowaensis thus shares a basal position in the Alleles Tree with most members of Patera (minus subpalliatus and wetherbyi, and plus clarki). Species not included in Appalachina, in the modified /nflectarius, or inthe modified Patera, are assigned temporarily to Mesodon s. str. by default. Each of these four nominal subgenera are treated in turn, from the most plesiomor- phic to the most apomorphic, with discussion of the evidence of the four Trees (Figs. 55-59) concerning its affinities and the evolutionary relationships of its component species. Patera, as modified above, is clearly the most plesiomorphic subgenus of the Mesod- ontini. It occupies the basal position in the Anatomy Tree (Fig. 55), the Alleles Tree (Figs. 56, 57), the Wagner-1 Tree (Fig. 58) and the Wagner-2 Tree (Fig. 59). Penial morphology is of little use in deter- mining evolutionary relationships within Pa- tera, because only two derived character states (Transformations 12, 30-32) occur within it. This latter character state, homopla- sious with Transformations 30 and 32 in a subset of Mesodon s. str. (Fig. 29), is unique to appressus. Transformation 12, however, links roemeri and leatherwoodi (the roemeri anatomical group). This roemeri group is also tightly linked in the Wagner-1 Tree, the only other Tree in which both these species occur, and therefore constitutes the most robust af- finity within Patera. The Alleles Tree is also unhelpful concern- ing Patera; it does not show any links among its species. The tight linkage between perigraptus and clenchi in the Wagner-1 Tree is suspect. The single specimen of “clenchi” that was elector- phoresed was a juvenile from a site (GS-97) at which perigraptus was common. Because juveniles of these two species are very diffi- Cult to distinguish, the electrophoresed spec- imen of clenchi in Table 2 and Figure 58 might actually be a specimen of perigraptus. In the Wagner-2 Tree, panselenus links to perigraptus at the same level as the two pop- ulations of perigraptus. This linkage is sup- ported by the adjacent positions of pansele- nus and perigraptus in the Wagner-1 Tree, and is therefore relatively robust. The linkage of binneyanus and indianorum in the Wagner-1 Tree is so tight that there can be little doubt that they are sister species. There is no test of this linkage in the Wagner- 2 Tree, however, because indianorum was not included in it. The linkage of appressus and laevior in the Wagner-1 Tree (weight, 1.0) slightly out- weighs their separation in the Wagner-2 Tree (weight, 0.7). The consensus of these two trees seems best expressed as a trichotomy among appressus, laevior and sargentianus, because sargentianus appears at the base of the appressus-laevior branch in the Wagner-1 Tree and paired with /aevior (two populations) in the Wagner-2 Tree. The consensus of the two Wagner Trees also seems to support a trichotomy among the panselenus-perigraptus branch, the appres- sus-laevior-sargentianus branch and clarki. This pattern is approximately that remaining in the lower portion of the Wagner-1 Tree upon removal of the roemeri-leatherwoodi branch and it is the pattern present in the middle por- tion of the Wagner-2 Tree upon removal of the laevior-sargentianus branch to a trichotomy with appressus, as discussed above. The reason for removing the roemeri-leath- erwoodi branch from the trichotomy just dis- cussed is the very different position of roemeri in the Wagner-2 Tree. In this tree, from which leatherwoodi is absent, roemeri is the most primitive species of the entire genus and, al- though linked to the remainder of Patera, lies well outside it. The two Wagner Trees also MESODONTINI EVOLUTION 175 differ in the position of the binneyanus-indian- orum branch. In the Wagner-1 Tree, this branch, along with kiowaensis, lies within Me- sodon s. str., fairly well isolated (but perhaps not significantly so) from the rest of Patera. In the Wagner-2 Tree, however, this branch (represented by binneyanus only) appears as a sister-group of the appressus-clarki-panse- lenus-perigraptus branch. The best resolution of these differences seems to be the topology for Patera presented in the Consensus Tree, in which kiowaensis is linked to the base of the binneyanus-indianorum branch, and in which this branch joins that of roemeri-leath- erwoodi. Owing to the high probability of mis- identification of the “clenchr’ tissue sample, noted earlier, clenchi is tentatively placed with kiowaensis in recognition of their very great conchological similarity. Inflectarius, as modified above, seems to be most closely allied to the revised Patera because three of its members (edentatus, verus and kalmianus) are in the perigraptus anatomical group, and because its other members show relatively limited derivation beyond this primitive kind of penial morphol- ogy (Anatomy Tree, Fig. 55). Analyses of the electrophoretic data (Figs. 56-59) together weakly support this alliance with Patera. In the Alleles Tree, three members of Patera (ap- pressus, clarki and kalmianus) appear within Inflectarius, whereas none of its members ap- pear within either Appalachina or Mesodon s. str. In the Wagner-2 Tree, Inflectarius clusters in a trichotomy with most species of Patera and one of the two species of Appalachina. In the Wagner-1 Tree, one member of /nflectar- ius (subpalliatus) clusters just above the most of Patera, and the rest of Inflectarius clusters just above three members of Patera (binney- anus, indianorum and kiowaensis), although at the same level as various members of Me- sodon s. str. and Appalachina. The relatively robust linkage between fer- rissi and subpalliatus was discussed above. A close linkage between inflectus and rug- eli occurs in all four Trees. In the Wagner-1 Tree, inflectus and rugeli form a trichotomy with approximans; this trichotomy is accepted in the Consensus Tree (Fig. 60), because the approximans group and inflectus can be paired in the Anatomy Tree without change in the consistency index (exchanging a ho- moplasy in Transformation 1 for a reversal in Transformation 3), and because approxi- mans, along with magazinenesis and the fer- rissi-subpalliatus branch, can be approxi- mated to the inflectus-rugeli-smithi branch in the Alleles Tree without change in the consis- tency index (exchanging a homoplasy in Transformation 2 for a reversal in Transfor- mation 38); approximans was not included in the Wagner-2 Tree. The relationship of verus to this inflectus- rugeli-approximans trichotomy is problematic. In the Wagner-2 Tree, verus pairs tightly with one population of inflectus; in the Anatomy Tree, verus groups with edentatus and kal- mianus in the primitive perigraptus anatomi- cal group. In the Alleles Tree, verus appears in two equally parsimonious trichotomies, both having one branch consisting of edenta- tus and a second branch comprising inflectus, rugeli and smithi; the third branch consists of either verus alone or verus together with ap- proximans, magazinensis and the ferrissi- subpalliatus branch. As a reasonable resolu- tion of these different positions, verus is placed in a dichotomy with the inflecus-rugeli- approximans branch in the Consensus Tree. Placement of the ferrissi-subpalliatus branch clearly should be within or near the verus-inflectus-rugeli-approximans branch. In the Anatomy Tree, as modified above from Figure 55, the ferrissi-subpalliatus branch, along with rugeli, forms a dichotomy with the inflectus-approximans group branch. In the Alleles Tree, as modified above from Figure 56, the ferrissi-subpalliatus branch arises from a branch on the same level with verus and approximans, although also with magazi- nensis. Inthe Wagner-1 Tree (Fig. 58), ferrissi clusters at the base of the inflectus-rugeli-ap- proximans branch. In the Wagner-2 Tree as well (Fig. 59), the ferrissi-subpalliatus branch arises between two populations each of in- flectus and rugeli, although sharing this posi- tion with smithi and edentatus. The best con- sensus of these topologies seems to be that expessed in the Wagner-1 Tree, with the ad- dition of verus: thus the ferrissi-subpalliatus branch in the Consensus Tree (Fig. 60) forms a dichotomy with the verus-inflectus-rugeli- approximans branch. The position of edentatus, magazinensis and smithi in the Wagner-1 Tree basal to the verus-inflectus-ferrissi lineage, discussed above, is supported both by their basal, although slightly separated, positions in the modified Alleles Tree, and by the basal posi- tions of edentatus (as a member of the peri- graptus anatomical group) and smithi—al- though not of magazinensis—in the Anatomy Tree, but is slightly contradicted by the clus- 176 EMBERTON pennsylvanicus binneyanus ındıanorum v—clenchi kıowaensıs гоетег! leatherwoodı panselenus perigraptus appressus laevıor sargentianus clarkı verus ınflectus approximans rugeli ferrissi subpalliatus edentatus magazinensis smithi o—kalmianus >— downieanus wetherbyı christy! wheatley jonesianus orestes archerı elevatus zaletus mitchellianus normalis andrewsae sayanus = nee altivagus sanus clausus trossulus thyroidus FIG. 60. Consensus Tree: alternative phylogenetic hypothesis for the Mesodontini, representing weighted consensus of Anatomy, Alleles, Wagner-1, and Wagner-2 Trees (Figs. 55-59). This tree is very similar to, but not identical with, preferred consensus tree (Fig. 52), which was used for taxonomic revision (Appendix 1). Gender endings conform to Pilsbry's usage (1940); revised endings shown in Table 6. tering of edentatus and smithi between pop- ulations of inflectus and rugeli, and on about the same level as the ferrissi-subpalliatus branch in the Wagner-2 Tree (Fig. 59). In view of the weights of these trees, the consensus topology seems to be with edentatus, maga- zinensis and smithi arising at the same level, basal to the verus-inflectus-ferrissi lineage. The position of kalmianus is somewhat problematic. It shares a primitive penial mor- phology with edentatus and verus (peri- graptus group in the Anatomy Tree), and it clusters with edentatus, magazinensis and smithi in the Wagner-1 Tree, but it is fairly isolated and highly derived in the Alleles Tree (kalmianus was not included in the Wagner-2 Tree). The best compromise, in view of the weights of the three trees, seems to be with kalmianus arising at the same level as eden- tatus, magazinensis and smithi as shown in the Consensus Tree. The position of downieanus in Inflectarius is highly problematic owing to the lack of elec- trophoretic data and the uncertainty about its penial morphology as a result of contractional distortion of the only specimens available for dissection. Its apparent, but questionable, membership in the approximans anatomical MESODONTINI EVOLUTION 177. group, which includes approximans and mag- azinensis, is unindicative of its topological po- sition in the Consensus Tree. Therefore downieanus, marked with question marks, is tentatively placed at the basal level of Inflec- tarius. Pilsbry’s (1940) Mesodon s. str. seems to be a catch-all subgenus for those species with globose shells and minimal apertural dentition. Mesodon s. str. was already re- duced in the present analysis by removing kalmianus and downieanus to Inflectarius, and is further expanded here by transferring wetherbyi from Patera. The Anatomy Tree in- dicates discrete clusters within Mesodon s. str. One of these anatomically differentiated groups (wetherbyi, christyi, wheatleyi, jone- sianus, orestes and archeri), henceforth called the wheatleyi group, is at least partly validated in two of the three electrophoretic trees. In the Alleles Tree the four electro- phoresed species of the wheatleyi group (wetherbyi, christyi, wheatleyi and orestes) group together either at the base of the rest of Mesodon s. str. (of Pilsbry, 1940) (Fig. 56; Fig. 57, left-most topology of “J”) or as an independent branch (branch “L”). In the Wag- ner-1 Tree, wheatleyi and orestes cluster to- gether, and wetherbyi and christyi cluster near each other and slightly removed from the wheatleyi-orestes branch. Only wheatleyi (two populations) occurs in the Wagner-2 Tree. Because of this validation in the other two allozymal trees, | have adopted un- changed in the Consensus Tree the topology of the wheatleyi group in the Anatomy Tree. Among the remaining species of Mesodon s. Str., a fairly robust group is that comprising thyroidus and the clausus anatomical group (clausus, trossulus and sanus), henceforth called the thyroidus group. These four spe- cies are tightly clustered in the Anatomy Tree. In the Wagner-1 Tree, thyroidus, clausus and trossulus cluster closely (along with Appa- lachina), but sanus is remote; this isolation of sanus in the Wagner-1 Tree is counterbal- anced, however, by its position in the Alleles Tree, in which sanus is separated from clausus and trossulus by only a single trans- formation (Transformation 40). The aberrant position of thyroidus—in a trichotomy with clarki and kalmianus within Inflectarius—in the Alleles Tree is counterbalanced by its con- sistent clustering with clausus (four and two populations respectively) in the Wagner-2 Tree. Thus the consensus of these trees seems to be the pairing of clausus and trossu- lus in a branch joining thyroidus to form a three-species branch joined to sanus, as shown in the Consensus Tree. The anatomical uniqueness of pennsylvan- icus is indisputable (see Transformations 6, 16) and is not contradicted by the position of this species in the Alleles and Wagner-1 Trees. Its position is unclear, but it is placed tentatively in this analysis at the base of Pa- tera. The remaining six species of Mesodon s. str. (of Pilsbry, 1940) (normalis, andrewsae, elevatus, zaletus, mitchellianus and altiva- gus), are grouped in the Anatomy Tree along a single lineage leading to the thyroidus group. This topology is supported by the Wag- ner-1 Tree in that these six species appear primitive with respect to the thyroidus group, but seems to be contradicted in this tree in that these six species are split into two iso- lated clusters, the first one close to the ma- jority of the thyroidus group and comprising andrewsae, elevatus and mitchellianus, and the second one distant from the majority of the thyroidus group and comprising normalis, zaletus and altivagus. This apparent contra- diction, however, is counterbalanced by the fact that the latter, distant group does indeed cluster near one member of the thyroidus group (sanus); thus the parallel isolation of sanus and of the normalis-zaletus-altivagus cluster might be an idiosyncracy of the Wag- ner-1 Tree that distorts true relationships. This view is supported by the topology of the Alleles Tree, in which the six species are ei- ther all on an equivalent branch level, or are primitive to the thyroidus group and are sep- arated, but into clusters different from those in the Wagner-1 Tree. The Wagner-2 Tree, although it differs from the other trees in in- termixing the three of these six species that it includes (altivagus, normalis and zaletus) with Patera and with the wheatleyi group, supports the other trees in placing these species prim- itive to the thyroidus group. In addition, it shows a great distance between the two pop- ulations of altivagus; such high intraspecific variation (or cryptic species) might cause other discrepancies among the four trees with re- spect to the positions of all six species. With regard to the interrelationships among these six species, | have chosen to follow the topology of the Anatomy Tree. The juxtaposi- tion of elevatus, zaletus and mitchellianus in this Tree is mirrored by their trichotomy in the Alleles Tree. The clustering of altivagus, nor- malis and zaletus in both the Wagner-1 and Wagner-2 Trees approximately supports their 178 EMBERTON proximity in the Anatomy Tree. The clustering of elevatus, mitchellianus and andrewsae in the Wagner-1 Tree likewise supports their rel- ative positions in the Anatomy Tree. Owing to the isolated, contradictory position of altiva- gus in the Alleles Tree, and owing to the fact that its position in the Anatomy Tree depends upon tentative decisions about the homology of its dorsal ridges, a question mark appears on the position of this species in the Consen- sus Tree, which otherwise exactly duplicates the topology of the Anatomy Tree for eleva- tus, zaletus, mitchellianus, normalis, andrew- sae and altivagus. In phylogenetic position Appalachina (say- anus and chilhoweensis) is very close to the thyroidus group, according to the Wagner-1 Tree and, to alesser extent, the Wagner-2 and the Alleles Trees. The consensus of these three trees (combined weight, 2.3) clearly out- weighs the Anatomy Tree (weight, 1.0). In the Consensus Tree, therefore, Appalachina ten- tatively appears at the base of the thyroidus group and altivagus combined, a position that only slightly decreases the parsimony of its position in the Anatomy Tree by exchanging one homoplasy (in Transformation 21) for two reversals (in Transformations 30, 22). The completed Consensus Tree for the Me- sodontini is presented in Figure 60. MALACOLOGIA, 1991, 33(1-2): 179-191 PHYSIOLOGICAL AND MORPHOMETRIC CHANGES IN DREISSENA POLYMORPHA (MOLLUSCA; BIVALVIA) DURING A STARVATION PERIOD Martin Sprung’ & Jost Borcherding Zoologisches Institut der Universität Köln, Lehrstuhl für Physiologische Ökologie, Weyertal 119, D-5000 Köln 41, Germany ABSTRACT Two experimental groups of the freshwater mussel Dreissena polymorpha were examined for oxygen consumption rates, carbohydrate and lipid content, morphometry of the gonads and the digestive gland, and for the number and size of the oocytes. Samples were taken over a starvation period of 31 days. Animals from one group had not spawned yet (from 9 m water depth), the other had, but still contained gametes (from 2 m water depth). Weight-specific oxygen consumption rates remained constant during the starvation period (9-m group) or even showed an increasing trend (2-m group). The digestive gland contributed to the decrease in weight mainly during the first 10 days of the starvation period. Lipids were predominantly used during this period. During the last part of the starvation period, both groups mainly respired the unidentified rest (probably proteins). Carbohydrates formed only a small fraction of the total tissue and contributed little to the metabolic demands of this species. A loss in gonad tissue was noted in the 2-m group during the whole time span, rendering it frequently impossible to identify the sex of the fresh material at the end of the experiment. In contrast, in the 9-m group a loss of gonad tissue was only observed during the last 3 weeks. In both groups, changes in oocyte number showed the same tendencies as in gonad tissue. Oocytes of all size fractions were resorbed in equal proportions, with a tendency for the 2-m animals to break down predominantly smaller oocytes. Key words: Dreissena polymorpha, starvation, oxygen consumption, gonads, oocyte size frequency distribution, digestive gland, biochemical composition. INTRODUCTION The freshwater mussel Dreissena polymor- pha has attracted the attention of scientists since it invaded the European freshwaters, coming from the Caspian area during the last 150 years (Stanczykowska, 1977). This inter- est has been reenforced by its recent pene- tration into the Great Lakes of North America (Hebert et al., 1989). The result of this is still to be seen. One reason why its spread has been so dramatic is linked with its rather ar- chetypical mode of reproduction by means of free-swimming larvae. The larvae develop from eggs, which are released into the water column in large numbers (Walz, 1978; Sprung, 1990). The species is iteroparous. Its repro- ductive period starts in autumn and ends in spring or summer, when the animals spawn at above 12°C (Antheunisse, 1963; Walz, 1978; Borcherding, 1989). Dreissena polymorpha is a filter feeder. Food accessable for these types of animals can vary greatly during the course of the year. Food conditions can be especially poor in winter (Epp et al., 1988), but also during sum- mer when the first plankton bloom is over (Schwoerbel, 1984), or when during late sum- mer dinoflagellates, which Dreissena cannot use, dominate the plankton (Stanczykowska, 1977). What happens when the animals enter a period of food shortage? The consequence must be degrowth in the widest sense. Rus- sell-Hunter et al. (1983) stated two hypothe- ses of how the animals may react: the first is a breakdown of all components the molluscan body offers. The second is that the animal can regulate its catabolism to achieve whatever pattern of differential depletion contributes most to its fitness. Whether material of the gonad is involved in the process of degrowth is of special interest in this context. We observed two groups of Dreissena poly- morpha, one which had already spawned, and the other which had not, over a period of 31 days at their natural temperature with the ab- “Present address: U. E. Ciéncias e Tecnologias dos Recursos Aquaticos, Universidade do Algarve, Campus de Gambelas, Apartado 322, P-8004 Faro Codex, Portugal. 180 SPRUNG & BORCHERDING TABLE 1. Analytical procedure of the starvation experiments described; after day 31, the mussels obtained a feeding pulse; for the estimation of the oxygen consumption rate the animals were split into five size groups except for the case indicated (*) where there were ten; the following symbols have been used for the estimates: BC: biochemical components (carbohydrates, lipid, ash); O: oxygen consumption; W: weight. Mussels processed during the following procedures Number of 2-m animals 9-m animals 33 di SiG ee Seren ar 18 11 Бо SuSE 569599 19 11 10 565539 sence of food. We report here the physiolog- ical and morphometric changes that were recorded. Histological and ultrastructural changes of the same material are described elsewhere (Bielefeld, 1991). MATERIAL AND METHODS Sampling and Storage The mussels were collected from two depths at the beginning of August 1987 by SCUBA diving from a lake close to Cologne (Fuhlinger See). (1) From 2 m depth: these animals had already spawned although the sex could still be recognized in the fresh ma- terial. At this depth, the water temperature was 17.1°C. (2) From 9 m depth: water tem- perature was 10.4°C, and the water had be- come anoxic because of a stratification of the lake. Anaerobic conditions probably had an effect on the composition of the animals (see Discussion and Bielefeld, 1991) but not on their respiration rate, which could be exam- ined only after two days in aerated water. These animals had not yet spawned. The mussels were transferred to the laboratory in insulated boxes containing water from the sampling sites. After cleaning the shells thor- oughly, they were stored at their original wa- ter temperature in deionized water with 1% seawater and an added CaCl, solution (final concentration 60 mg Ca/l). The water was changed daily during the first week, then three times a week in order to remove all par- ticles from the aquaria. No mortality and no spawning activity was observed. The mussels were separated into five size groups of one to Days after sampling 9 31 32 10 31 32 O;W BC O O;W BC five animals before storage and were pro- cessed according to the schedule outlined in Table 1. Animals for morphometry and oocyte measurements were examined on the sam- pling date, day 10 and day 31 of the starvation period. Oxygen Consumption Oxygen consumption was recorded in a flow-through system: five chambers of about 30 ml capacity were run in parallel; a peristal- tic pump sucked water through them; the flow rate of every chamber (typically about 60 ml/ h) was checked by a graduated cylinder be- fore and after the experiment. The water stream was regulated by electrical valves so that every five minutes the outflow of another chamber passed the oxygen electrode of a recording device of the type Radiometer Copenhagen (PHM 72). Oxygen partial pres- sure in mm Hg of five consecutive cycles was registered on a chart recorder. The system was Calibrated with deoxygenated water (5% Na, SO, solution supplied to the electrodes by means of a bypass) and aerated water pro- vided to the electrodes between each cycle. In addition, at the end of each experiment, checks for a possible blank respiration rate was performed with empty chambers. The rate of oxygen consumption was calculated by applying the formula: VO, = Aimm:Hg “fl 6 А mm Hg: difference in oxygen partial pres- sure between the chamber and the bypass corrected by a blank run CHANGES IN DREISSENA DURING STARVATION 181 fl: flow rate (average of initial and final rate) | B: absorption coefficient of oxygen in water calculated from Benson & Krause (1980). Weight and Shell Length Determination Individual body weight was estimated by re- moving the soft parts with a scalpel and sep- arating the soft parts in the visceral sack (go- nads, digestive gland, stomach, parts of the digestive tract, parts of the adductor muscle, byssus gland) and non-visceral sack tissues (gills, mantle, kidneys, parts of the adductor muscle). These were transferred to pre- weighed aluminium trays, dried for two days at 105°C and weighed again (dry weight). Ash content was estimated after three hours at 450°C in a muffle furnace. The shell length of each mussel used in the experiments was re- corded to + 0.1 mm by means of a calliper rule. Estimations of Lipid and Carbohydrate Content Estimates were made from the digestive gland, gonads and the total soft parts. The latter were homogenized in a Worrax blender, which was rinsed out with distilled water. Samples were dried at 105°C overnight and stored at —18°C until analysis. Samples were pooled from five females and five males sep- arately, if sex was detectable. Sex was deter- mined by pricking the animal with a needle and examining the adherent cells on a slide under a compound microscope. Analysis was made in triplicate, if sufficient material was available; the means of the three estimates were used for further evaluation. Two to 5 mg of this redried tissue (105°C overnight) was dissolved for analysis by boil- ing it for 10 min in 2 ml 90% H,SO,. After cooling, 100 HI was added to 2 ml sulpho- phospho-vanillin colour reagent (kit: total lip- ids, MERCK) to estimate the lipid content af- ter 40 to 50 min at 530 nm wave length. For estimating the carbohydrate content, 500 pl of the wet ashed sample was trans- ferred to 2 ml TCA; proteins were spun off after five minutes. One ml of the supernatant was added to 5 mi anthrone reagent and boiled for 15 min. Carbohydrate content was estimated at 630 nm wave length after cool- ing. The lipid calibration curve was established by using a standard provided with the test; the carbohydrate estimates were calibrated with glucose solution processed according to the analytical procedure outlined above. Conversions Following Wieser (1986), the subsequent conversions to energy units have been used: lipid: 39.2 kJ/g carbohydrates: 17.5 kJ/g rest (estimate for protein): 24.5 kJ/g ash-free dry weight: 22.9 kJ/g oxygen: 20.3 J/ml Body Compartments The gonads of Dreissena show a close in- tergrowth with other tissues of the visceral sack. That is why the size of the gonads and of the digestive gland were estimated by mea- suring the area of each on various sections taken over the whole visceral sack (Borcher- ding, 1986, 1989; Morvan & Ansell, 1988). This was done according to the following schedule: The visceral sack of 20 mussels from each experiment and sampling date were separated from the soft parts, fixed in Bouin-Allen’s fluid, dehydrated in an ascend- ing alcohol series and methylbenzoate, then embedded in Paraplast Plus (melting point 56°C). Each visceral sack was cut completely in transverse sections with a Leitz microtome (Typ 1300; 10 вт). The distance from the first to the last section was assumed to be the length of the visceral sack. During this pro- cess, up to 20 sections were taken at intervals through the visceral sack, stained with May- er's haemalaun and eosin (Adam & Czihak, 1964), and mounted in Canada Balsam. For each section, the areas of the gonad, the digestive gland and the whole visceral sack were measured using an image analys- ing system (Soft Imaging GmbH, Munster). Each mean tissue area of each mussel was multiplied with the length of the visceral sack to calculate the volume of the gonads, diges- tive gland and visceral sack, respectively. By relating the dry weight to the volume of the visceral sack (Ansell & Trevallion, 1967), the dry weight of the different tissues was calcu- lated. Oocyte Measurements The diameter of 120 to 150 oocytes per fe- male was estimated with the image analysing 182 SPRUNG & BORCHERDING system. Only oocytes with a clearly visible nu- cleus were selected to ensure a central sec- tion. Mature oocytes were clearly round and had no attachment sites to the follicle wall. Only for this fraction the mean diameter was computed for all values greater than 40 um. These oocytes generally form a distinct peak in a histogram, as indicated by the shaded bars in Figure 3. For mussels for which no clear measure- ment of the oocytes was possible (e.g. oocytes that were too small to be precisely detected), only an approximate estimation of the size and the distribution is given. They are indicated by hatched lines in the histogram. (For details of the oocyte measurements and evaluations, see Borcherding, 1989.) Evaluations In order to obtain an estimate for a standard size, regressions have been fitted for tissue weight, tissue volume, oocyte numbers and oxygen consumption (y) versus respective tis- sue weight or shell length (x), using the power function: y=a a and b: fitted parameters A shell length of 20 mm (range 12.7 to 27.9 mm) and 20 mg tissue weight (range 7.1 to 42.5 mg) represents a typical value for the animals from 2 m depth, as did 25 mm shell length (range 18.3 to 34.6 mm) and 60 mg tissue dry weight (range 33.0 to 185.3 mg) for the 9 m animals. For these standard sizes, estimates and 95% confidence intervals have been calculated using the formula given by Sachs (1978, p. 342). Two samples were compared with the U-test of Mann-Whitney- Wilcoxon. All calculations were made with a personal computer using self-written BASIC progams. In the 9-m group, the sample size was low; above this, males and females were not present in equal proportions. This led to un- realistic trends in the 9-m group between days two and ten, when separate weights for males and females were calculated. That is why in this case different weight data were pooled for both sexes. They were assumed to represent the weight on day six. Weight on the original dates was recalculated by means of the weight-specific oxygen consumption rates (see “Conversions”). RESULTS Oxygen Consumption Weight-specific oxygen consumption rates remained roughly constant in the 9-m group or even showed an increasing trend over the starvation period in the 2-m group (Fig. 1). It did not change significantly after refeeding. In both groups, there is agreement from the quantitative point of view that no starvation metabolism exists in Dreissena, which devi- ates from that already established on the first day after food deprivation. Weight Changes Mussels of the standard size from 9 m were nearly three times heavier than those from 2 m. This is in part due to the different standard length referred to, but also to the different de- velopmental stage of the gonads. As is to be expected, the total dry weight at standard shell lengths decreased in general (Table 2). During the first ten days, in both groups, weight loss of digestive gland was higher than of other organs: for the 2-m and 9-m animals, the quotient of digestive gland to body size decreased significantly (p < 0.05). But also the gonads of the 2-m animals were affected. During the following three weeks, in both groups, the gonads contributed most to the weight loss rendering it impossible to identify the sex in the fresh material of animals of the 2-m group. During this period, substance from the digestive gland played only a minor role. Biochemical Components The results of the biochemical analysis of the dried homogenate are presented in Table 3. Carbohydrates formed only a relatively small part of the dry tissue in total and was consistently higher in males. Particularly low was the carbohydrate content of the gonads and of the digestive gland when compared with the content of the total soft parts. This agrees with the observation that carbohy- drates are mainly found in the mantle tissue (Borcherding, 1989). Females have a higher lipid fraction than males, probably due to the lipid reserves of the mature oocytes; a high lipid content was noted not only in the gonads but also in the digestive gland of the 9-m females. Ash content was always higher in the tissue of male animals. CHANGES IN DREISSENA DURING STARVATION 183 1.4 20 mg afdw; 17.1° C -^ 1.2 Mo) 1.0 Oxygen consumption (ml Os К 0.8 0.6 0.4 0.2 1 9 31 32 0.5 5 60 mg afdw; 10.4 C 0.4 0.3 0.2 0.1 2 10 31 32 Length of starvation period (d) FIG. 1. Dreissena polymorpha: weight-specific rates of oxygen consumption on distinct days of a starvation period; the vertical bars indicate 95% confidence intervals, the arrow a feeding pulse (sample size, see Table ae Changes in Biochemical Composition lipid. For metabolic demands, carbohydrates only played a minor role in both groups. Inde- Combining data of weight and of relative pendent of sex the main fraction lost during biochemical composition with its energy con- the last three weeks of the experiment was tent, absolute changes can be estimated (Fig. the “unidentified rest,” which should contain 2). The main loss during the first week was all the proteins. In Table 4, changes are doc- 184 SPRUNG & BORCHERDING TABLE 2. Dreissena polymorpha: Composition and weight of the soft body of two experimental groups on distinct days of the starvation period; sample size for the quotients 20, for the weight see Table 1; values in brackets: 95% confidence intervals Quotient of Quotient of Tissue Estimated Estimated Experiment Day of gonad digestive gland dry gonad digestive gland (shell starvation to body size to body size weight weight weight length) period (%) (%) (mg) (mg) (mg) 0 20.6 15.2 28.5 5.9 4.3 (17.2-24.5) (13.0-17.8) (25.8-31.4) Mussels from 10 17.5 10.5 25:2 4.4 2.6 2 т depth (14.8—20.9) (8.6-12.9) (22.0-28.7) (20 mm) 31 12:2 MA 18.8 2.3 2.2 (10.3-14.6) (9.6-14.2) (17.0-20.8) 0 35.1 11.2 72.0 25:3 8.1 (33.0-37.1) (10.0-12.4) (56.1-92.4) Mussels from 10 38.5 9.1 7A 27.4 6.5 9 m depth (35.8-41.1) (7.7-10.8) (56.6-89.6) (25 mm) 31 32:3 9.3 55 17.8 5.1 (29.3-35.6) (7.9-11.0) (44.3-64.8) TABLE 3. Dreissena polymorpha: relative content of carbohydrates, lipid and ash in the homogenized soft parts, digestive gland and gonads on distinct days of a starvation period; for sample size, see Table 2-m animals 9-m animals Group 1 9 31 2 10 31 Day $ 3 $ 3 Аа $ 3 $ 3 Q 3 Carbohydrate content (%) Total tissue 5:5 10.6 53 9.3 6.8 5.8 8.8 6.3 9.3 5.6 8.5 Digestive gland 3.2 4.4 1.6 4.0 2.5 3.2 3.4 4.4 4.1 2.8 3.1 Gonads 219 3.8 2.0 4.5 — 3.6 2.9 2.2 25 2.6 3.9 Lipid content (%) Total tissue 157 14.4 14.9 127 11.0 1728 041572 17.0 13.9 18.9 15% Digestive gland 10.5 9.6 12.3 12.0 14.1 18.6 15:2 14.5 10:9 зоб Gonads 14.7 14.3 23.9 11.0 — 271 18.97 228 12:9 227 2:0 Ash content (%) Total tissue 16:27 A910)" 13:9 14.5 13:3 11.9 1827 10:2; 501651 11.272168 Digestive gland 13.0 10.9 aa 13.8 19.4 MEA Ale “ANS 7/ 1802 ES LA Gonads 7.4 13:3: = 214:4 18.5 — 12:9 19.2 146 223 7152551672 umented for different body compartments. gonads of the females and the remaining The following tendencies could be observed: body tissues of the males contributed to it. For the mussels from 2 m depth, the amount However, the main loss in the digestive gland of lipid in female gonads increased during the was from the “rest.” Because of a gain in first ten days of starvation, while it was de- “rest,” especially in the gonads, this loss was creasing in male gonads. A distinct loss of not noted for the whole animal. The body “rest” (probably protein) in the digestive gland compartments contributing to the loss of and the gonads coincided with a clear in- “rest” from days 10 to 31 of the experiment crease of “rest” in the remaining body tis- were the gonads and the remaining body tis- sues. sues. Loss of lipids in the gonads was com- For the mussels from 9 m depth, the lipids pensated by a gain in lipids in the remaining respired during the first ten days were derived body tissues. During this period, the carbohy- from the digestive gland. Also lipids from the drates from the mantle tissue also contributed CHANGES IN DREISSENA DURING STARVATION 185 |20 mm shell length; REN e] 60 30 O carbohydrates | YA пра № rest 60 mu da 30 + с" C+ oS 0 Mm Biochemical components respired (J) 1 9 31 200 100 о 0 Vm. 200 100 J о’ 0 287% |); CON ive aa 2 10 31 Length of starvation period (d) FIG. 2. Dreissena polymorpha: contribution of biochemical components to metabolic activity at certain intervals of a starvation period (sample size, see Table 1). 186 SPRUNG & BORCHERDING 22 14 70 n=3 n=3 =a ==> (ir Zn Den 0.1 0.2 0.1 0.2 [177] n=5 n=9 a= 51/6 а=475 g = 301.000 g = 286.000 7] Zu Th, Oocyte size classes (diameter; ит) V} HH, WHEN, ПИ ИЛИ ИИ HUHN, n=8 d=50.2 — g = 995.000 — MA 0.1 0.2 0.1 0.2 day O day 10 AMA 0.1 0.2 Чау 31 FIG. 3. Dreissena polymorpha: oocyte size frequency distributions on distinct days of a starvation period; п: number of mussels; d: mean diameter for the oocytes of the ripe fraction (stripped bars); g: total numbers of oocytes per mussel; x-axis: oocyte number per size class ( 10°); for those animals to which no exact analysis was possible, data were only estimated (hatched lines). CHANGES IN DREISSENA DURING STARVATION 187 TABLE 4. Dreissena polymorpha: changes of biochemical components in mg during starvation; the values are calculated with data from Table 2 and from Table 3 for mussels of 20 mm shell length (animals from 2 m depth) and of 25 mm shell length (animals from 9 m depth). Group 2-m animals Period (days) 0-10 Tissue Component male Lipid —0.07 Digestive Carboh. —0.09 gland Rest — 1.38 Lipid — 0.43 Gonad Carboh. —0.04 Rest —1.48 Remaining Lipid —0.50 body Carboh. 05% tissue Rest + 2.96 female OH — 0.08 — 1.04 + 0.39 — 0.03 EL — 0.88 —0.09 +1.35 9-m animals 0-10 10-31 male female male female —0.54 —0.59 —0.04 — 0.22 + 0.01 —0.01 —0.09 —0:12 0.87 — 1.21 — 0:31 — 0.55 880) 1:03 1108 —1|,70 —0!05 — 0.38 +0.04 —0.08 + 0.66 + 1.68 — 3.64 —4.71 —0.86 +0.42 + 0.32 +1.04 +0.14 +0.59 EOS 1 00 +0.11 —1.57 —4.35 —3.94 in part to the metabolic demands of the ani- mal. Oocytes The size-frequency distribution of the oocytes showed similar tendencies to those found in the gonad analysis in Table 2 (Fig. 3). There was an almost constant decrease in oocyte numbers during the whole observation period in the 2-m animals, but in the 9-m an- imals only from day 10 to 31. Oocytes of all size classes were resorbed in equal propor- tions; the diameter of the ripe oocyte fraction stayed about the same. When only considering the 2-m animals with intact gonads, another trend was found: Oocyte diameter of the ripe fraction de- creased significantly (p < 0.05), and at the same time the relative proportion of immature oocytes (< 40 um diameter) decreased from day 0 to day 31 from 33.4% to 25.6%. This could suggest that ripe oocytes were main- tained at the expense of the immature. How- ever, we cannot be certain of this because of the low number of females with intact gonads at the end of the experiment. Also many fe- males had already formed small oocytes of which we could only get an approximate es- timate. DISCUSSION Molluscs have evolved a lot of mechanisms to survive periods of food shortage without affecting its residual reproductive capacity (in the sense of Fisher, 1930). They can encom- pass the following components: (1) A reduction of the respiratory rate during starvation Evidence for this has been provided by var- ious authors for different species of molluscs, e.g. the pulmonates Australorbis glabratus (by Brand et al., 1948) and Lymnea palustris (by Duerr, 1965) and the prosobranch Thais lapillus (by Bayne & Scullard, 1978). This is also true for many bivalve molluscs (Newell, 1970). With respect to Mytilus edulis, Thompson & Bayne (1972) characterized its metabolism according to the state of nutrition as an active metabolic rate when the mussel was feeding vigorously; a decrease to a standard meta- bolic rate characterizes a starving animal. The time span in which it changes over varies with the season and with the state of nutrition from a few days up to several weeks. Be- tween these extremes, Mytilus can build up different levels of a routine metabolism. The respiratory rate of Dreissena recorded here is not unusual when compared to that of Mytilus or other bivalves (data from Bayne & Newell, 1983). However, quite obviously this classifi- cation of various levels cannot be applied to Dreissena. One reason may probably be looked for in changes in activity patterns. Bayne et al. (1973) noted a two- to three-fold increase in the oxygen consumption due to the state of activity in Mytilus. Other authors have demonstrated similar correlations be- tween the state of activity and oxygen con- sumption (see references in Jorgensen et al., 1986). Despite this, Jorgensen et al. (1986) 188 SPRUNG & BORCHERDING argued that directly in the process of filtration only a small amount of metabolic energy should be involved. In agreement with this and our data, MacKay & Shumway (1980) found no elevated level of oxygen consump- tion when food was offered to Chlamys deli- catula after a starvation period. They inter- preted this as due to the fact that Chlamys lives in an environment with a continuous food supply. In contrast, Mytilus is adapted to discontinuous feeding in the tidal zone. Con- sequently, it reacts more distinctly to a differ- ent food supply with respect to feeding and digestion. However, in small Mytilus edulis, an elevation in respiratory rate after a feeding pulse lasts only a few hours (Gaffney & Diehl, 1986). This time interval had not been under consideration in our observations. Fairly constant or even increasing meta- bolic activity over a starvation period could be observed not only in the pulmonate snail He- lisoma trivolis (Russell-Hunter et al., 1983) and the welk Thais lamellosa (Stickle & Duerr, 1970; Stickle, 1971), but also in Thais lapillus (Stickle & Bayne, 1982). This contrasts to the earlier findings with Thais lapillus mentioned above (Bayne & Scullard, 1978). The authors explain this inconsistency in part by the fact that Thais was not prefed before the mea- surements for the later observations. There could be an analogy in the results obtained here for Dreissena, because they were also examined without an extra pretreatment. (2) A preferential utilization of stored sub- stances to evitate degrowth of structural pro- teins Giese (1969) and Gabbott (1976, 1983) pointed out the importance of carbohydrates for the metabolism of bivalves. De Zwaan & Wijsman (1976) reported seasonal oscilla- tions especially for glycogen in the range of 5% to 40% of the dry body weight (lipids should account for 1.5% to 10% and protein for 40% to 70% of the dry body weight). There could be a possible connection with the ne- cessity for an anaerobic metabolism in bi- valves: for example, Crenshaw & Neff (1972) observed that as soon as 25 min after shell closure, the extrapallial fluid of Mercenaria mercenaria was free of oxygen. It is then that carbohydrates play an outstanding role, be- cause they can be degraded without the Krebs cycle. Respectively lipids and protein can only be utilized under these conditions if fumarate from the carbohydrates is supplied to the Krebs cycle (de Zwaan & Wijsman, 1976). Although it may be a seasonal phe- nomenon, the low carbohydrate content in the soft parts of Dreissena is peculiar, but is con- firmed by histological examinations by Tourari et al. (1988) on the same species. The generally high lipid content of the soft parts when compared to ranges reported for other bivalves (see above) is also peculiar. Elevated lipid levels in the digestive gland are quite typical for molluscs (Giese, 1969). Nevertheless, the high lipid content in the di- gestive gland of the 9-m animals when com- pared to that of the 2-m animals can be the consequence of the anaerobic conditions at the time of sampling, which may have re- stricted the utilization of lipids. Despite the outstanding role carbohydrates obviously have as an energy store, especially in many bivalves, it must not be inferred that they always dominate metabolic events dur- ing starvation. Data from observations of var- ious mollusc groups during starvation show little regularity. For example, Giese (1969) re- ported a respiration of lipids for chitons, as did Emerson (1967) for the prosobranch Littorina planaxis and Holland (1978) for planktonic mollusc larvae in the initial phase. Emerson & Duerr (1967) demonstrated the importance of carbohydrates during starvation for the pul- monate Planorbis corneus. Epp et al. (1988) drew attention to the importance of protein during overwintering for Argopecten irradians irradians. For the prosobranch Thais lapillus, Stickle & Bayne (1982) described a protein- orientated metabolism during — starvation, which they thought to be characteristic for all carnivors. Closer examination also demon- strated that the type of substrate metabolized during starvation varies with the season: Gab- bott & Bayne (1973) observed that in Mytilus edulis the whole energy loss during a starva- tion period of 41 days at the beginning of the new reproductive cycle in late summer and early autumn was derived from carbohy- drates. In autumn, there was a marked in- crease in the utilization of lipid reserves for a 60-day starvation period (Bayne, 1973); in winter, there was a change over to protein as the main respiratory substrate (Bayne & Thompson, 1970). The substrate metabolized can also vary with the length of the starvation period: Riley (1976) showed that in the oyster Crassostrea gigas during the first 25 days, each biochem- ical component (protein, lipids, carbohydrate) declined to roughly 80% of its initial value. During days 25 to 125, loss of carbohydrates continued steadily while the decline in lipids CHANGES IN DREISSENA DURING STARVATION 189 was slowed and little protein was lost. During the terminal phase (days 125 to 175), all com- ponents declined rapidly. Especially for over- wintering freshwater pulmonates, there is ev- idence that their energy stores frequently do not meet their metabolic demands and that they show true degrowth of structural proteins (Russell-Hunter & Buckley, 1983). (3) A depletion of storage sites Giese (1969) may be right when he stated that molluscs have no special storage site for reserve substances, as, for example, fat tis- sue in mammals. It is, however, not true that molluscs in general do not have energy stores to survive periods of negative energy bal- ance, for example those caused by food shortage or reproductive demands. Well- known examples are the mantle tissue of Mytilus edulis (Zandee et al., 1980) and the adductor muscle of pectinids (Barber & Blake, 1981; Gould et al., 1988). The digestive gland plays a special part during short periods of food deprivation. Al- though its primary task is the digestion of food (Purchon, 1971), it also acts as a food distrib- utor (Barber & Blake, 1985) and as a storage organ (Bayne et al., 1976). As a storage or- gan, it is the most important energy source during gametogenesis (Gabbott & Bayne, 1973). But material from the digestive gland is also rapidly utilized during nutritive stress, as demonstrated for Mytilus edulis (Thompson et al., 1974), for Argopecten irradians concentri- cus (Barber & Blake, 1983, 1985) and for Dre- issena in this study. Shortly before spawning, a great part of the molluscan body is comprised of gonads. The roughly 40% found here are by no means unusual for bivalves. For example, Giese et al. (1967) recorded 27% for Tivela stultorum and Thompson (1979) up to 59% for Mytilus edulis. The gonads should be the primary target for tissue degrowth when the reserves from other storage sites, such as the diges- tive gland, are depleted. In fact, resorption of gonad material has been demonstrated for Crassostrea virginica (Loosanoff & Davis, 1951), for Argopecten irradians (Sastry, 1966), and for Mytilus edulis (Bayne et al., 1978). Walz (1978) denied this option for Dreis- sena polymorpha. However, Tourari et al. (1988) demon- strated that Dreissena can do so under ther- mal stress (exposure to 30°C). In the present study, evidence is provided that degrowth of gonad tissue can occur even under moderate temperature conditions. However, when it has not yet spawned, the animal is quite reluctant to make use of the gonad material for meta- bolic purposes, as indicated by the results of the mussels from 9 m depth. The present results thus support the sec- ond hypothesis put forward by Russell-Hunter et al. (1983): a regulation of the catabolism. This was not achieved on a quantitative level (reduction of the respiration rate), but on a qualitative level, by differential depletion of stored substances (primarily lipids in this case). Furthermore, animals that have spawned react in a different manner from those that had not by a deliberate degrowth of gonad tissue. This is quite a meaningful be- haviour: for these animals, the spawning sea- son is over, and it is now essential to survive until the next season. ACKNOWLEDGEMENTS Our most cordial thanks are due to Prof. Dr. D. Neumann for critically reading the manu- script, to two anonymous reviewers for their most valuable comments, and to Mrs. P. Horta e Costa for correcting the English. The work was funded in part by the Deutsche For- schungsgemeinschaft (contract No. Ne 72/ 26-1-1378/87). LITERATURE CITED ADAM, H. & G. CZIHAK, 1964, Arbeitsmethoden. Gustav Fischer Verlag, Stuttgart. ANSELL, A. D. & A. TREVALLION, 1967, Studies on Tellina tenuis da Costa. |. Seasonal growth and biochemical cycle. Journal of Experimental Marine Biology and Ecology, 1: 220-235. ANTHEUNISSE, L. J., 1963, Neurosecretory phe- nomena in the zebra mussel Dreissena polymor- pha Pallas. Archives Néerlandaises de Zoologie, 15: 237-314. BARBER, B. J. & N. J. BLAKE, 1981, Energy stor- age and utilization in relation to gametogenesis in Argopecten irradians concentricus (Say). Journal of Experimental Marine Biology and Ecology, 52: 121-134. BARBER, B. J. & N. J. BLAKE, 1983, Growth and reproduction of the bay scallop, Argopecten irra- dians (Lamarck) at its southern distributional limit. Journal of Experimental Marine Biology and Ecology, 66: 247-256. BARBER, B. J. & N. J. BLAKE, 1985, Inter-organ biochemical transformations associated with oo- genesis in the bay scallop Argopecten irradians concentricus (Say), as indicated by '*C incorpo- 190 SPRUNG & BORCHERDING ration. Biological Bulletin (Woods Hole), 168: 39— 49. BAYNE, B. L., 1973, Aspects of the metabolism of Mytilus edulis during starvation. Netherlands Journal of Sea Research, 7: 399-410. BAYNE, B. L., D. L. HOLLAND, M. N. MOORE, D. M. LOWE 4 J. WIDDOWS, 1978, Further studies on the effects of stress in the adult on the eggs of Mytilus edulis. Journal of the Marine Biological Association of the United Kingdom, 58: 825—841. BAYNE, B. L. & R. C. NEWELL, 1983, Physiologi- cal energetics of marine molluscs. Pp. 407-515, in: А. 5. М. SALEUDDIN & К. М. WILBUR, eds., The Mollusca. Vol. 4. Physiology I. Academic Press, New York. BAYNE, B. L. & C. SCULLARD, 1978, Rates of oxygen consumption by Thais (Nucella) lapillus (L.). Journal of Experimental Marine Biology and Ecology, 32: 97-111. BAYNE, B. L. & R. J. THOMPSON, 1970, Some physiological consequences of keeping Mytilus edulis in the laboratory. Helgolander Wissen- schaftliche Meeresuntersuchungen, 20: 526- 552. BAYNE, В. L., В. J. THOMPSON & J. WIDDOWS, 1973, Some effects of temperature and food on the rate of oxygen consumption by Mytilus edulis L. Pp. 181-193 in: W. WIESER, ed., Effects of temperature on ectothermic organisms. Springer, Berlin, Heidelberg, New York. BAYNE, B. L., R. J. THOMPSON & J. WIDDOWS, 1976, Physiology |. Pp. 121-206, in: B. L. BAYNE, ed., Marine mussels: their ecology and physiology. Cambridge University Press, Cam- bridge. BENSON, B. B. & D. JR. KRAUSE, 1980, The concentration and isotopic fraction of gases dis- solved in freshwater in equilibrium with the atmo- sphere. 1. Oxygen. Limnology and Ocean- ography, 25: 662-671. BIELEFELD, U., 1991, Histological observations of gonads and digestive gland in starving Dreissena polymorpha. Malacologia (in press). BORCHERDING, J., 1986, Gonadenreifung und Fortpflanzungszyklen bei Dreissena polymorpha (Pall.). Diplomarbeit, University of Cologne. BORCHERDING, J., 1989, Die Reproduktionsleis- tung der Wandermuschel Dreissena polymorpha. Ph.D. Thesis, University of Cologne. BRAND, T. v., P. MCMAHON & M. O. NOLAN, 1957, Physiological observations on starvation and desiccation of the snail Australorbis glabra- tus. Biological Bulletin (Woods Hole), 113: 89— 102. CRENSHAW, M. A. & J. M. NEFF, 1969, Decalci- fication at the mantle-shell interface in molluscs. American Zoologist, 9: 881-889. DE ZWANN, A. & T. C. WIJSMAN, 1976, Anaerobic metabolism in bivalvia (Mollusca)—Characteris- tics of anaerobic metabolism. Comparative Bio- chemistry and Physiology, 54B: 313-324. DUERR, Е. G., 1965, Some effects of diet on the respiration rate of the fresh water pulmonate snail Lymnaea palustris. Proceedings of the South Dakota Academy of Science, 44: 245. EPP, J., V. M., BRICELJ & R. E. MALOUF, 1988, Seasonal partitioning of energy reserves in two age classes of the bay scallop Argopecten irra- dians irradians (Lamarck). Journal of Experimen- tal Marine Biology and Ecology, 121: 113-136. EMERSON, D. N., 1967, Carbohydrate orientated metabolism of Planorbis corneus (Mollusca, Planorbidae) during starvation. Comparative Bio- chemistry and Physiology, 22: 571-579. EMERSON, D. N. & F. G. DUERR, 1967, Some physiological effects of starvation in the intertidal prosobranch Littorina planaxis (Philippi, 1847). Comparative Biochemistry and Physiology, 20: 45-53. FISHER, R. A., 1930, The genetical theory of nat- ural selection. Oxford University Press, London. GABBOTT, P. A., 1976, Energy metabolism. Pp. 293-356, in: B. L. BAYNE, ed., Marine mussels: their ecology and physiology. Cambridge Univer- sity Press, Cambridge. GABBOTT, P. A., 1983, Developmental and sea- sonal activities in marine molluscs. Pp. 165-217, in: Р. W. НОСНАСНКА, ed., The Mollusca, Vol. 2, Environmental biochemistry and physiology. Academic Press, New York. GABBOTT, P. A. & B. L. BAYNE, 1973, Biochemi- cal effects of temperature and nutritive stress on Mytilus edulis L. Journal of the Marine Biological Association of the United Kingdom, 53: 269-286. GAFFNEY, P. M. & W. J. DIEHL, 1986, Growth, condition and specific dynamic action in the mus- sel Mytilus edulis recovering from starvation. Ma- rine Biology, 93: 401—409. GIESE, A. C., 1969, A new approach to the bio- chemical composition of the mollusc body. Oceanography and Marine Biology Annual Re- view, 7: 175-229. GIESE, A. C., M. A. HART, A. M. SMITH & M. A. CHEUNG, 1967, Seasonal changes in body component indices and chemical composition in the pismo clam Tivela stultorum. Comparative Biochemistry and Physiology, 22: 549-561. GOULD, E., D. RUSANOWSKY & D. A. LUEDKE, 1988, Note on muscle glycogen as an indicator of spawning potential in the sea scallop Pla- copecten magellanicus. U. S. National Marine Fisheries Service, Fishery Bulletin, 86: 597-601. HERBERT, P. D. N., B. W. MUNCASTER & G. L. MACKIE, 1989, Ecological and genetic studies on Dreissena polymorpha (Pallas): a new mol- lusc in the Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences, 46: 1587-1591. HOLLAND, D. L., 1978, Lipid reserves and energy metabolism in the larvae of benthic marine inver- tebrates. Pp. 85-123, in: D. C. MALINS & J. R. SARGENT, eds., Biochemical and biophysical perspectives in marine biology, Vol. 4. Academic Press, New York. JORGENSEN, C. B., F. MOHLENBERG & O. STEN-KNUDSEN, 1986, Nature of relation be- tween ventilation and oxygen consumption in fil- CHANGES IN DREISSENA DURING STARVATION 191 ter feeders. Marine Ecology Progress Series, 29: 73-88. LOOSANOFF, V. L., 4 H. C. DAVIS, 1951, Delay- ing spawning of lamellibranchs by low tempera- ture. Journal of Marine Research, 10: 197-202. MACKAY, J. & S. E. SHUMWAY, 1980, Factors affecting oxygen consumption in the scallop Chlamys delicatula. Ophelia, 19: 19-26. MORVAN, C. & A. D. ANSELL, 1988, Stereological methods applied to reproductive cycle of Tapes rhomboides. Marine Biology, 97: 355-364. NEWELL, R. C., 1970, Biology of intertidal animals. Logos Press, London. PURCHON, R. D., 1971, Digestion in filter feeding bivalves—a new concept. Proceedings of the Malacological Society of London, 39: 253-262. RILEY, R. T., 1976, Changes in the total protein, lipid, carbohydrate, and extracellular body fluid free amino acids of the Pacific oyster, Crassos- trea gigas, during starvation. Proceedings of the National Shellfisheries Association, 65: 84—90. RUSSELL-HUNTER, W. D. & D. E. BUCKLEY, 1983, Actuarial bioenergetics of nonmarine mol- luscan productivity. Pp. 463-503, in: W. D. RUS- SELL-HUNTER, ed., The Mollusca, Vol. 6, Ecol- ogy, Academic Press, New York. RUSSELL-HUNTER, W. D., D. W. ALDRIDGE, J. S. TASHIRO & B. S. PAYNE, 1983, Oxygen up- take and nitrogenous excretion rates during over- winter degrowth conditions in the pulmonate snail, Helisoma trivolis. Comparative Biochemis- try and Physiology, 74A: 491-497. SACHS, L., 1978, Angewandte Statistik, Planung und Auswertung, Methoden und Modelle, 5th ed., Springer, Berlin, Heidelberg, New York. SASTRY, A. N., 1966, Temperature effects in re- production of the bay scallop, Aequipecten irra- dians Lamarck. Biological Bulletin (Woods Hole), 130: 118-134. SCHWOERBEL, J., 1984, Einführung in die Lim- nologie, 5th ed., Fischer, Stuttgart. SPRUNG, M., 1990, Costs of reproduction: a study on metabolic requirements of the gonads and fe- cundity of the bivalve Dreissena polymorpha. Malacologia, 32: 267-274. STANCZYKOWSKA, A., 1977, Ecology of Dreis- sena polymorpha (Pall.) (Bivalvia) in lakes. Pol- skie Archivum Hydrobiologii, 24: 461—530. STICKLE, W. B., 1971, The metabolic effects of starving Thais lamellosa immediately after spawning. Comparative Biochemistry and Phys- iology, 40A: 627-634. STICKLE, W. B., & B. L. BAYNE, 1982, Effects of temperature and salinity on oxygen consumption and nitrogen excretion in Thais (Nucella) lapillus (L.). Journal of Experimental Marine Biology and Ecology, 58: 1-17. STICKLE, W. В. & Е. G. DUERR, 1970, The effects of starvation on the respiration and major nutrient stores of Thais lamellosa. Comparative Biochem- istry and Physiology, 33: 689-695. THOMPSON, R. J., 1979, Fecundity and reproduc- tive effort of the blue mussel (Mytilus edulis), the sea urchin (Strongylocentrotus droebachiensis), and the snow crab (Chionoectes opilio) from pop- ulations in Nova Scotia and Newfoundland. Jour- nal of the Fisheries Research Board of Canada, 36: 955-964. THOMPSON, R. J. & B. L. BAYNE, 1972, Active metabolism associated with feeding in the mus- sel Mytilus edulis L. Journal of Experimental Ma- rine Biology and Ecology, 9: 111-124. THOMPSON, R. J., N. A. RATCLIFFE & B. L. BAYNE, 1974, Effects of starvation on structure and function in the digestive gland of the mussel (Mytilus edulis L.). Journal of the Marine Biolog- ical Association of the United Kingdon, 54: 699— 742% TOURARI, A. L., C. CROCHARD & J. C. PIHAN, 1988, Action de la temperature sur le cycle de reproduction de Dreissena polymorpha (Pallas). Etude “in situ” et au laboratoire. Haliotis, 18: 85— 98. WALZ, N., 1978, The energy balance of the fresh- water mussel Dreissena polymorpha Pallas in laboratory experiments and in lake Constance. Il. Reproduction. Archiv fúr Hydrobiologie/Supple- ment, 55: 106-119. WIESER, W., 1986, Bioenergetik—Energietrans- formationen bei Organismen. Thieme, Stuttgart, New York. ZANDEE, D. I., J. Н. KLUYTMANS, W. ZURBURG 8 H. PIETERS, 1980, Seasonal variation in bio- chemical composition of Mytilus edulis with ref- erence to energy metabolism and gametogene- sis. Netherlands Journal of Sea Research, 14: 1-29. Revised Ms. accepted 26 November 1990 MALACOLOGIA, 1991, 33(1-2): 193-198 PHYSIOLOGICAL AND BEHAVIOURAL COMPENSATION FOR FOOD QUALITY AND QUANTITY IN THE SLUG LEHMANNIA MARGINATA Michele K. Surbey' & C. David Rollo? ABSTRACT The growth rates, dry mass budgets and behavioural time budgets of juvenile Lehmannia marginata were examined during and after 21 d of starvation or on a diet 90% diluted with cellulose. Controls were fed a high quality diet originally developed for rearing herbivorous insects. Slugs lost weight in both experimental regimes, but marked compensatory adjustments were evident for those fed low-quality food (consumption rate 256% of controls, respiration rate 233% of controls, total activity 160% of controls). Starved animals lost weight more rapidly and total activity was only 59% of controls. When switched to the high quality diet, animals previously fed low quality food showed little evidence of compensatory “catch up” growth or feeding. This appeared to be due to a carry-over of their high respiration rates. Starved slugs showed a post-starvation growth rate significantly greater than controls, but the degree of compensation was less than observed in published studies on other animals. The results document that species capable of strong adjustments to inadequate diets may not necessarily have strong post-treatment compensatory abilities. This suggests that these phenomena involve separate mechanisms. Key words: Lehmannia, slugs, diet, starvation, feeding. INTRODUCTION There is a large literature concerned with feeding regulation in vertebrates and insects (see Toates, 1980; Pyke, 1984; Rollo, 1984; Slansky & Scriber, 1985), but relatively little is known about molluscs (reviewed by Rollo, 1988a, b; Rollo & Hawryluk, 1988). A major aspect of such regulation involves compensa- tory responses (e.g. Simpson & Abisgold, 1985). These entail various physiological or behavioural adjustments that offset inade- quate nutrition or periods of deprivation. Most studies address either adjustments for inade- quate (i.e. diluted) diets or post-starvation re- sponses. To our knowledge, a comparison of such adjustments has never been carried out in a controlled study with a single species. Moreover, most studies address physiological adjustments (e.g. “catch-up” growth, in- creased consumption, changes in assimila- tion or production efficiencies) or behaviour (e.g. relative amounts of travel, feeding) but not both. Rollo (1984) examined both behav- ioural and physiological aspects of compen- sation for starvation in an insect, but no such holistic approach is available for a mollusc. Whether an organism has strong compen- satory ability has implications for many eco- logical and behavioural aspects. Firstly, compensation might allow maintenance of relatively targeted growth rates despite varia- tions in food quality and quantity. Thus, a day- degree approach to development could be applied. Alternatively, a relatively complete knowledge of microclimatic responses (e.g. Rollo, 1991) might not allow direct prediction of growth and reproduction because short pe- riods of constraint on activity can be offset later. Given that behaviour is part of the homeostatic adjustment, compensatory re- sponses would imply that foraging patterns may vary widely depending on the quality of available food, or the nutritional history of the population. Similarly, the nutritional quality of food might influence numerous variables, such as rates of bioaccumulation of pollut- ants, secondary compounds or pesticides. In general, there are very few studies of com- pensatory mechanisms in molluscs (Calow, 1975; Rollo & Hawryluk, 1988; Vianey-Liaud, 1984; Rollo & Shibata, 1991; Rueda et al., 1990). The present study examines both be- havioural and physiological adjustments by the slug Lehmannia marginata to dietary dilu- tion and starvation. In addition, responses of “Department of Psychology, University of Toronto, Erindale College, Mississauga, Ontario, Canada L5L 1C6 “Department of Biology, McMaster University, Hamilton, Ontario, Canada L8S 4K1 (address for correspondence) 194 FEEDING IN LEHMANNIA animals fed a nutritionally complete diet fol- lowing complete starvation or a period on di- luted food are examined. MATERIALS AND METHODS Juvenile Lehmannia marginata approxi- mately one month old were obtained from rearing cultures at McMaster University orig- inally established with slugs collected in local greenhouses. Colonies were reared in plastic containers with screened tops containing a layer of moist vermiculite. Carrots, lettuce, po- tatoes and dog chow were provided, ad libi- tum as food. Initially 45 juvenile specimens of approximately 50 mg live weight each were placed individually in jars 5 cm wide and 7 cm deep covered with nylon mesh and lined with moist paper towels. The containers were placed in an environmental chamber on an 8:16 light:dark photoperiod at 18°C. A relative humidity of 100% was ensured by covering the jars with layers of wet cloth. Slugs were fed a high quality artificial diet (see Shibata & Rollo, 1988) for at least 5 d prior to the exper- iment. All slugs were individually weighed and assigned to one of three groups such that the average weight per group was equivalent. Slugs in the control group were maintained on the standard diet for 42 d. At the end of two weeks, a dry mass budget was calculated based on consumption and defecation over a three-day period. Slugs in the Low Quality group were given the standard diet diluted by 90% with cellulose powder. During the third week of this regime, a dry mass budget was calculated over a three-day period. The slugs were then switched to the high quality stan- dard diet, and three days later a second dry mass budget was calculated. Slugs in the starvation group were deprived for three weeks and were then given the high quality standard diet. No mortality associated with deprivation occurred, although two slugs died when they escaped from their containers and one control animal that appeared diseased was removed. During food deprivation the moist paper towels were replaced with moist vermiculite because starved slugs may ingest paper. Faeces were removed from the jars to prevent coprophagy. Complete food depriva- tion precluded the determination of a dry mass budget for this group. After three days on the high quality diet, however, a dry mass budget was calculated over a period of three days. Throughout this experiment all slugs were checked daily. Jars were cleaned, food re- placed and slugs were weighed every 3-4 days. At the conclusion of the experiment, five specimens from each group were sacrificed and their dry weight to wet weight ratio was determined. To determine growth rates for each individual, dry weights (estimated from the above sample) were transformed logarith- mically and plotted against time to obtain a linear relationship (Shibata & Rollo, 1988). Linear regression was performed, and the slope of the best-fitting regression line taken as an estimate of individual growth rate. An ANOVA and several planned comparisons (t- tests) were employed to compare the growth rates of the slugs under the different condi- tions. Growth rates were determined for the Low Quality and Starvation groups both be- fore and after the switch to high quality food. All measurements, except the behavioural observations, were recorded during the pho- tophase. Dry mass budgets for individual slugs were calculated as follows: Initially each animal was weighed and placed in a clean jar with fresh paper towels and a pre-weighed portion of the appropriate diet. A wet-weight-to-dry- weight conversion factor for the food was es- tablished by drying pre-weighed samples of the diet to constant weight at 60°C. At the end of three days, the slug was reweighed, and the food remaining was oven dried and re- weighed. The faeces were also recovered, dried and weighed. For each slug, consumption was calculated by subtracting the dry weight of the remaining food from the estimated dry weight of the orig- inal portion using the conversion factor estab- lished from the samples. Assimilation was calculated by subtracting the dry weight of the faeces from the dry weight of food consumed. Since there was no reproduction, production consisted only of somatic growth. The wet weight gain was converted to dry mass on the basis of the conversion factor obtained from samples. Respiration, which for the purposes of this study included both slime and urine, was estimated by subtracting production from assimilation. Assimilation, production and respiration values were divided by the num- ber of days in the mass budgeting period (3) so all final measurements were in units of dry mg/mg dry slug/d. From these initial mea- sures, assimilation efficiency (A.E. = assim- ilation/consumption X 100), gross production efficiency (G.P.E. = production/consumption SURBEY & ROLLO 195 TABLE 1. Growth rates of L. marginata fed a high quality diet (Control), a diet 90% diluted with cellulose (Low Quality), starved 21 d (Starvation) and following a switch to control diet in both Starvation and Low Quality groups. Growth Rate Group N (slope + s.e.) Control 13 0.027 + 0.002 Low Quality 15 —0.006 + 0.002 Starvation 14 —0.017 + 0.002 Low Quality switched to 15 0.030 + 0.003 High Quality Starvation switched to 14 0.036 + 0.002 High Quality Low Quality and Starvation groups lost weight and differed significantly from controls (p < 0.001 in both cases) and from one another (p < 0.01). Following a switch to the high quality control diet, the previously starved slugs grew at a significantly faster rate than controls (p < 0.01) whereas those previously fed low quality food also grew faster than controls, but this not significant (p > 0.10). X 100) and net production efficiency (N.P.E. = production/assimilation X 100) were calcu- lated for each group. In addition to the physiological measures described above, behavioural observations were recorded during the week of the first and second mass budgets. The behaviour of the slugs was observed in the first four hours of the scotophase (the peak activity period) for two evenings at half-hour intervals. The be- havioural categories included travelling, feed- ing, defecating, in contact with their food dish (but not eating) and resting. For the purposes of analysis, behaviour was divided into either resting or active categories, and an ANOVA was performed to compare their relative fre- quencies in the different conditions before and after the switch to the high quality diet. RESULTS Growth Rate The mean growth rates for the different groups of slugs during both phases of the ex- periment are summarized in Table 1. An ANOVA indicated significant differences in growth rate among groups (F = 110.8, d.f. 4 p < 0.0001). Planned pairwise comparisons between specific groups were made using t tests. Before the switch to high quality food, animals in the Low Quality and Starvation groups exhibited negative growth rates (slopes = —0.006 and —0.017, respectively), significantly different (p < 0.001) from the growth rate of controls (slope = 0.027). Slugs inthe Starvation group lost weight significantly faster than those in the Low Quality group (p < 0.01). There was considerable individual variation in rates of weight loss. Some animals inthe Low Quality group maintained their initial weight, but no animals in the Starvation group did so. After the switch to high quality food, both groups exhibited growth rates in excess of the controls, but this compensatory re- sponse was only significant in starved slugs (p < 0.01). Previously starved animals gained weight faster than those previously on low quality food, but this was not statistically re- solved (p < 0.10). The trends strongly suggest that a larger sample size may have revealed that slugs previously fed low quality food have a slight, but significant compensatory re- sponse when fed the high quality diet. Dry Mass Budgeting Dry mass budget results (Table 2) revealed further group differences underlying the re- sponses. Animals fed low quality food ate 2.5 times more food than control animals. When switched to the high quality diet, however, these animals reduced consumption to within 12% of controls. The starved animals, when fed the high quality diet, exhibited a similar level of consumption (110% of controls). As- similation rate was elevated in all three treat- ment groups compared to control. Animals fed a low quality diet had assimilation rates almost 26% higher than controls, but this rate dropped to 11% above the controls after ob- taining the high quality diet. Previously starved animals exhibited a rate 13% above controls. Faecal production in the Low Quality group before the switch to the control diet was 5.7 times greater than controls. When re- turned to a normal diet, both Starvation and Low Quality animals continued to produce more faeces (15%-20% more than controls). Assimilation efficiency was halved in animals fed a low quality diet. The assimilation effi- ciency of all groups eating the high quality diet was essentially identical (~70%). Production (growth) was slightly negative in animals fed the diluted diet. Growth rate in- creased when fed high quality food to 79% of controls during the mass budgeting period. Production efficiencies improved, but re- mained only 70% that of controls, after the return of a high quality diet. In contrast, pre- viously starved animals exhibited a produc- 196 FEEDING IN LEHMANNIA TABLE 2. Dry mass budgets of L. marginata on a high quality diet (Control), on a diet diluted by 90% with cellulose (Low Quality) and after the switch to a high quality diet for both Low Quality and Starvation groups. All measurements are in units of dry mg/mg dry slug/day (group means +/— S.E.). Bracketed values indicate the percentage of treatment values relative to Control values. Low Quality Starvation (switched to (switched to Component Control Low Quality high quality) high quality) Consumption 0.222 = 0187 0.569 + 0.056* 0.249 + 0.002 0.254 + 0.001 * (256) (112) (110) Assimilation 0.157 + 0.011 0.198 + 0.018 0.175 + 0.009 0.177 + 0.001 (126) (111) (113) Faeces 0.065 + 0.005 * 0.371 + 0.004 0.075 + 0.005 0.078 + 0.004 * (571) | (115) (120) Production 0.070 + 0.007 —0.005 + 0.007 0.055 + 0.007 * 0.077 + 0.001 * (0) (79) (110) Respiration 0.087 + 0.002 * 0.203 + 0.019 0.119 + 0.005*° 0.100 + 0.001° (233) (137) (115) Assimilation 70.7 34.8 70.3 712.2 Efficiency (%) (49) (99) (102) Gross Production 31.5 —0.9 22.1 31.4 Efficiency (%) (0) (70) (100) Net Production 46.6 215) 31.4 43.5 Efficiency (%) (0) (67) (93) A series of a priori planned comparisons was made using t tests. Within a given factor (e.g. consumption), a ия’ treatments that differed at the 0.001 level, and a "+" or a “o” denotes treatments that differed at the 0.05 level. tion 10% above controls after fed the high quality diet and their G.P.E. and N.P.E. were equivalent to controls. Respiration was 2.3 times greater than con- trols in the slugs fed a low quality diet. Even when they were switched to the high quality diet their respiration remained 37% higher than controls. Starved animals switched to high quality food had a respiration rate only 15% higher than controls. Behaviour An ANOVA was performed on the total fre- quencies of active and resting behaviours in the three groups before and after the switch to a high quality diet (Table 3) (F = 7.107, df = 2, p < 0.005). In particular, animals with Low Quality food fed eight times as frequently as controls. They also remained in contact with the food dish (but not eating) ten times as much as controls. After Low Quality and Star- vation groups were switched to the high quality diet there was a general increase in activity levels in all groups (including controls, which nearly doubled their activity levels). Group dif- ferences in the frequency of total active or resting behaviours after the switch were not significant (F = 1.537, df = 2, p < 0.25). However, those animals previously starved exhibited lower activity levels and remained in contact with food considerably more than the Control or Low Quality animals. A larger sam- ple would probably have allowed statistical resolution. DISCUSSION The ability of a species to compensate for variations in the quantity and/or quality of food determines how sensitive growth rate will be to fluctuations in dietary resources. It is not sur- prising that slugs fed such low quality food or those starved completely had negative growth rates (Table 1). Given that the low quality diet was diluted by 90% with cellulose, it was re- markable that growth was not depressed even further in that treatment. These slugs nearly tripled their consumption rate (256% of con- trols). Such adjustments in response to low quality diets were also documented in mol- luscs by Calow (1975), Rollo & Hawryluk (1988) and Rueda et al. (1990, in press). The assimilation efficiency of L. marginata on the low quality diet, although reduced, was better than expected from the dietary dilution with cellulose (i.e., 49% of controls rather SURBEY & ROLLO 197 TABLE 3. Total frequencies of behaviours observed in groups of L. marginata in Control, Low Quality and Starvation treatments before and after the latter two groups were switched to the control diet. R ACTIVE ESHING Before Number of In contact Switch Animals Travelling Defecating Feeding Total Total with food dish Control 13 42 0 5.4 47.4 204.6 6.5 Low Quality 15 30.8 0.9 43.9 75.6 176.4 69.1 Starvation 14 28 — =- 28 224 — After Switch Control 13 73"? eu 19.4 93.7 158.3 14 Low Quality 15 64.4 0.9 14 79.3 172.7 8.4 Starvation 14 47 2 17 66 186 24 Because the number of animals varied among groups, all frequencies were adjusted to reflect a group size of 14 (i.e. controls were multiplied by 14/13). than an expected 10%, Table 2). Similar re- ductions of assimilation efficiency with in- creased consumption of lower quality food were documented by Calow (1975) and Rollo & Hawryluk, (1988). This assimilation effi- ciency combined with increased consumption actually allowed L. marginata to assimilate re- sources at a faster rate than controls (see also Rollo & Hawryluk, 1988). Despite this greater assimilation, respiration rate was ele- vated to 233% of controls, resulting in a slow weight loss. Part of this respiratory cost was undoubtedly associated with digestive pro- cessing, but a substantial portion must have been incurred in behavioural costs because these animals were 1.6 times more active than controls (Table 3). Most of this increased activity was associated with feeding (elevated eight times higher than controls) because travelling was actually reduced in the low quality treatment (73% of controls). This was achieved by remaining in direct association with the food, even when not eating (10.6 times over controls) (Table 3). Rollo (1988a) previously observed such behaviour in D. re- ticulatum and suggested that food might re- main attractive even when slugs are replete if they have been malnourished. In well nour- ished animals, the sense organs may show more rapid adaptation to food stimuli or food may lose its attractiveness upon satiation. Starved animals lost weight relatively rap- idly. Rollo (1988a) previously showed that slugs do not have large physiological re- serves, but decrease their body size to adjust for nutritional shortages. It was not possible to calculate a mass budget for starved animals, but total activity of starved slugs was only 59% that of controls. Thus, the respiratory rate of starved slugs was considerably re- duced. There was slight evidence for compensa- tory adjustments when slugs fed low quality food were switched to the high quality diet. Growth rate was not statistically different from controls (Table 1). The mass budget indicated that the high respiratory rate induced by the low quality food carried over on the new diet (37% higher than controls) (Table 2). This off- set slightly greater feeding and assimilation rates (about 10% above controls). The mass budget indicated a lower production rate than was obtained by the overall growth equations. This was probably because the mass budget was calculated soon after the shift, whereas the growth equation was fitted over a longer time frame. The carry-over of high respiration soon after the switch probably explains the difference. Despite this small discrepancy, both tables indicated very little post-treatment compensation. Assimilation rate returned to control levels and also showed no compen- satory modification (Table 2). Although significant, the growth rate of starved L. marginata when resupplied with food was only 1.3 times higher than controls. Juvenile Deroceras reticulatum starved 36 d showed a 20 d compensatory growth rate three times that of controls when re-supplied with food (Rollo, 1988a). In Deroceras laeve, slugs fed following 14 d of starvation engaged a compensatory growth rate 1.6 times higher than controls (Rollo & Shibata, 1991, in press). Rollo (1988a) documented that the presence of compensatory responses varies widely among molluscan species and devel- opmental stages. For example, although ju- venile D. reticulatum compensated strongly 198 FEEDING IN LEHMANNIA following starvation, adults showed no re- sponse at all. Following starvation, consumption of L. marginata on the control diet was only slightly higher than controls. Assimilation efficiency was essentially identical to controls, in con- trast to increases observed in aquatic snails (Calow, 1975). Slightly greater production was achieved via an increased assimilation rate (Table 2), but this was a relatively weak response compared to post-starvation com- pensatory adjustments common in insects. For example, the cockroach Periplaneta americana showed post-starvation consump- tion rates five times higher than normal (Rollo, 1984). An unanticipated problem was evident in the behavioural analysis in that the overall ac- tivity of the control animals increased. This was probably associated with maturation. Rollo (1982) previously documented that ju- venile Limax maximus were much less active than adults. Because the low quality and star- vation treatments undoubtedly retarded de- velopment, it is difficult to Know which control data are most appropriate to use for compar- ison. If development was strongly affected by malnutrition, the initial control data might be more appropriate. This would suggest a stronger post-treatment compensatory re- sponse than is indicated by comparison to controls of similar age. Using the latter com- parison, there was no significant difference in activity among groups, although the previ- ously starved animals still had relatively low activity (Table 3). A major insight gleaned from these results is that a strong compensatory response to low quality food may be engaged by a species that has relatively weak post-deficiency com- pensation for low quality or low quantity diets. Thus, the mechanisms involved in adjusting to diluted diets are probably different than those involved in post-deficiency adjust- ments. It would be most worthwhile to repeat this study with other species that have strong- er or weaker compensatory responses. Such comparisons might shed light on the ecolog- ical factors shaping the evolution of such abil- ities. LITERATURE CITED CALOW, P., 1975, The feeding strategies of two freshwater gastropods Ancylus fluviatilis Mull. and Planorbis contortus Linn. (Pulmonata), in terms of ingestion rates and absorption efficien- cies. Oecologia, 20: 33-49. PYKE, A. J., 1984, Optimal foraging theory: a crit- ical review. Annual Review of Ecology and Sys- tematics, 15: 523-575. ROLLO, C. D., 1982, The regulation of activity in populations of the terrestrial slug Limax maximus (Gastropoda: Limacidae). Researches on Popu- lation Ecology, 24: 1-32. ROLLO, C. D., 1984, Resource allocation and time budgeting in adults of the cockroach, Periplaneta americana: the interaction of behaviour and met- abolic reverses. Researches on Population Ecol- ogy, 26: 150-187. ROLLO, C. D., 1988a, The feeding of terrestrial slugs in relation to food characteristics, starva- tion, maturation and life history. Malacologia, 28: 29-39. ROLLO, C. D., 1988b, A quantitative analysis of food consumption for the terrestrial Mollusca: al- lometry, food hydration and temperature. Mala- cologia, 28: 41-51. ROLLO, C. D., 1991, Endogenous and exogenous regulation of activity in Deroceras reticulatum, a weather-sensitive slug. Malacologia 33: 199— 220. ROLLO, C. D. & M. D. HAWRYLUK, 1988, Com- pensatory scope and resource allocation in two species of aquatic snails. Ecology, 69: 146-156. ROLLO, C. D. & D. M. SHIBATA, 1991, Resilience, robustness and plasticity in a terrestrial slug with particular reference to food quality. Canadian Journal of Zoology (in press). RUEDA, A. A., F. SLANSKY & G. S. WHEELER, 1990, Compensatory feeding response of the slug Sarasinula plebeia (Soleolifera: Veronicell- idae) to dilution of two different diets. Journal of Experimental Biology (in press). SHIBATA, D. M. & C. D. ROLLO, 1988, Intraspe- cific variation in the growth rate of gastropods: five hypotheses. Memoirs of the Entomological Society of Canada, 146: 199-213. SIMPSON, S. J. & D. ABISGOLD, 1985, Compen- sation by locusts for changes in dietary nutrients: behavioural mechanisms. Physiological Ento- mology, 10: 443-452. SLANSKY, F., JR. & J. M. SCRIBER, 1985, Food consumption and utilization. In: G. A. KERKUT 8 Е. 1. GILBERT, eds., Comprehensive insect phys- iology, biochemistry and pharmacology, Perga- mon Press, Oxford, New York, Paris 4: 88-163. TOATES, Е. M., 1980, Animal behaviour—a sys- tems approach. John Wiley and Sons. VIANEY-LIAUD, M., 1984, Effects of starvation on growth and reproductive apparatus of two imma- ture freshwater snails Biomphalaria pfeifferi and Biomphalaria glabrata (Gastropoda: Planor- bidae). Hydrobiologia, 109: 165-172. Revised Ms. accepted 23 October 1990 MALACOLOGIA, 1991, 33(1-2): 199-220 ENDOGENOUS AND EXOGENOUS REGULATION OF ACTIVITY IN DEROCERAS RETICULATUM, A WEATHER-SENSITIVE TERRESTRIAL SLUG C. David Rollo Department of Biology, McMaster University, 1280 Main Street West, Hamilton, Ontario, Canada L8S 4K1 ABSTRACT The endogenous and exogenous regulation of activity was studied in the terrestrial slug Deroceras reticulatum. Hourly observations of activity and microclimate were made on 21 nights from May until October 1976. The data were transformed to account for non-linearity and analyzed using correlation and step-down multiple regression. Analyses were also performed using weather data from the previous hour. Regression models explained 80% to 87% of the variation in activity, a vast improvement on previous attempts to predict behaviour in the field. A model was also developed based on key limiting factors. This was based upon upper limits of activity associated with various factors, and the model explained 85% of the variation in the data. Laboratory experiments identified factors causally related to activity. The basic activity pattern arose from a circadian rhythm, entrained by a combination of light and temperature cycles. There was an inverse linkage between foraging and homing motivations. The clock ensured that activity rarely occurred in dangerous weather. Shelters were required in laboratory experiments to prevent artifacts associated with homing. Behaviour can completely reverse depending on whether shelters are available or not. A complex of environmental factors constrained activity. Unfavourable microclimates induced homing, via olfactory markers deposited inside shelters, rather than directly to limiting factors. Temperature was a key factor regulating activity. Light and wind were the next most important, respectively, and barometric pressure possibly exerted some influence. Evaporation depressed activity in the laboratory, but never completely suppressed activity of slugs with acces to moist shelters. The conditions constraining slug foraging were considerably less severe than those limiting locomotor ability. This safety margin between physiological and activity limits, the an- ticipatory nature of the circadian rhythm, and linkage of the clock to accurate homing combined to reduce the risks associated with foraging. Key words: circadian rhythms, microclimate, activity patterns, light, temperature, evaporation, modelling, Deroceras reticulatum, slugs, foraging. INTRODUCTION There has been intense interest in the en- dogenous control of activity, particularly drink- ing, eating, mating (Toates, 1980; Simpson & Bernays, 1985) and rhythmicity (Saunders, 1976; Beck, 1980). Such studies usually focus on regulating mechanisms or clocks, and ig- nore tradeoffs among behaviours (McFarland, 1977) or environmental constraints. There has also been considerable work relating the environment to activity, but these studies usually assume that behaviour reflects the im- mediate environment and ignore innate fac- tors (Johnson, 1969; Enright, 1970). Conse- quently, attempts to predict the activity of field populations have been relatively unsuccess- ful. Success requires a conceptual model that 199 recognizes that the basic pattern is endoge- nous, and that while certain factors modulate this pattern via innate clocks, others such as microclimate act as direct constraints. Very few studies have simultaneously addressed the integration of endogenous and exogenous control of activity (Bailey, 1975; Rollo, 1982; Cook & Ford, 1989). The key literature on field activity is tar- geted on such weather-sensitive organisms as molluscs or woodlice. The terrestrial slug Deroceras reticulatum (Müller) must secrete mucus to move, and it cannot prevent epider- mal evaporation (Machin, 1975; Prior, 1985). Abrupt changes in microclimate can be lethal, and given their renowned lack of fleetness, slugs are extremely sensitive to weather. Consequently they provide an excellent sys- 200 ROLLO tem for activity studies. The present study synthesizes laboratory experiments and field observations of key factors influencing activ- ity, and it incorporates these into predictive models. METHODS A holistic consideration of key endogenous factors and the numerous environmental vari- ables influencing activity is necessarily com- plex. The mathematical approach required for analysis makes the problem even more diffi- cult. Consequently, | provide a brief guide for reading and understanding this paper. This study is comprised of three separate ap- proaches. Firstly, activity on the soil surface in large field cages with underground shelters was monitored on nights spanning an entire season. This provided information on natural activity patterns where all variables act simul- taneously. Conventional multiple correlation- regression analysis was applied to these data to obtain empirical relationships. Rollo (1989) and Cook & Ford (1989) provide additional information on methodology. This analysis provided: (a) a large table of simple correla- tions among various variables and (b) regres- sion models incorporating these variables in descriptive equations. The second approach, limit modelling, was an innovative analysis applied to the same field data. This model was based on the sim- ple observation that at any given time, a sin- gle factor seemed to dominate with respect to limiting activity. A conceptual analogy might be limitation of animal growth by the least available essential nutrient (e.g. limiting amino acids). A similar view has been devel- oped regarding limitation of phytoplankton growth via various key factors (light, temper- ature, phosphorus, nitrogen, etc.). As one constraint is removed, the next sets the new upper limit. In studies on population dynam- ics, models based on dominant “key factors” provide another analogy (winter kill, preda- tors, parasites, etc.). It is important to realize that limit modeling is not a statistical tech- nique but an explanatory model. The third approach was a series of experi- ments aimed at demonstrating the causal re- lationship between activity and several key factors. Most of these experiments utilized an activity chamber that had a lower dark shelter and upper activity arena. Thus, it exactly par- alleled the structure of the field experiment. A general description of this approach is out- lined and deviations from standard protocol are elaborated for each experiment. Presen- tation of the results and discussion is non- standard. If each of the three approaches was discussed separately, information related to particular factors would be spread over three or four separate portions of text. To focus the paper, it was necessary to discuss each vari- able sequentially in the results and discus- sion. Under each heading, | group and syn- thesize the relevant results and discussion from the correlation matrix, regression mod- elling, and laboratory experiments. This is fol- lowed by a summary integrating these com- ponents and providing an overview of the major conclusions. Field Methods Activity was monitored in four 1.4 by 5.0 m outdoor cages at the University of British Co- lumbia. Each cage contained 10 underground shelters (19 cm diameter ceramic flower pots covered with 5 cm thick wooden planks with central entrance holes). Activity involved for- aging on the soil surface and accurate hom- ing (Gelperin, 1974; Rollo & Wellington, 1981). Polyester canopies prevented escape. Adult slugs (> 200 mg) were collected locally and introduced to the cages in May. The num- ber of animals outside of shelters was moni- tored by dim flashlight each hour between 4:00 p.m. and 9:00 a.m. on 21 nights (358 hourly observations) from May until October 1978. Brief illumination with dim light had no effect on immediate behaviour or the overall activity pattern. Microclimatic measurements were made simultaneously. The population was censused in the morning. Numbers var- ied (40 to 100 slugs) due to mortality but den- sity was comparable to local populations. Slugs immediately established normal hom- ing and foraging patterns but were allowed 5 d to adapt to the cages before observations. Mixed piles of lettuce and carrots were dis- tributed equidistantly from shelters. Shelters were watered regularly, so that slugs were always fully hydrated. The rationale and methodology for the factors investigated are detailed by Rollo (1982, 1989). Briefly, the variables were: (1) time of day (counted from noon so that 1:00 am = hour 13); (2) time of sunset (= phase), measured as hours and 10ths of hours; (3) scotoperiod (length of the night in minutes); (4) light intensity ‘(a Kahlsico® light meter scaled in foot candles); REGULATION OF ACTIVITY IN DEROCERAS 201 (5) changes in light intensity (change from previous to current hour); (6) phase of the moon (% of full moon); (7) air temperature (°C) recorded at 0, 10 and 30 cm above the ground with a YSI® telethermometer; (8) shelter temperature (19 cm below ground ina shelter); (9) temperature gradient (a differ- ence between shelter and surface tempera- ture in °C; (10) temperature changes (current surface temperature minus the previous hour’s); (11) ambient barometric pressure (millibars with a Belfort® microbarograph); (12) barometric pressure changes (current value minus the previous hour's); (13) percent relative humidity (R.H.), measured with a Bendex® psychrometer at ground level; (14) vapor pressure deficit (V.P.D.), calculated from tables using R.H. and temperature; (15) evaporation rate (mm*/minute), the evapori- meter a scaled 1 mm bore capillary tube with a filter-paper evaporating surface on one end; (16) wind measured in feet/h with a Casella® 3-cup anemometer located 1 m above short- cut sod. Light, temperature and moisture were measured inside cages. Correlation-Regression Analysis Rollo (1982, 1989) detailed the analytical methodology, and Cook & Ford (1989) sug- gested some extensions. Briefly, the propor- tion of the population active was used as the dependent variable to allow comparison among sampling dates when numbers varied. Proportions were transformed as suggested by Zar (1974): Y = ARCSIN (SQRT (Y+0.001)), where Y, the dependent factor, was the proportion of slugs active. Double precision correlation and step-down multiple regression analyses were performed. Scatter diagrams were examined to detect curvilinear relationships. Independent factors (e.g. time, light, temperature) were then transformed to obtain the best linear approximations, a step necessary to meet the assumptions of regres- sion analysis. Rather than submitting all pos- sible factors and their transformations for analysis, the models were generated using only a single representation or transformation of each, and only those deemed to be of pos- sible causal importance were used. In this way, | hoped to obtain models that could be related more directly to underlying mecha- nisms. For example, some factors were mea- sured in several ways (e.g. temperature at three heights, and water stress by R.H., V.P.D. and evaporation). Only a single factor, that with the best correlation with activity, was included for regression analysis. Similarly, only one transformation of a given variable was used (again that with the best correlation with activity). Polynomial expansions were used to account for non-linearity in some vari- ables. For example, temperature to the first, second and third powers (three factors) was included because this best accounted for the parabola-shaped relationship of activity to temperature (Fig. 8). Factors used are indi- cated Буа * in Table 1 and are summarized in Table 3. The total number of factors submitted for a single regression analysis, including those needed for polynomial expansions was 30, leaving 326 degrees of freedom. Because slugs travel slowly, current activity was also analysed using measurements from the pre- vious hour. Time and factors representing changes (e.g. temperature changes) were not altered for lagged analyses. Sometimes lin- earity required logarithmic transformation of the dependent variable. Therefore, models with logarithmic dependent variables were also generated for both current and lagged weather. Limit Modelling It appeared that at any given time, activity was limited by a key constraining factor. For example, light and evaporation might be ideal, but it was too hot or cold, or tempera- ture might be optimal, but it was too bright. Weather might even be unconstraining, but it might be the wrong time of day (i.e. the circa- dian rhythm limited activity). A model was suggested in which the innate clock gener- ated a basic activity pattern and weather im- posed constraints on its expression. When a given weather factor was graphed against ac- tivity, most showed a clear outer boundary that possibly represented levels at which this variable acted as the key limiting factor. Boundaries for the innate clock could also be estimated from plots against time. To obtain equations for these boundaries, the points along the boundary were submitted to regres- sion analysis. Thus, for any variable, there was a maximum amount of activity that would be allowed for any given value. For example, given a particular value of light intensity, the boundary equation would predict the maxi- mum activity that could occur. Other vari- ables, such as temperature, might predict higher or lower values. The limiting factor, 202 ROLLO GLASS LID DRYING OR HUMIDIFYING AGENTS UPPER TRANSPARENT - | FORAGING AREA MOIST PAPER OVER OPAQUE LID SHELTER ENTRANCE MOIST PAPER EEG RUBBER FIG. 1. Standard cage used in most laboratory ex- periments examining the pattern of activity and homing of Deroceras reticulatum. and the model’s overall prediction of activity, was the lowest value of activity predicted by any one of the considered variables. Rollo (1989) and Cook & Ford (1989) discuss fur- ther extensions and details. The model is not a statistical technique. Re- gression was employed merely as a conve- nient way of obtaining quantitative estimates of the outer boundaries. The validity of the model rests on its ability to predict activity. Clearly, if the assumptions of the model are incorrect, it should vastly overestimate activ- ity. Laboratory Experiments Correlation-regression detects associa- tions, but experiments were required to con- firm causality. Most used a standard appara- tus with a dark shelter on the bottom and an upper transparent foraging arena that could be exposed to light cycles (Fig. 1). The shelter was a Straight-sided bowl, 9 cm deep and 18 cm wide lined with moist paper towels to en- sure 100% R.H. A circular lid of black plexi- glas with a 2.5 cm entrance hole covered the top of this bowl and served as the floor for the foraging area. A tube of 1 mm thick transpar- ent plastic was fitted around the bowl so that 15 cm extended above the surface of the lid. The tube surrounding the bowl was painted black to keep the shelter dark, leaving the up- per 15 cm transparent. Another transparent bowl inverted over the top of the tube pre- vented escape. The floor of the upper forag- ing area was covered with moistened paper, and sliced carrot was provided as food. Cages were set inside environmental cham- bers upon 3 cm of foam rubber to minimize vibrations. A sheet of transparent plastic was attached across the door frame of the envi- ronmental chambers so that they could be opened for observations (with a dim flash- light) without exchange of air or fluctuations in temperature. Fifteen to 20 slugs were used in each experiment. The chambers were pro- grammed with the appropriate photoperiod and temperature, and the slugs were allowed at least 5 d for temperature adaptation and entrainment of their circadian rhythms. Slugs immediately established nocturnal feeding and homing patterns, with little activity in the light. Observations were made at half hour intervals throughout the activity period. Circadian Rhythms The basic activity pattern is generated by the endogenous clock, and it was necessary to characterize this before interactions with prevailing weather could be understood. Light cycles commonly entrain rhythms, so the ac- tivity of D. reticulatum was ascertained in standard activity cages following entrainment with various photoperiods and light intensities at 15°C. Temperature cycles also commonly entrain rhythms. Because Dainton (1954a) identified 21°C as a critical temperature for changes in activity, slugs were subjected to darkness and a temperature cycle of 12 h at 15°C and 12 h at 23°C to bracket 21°C. Con- trols were kept in darkness at a constant 15°C. Light and temperature cycles may in- teract to determine activity, so another exper- iment was performed with a photoperiod of L:D 12:12 h at 300 ft-c and a corresponding temperature cycle at 23° to 15°C. Activity was monitored over 24 h after two weeks in each regime to see if the typical activity pattern was expressed or whether it followed the temper- ature cycle. Food Distribution and Surface Activity Activity is largely foraging, so food available inside shelters might modify behaviour. This was tested using standard activity cages (Fig. 1) in a L:D 12:12 photoperiod at 15°C. Carrot slices were provided either inside shelters or REGULATION OF ACTIVITY IN DEROCERAS 203 in the upper foraging areas, and subsequent activity was compared. Ambient Light Intensity A light-choice chamber was constructed from a water-tight plastic box 16 by 20 by 4 cm deep. Half of the box was painted black, and the other half was transparent. The bot- tom was lined with moist paper to facilitate locomotion. After introducing the animals, the box was submerged 10 cm in a water bath at 15°C. A rheostat allowed variable control of a 200-watt incandescent light suspended 10 cm above the water and directly over the trans- parent section of the chamber. Thermocou- ples did not detect any variation in tempera- ture between the dark or light sections. Light intensities of 10, 240, and 2,838 ft-c (mea- sured inside the chamber) were employed. Deroceras reticulatum were collected from the field and kept in total darkness at 15°C for one week. Slugs were removed from these conditions and immediately placed in groups of ten directly in the center of the transparent section. Their behaviour was observed and after 1 h their positions were recorded. A few experiments were run for 8 to 12 h to detect adaptation. Atmospheric Moisture Sixteen D. reticulatum were placed in each of four standard chambers and entrained to a photoperiod of L:D 12:12 (800:0 ft-c) at 15°C for 5 d. An 8 cm petri dish containing calcium chloride was suspended in the upper area of two chambers. Dishes of distilled water were similarly suspended in two control cages. Each dish was covered with a screen to ex- clude slugs (Fig. 1). Carrot slices were pro- vided in the upper area as usual, but the sub- strate was not moistened. One day was allowed for the calcium chloride to dry the air. The shelters in all treatments were kept moist. The experiment was repeated using silica gel as a drying agent to reduce the possibility of confounding olfactory responses. RESULTS AND MODELLING Correlation-Regression Analysis Simple correlations between activity and in- dependent factors were used to select those for stepdown regression (Table 1). The four regression models (i.e. log and linear depen- dent variables versus current and lagged in- dependent variables) explained between 80% to 87% of the variation in season-wide activity (Table 2 provides an example). Independent factors included in models varied, particularly between those with different transformations of the dependent variable (Table 3). Simple correlations with current weather were better than for lagged data, except for barometric pressure and wind speed (Table 1). Regres- sion using lagged data explained only 1% more variation than models using current weather. Time and Circadian Rhythmicity Time was included in the analysis of field data to account for endogenous rhythmicity. The relationship was non-linear (Fig. 2), sug- gesting a quartic fit. The best relationship was obtained with the logarithm of time to the first, second and fourth powers (Table 1). All poly- nomials were selected in every model (Table 3). Sunrise and scotoperiod were included to account for seasonal changes in the phase and period of the rhythm (Fig. 12, Table 1). Scotoperiod was included in two models and the best-fitting model included scotoperiod and phase (Table 3). Field activity entailed an early evening peak followed by a gradual decline throughout the night (Fig. 2). Laboratory experiments dem- onstrated that this pattern arose endoge- nously. High light intensity was needed to en- train the circadian rhythm. With a light cycle employing 300 ft-c, D. reticulatum showed lit- tle pattern to its activity during the dark, and was frequently active in the light (Fig. 3). With an intensity of 800 ft-c, activity was restricted to the dark, but was evenly distributed throughout the night instead of showing the typical field pattern (Fig. 4). When photope- riod and temperature cycles were coupled, slugs displayed activity patterns identical to the field. Even a light intensity of 300 ft-c was effective if coupled with temperature cycles (compare Figs. 2 and 5). Animals maintained in darkness with an identical thermoperiod, however, had activity randomly distributed over 24 h and showed no changes at the tem- perature transitions. Slugs held in darkness and constant temperatures also showed a random pattern after the two-week period of exposure. 204 ROLLO TABLE 1. Correlation coefficients relating the proportion of Deroceras reticulatum active on the soil surface (Y) to environmental factors. Dependent Variable Y Log Y Log Y Log Y Log Y Log Y Log Y Log Y Log Y Log Y Log Y * Log-Time ы Time Log-Time Polynomial Phase Scotoperiod 0.1834 0.3505 0.7970 —0.0603 0.0597 0.3142 0.4822 0.8806 —0.0632 0.0579 Log Exponent Lag Log Scotoperiod Scotoperiod Light Light Light 0.0615 0.0579 —0.6146 —0.6508 —0.7973 0.0574 0.0578 —0.6756 —0.7167 —0.7620 Lag Log Light Light Changes Moon Log Light Changes Polynomials Phase Moon —0.6972 0.0546 0.4565 —0.0536 —0.0454 —0.6803 0.0818 0.5011 —0.0094 —0.0227 > $ É Changes in Barometric Exponent Barometric Lag Barometric Barometric Changes Moon Pressure Pressure Pressure Polynomial —0.0543 —0.0881 —0.1022 0.2040 0.3826 —0.0077 —0.0237 —0.0371 0.2165 0.2977 Wind Log Wind Exponent Lag Lag Log Speed Speed Wind Speed Wind Speed Wind Speed —0.2869 —0.3338 —0.2810 —0.2947 —0.3484 —0.2441 —0.2984 —0.2380 — 0.2533 —0.3047 Lag E Temperature Temperature Exponent Shelter Surface at at Wind Temperature Temperature 10 ст 30 cm 0.2886 0.0704 0.5840 0.5783 0.5724 0.2471 0.0947 0.7012 0.6735 0.6648 Lag Surface Lag Surface Temp. Surface Temperature Temperature Temp. Changes Temperature Polynomial Polynomial Changes Polynomial 0.5053 0.6802 0.6128 0.0053 0.4473 0.6053 0.7726 0.6719 0.3152 0.3697 Lag Temperature Lag Temperature Temperature Gradient Gradient Relative Gradient Gradient Polynomial Polynomial Humidity 0.6891 0.6067 0.7819 0.6979 0.6028 0.7954 0.6809 0.8484 0.7245 0.6087 Log Exponent Lag Exponent Vapor Lag Vapor Relative Relative Relative Pressure Pressure Humidity Humidity Humidity Deficit Deficit 0.6005 0.6016 0.5210 0.6419 0.5625 0.6140 0.6017 0.5137 0.6870 0.5969 * Evaporation * Lag Evaporation 0.6523 07191 —0.6216 —0.6535 *Indicates a factor included for selection in multiple regression analysis. Dependent variables were first transformed to their arcsine as detailed in the text. Polynomial values were obtained by taking the square root of the r? value for the best fitting regression model. Small numbers were added to original observations to allow logarithmic transformations when zero values were possible (light = .01, moon = 0.001, wind = 0.000001, temperature = 0.01). When exponential transformations or polynomials were calculated, some variables were scaled down by multiplying by fractions (scotoperiod = 0.001, light changes = 0.001, moon = 0.01, temperature = 0.01, relative humidity = 0.01). REGULATION OF ACTIVITY IN DEROCERAS 205 TABLE 2. Regression analysis relating the activity of Deroceras reticulatum (Y) to concurrent environmental factors. Superscripts indicate polynomials. Variable Coefficient Log Time — 19138323 Log Time ? 12.84341 Log Time * —2.76423 Log Scotoperiod —0.39988 Log Light —0.04566 Surface Temperature 7.05774 Surface Temperature * —43.62514 Surface Temperature ° 69.36275 Temperature Gradient 3 —239.23158 Temperature Gradient * —1380.73130 Log Wind 052017 Barometric Pressure —0.00419 Constant 8.92756 Standard Error Probability 2.880 0.0000 2.453 0.0000 0.469 0.0000 0.152 0.0088 0.011 0.0000 2.285 0.0030 13.790 0.0022 26.700 0.0106 51.060 0.0000 429.600 0.0018 0.021 0.0000 0.002 0.0481 2.387 0.0003 r = 0.7989, p < 0.00001. Temperature data were scaled by multiplying by 0.01 before analysis. Base 10 logarithms. Small numbers were added to time (.01), light (.01) and wind (.000001) so logarithmic transformations could be taken with zero values. An arcsine transformation was performed on Y as described in text. TABLE 3. Summary of factors included in four separate multiple regression models describing the activity of Deroceras reticulatum. Number of Factors in Factor One Model Time Wind Surface Temperature Temperature Gradient Scotoperiod Ambient Barometric Pressure Barometric Pressure Changes Changes in Light Temperature Changes Shelter Temperature Ambient Light Phase (Sunset) Moon Phase Evaporation `` ot ot ot AA A | | ADO = |W Potential Number Total Number of Factors of Factors Available For Selected In Percent Selection Four Models Selected 12 12 100 4 4 100 12 6 50 16 8 50 4 2 50 4 2 50 16 6 38 16 6 38 16 5 31 4 1 25 4 1 25 4 1 25 4 0 0 4 0 0 The four models fitted were for logarithmic and non-logarithmic dependent factors (activity) each used with lagged and current independent factors (e.g. light, temperature, humidity). Food Fresh carrot slices placed inside shelter (Fig. 1) greatly reduced activity, but as food deteriorated, slugs emerged more frequently (Fig. 6). Full activity only occurred with all food on the surface. Ambient Light Intensity Many slugs sheltered in the corners of the transparent end of the light-choice apparatus, instead of finding the dark. Consequently, number active in the light was a better mea- sure of the influence of light. Deroceras retic- ulatum strongly responded to light at constant temperature (Fig. 7, Table 5) and showed no evidence of adapting, even after several hours. At 10 ft-c the repellent/inhibitory effect was evident. At the highest intensity, all ac- tivity ceased. Slugs in strong light withdrew their heads beneath their mantles and often buried the front of the body in that of another slug or in corners of the apparatus. Individu- als in the dark remained active or rested nor- mally. In bright light, most individuals moved 206 ROLLO 25 — 3 5 1.00 100 D D > ® 15007 |, > 15 а ES 2 0.80 086 a о 5 2 En = = O = 10004 < 0.50 0.6 3 [10 © = z = a © 5 р © E = Œ 0.40 0.4 =. 500 À © = [bs © 0.20 0.2 Ice | — | 0 0 0.0 0 4 5 6 7 8 9 10 11 121am2 3 4 5 6 7 8 9 10 TIME FIG. 2. Activity pattern of Deroceras reticulatum on May 27-28, 1976, illustrating the typical integration of activity with light, temperature and evaporation. LIGHT 0.90 | (300 FT-CANDLES) u = - о =. 0-80 a © 070 + 3 3 060 a O a 0:50 u O м z 0-40 O= CAGE |, 12 SLUGS © = m= CAGE 2, I2 SLUGS x 0-30 O a e) a 0-20 A E a E E Z < a a 0-10 o о Е > 0.00 о о о о о о о о O o о о о о о 3 о o о = о o о o E © o o = a 2 a 2) + a) © - © 9 = a m TIME FIG. 3. Typical activity pattern of Deroceras reticulatum in a photoperiod of 16 h light (300 ft-c) and 8 h dark at 15°C. REGULATION OF ACTIVITY IN DEROCERAS 207 2) о O PROPORTION OF POPULATION ACTIVE (N O EZ Ipm 2 LIGHT 3 Mar 33% 672270208 9 TIME FIG. 4. Typical activity pattern of Deroceras reticulatum entrained to a photoperiod of 12 h light (800 ft-c) and 12 h dark at 15°C. to the extreme end of the dark section. Those that ventured into the light, turned sharply af- ter a few cm and returned to the dark. Such avoidance and the inhibition of activity attest to strong, long-term constraint by light. There was a strong negative correlation between light intensity and field activity (Table 1), and this was included in one model (Table 3). Changes in Light Intensity Activity was weakly correlated with changes in light intensity, but non-linearity was ac- counted for by a quartic expansion (Table 1). Scatter diagrams indicated rapid decreases in activity with increasing light, rapid increases with decreasing light and maximal activity in darkness. Polynomials of changes in light were included in three models (Table 3). Ambient Temperature Activity was highly correlated with surface temperature (Table 1). Scatter diagrams sug- gested a quartic fit (Fig. 8). A third order ex- pansion was included in the analysis (Table 1), and polynomials were included in 3 mod- els (Table 3). Deroceras reticulatum's activity peaked at 15°C and fell rapidly above 20°C (Fig. 8). Slugs active at higher temperatures were homing. Activity declined less precipi- tously with temperatures below 15°C, and no slug emerged below 3°C. At temperatures of 5°C and less, slugs clustered at the shelter entrance and those emerging returned to shelter after travelling only 5 to 20 cm. Temperature Gradients The gradient between the shelter and the soil surface was highly correlated with activ- ity. A quartic polynomial accounted for non- linearity (Table 1), and these were included in every model (Table 3). Temperature Changes. Rapidly increasing the temperature during the normal activity period in standard labora- 1-00 2 ZELL LIGHT DARK LIGHT 0-90 (300 FT-CANDLES) PROPORTION OF POPULATION ACTIVE O oO O 0-40 0-30 0:20 = < < < o oO 0-10 5 o > 75 0.00 о о о о о о о о о о о о о о Oo o o o o ©) о Oo o o N |) о = a = © m + (19) © N = = = pm TIME FIG. 5. Typical activity pattern of Deroceras reticulatum entrained to a photoperiod of 16 h light (300 ft-c) and 8 h dark in a thermoperiod of 16 h at 23°C and 8 h at 15°C. TABLE 4. Regression model estimating the expected activity of Deroceras reticulatum (Y) due to the circadian rhythm when weather was not constraining. Variable Coefficient Standard Error Probability Time 5.9237 0.5380 0.00001 Time ? ESAS 0.1454 0.0001 Time * 0.3718 0.0507 0.0001 Time * Scotoperiod 0.0020 0.0003 0.0001 Time * Temperature 0.0676 0.0186 0.001 Phase 0.5784 0.0958 0.0001 Phase * Temperature 0.0148 0.0035 0.0001 Constant 0.7287 0.3669 0.0510 Time was scaled originally by multiplying by 0.1. Temperature = shelter temperature г? = 0.879, р < 0.0001. Y = the proportion of the population active. tory cages caused slugs to home, showing that 9). The form of the activity pattern was normal they respond to temperature changes. The but numbers were reduced throughout the simple correlation of temperature changes night by dry air. Slugs in dry air ate quickly, and with activity was greatly improved by a quartic then returned directly home. Slugs in moist air expansion (Table 1). Three regression models moved and ate more slowly, and frequently included polynomials for temperature changes wandered in exploratory patterns or rested. (Table 3). Evaporation was highly correlated with field activity, V.P.D. was slightly inferior and R.H. was worst (Table 1). R.H. was highly corre- Laboratory experiments using different dry- lated with V.P.D. (г? = 0.87). The correlation ing agents were similar and were pooled (Fig. between V.P.D. and evaporation was surpris- Atmospheric Moisture REGULATION OF ACTIVITY IN DEROCERAS 209 TABLE 5. The influence of light intensity on the activity of Deroceras reticulatum in a light-choice chamber at 15°C. Number Light Intensity of Slugs (ft-candles) Used 0 0 10 30 240 50 2838 50 1:00 0:90 0:80 0:70 EIICHTS=ZORE 0-60 0:50 0.40 0-30 0.20 0:10 PROPORTION OF POPULATION ACTIVE Proportion Proportion Resting in the Active in the Light Light 0.50 (expected) 0.50 (expected) 0.267 0.267 0.240 0.102 0.102 0.000 iz | O Io IH Wen = | © Lier ©7200 2:0 HO оо ооо OO ION or 292 оо ¡OO MO” COMMONS ON N il) A A © NO о pm TIME @= FOOD OUTSIDE SHELTER (5 DAYS OLD) (N= 27) v=FOOD INSIDE SHELTER (14 DAYS OLD) (N= 15) и = FOOD INSIDE SHELTER ( 5 DAYS OLD) (N= 15) FIG. 6. Changes in the activity pattern of Deroceras reticulatum in response to the availability of food within shelter. Symbols represent e and 8 food inside shelter, 14 days old. ingly poor (г? = 0.69), and that between В.Н. and evaporation was lowest (r? = 0.58). There was little activity below 65% R.H. Slug activity decreased exponentially with increas- ing evaporation (Fig. 10). Despite the clear laboratory results (Fig. 9), evaporation was food outside shelter, 5 days old; Y food inside shelter, 5 days old; not included in any models of field activity. No severe evaporation was measured dur- ing 1976, so activity was seldom restricted by this factor. During 1977, however, dry windy weather was associated with very low activity. 210 ROLLO 1-00 uw e > e = . LIMIT MODEL o = N —--- У=-0:00056Х + 0:8189 \ = 0:75 N = pe REGRESSION MODEL \ 5 \ ARCSIN(VY_ )= — 0:1020(LOG(X) )+ 0-5005 à x 2- 0:64, p<0-0000 a | : u 0:50 O = о + x oO S a \ e o e o 0:00 © sun co de e o 90 ооо в © . o 500 1000 1500 2000 2500 3000 3500 LIGHT INTENSITY (FT-CANDLES) FIG. 7. Activity of Deroceras reticulatum in relation to light intensity illustrating the best-fitting regression line and the outer activity limit. Wind Speed Wind was negatively correlated with activity (Fig. 11), and its logarithmic transformation was included in every model (Table 3). No clear upper limit of activity was discernible (Fig. 11), so this factor was not included in the limit model. The influence of wind may be un- derestimated because the canopies reduced air currents and few windy nights occurred. In 1977, winds above 40 km/h drastically cur- tailed activity. On windy nights, slugs re- mained active in sheltered areas but avoided open spaces. Ambient Barometric Pressure On two occasions, large numbers of D. re- ticulatum became active in mid-day tempera- tures and bright sunlight about 10 minutes prior to thunderstorms. Prevailing microcli- mate did not reflect the approach of rain, but barometric pressure changes are known to precede storms. Ambient barometric pres- sure was only weakly correlated with activity (Table 1), but was included in two models (Table 3). Barometric Pressure Changes Change in barometric pressure was better correlated with activity than was ambient barometric pressure, and a quartic fit best ac- counted for non-linearity (Table 1). Polynomi- als were included in three models (Table 3). Limit Model For the limit model, predictions of activity based upon the circadian rhythm when un- constrained by weather were required. Activ- ity on 15 days over the season was plotted REGULATION OF ACTIVITY IN DEROCERAS 211 oe / \ / \ [LOOM À re / : / ! : / q : | qe \ e z o ш eee = | e = \ e : nN © 075 Vee acces Gi 72 / © 3 \ e ol © . % > © = | 4 1% SEE < ee “o © % 0 =) el e e ge a e > Ш 5 o A E a le 2 e wo ee < + 5 ав о о ee e . a | ® eme © O° \ : Ir „ 0:50 Dede = AN о es e? e e re | %e x ae LIMIT MODEL === = о: o E E \ : = REGRESSION MODEL—— о 8 .:3 © \ а 0:25 CA EN N xo. N N vs e 0-00 o / mn be cs Фо ce о =2 O 2 4 6 8 10 12 14 16 1827207 (22) 24 26" 828)" 30) 73275347 736 SURFACE TEMPERATURE (°C) FIG. 8. The activity of Deroceras reticulatum in relation to surface temperature illustrating the best-fitting regression line and the outer activity limit. (e.g. Fig. 12). By comparing field patterns on successive nights, and by examining patterns obtained in laboratory experiments with vari- ous photoperiods (e.g. Fig. 5), the innate ac- tivity pattern that would be freely expressed at any given time or date was estimated. A re- gression model was fitted to these estimates using scotoperiod, time of sunset, a quartic expansion of time, and interaction terms. Shelter temperature was included because temperature may influence the phase and pe- riod of rhythms. The model fitted the esti- mated points well (г? = 0.88) and was used to represent endogenously generated activity in the limit model (Table 4). Although emergence was largely nocturnal (Figs. 2, 7), D. reticulatum was often active on cool or dull days in the spring and fall. Al- though the best correlation of activity was with the logarithm of light, a simple linear model was a reasonable approximation for the limit model (Fig. 7). Points along the outer bound- ary were estimated by eye, and a regression line was fitted where: Y = 0.8189-0.00555(L), © = 0.96, р < 0.001, and Y = proportion of slugs active and L = light intensity in ft-c. A discrete upper boundary to activity for any given temperature was apparent (Fig. 8). Points delineating this outer limit were esti- mated by eye, and fitted to a fourth-order polynomial. The resulting equation used in the limit model was: Y 0.204417 (T) — 0.007304(T*)-0.556349, г? = 0.89, р < 0.001 where Y = proportion of slugs active and T = surface temperature (°C) For evaporation a linear equation ade- quately represented the boundary (Fig. 10): Y = 0.9397—0.6696(Е), г? = 0.98, р < 0.001 where Y = proportion of slugs active and E = evaporation rate (mm*/min). 212 ROLLO WET AIR NO DATA DRY AIR PROPORTION OF POPULATION ACTIVE NO DATA TIME FIG. 9. The influence of evaporative stress on the activity of Deroceras reticulatum. Slugs were entrained to a 12 h photoperiod (800 ft-c) at 15°C. Clearly, the equations generated for the limit model are simply a way of quantifying the subjective estimate of the outer boundaries. The effectiveness of this exercise is best ap- preciated by examining the relevant figures. Activity was calculated for each of the above factors, and the lowest estimate was selected as the model’s prediction for that hour. Rollo (1982, 1989) provides a flowchart and more detail on the procedure. Values from the best-fitting regression model, the limit model and actual activity for each hour are illustrated for five nights in Figure 12. Re- gression of the limit model predictions on ac- tual activity for the five days gave an г? of 0.85 (p < 0.001), equivalent to the statistical mod- els. Considering that only four factors were included, this reflects remarkable accuracy and suggests that a key limiting factor ap- proach is warranted. The circadian rhythm limited activity in 71% of observations. Temperature was limiting in 19% of observations, while light and evapo- ration were only constraining in 6.4% and 3.2% of cases, respectively. Values sufficient to prevent activity occurred in 14% of cases for time, 13% for temperature, 11% for light and 10% for evaporation. Thus, although all of these factors had similar potential for limit- ing activity, it was usually the endogenous rhythm that acted. DISCUSSION Circadian Rhythms The results illustrate that activity has an en- dogenous pattern (Figs. 2, 12) entrained: by light cycles (Figs. 3, 4, 5). Considerable re- REGULATION OF ACTIVITY IN DEROCERAS 213 1-00 \ a. ao \ > À 2 ш 075 j:- \ ei = N 5 ь. ; < Е: e N z N о oe? SS N < CN = f «e... © a 050 ES N в + u o = o = x = 0-25 о x a 0-00 г2 = 0-5172, LIMIT MODEL Y= 0.9397 — 0: 6696(x) REGRESSION MODEL —— LOGIO(ARCSIN(V Y+0-001)=- 0:2256— 0-7183(X) P< 0-0000 EVAPORATION (mm minute) FIG. 10. The activity of Deroceras reticulatum in relation to evaporation rate in the field illustrating the best-fitting regression line and the outer activity limit. search supports this view (Newell, 1965, 1968; Lewis, 1969a,b; Daxl, 1969; Pinder, 1969; Sokolove et al., 1977; Morton, 1979; Beiswanger et al., 1981; Rollo 1982; Hess & Prior, 1985; Wareing & Bailey; 1985; Ford & Cook, 1987, Cook & Ford, 1989). Low photo- phase intensities were ineffective, which ex- plains why Dainton (1954b) could not entrain D. reticulatum using 40 ft-c. Similarly, Ware- ing & Bailey (1985) found that activity was never completely nocturnal with 10 ft-c. Al- though slugs sheltering underground could not detect light, they were activated by their clocks and waited at their shelter entrance for favourable conditions. This would expose them to intensities sufficient to entrain and shift the rhythm seasonally. Dainton (1954a) and Pinder (1969) previ- ously concluded that temperature cycles en- trained activity. Slugs may require light to rec- ognize shelters, so complete darkness may have disrupted homing in my experiments. Moonlight might serve for field orientation. Re- gardless, entrainment was improved when cy- cles of temperature and light were coupled (Fig. 5). Wareing & Bailey (1985) also showed that temperature cycles reinforced light. They obtained activity close to the field pattern when temperature fell at dark and rose at dawn by 2°C/h. Ford & Cook (1987) obtained only weak entrainment of Limax pseudoflavus with tem- perature cycles, but these also strongly en- hanced entrainment by light. Photoperiod ap- pears to be the key factor entraining D. reticulatum, with temperature cycles reinforc- ing and perhaps modifying the response. Activity was largely concerned with forag- ing (Fig. 6). With subterranean food, slugs may remain underground. That microclimate constrains but does not determine activity patterns contrasts with earlier views that rhythms initiate activity which then responds directly to microclimate (Newell, 1965, 1966, 1968; Lewis, 1969 a,b). A critical aspect of activity is diurnal homing (Rollo, 1982) using pheromone markers (Gelperin, 1974; Cook, 1980; Rollo & Wellington, 1981). Thus, activ- ity does not conform to the Fraenkel & Gunn (1961) interpretation of orientation to the lim- iting factors themselves. Such taxes or kine- sis may be employed only when lost, during dispersal, or when shelter is disturbed. The circadian rhythm partially explains the high correlation of activity with time (Table 1), 214 1.00 LL) > > O < 0.75 zZ je) > < 1 5 & 0.50 oO 5. LL. oO г О 0.25 be oc O a O œ RE оф . > z LE L 1 2 0.00 gir tele NE pee UA re ma? . © ROLLO г2 = 0.1039, P<0.00000002 10 15 20 25 WIND SPEED (km/h) FIG. 11. The activity of Deroceras reticulatum in relation to wind speed in the field illustrating the best-fitting regression line. and why time was included in all of the re- gression models (Table 3). Most microcli- matic factors were also highly correlated with time. It is this very fact that makes the rhythm so valuable, ensuring that slugs anticipate fa- vourable microclimate (Fig. 2), and return home prior to danger. Such anticipation is es- sential for animals that are slow, highly vul- nerable and dependent on capricious olfac- tory navigation (Rollo et al., 1983a). The limit model suggested that in 71% of cases, activ- ity was constrained by the clock rather than directly by climate. The greatest danger, evaporation, was only directly constraining in 3.2% of cases. It may come as some surprise to many (e.g. Dainton, 1989) that the activity and orientation of slugs, although largely adapted to deal with microclimatic problems, largely achieves this indirectly by reliance on innate clocks and pheromones. Further fine tuning of these tactics involves direct re- sponses to key environmental factors as dis- cussed below. Ambient Light Intensity Activity of D. reticulatum was negatively correlated with light (Table 1) and as in other species, the response was logarithmic (Fig. 7) (Frandsen, 1901; Lewis, 1967; Poulin, 1967; Rollo, 1982). Whether light regulates activity of D. reticulatum has been controversial. Dark-adapted molluscs initiate movement when illuminated (Dainton, 1954b; Sokolove et al., 1977; Gelderloos, 1979; Beiswanger et al., 1981; Wareing & Bailey, 1985), but be- cause this only lasts 1 h, Dainton (1954b) concluded that light could not sustain locomo- tion. Even the heat from a fluorescent lamp may affect locomotion (Dainton & Wright, 1985), and only Dainton (1954b) previously employed a water bath to control this. My re- sults show that D. reticulatum is repelled or inhibited by light (Table 5) and this was effec- tive even after several hours. In other exper- iments, slugs occasionally failed to home be- fore the lights came on. Following a burst of searching, such individuals adopted the con- tracted resting position for up to 12 h. Thus, light may stimulate homing of active slugs or escape if shelters are disturbed. Following such responses it strongly inhibits activity. Supporting this conclusion, Newell (1965) ob- served that bright light prevented activity. — Dax! (1969) found that continuous light de- | REGULATION OF ACTIVITY IN DEROCERAS 215 too | JUNE 17 | 1:00 OBSERVED ‘50 REGRESSION======= PROPORTION OF POPULATION ACTIVE 100 | JULY 13 al PACIFIC STANDARD TIME + I h FIG. 12. Activity of Deroceras reticulatum over the 1976 season with predicted activity from the best-fitting regression model and the limit model. Arrows mark the time of sunset (S) and sunrise (R). layed emergence by 2 h, and Ford & Cook (1987) found that bright light curtailed activity for 17 days. Some conflict among studies arises due to perspectives. Dainton (1954b) was looking for factors stimulating activity. Light only did so briefly, so she interpreted subsequent inactiv- ity as lack of stimulation rather than as inhi- bition. Subsequent authors suggested that slugs were activated by other factors, but light inhibited activity (Karlin, 1961; Getz, 1963; Newell, 1965, 1966, 1968; Dundee, 1977). Deroceras reticulatum is considerably less sensitive to light than other molluscs, such as L. maximus (Rollo, 1982). This accounts for differences among studies using various in- tensities, as well as frequent daylight activity in cool, moist microclimates (Ingram, 1941; Dainton, 1954a; Getz, 1959; Rollo, 1974). Daylight activity was most frequent in spring and fall, suggesting that animals may be more nocturnal at higher temperatures (Cameron, 1970b; Richter, 1976). Changes in Light Intensity Change in light intensity was included in three models (Table 3). Trends in illumination are highly correlated with pending microcli- mate, and it would be adaptive to differentiate the degree and direction of such changes. Certainly the circadian rhythm is differentially affected by light-to-dark versus dark-to-light transitions. Moisture All field observations were of fully hydrated animals. Slugs regulate their hydration by se- lecting appropriate substrates and actively 216 ROLLO absorbing water through the foot (Prior, 1985, 1989). As soils dry and molluscs dehydrate, activity is severely constrained (Bailey, 1975; Smith, 1981; Rollo et al. 1983b; Phifer & Prior 1985; Young & Port, 1989). The importance of hydration was illustrated by the observation that activity in watered field cages continued unabated, whereas adjacent natural popula- tions ceased activity altogether during ex- tended dry periods. Despite numerous claims that molluscs need moist substrates for loco- motion, hydrated D. reticulatum readily for- aged across dry surfaces. This requires greater mucus production however (Blinn 1963, Judge 1972), and presumably in- creases the risk of being immobilized. Dead, dehydrated slugs are commonly found on the soil surface, attesting to the high risks of foraging. Surprisingly, dry air did not deter activity (Figs. 9, 10). The circadian rhythm effectively limited field activity to peri- ods of low evaporation (Fig. 2), and evapora- tion was not included in any models. The lab- oratory experiments showed that slugs reduce their exposure to dry air, but even then there was no obvious difference in the amount of food eaten. Slugs simply changed their rates and types of activities. Other authors have documented that gas- tropods are inhibited by or attempt to escape desiccation (Lewis, 1969a; Cameron, 1970a; Abdel-Rehim, 1983), but Dainton (1954a) concluded that R.H. did not stimulate activity. Her animals, however, were probably not de- hydrated. Dehydrated gastropods are stimu- lated by water uptake (Herreid & Rokitka, 1976; Takeda & Ozaki, 1986). Hess & Prior (1985), using activity wheels, found that L. maximus increased its intensity and duration of locomotion in dry conditions. This was probably an escape response similar to the “lights on” burst of activity common in exper- iments lacking shelters (Ford & Cook, 1987). Following initial escape responses, slugs ag- gregate and “huddle” to reduce evaporation (Rollo, 1974; Richter, 1976; Cook, 1981; Waite, 1988). Moist air induces activity of hud- dling slugs (Rollo 1974), and | observed that they can locate moistened pieces of paper from a distance. Thornthwaite (1940) pointed out that R.H. and V.P.D. inadequately esti- mate evaporation and that R.H. and V.P.D. were in fact less correlated with activity than evaporation (Table 1). Neither R.H. nor V.P.D. is applicable if body and air temperatures dif- fer and evaporation drastically alters the tem- perature of slugs (Hogben & Kirk, 1944). Sim- ilarly, comparing fixed humidities at different temperatures is a faulty experimental design (e.g. Dainton, 1954a) because evaporation then differs markedly at each temperature. Such problems, and the variation of response with hydration, probably accounts for contra- dictory observations among studies (Webley, 1964; Hunter, 1968; Crawford-Sidebotham, 1972; Baker, 1973; Stephenson, 1973). In summary, fully hydrated slugs may be far less sensitive to humidity than previously sus- pected, and constraining rates of evaporation may be seldom encountered during nocturnal activity. Shelter or soil moisture is much more important, and dehydrated molluscs may re- spond strongly to evaporation and orient di- rectly to humidity gradients. Ambient Temperature Temperature strongly influenced activity, and polynomials were included in most mod- els (Tables 3). Temperature was next most important after the circadian rhythm in the limit model. Although D. reticulatum can crawl near 0°C and into the lethal range above 30°C, slugs active below 5°C or above 21°C were homing. Maximum activity occurred be- tween 9°C and 20°C, close to previous obser- vations (Carrick, 1942; Newell, 1965; Ware- ing & Bailey, 1985). Activity declined slowly from 16°C to 19°C and then fell precipitously (Fig. 8). Although activity occurs below 5°C (Carrick, 1942; Mellanby, 1961), this is infre- quent (White, 1959; Karlin & Naegel, 1960). A key insight is that activity is much more con- strained than physiological or locomotor ca- pabilities, a feature that adaptively avoids risk. Temperature Gradients Deroceras reticulatum shows strong prefer- ences in temperature gradients, with most choosing 13°C to 19°C (Dainton, 1954a, 1989; Chichester, 1968). This matches the ambient temperature selected for activity (Fig. 8). Getz (1959) found aggregation at a higher range (18°C to 24°C) perhaps due to acclima- tion. The temperature gradient polynomial had an even higher r value than that for am- bient surface temperature (Table 1). Dainton (1954a, 1989) suggested that diurnal activity involved movements along temperature gradients, and polynomials were indeed se- lected in all models (Table 3). Dainton (1989) still supports her theory that temperature re- REGULATION OF ACTIVITY IN DEROCERAS 217 sponses explain all activity and that clocks are unnecessary. The close agreement of the literature with the field results (Fig. 8) indeed suggests that the distribution of slugs in space and time largely involves tracking optimal temperatures. Vertical movements may be a widespread strategy for therm- oregulation (Jaremovic & Rollo, 1979). Max- imum shelter temperatures occurred in early evening, when surface temperatures plunged. The activity pattern of D. reticulatum avoided maximum temperatures both on the surface and underground. Slugs have a geo- tropic rhythm that may be relevant here (Cro- zier & Wolf, 1929). Death from heat was im- portant during summer, so timed vertical movements would be adaptive. Temperature Changes Dainton (1954a) showed that temperatures falling below 21°C stimulated movement of D. reticulatum, as did temperatures rising above 21°C. Constant temperatures or those in the opposite directions had no effect or were in- hibitory. Subsequently, molluscan activity was interpreted in relation to temperature changes, often with scanty evidence (Karlin, 1961; Blinn, 1963; Henne, 1963; Arias & Crowell, 1963; Hunter, 1968; Daxl, 1969; Baker, 1973; Bailey, 1975; Machin, 1975; Herreid & Rokitka, 1976; Dundee, 1977). Newell (1968) and Ford & Cook (1987) con- cluded that the diurnal pattern of temperature could not possibly explain the form of field activity patterns, and | agree. Lewis (1969a) found no evidence that activity of Arion ater was related to temperature changes. | found, however, that rapid increases in temperature above 21°C induced homing, and Dainton & Wright (1985) also showed that A. ater re- sponds to temperature changes. The field results (Fig. 7) confirm the signif- icance of 21°C, in that slugs were inactive or homed above this point. Dainton’s (1954a,b) slugs had no shelters, and extrapolating her results to the field is difficult because locomo- tor responses are involved in both homing and foraging. Possibly temperatures falling below 21°C (evening) initiate emergence from shelters, in concert with the clock. Tempera- tures rising below 21°C (morning) or falling above 21°C (late afternoon) may inhibit emer- gence or induce homing. Temperatures rising above 21°C (late morning) may induce hom- ing in exposed slugs or escape responses if shelters overheat. Thus, 21°C may represent a set point for switches between activity and homing regulated by temperature changes. As discussed, cycles of such changes rein- force the innate rhythm. Thus, temperature changes may be important in regulating ac- tivity, but the pattern of activity is endoge- nous. This interpretation differs from those of earlier authors who did not consider homing. | predict that pronounced changes in behaviour may occur near 21°C and that a similar threshold may exist near 5°C. This variable was calculated simply as the hourly change in temperature with no con- sideration of the 21°C threshold. Dainton (1989) showed that the rate of change is di- rectly related to the degree of induced activity, and this is supported here. Better results might be obtained by linking the degree of change to its proximity to 21°C. The polyno- mial for temperature changes had a lower value than that for ambient temperature or the temperature gradient (Table 1), but tempera- ture change polynomials were included in three models (Table 3). Possibly temperature change and gradient polynomials measured the same thing. Slugs responding to temper- ature changes by homing would appear to be orienting to temperature gradients because of precipitous temperature changes at shelter entrances. Alternatively, slugs rested at pro- gressively lower depths to avoid surface tem- perature extremes. Thus, appropriate thermal settings may be achieved by indirect and di- rect responses. Wind Speed Evaporation is greatly increased by con- vection (Thornthwaite, 1940; Machin, 1964), which explains why strong winds repel gas- tropods (Kalmus, 1942; Dainton, 1943, 1954b; Richter, 1976; Dundee, 1977). In ad- dition, olfactory orientation to shelter and food (Kalmus, 1942; Kittel, 1956; Gelperin, 1974; Cook, 1979; Rollo & Wellington, 1981), is dis- rupted by turbulence. Although wind inhibited activity, there was high variation (Fig. 11), possibly associated with variable air masses. Webley (1964) concluded that activity was only reduced by winds above 20 km/h. De- spite the high variation, wind was included in every model (Table 3) attesting to its impor- tance as a key factor. Barometric Pressure Activity in hot, dry weather preceding storms suggested that D. reticulatum detects 218 ROLLO barometric pressure. Ambient barometric pressure or polynomials for barometric pres- sure changes were included in every model (Table 3). With their hydrostatic skeletons, slugs may be ideally suited to detect it. SUMMARY AND SYNTHESIS Behaviour must be interpreted relative to the activity framework of these animals: noc- turnal activity on the soil surface followed by homing and inactivity inside shelters. Animals lacking shelters may orient directly to micro- climatic gradients to escape, and such re- sponses may be exactly opposite to those ex- hibited by animals with shelters. Bright light, for example, briefly stimulates exposed slugs but inhibits those with shelters. This is a gen- eral point because even cockroaches and woodlice, which served as “models” for Fraenkel & Gunn’s (1961) interpretations of direct orientation, are shelter dependent. In addition, experiments without shelters mea- sure mobility or physiological tolerance rather than activity. The range of conditions that an- imals venture into from shelters is much more restricted. This study demonstrated the priority of cir- cadian rhythms in regulating foraging and homing. The clock ensures that activity rarely occurs in risky microclimates so that direct avoidance is seldom needed. A combination of the circadian rhythm, accurate homing, and safe activity limits buffer slugs against poten- tially lethal conditions. A complex of environ- mental variables further constrains activity. Soil moisture, particularly that of shelters, is probably most important. Given suitable sites for rehydration, evaporation may not be as limiting as was previously assumed. Temper- ature was a key factor, and future experi- ments should address the relationship be- tween responses to temperature change and orientation to gradients. Light and wind were also key factors, and barometric pressure may influence this species. This study showed that it is indeed possible to quantify the expression of general activity, and this could be extended to examine the occurrence and rate of particular behaviours within the activity framework. If indeed the key limiting factor approach used in the limit model is gen- erally applicable, rather simple models might be capable of describing the complex inter- play of endogenous and exogenous factors. ACKNOWLEDGMENTS | wish to thank Mr. R. Bereska and Ms. R. Jaremovic for help with the field work. Dr. G. Eaton provided invaluable statistical advice, and Mr. D. Pierce provided some meteorolog- ical data. Dr. W. G. Wellington contributed nu- merous suggestions and insights for this project. The paper was improved by com- ments and an unpublished manuscript sup- plied by Dr. A. Cook. Financial assistance was provided by a grant from the Natural Sci- ences and Engineering Research Council of Canada. | was supported for part of this study as an N.S.E.R.C. Fellow. LITERATURE CITED ABDEL-REHIM, A. H. 1983, The effects of temper- ature and humidity on the nocturnal activity of the different shell colour morphs of the land snail Ari- anta arbustorum. Biological Journal of the Lin- nean Society, 20: 385-395. ARIAS, В. & H. H. CROWELL, 1963, A contribution to the biology of the grey garden slug. Bulletin of the California Academy of Sciences, 62: 83-97. BAILEY, S. E. R., 1975, The seasonal and daily patterns of locomotor activity in the snail Helix aspersa Müller, and their relation to environmen- tal variables. Proceedings of the Malacological Society of London, 41: 415-428. BAKER, A. N., 1973, Factors contributing to the initiation of slug activity in the field. Proceedings of the Malacological Society of London, 40: 329— 333. BECK, S. D., 1980, Insect photoperiodism. Aca- demic Press, New York, USA. BEISWANGER, C. M., P. G. SOKOLOVE & D. J. PRIOR, 1981, Extraocular photoentrainment of the circadian rhythm of the garden slug Limax. Journal of Experimental Zoology, 216: 13-23. BLINN, W. C., 1963, Ecology of the land snails Me- sodon thyroidus and Allogona profounda. Ecol- ogy, 44: 498-505. CAMERON, R. A. D., 1970a, The survival, weight loss and behaviour of three species of land snail in conditions of low humidity. Journal of Zoology (London), 160: 143-157. CAMERON, R. A. D., 1970b, The effect of temper- ature on the activity of three species of helicid snail (Mollusca: Gastropoda). Journal of Zoology (London), 162: 303-315. CARRICK, R., 1942, The grey field slug Agriolimax agrestis L., and its environment. Annals of Ap- plied Biology, 29: 43-55. CHICHESTER, L. F., 1968, The zoogeography, ecology and taxonomy of arionid and limacid slugs introduced into northeastern North Amer- ica. Ph.D. Thesis, University of Connecticut, U.S.A. р REGULATION OF ACTIVITY IN DEROCERAS 219 COOK, A., 1979, Homing by the slug Limax pseu- doflavus. Animal Behaviour 27: 545-552. COOK, A., 1980, Field studies of homing in the pulmonate slug Limax pseudoflavus (Evans). Journal of Molluscan Studies, 46: 100-105. COOK, A., 1981, Huddling and the control of water loss by the slug Limax pseudoflavus Evans. An- imal Behaviour, 29: 289-298. COOK, A. & D. J. G. FORD, 1989, The control of activity of the pulmonate slug, Limax pseudofla- vus by weather. British Crop Protection Council Monograph, 41: 337-342. CRAWFORD-SIDEBOTHAM, T. J., 1972, The in- fluence of weather upon the activity of slugs Oecologia, 9: 141-154. CROZIER, W. J. & E. WOLF, 1929, Dark adapata- tion in Agriolimax. Journal of General Physiology, 12: 83-109. DAINTON, B. H., 1943, Effects of air currents, light, humidity and temperature on slugs. Nature (Lon- don), 151: 25. DAINTON, B. H., 1954a, The activity of slugs I. The induction of activity by changing temperatures. Journal of Experimental Biology, 31: 165-187. DAINTON, В. H., 19546, The activity of slugs Il. The effect of light and air currents. Journal of Experimental Biology, 31: 188-197. DAINTON, B. H., 1989, Field and laboratory obser- vations on slug and snail behaviour. British Crop Protection Council Monograph, 41: 201-207. DAINTON, B. H. & J. WRIGHT, 1985, Falling tem- perature stimulates activity in the slug Arion ater. Journal of Experimental Biology, 118: 439—443. DAXL, R., 1969, Beobachtungen zur diurnalen and saisonellen Aktivität einiger Nacktschnecke- narten. Zeitschrift für Angewandte Zoologie, 56: 357-370. DUNDEE, D. S., 1977, Observations on the veroni- cellid slugs of the southern United States. Nau- tilus, 91: 108-114. ENRIGHT, J. T., 1970, Ecological aspects of en- dogenous rhythmicity. Annual Review of Ecology and Systematics, 1: 221-238. FORD, D. J. G. & A. COOK, 1987, The effects of temperature and light on the circadian activity of the pulmonate slug Limax pseudoflavus Evans. Animal Behaviour, 35: 1754-1765. FRAENKEL, G. & D. L. GUNN, 1961, The orienta- tion of animals: kineses, taxes and compass re- actions. Dover Publications, London. FRANDSEN, P., 1901, Studies on the reaction of Limax maximus to directive stimuli. Proceedings of the American Academy of Arts and Science, 37: 185-227. GELDERLOOS, O. G., 1979, Locomotor activity pattern of the Roman garden snail (Helix pomatia L.) in light and dark conditions. American Midland Naturalist, 101: 218-222. GELPERIN, A., 1974, Olfactory basis of homing be- haviour in the giant garden slug, Limax maximus. Proceedings of the National Academy of Science USA, 71: 966-970. GETZ, L. L., 1959, Notes on the ecology of the slugs: Arion circumscriptus, Deroceras reticula- tum, and D. laeve. American Midland Naturalist, 61: 485-498. GETZ, L. L., 1963, Light preferences of molluscs. Ecology, 44: 612-613. HENNE, F. C., 1963, The effect of light and tem- perature on the locomotory activity of Polygyra albolabris (Say). Bios, 34: 129-133. HERREID, C. F. 8 M. A. ROKITKA, 1976, Environ- mental stimuli for arousal from dormancy in the land snail Otala lactea. Physiological Zoology, 49: 181-190. HESS, S. D. & D. J. PRIOR, 1985, Locomotor ac- tivity of the terrestrial slug, Limax maximus: re- sponse to progressive dehydration. Journal of Experimental Biology, 116: 323-330. HOGBEN, L. & R. L. KIRK, 1944, Studies on tem- perature regulation I. The Pulmonata and Oli- gochaeta. Proceedings of the Royal Society (London), 132B: 239-252. HUNTER, P. J., 1968, Studies on slugs of arable ground Ill. Feeding habits, Malacologia 6: 391— 399. INGRAM, W. M., 1941, Daylight activity of land mol- lusks. Nautilus, 54: 87-90. JAREMOVIC, R. 8 C. D. ROLLO, 1979, Tree climb- ing by the snail Cepaea nemoralis (L.): a possible method for regulating temperature and hydration. Canadian Journal of Zoology, 57: 1010-1014. JOHNSON, G.C., 1969, Migration and dispersal of insects by flight. Methuen Publishers, New York, USA. JUDGE, F.D., 1972, Aspects of the biology of the gray garden slug Deroceras reticulatum (Múller). Search, 2: 1-18. KALMUS, E. J., 1942, Anemotaxis in soft-skinned animals. Nature (London), 150: 524. KARLIN, E. J., 1961, Temperature and light as fac- tors affecting the locomotor activity of slugs. Nau- tilus, 74: 125-130. KARLIN, E. J. & 4. A. NAEGELE, 1960, Biology of the Mollusca of greenhouses in New York state —Cornell University Agricultural Station Memoir, 372: 35 pp. KITTEL, R., 1956, Untersuchungen Uber den Geruchs—und Geschmacksinn bei den Gattun- gen Arion and Limax (Mollusca: Pulmonata). Zo- ologischer Anzeiger, 157: 185-195. LEWIS, R. D., 1967, Studies on the activity of the slug Arion ater (L.). Ph.D. Thesis, University of Wales, Cardiff, U.K. LEWIS, R. D., 1969a, Studies on the locomotor ac- tivity of the slug Arion ater (Linnaeus) |. Humidity, temperature and light reactions. Malacologia, 7: 295-306. LEWIS, R. D., 1969b, Studies on the locomotor ac- tivity of the slug Arion ater (Linnaeus) Il. Loco- motor activity rhythms. Malacologia, 7: 307-312. MACHIN, J., 1964, The evaporation of water from Helix aspersa Il. Measurement of air flow and the diffusion of water vapour. Journal of Experimen- tal Biology, 41: 771-781. MACHIN, J., 1975, Water relationships. Pp. 105— 220 163 in: V. FRETTER & J. PEAKE, eds., Pulmo- nata Volume 1. Academic Press, New York, USA. MCFARLAND, D. J., 1977, Decision making in an- imals. Nature (London), 269: 15-21. MELLANBY, K., 1961, Slugs at low temperatures. Nature (London), 189: 944. MORTON, B., 1979, The diurnal rhythm and the cycle of feeding and digestion in the slug Dero- ceras caruanae. Journal of Zoology (London), 187: 135-152. NEWELL, P., 1965, The behaviour and distribution of slugs. Ph.D. Thesis, University of London. NEWELL, P., 1966, The nocturnal behaviour of slugs. Medicine and Biology Illustrated, 16: 146— 159. NEWELL, P., 1968, The measurement of light and temperature as factors controlling the surface ac- tivity of the slug Agriolimax reticulatus (Muller). Pp. 141-146 in: R. M. WADSWORTH, ed., The measurement of environmental factors in terres- trial ecology. Blackwell, Oxford. PHIFER, C. B. & PRIOR D. J., 1985, Body hydra- tion and haemolymph osmolality affecting feed- ing and its neural correlate in the terrestrial gas- tropod, Limax maximus. Journal of Experimental Biology, 118: 405-421. PINDER, L. C. V., 1969, The biology and behaviour of some slug of economic importance, Agriolimax reticulatus, Arion hortensis, and Milax budapes- tensis. Ph.D. Thesis, University of Newcastle Upon Tyne, U.K. POULIN, G., 1967, Ecologie d’ Arion ater (L.). M.Sc. Thesis, University of Sherbrooke, Quebec, Canada. PRIOR, D. J., 1985, Water regulatory behaviour in terrestrial gastropods. Biological Reviews, 60: 403-424. PRIOR, D. J., 1989, Contact-rehydration in slugs: a water regulatory behaviour. British Crop Protec- tion Council Monograph, 41: 217-223. RICHTER, K. O., 1976, The foraging ecology of the banana slug Ariolimax columbianus Gould (Ari- ondiae). Ph.D. Thesis, University of Washington, Seattle, USA. ROLLO, C. D., 1974, Ecology of the slugs Dero- ceras reticulatum D. laeve, and Arion fasciatus in Ontario corn fields. M.Sc. Thesis, University of Guelph, Guelph, Canada. ROLLO, C. D., 1982, The regulation of activity in populations of the terrestrial slug Limax maximus (Gastropoda: Limacidae). Researches on Popu- lation Ecology, 24: 1-32. ROLLO, C. D. 1989, Experimental and analytical methodologies for studying molluscan activity. British Crop Protection Council Monograph, 41: 343-348. ROLLO, С. D., 1. В. VERTINSKY, W. G. WELLING- TON, & V. K. KANETKAR, 1983a Alternative risk-taking styles: the case of time-budgeting strategies of terrestrial gastropods. Researches on Population Ecology, 25: 321-335. ROLLO ROLLO, C. D., |. B. VERTINSKY, W. G. WELLING- TON, W. A. THOMPSON & Y. KWAN, 1983b, Description and testing of a comprehensive sim- ulation model of the ecology of terrestrial gastro- pods in unstable environments. Researches on Population Ecology, 25: 150-179. ROLLO, C. D. & W. G. WELLINGTON, 1981, En- vironmental orientation by terrestrial Mollusca with particular reference to homing behaviour. Canadian Journal of Zoology, 59: 225-239. SAUNDERS, D.S., 1976, Insect clocks. Pergamon Press, Oxford, U.K. SIMPSON, S. J. & E. A. BERNAYS, 1985, The reg- ulation of feeding: locusts and blowflies are not so different from mammals. Pp. 313-346 in: Vol- ume 4. G. A. KERKUT & E. A. BERNAYS, eds., Comprehensive insect physiology, biochemistry and pharmacology. Pergamon Press, Oxford, U.K. SMITH, L. H. Jr., 1981, Quantified aspects of pallial fluid and its affect on the duration of locomotor activity in the terrestrial gastropod Tridopsis al- bolabris. Physiological Zoology, 54: 408-414. SOKOLOVE, P. G., C. M. BEISWANGER, D. J. PRIOR & A. GELPERIN, 1977, A circadian rhythm in the locomotor behaviour of the giant garden slug Limax maximus. Journal of Experi- mental Biology, 66: 47-64. STEPHENSON, J. W., 1973, Effect of atmospheric humidity on feeding. Report of the Rothamsted Experimental Station, 1972: 215. TAKEDA, N. & T. OZAKI, 1986, Induction of loco- motor behaviour in the giant African snail, Acha- tina fulica. Comparative Biochemistry and Phys- iology, 83A: 77-82. THORNTHWAITE, C. W., 1940, Atmospheric mois- ture in relation to ecological problems. Ecology 21: 77-82. TOATES, F. M., 1980, Animal behaviour—a sys- tems approach. John Wiley and Sons, New York, USA. WAITE, T. A., 1988, Huddling and postural adjust- ments in response to dessicating conditions in Deroceras reticulatum (Muller). Journal of Mol- luscan Studies, 54: 249-250. WAREING, D. R. & S. E. R. BAILEY, 1985, The effects of steady and cycling temperatures on the activity of the slug Deroceras reticulatum. Jour- nal of Moluscan Studies, 51: 257-266. WEBLEY, D., 1964, Slug activity in relation to weather. Annals of Applied Biology, 53: 407- 414. WHITE, А. R., 1959, Observations on slug activity | in a Northumberland garden. Plant Pathology, 8: | 62-68. YOUNG, A. G. & G. R. PORT, 1989, The effect of microclimate on slug activity in the field. British — Crop Protection Council Monograph, 41: 263- | 269. ZAR, J. H., 1974, Biostatistical analysis. Prentice Hall, Englewood Cliffs, New Jersey, U.S.A. Revised Ms. accepted 10 December 1990 | | | MALACOLOGIA, 1991, 33(1-2): 221-240 UTILIZATION OF A SIMPLE MORPHOSPACE BY POLYPLACOPHORANS AND ITS EVOLUTIONARY IMPLICATIONS G. Thomas Watters The Museum of Zoology, The Ohio State University, Columbus, Ohio 43210-1394 USA ABSTRACT Nominal species of 408 polyplacophorans were measured for vermiformity and tegmental coverage; the results were fit to the allometric equation. The families Leptochitonidae, Ischno- chitonidae, and Chitonidae (termed here the Main Trajectory, MT) are isometrically related for these two attributes and maintain geometric similarity regardless of size. They display phyletic size increase as a consequence of stochastic and ecological processes. The genera of the families Mopaliidae and Cryptoplacidae have deviated from the MT in both vermiformity and tegmental coverage. They have accomplished this primarily by the transposition of the scaling factor (b), although some change in the growth rate (k) has also occurred. Key words: Polyplacophora, chiton, allometry, morphospace, ontogeny, phylogeny. INTRODUCTION The study of the relationship between body form and size has received renewed interest (Gould, 1966, 1971, 1974, 1977; Stanley, 1973a, 1973b), and models have been devel- oped to explain the interactions between shape and rates of growth and size (Alberch et al., 1979). Chitons have a relatively simple body shape at the gross morphological level, and in this study, measurements of the most elementary aspects of chiton morphology, overall shape and amount of tegmental (valve) coverage, are evaluated by inter- and intraspecific allometric plots in the light of the proposed models. It is hoped that this study will reveal some insight into the nature of the polyplacophoran morphospace, and how well the proposed models can elucidate data of this sort. The terms “ovate” and “vermiform” are used in this study in relation to the line of isometry. Points (species) lying below the line of width/length isometry are considered ver- miform. Those points lying above the line are ovate. MATERIALS AND METHODS This study used 408 nominal species and 39 genera of Recent chitons; the higher taxa are those recognized by Watters (in prep.) for the Cryptoplacidae and by Van Belle (1978) for the remaining families. The 40 nominal species of mopaliids used were placed in the 221 following genera: Amicula, Katharina, Mopa- lia, Plaxiphora (ail Gray, 1847), and Placi- phorella Dall, 1879. | have considered Pla- cophoropsis Pilsbry, 1893, a subgenus (at best) of Placiphorella. Basiliochiton Berry, 1918, and Dendrochiton Berry, 1911, have been included in the Mopaliidae on occasion, although Van Belle (1978) placed them under Lepidochitona in the Ischnochitonidae. Put- man (1980) created the new family Dendro- chitonidae in the Lepidochitonina for these two genera, but inasmuch as he apparently did not fulfill the requirements for availability (ICZN Art. 13), that taxon cannot be used. | have followed Van Belle and placed them in the Ischnochitonidae. Katharina has been placed by Lyons (1988b) in its own family, the Katharinidae (containing as its sole member Katharina tunicata (Wood, 1815)); although this is not without its merits, that genus is here retained in the Mopaliidae. The systematic treatment of the 85 nominal species of cryp- toplacids used here is different from that of both A.G. Smith (1960) and Van Belle (1978). | have restricted Craspedochiton s. s. Shut- tleworth, 1853, to those few forms with an an- teriorly dilated girdle and have recognized Amblyplax Ashby, 1926, and Craspedoplax Iredale & Hull, 1925, as valid subgenera of Craspedochiton. The genus Cryptoconchus Burrow, 1815, is probably polyphyletic, and the high correlation of the regression lines found in this study are probably the result of the two disparate size classes found in that genus. The monotypic genus Cryptochiton has 222 been variously placed in the Mopaliidae and the Cryptoplacidae, the great majority of mod- ern workers assigning it to the latter family. | believe that placement is in error. However, in lieu of a more comprehensive study of that genus, | have excluded it from this study. As the largest living chiton, its misplacement to any family considered here would have dis- proportionate consequences. For reasons that will be outlined below, only the largest and least distorted individuals were measured whenever possible. For those taxa that were not available as actual specimens, measurements were derived from literature il- lustrations, which were screened for accep- tance. Many otherwise important works on the chitons were not used, either because the il- lustrations were too unreliable (Reeve, 1847— 1848, for example) or did not show the spec- imens in the required orientation (as in Leloup, 1956). Published illustrations were taken from the following sources, listed alphabetically by author: Abbott (1974), Ang (1967), Ashby (1928, 1931), Barash & Danin (1977), Berry (1917), Bullock (1972a, b), Burghardt & Burghardt (1969), Castellanos (1948, 1956), Cotton (1964), Dall (1909), Ferreira (1974, 1976, 1978a-d, 1979a, b, 1980), Ferreira & Bertsch (1979), Habe & Ito (1965), Haddon (1886), Hull & Risbec (1930, 1931), Iredale & Hull (1927, 1929), Kaas (1976, 1979), Kaas & Van Belle (1985a, b, 1987), Keen (1971), Leloup (1940a, b, 1941, 1952, 1961, 1965, 1966), MacPherson & Gabriel (1962), Marin- covich (1973), Milne (1963), Nierstrasz (1905), Ried! (1963), Sirenko (1976), А.С. Smith (1961), Smith & Cowan (1966), Smith & Ferreira (1977), Smith & Hanna (1952), Taki (1953), Thiele (1909-10), Utunomi (1968), Watters (1981, 1990), and Yakovleva (1965). By using a large undistorted individual as representative of a species, | have removed, as far as is possible with the available data, the problem of mixing age classes in this al- lometric study (see Gould, 1972, 1974). Ina group of organisms whose life histories are as little understood as the chitons, it is difficult to determine a comparable age class with any reliability; developmental stages would be le- gitimate criteria, but the embryology of all but a few species is unknown. Sexually mature individuals would also be useful, but the ac- tual dissection of specimens to determine go- nadal development is impossible when using illustrations and impractical even for museum work. An alternative is to measure large indi- viduals, because the largest specimen is the WATTERS most likely member of that set to actually be sexually mature. By selecting only these sam- ples, distortion caused by ontogenetic allom- etry is minimized. It may be argued that the range of sizes of sexually mature chitons within any species may be extensive and therefore, by selecting only the largest specimens, the results have been biased towards unnaturally large indi- viduals. However, the records used, whether from collections or published works, actually represent only the largest specimens known to me; this is not necessarily the greatest at- tainable size for that species. Furthermore, many rare species are known from only a small number of specimens, and we lack any knowledge of their possible size range. Mea- surements for these rare species are often from the only available specimen or illustra- tion. A uniform bias towards large specimens would change the intercepts of the allometric lines but not the conclusions drawn from a comparison between them. It is felt that the benefits of removing ontogenetic factors out- weigh the bias towards large individuals. Each specimen or illustration of a specimen was checked for distortion, that is, curling of the body, non-extension of the girdle, terata- logical individuals, etc. On each individual the following measurements were taken: (1) total length, including girdle; (2) maximum width, including girdle; (3) maximum width of the tegmentum of valve II (Fig. 1). The second valve was chosen because valves III-VII be- come increasingly separated and vary in size during ontogeny in the genus Cryptoplax Blainville, 1818, and are not suitable. Data were fitted to the allometric equation: Ух. where x represents a measure of total body size and y is some morphological or physio- logical characteristic relative to x. The value k is the ratio of the growth rates of x and y, the “constant differential growth rate” of Huxley (1932). The allometric coefficient b repre- sents the value of y when x = 1, or the inter- cept of y of the logarithmic transformation of the allometric equation: log y = k log x + log b. This coefficient essentially was ignored by | earlier workers until White & Gould (1965) and Gould (1971) demonstrated its biological significance and devised a method of deter- POLYPLACOPHORAN MORPHOSPACE 223 tegmentum width total length total width FIG. 1. Measurements used in this study. mining the amount of similarity between com- parable data based upon b. RESULTS Main Trajectory The adults of 283 nominal species of Re- cent chitons comprising the families Leptochi- tonidae, Ischnochitonidae, and Chitonidae are herein considered the constituents of the main trajectory (abbreviated to MT for the re- mainder of the article), a morphospace that includes the majority of living chitons. Mem- bers of this group show very similar degrees of vermiformity/ovateness and tegmental re- duction and are best discussed as a unit. The Cryptoplacidae and the Mopaliidae deviate from the MT and will be considered sepa- rately. The MT species also fall upon a well-de- fined, isometric line (Fig. 2) of width/length of the equation: width = 0.55 length'°' (г = 0.97; p<0.00). The species of the MT show a nearly isomet- ric relationship (Fig. 3) between width and tegmental width: tegmental width = 0.74 width? (г = p<0.00). 0.98; Mopaliidae Because of the great diversity in body shapes found in this family, no single regres- sion line for all species can be drawn for ei- ther width/length or tegmental width/width. With the exception of Plaxiphora, which is slightly more vermiform at large sizes than MT species, all other genera are more ovate than MT taxa (Fig. 4). The correlation coeffi- cient of the line for Amicula is not statistically significant. For tegmental coverage, all have less coverage than comparable MT species, except for Placiphorella, which has greater coverage (Fig. 5). Equations for the lines are given in Table 1. Cryptoplacidae As with the mopaliids, the cryptoplacids show great variety in body shapes and de- gree of tegmental coverage; no single regres- sion line can be drawn for all species in the family for either width/length or tegmental width/width (Figs. 6, 7). The genus Craspe- dochiton has regression lines most similar to the MT for vermiformity and tegmental cover- age. All cryptoplacid genera are more vermi- form throughout most of their size ranges; all also have reduced tegmental coverage rela- tive to MT species. Equations for the lines are given in Table 1. DISCUSSION Main Trajectory It is clear that in these chitons the relation- ship between total width and tegmental width is correlated with total length. The MT species display isometry, a special case of allometry in which the rate of change between the two “parts” under consideration remains constant over a range of sizes and geometric similarity 224 WATTERS log (width) (mm) E 0.4 - a? 0:2 | 0.0 = O A ER CE 0.4. 016: G8 10 12 14 _ ——— 7" 1.6 | | | | | | | | | | | | | | | | | | A O A A А T J ÍA 18 2.0 22 24 26 28 3 log (length) (mm) FIG. 2. Plot of width vs. length for species comprising the MT. Equation for line: width = 0.55 length'°! (r = 0.97; p<0.00). is maintained (Gould, 1966). In this trajectory, large individuals are nothing more than scaled-up replicas of smaller ones (or vice versa). Because the relationships between width/length and tegmental width/width are isometric, no allometric transposition is nec- essary to ensure geometric similarity at larger sizes. Presumably some instances of phyletic size increase involve a transposition (via b, the scaling factor) to maintain a similar shape across a broad range of sizes due to the fact that the intraspecific plots are themselves al- lometric (Gould, 1971). This is not the case for the MT species. Deviations within the MT are rare and ap- parently arose in response to new niches. Two ischnochitonid genera have deviated towards vermiformity: Stenochiton Adams & Angas, 1864, and Stenoplax Dall, 1878. Stenochiton lives attached to sea-grasses, such as Zostera and Cymodocea, and its vermiform shape may be an adaptation to living on the narrow blades (Ashby, 1926; Iredale & Hull, 1927). Other species, such as Lepidochitona dentiens (Gould, 1846), may also live on algae or hold- fasts (Kues, 1974) but are not vermiform. The difference lies in the fact that the vermiformity of Stenochiton allows a larger species to oc- cupy a niche where one major constraint is size (Stenochiton may reach a length of 45 mm, but holdfast specimens of L. dentiens may only attain a length of 19 mm). On the other hand, the vermiformity of Stenoplax may be related to streamlining. Stenoplax floridana (Pilsbry, 1892) have been found moving across and between rocks completely buried in sand (pers. obs.). Other members of the genus may behave in a sim- ilar manner. An increase in vermiformity would reduce the resistance to forward mo- tion through this substrate. The plot of adult species of Stenoplax s. s. for width/length (Fig. 8) reveals that within the genus, as within the Mopaliidae and Crypto- placidae, discussed later, vermiformity is ob- tained primarily by a change in b. The change of shape in Stenoplax occurs by a lowering of the scaling factor, b, from the MT value of 0.55 to 0.48 (Table 1). These vermiform adults may be derived by two means. A change in the initial proportions of the post-trochophores (a transposition of POLYPLACOPHORAN MORPHOSPACE 225 ЕЕ А = ae a | a | — | 2 74 i = 1.6 - = + = | р . р = 1.4 = ee 71 = | a + 3 | D 5 E E 1.24 NS . | | aoe 7 | E 1 o soc et” = | р 5 | =~ | 4 087 a A a ae O 0.6 - ue | = | = © | Bee) Ve | © м | 024, 0 0 JET | faz == A E Mr cae (AE A re af 5 She == 027 04 0608 TO. 12 14 16 18 20 22 24 26 28 30 log (width) (mm) FIG. 3. Plot of tegmental width vs. width for species comprising the MT. Equation for the line: tegmental width = 0.74 width? (г = 0.98; p<0.00). b), such that they are essentially “born” more vermiform at the outset, would not entail any allometric ontogenetic trajectories to produce a series of isometrically related adults. Alter- natively, a MT species may gradually ap- proach vermiformity as an adult by allometric (non-isometric) growth in its ontogeny. A growth series of Stenoplax floridana (Pilsbry, 1892) (N = 31) reveals that this species’ on- togenetic trajectory (Figs. 8, 9) is very similar to that of all adult Stenoplax. | conclude that Stenoplax (and probably Stenochiton) origi- nated from an ischnochitonid of the MT by a change of b, and that both genera have adults whose forms remain nearly isometrically re- lated to each other (see Alberch et al., 1979). This has resulted in two morphologically sim- ilar genera that may be vermiform for different reasons, however. Most of the species comprising the MT lie _ along an isometric plot where one species is nothing more than a scaled-up (or down) rep- | | | | | _lica of another in terms of body shape. We ' may speculate that the greatest impetus for size increase in chitons was the colonization of the open littoral zone, a habitat that made available unused food resources and sub- strates. Although exposure to desiccation, temperature extremes, and wave action all preclude a successful invasion by small forms, any increase in body size would lower the surface-area-to-volume ratio, diminishing the effects of evaporation and temperature. Furthermore, the ability to withstand wave ac- tion, and to a certain extent predation, is pro- portional to the adhesive area of the foot, which goes as the square of its linear dimen- sion. Larger forms have greater adhesive powers, all other things being equal. There may be a selective advantage for increased size among chitons along the MT as a pre- requisite to intertidal colonization. Linsenmayer (1975) has demonstrated that the force required to dislodge various species of chitons was proportional to the degree of exposure to wave action in the natural habitat; more force was required to remove intertidally exposed species than was necessary to re- move a subtidal form. Unfortunately, the ex- periment apparently did not use chitons of uniform size in all classes or give the sizes of individuals. Thus, it cannot be determined 226 WATTERS E 4 | 2 ‘ | a dl | ] Ay mE Я Е | я | Е 10 - 2 | = и’ | e 34 ES | = = oo | = = | | y | 7 | 1 | i is T ss een + 0 1 10 100 1000 length (mm) FIG. 4. Plot of width vs. length for genera of Mopaliidae and MT. (x )—MT; (m)—Amicula; (e)— Mopalia; (*)—Placiphorella; ( +)—Plaxiphora. See Table 1 for equations for lines. tegmentum width (mm) 100 1000 width (mm) FIG. 5. Plot of tegmental width vs. width for genera of the Mopaliidae and MT. (x )—MT; (m)—Amicula; (e)—Mopalia; (*)— Placiphorella; (+ )—Plaxiphora. See Table 1 for equations for lines. : POLYPLACOPHORAN MORPHOSPACE 227 TABLE 1. Equations and correlations for MT species and the genera of Mopaliidae and Cryptoplacidae. Vermiformity MT width = 0.55 length! °' г 20.97 p<0.00 N = 283 Stenoplax 5.5. width = 0.48 length??? r = 0.98 p<0.00 Ni=6 Amicula width = 1.07 length? 9° r = 0.66 p= 0:34 NETA Mopalia width = 0.43 length**9 r = 0.99 p<0.00 М1 Plaxiphora width = 0.83 length? 9" r = 0.98 p<0.00 NN? Placiphorella width = 0.60 length! 9 r = 0.99 p<0.00 МО Acanthochitona width = 0.98 length? 75 Г 10:83 p<0.00 N = 40 Craspedochiton width = 0.51 length'°' р = 093 p<0.00 М2 Cryptochonchus width = 0.37 length" °* r= 0.97 p = 0.03 N=4 Cryptoplax width = 0.78 length°:5° 70:84 p<0.00 N = 14 Notoplax width = 0.89 length 7° r= 0:92 p<0.00 N= 15 Tegmental coverage (tegw—width of tegmentum of valve Il) MT tegw = 0.74 width0-2 r = 0.98 p<0.00 N = 283 Amicula tegw = 0.20 width ®' r = 0.86 р —0. 14 М = 4 Mopalia tegw = 0.68 width-9* 2.0.97 p<0.00 МЕН Plaxiphora tegw = 0.87 width? 8° r = 0.99 p<0.00 №12 Placiphorella tegw = 0.76 width0-98 r = 0.98 p<0.00 М Acanthochitona tegw = 0.52 width0-93 r= 0.87 p<0.00 N = 40 Craspedochiton tegw = 0.65 width9-90 r = 0.96 p<0.00 Nie Cryptoconchus tegw = 0.18 width‘ 77 r = 0.79 р — 0:21 №4 Cryptoplax tegw = 0.16 width' 2? Г — 0.95 p<0.00 N = 14 Notoplax tegw = 0.44 width? 8° n= O83 p<0.00 №15 100: width (mm) 10 100 1000 length (mm) FIG. 6. Plot of width vs. length for genera of the Cryptoplacidae and MT. (#)—MT; ( x )— Acanthochitona; (m)—Craspedochiton; (+)— Cryptoconchus; (e) —Cryptoplax; (*)—Notoplax. See Table 1 for equations for lines. if the increased adhesive force is a simple One would expect intertidal forms to be consequence of increased size or if this force larger, regardless of taxonomic position, than is disproportionately large in intertidal forms. sublittoral species, if increased size has less 228 WATTERS 100 Е Е 210 = YA | m TEE | 3 4 Bey | 4 MS ae SS | Е - ов à | 5 | ESE E a | + > Е 13 > RE 8 | Ф 4 I a Е 4 de [e) u 2 | | | 1 10 100 1000 width (mm) FIG. 7. Plot of tegmental width vs. width for genera of the Cryptoplacidae and MT. (4 )—MT; (x )—Acan- thochitona; (m)— Craspedochiton; (+ )—Cryptoconchus; (e)— Cryptoplax; (*) —Notoplax. See Table 1 for equations for lines. 100 a | : = | L | | a | т = | | a E — | — | a | 10E Par | = a | D В о | | иж | = | Pr | | e + | E = | | | 1 [ eye CSN O E LS 1 10 100 length (mm) FIG. 8. Plot of width vs. length for species of adult Stenoplax (N = 6) (+), individuals of Stenoplax floridana (*), and species of MT (m). Equation for adult Stenoplax line: width = 0.48 length 0.93 (r = 0.98; p<0.01). selective advantage in sublittoral taxa. This is niches are exposed to the same extremes in indeed the case for Recent species (Figs. 10, environment. Crevices, potholes, and other ir- 11). Not all intertidal taxa are larger than their regularities offer sufficient protection to en- sublittoral counterparts, but not all intertidal able small species to exist intertidally. At one POLYPLACOPHORAN MORPHOSPACE 229 1.4 y =. | ei a | = 1.2 | £ Wi | ni A” | fa „м. | 5 DE a | Be . > a 07: | =< Le | © ¡ES © 0.6 if : lis 212: 14 16. 48 2 22954 log (length) (mm) FIG. 9. Plot of width vs. length for individuals of Stenoplax floridana (М = 31). Equation for the line: width = 0.42 length® °° (г = 0.92; p<0.00). 7 | 60% -” Offshore species 50% N = 56 nn 40% 2 O Ф 30% - о. 0) = 20% 10% : PEER REE 0% = | er 60 70 80 90 100 110 120 length (mm) FIG. 10. Percent of species studied reported from offshore or sublittoral habitats grouped in 10 mm size classes. extreme, Lepidochitona caverna Eernisse, lina californica (Reeve, 1847), a larger inter- 1986, a small ischnochitonid, occasionally tidal species (Gömez, 1975). There is also lives sheltered in the pallial groove of Nuttal- evidence to suggest that the juveniles of 230 WATTERS 25% т % Species = A 5% - A == == A CTA 10 20 30 40 50 Intertidal species N = 114 ЕЕ EA N CA 90 100 110 120 | 70 80 FIG. 11. Percent of species studied reported from intertidal habitats grouped in 10 mm size classes. some intertidal species are subtidal (Demo- poulos, 1975) or sheltered in intertidal crev- ices (Glynn, 1970) until some critical size is reached. This sheltered group of intertidal chi- tons is responsible for the smaller size classes seen in Figure 11. The colonization of the intertidal zone may have increased competition, both among and between chitons and patellaceans. It is plau- sible that the increase in size needed to col- onize this region required an even greater in- crease in food intake. Inasmuch as most chitons are grazers, this translates into an in- crease in the absolute size of the feeding range. It is therefore not surprising that some intertidal chitons have been shown to pos- sess a home territory from which other chi- tons and limpets are excluded (Connor, 1975; S. Y. Smith, 1975). Tendencies for increased size must be balanced by the expenditure of energy needed to maintain a home range or feeding territory; this expenditure may be the limiting factor in determining the upper limits of size in chitons. Stanley (1973b) and Hallam (1974) have suggested that increased competition may drive evolutionary rates faster than non-com- petitive situations. If this hypothesis is correct, one would expect to find higher rates of spe- ciation in intertidal forms than in sublittoral ones. In the paleontological record, there is a marked increase in diversity in all neolori- cates in the Cenozoic, partially due to a more complete fossil record and partially due to the continued fragmentation of Pangaea Il. The habitats of extinct species are unknown, and it is impossible to be certain if ancestral forms were predominantly sublittoral or littoral. How- ever, the largest family to survive the Permo- Triassic extinction was the Leptochitonidae, the ancestral stock of most Recent forms. These chitons are typically small, sublittoral species, some occurring at great depths. With the breakup of Pangaea II and an increase of the littoral zone (Valentine, 1973), descen- dants of this offshore remnant colonized the new continental coasts. Two of the largest liv- ing families, the Chitonidae and the Mopali- idae, experienced a pronounced increase in diversity in the Cenozoic, and both are com- posed of predominantly large, intertidal spe- cies. Dawkins & Krebs (1979:502) believed that intraspecific competition was “the primary driving force of Darwinian evolution” and sug- gested that it was a possible mechanism for phyletic size increase. According to these au- thors, interspecific competition would most likely result in niche partitioning rather than escalated competition and would not neces- POLYPLACOPHORAN MORPHOSPACE 231 % species | Leptochitonidae | | N = 33 т 2 un T | 70 80 90 100 110 120 length (тт) FIG. 12. Percent of species of Leptochitonidae in 10 mm size classes. sarily lead to phyletic size increase. The ex- plosion of chiton diversity in the Cenozoic is tied to a trend in size increase that enabled some forms to colonize the intertidal zone. An adaptive radiation in this area may have been driven by the sudden appearance of intraspe- cific competition, which was itself the conse- quence of increased food resources required by a large animal. In addition to minimizing surface area/vol- ume effects, size increase may permit the in- troduction of new anatomical structures and the elaboration of existing ones (Gould, 1966). One indication of this increased ana- tomical complexity found in chitons is the es- thete system. At least one extinct genus (Hoare et al., 1972) and all Recent species possess esthetes or “shell-eyes” of some sort, though not all esthetes are photorecep- tors (Fischer, 1988). However, it is only in the Callochitonidae, Schizochitonidae, and Liolo- phura (Pilsbry, 1893), Enoplochiton, Onitho- chiton, and Tonicia (all Gray, 1847) of the Chi- tonidae that macresthetes are transformed into photoreceptive structures complete with cornea, lens, and possible retinal cells (Fisch- er-Piette & Franc, 1960; Hyman, 1967). With the exception of a few callochitons, all of these forms are of large size. Larger size also may entail a change of diet. Presumably larger species can accom- modate larger food items that could not be used by smaller ones. It is in the large mo- paliids of Placiphorella that we see a radical change to active carnivory (Barnawell, 1959). Phyletic size increase was already operat- ing on a major sublittoral group, the ischno- chitons, prior to the invasion of the intertidal zone. This suggests the presence of a second mechanism for size increase. The size distri- bution of the three families depicted in Fig- ures 12-14 are not symmetric about the mean but are right skewed. This phenomenon was discussed by Stanley (1973a), who envi- sioned phyletic size increase not as a trend towards larger size but as a trend away from smaller ones in which no intrinsic selective advantage to larger descendants need be in- voked. Large species are often more special- ized than smaller ones and do not represent good evolutionary material for new forms; Hallam (1975, 1978) and Dawkins & Krebs (1979) have demonstrated that larger individ- uals are more susceptible to extinction for that reason. Therefore, most new taxa are derived from small forms from the ancestral size range. This new taxon will then radiate within its own new morphospace size range to cre- ate both larger and smaller forms. However, as shown by Stanley (1973a), a comparison E 120 120 EN | 110 94 110 100 N Ф © 32: = O Es is O 100 90 90 80 80 JE = = F fi = = at ae = AN NN on Ze = NN = WATTERS Г tees Ali | IN | 1 a 8? г? a 2? = ae г wo O LO O ve) о o) о © 0) N N = г sal9ads % 232 70 70 length (mm) FIG. 14. Percent of species of Chitonidae in 10 mm size classes. and an equal 25% increase from 10 mm re- sults in a 12.5 mm descendant. Four such 25% decrease results in a form 7.5 mm long, 60 length (mm) 60 5 40 = = = i HR FIG. 13. Percent of species of Ischnochitonidae in 10 mm size classes. 10 5% 7 0% + creases results in disparate changes. For ex- of equally proportional size decreases and in- ample, consider a taxon of length 10 mm: a POLYPLACOPHORAN MORPHOSPACE 233 proportionately equal changes produce a 3.16 mm and a 24.4 mm form, respectively. With each decrease the size asymptotically approaches zero, while size increases result in ever more pronounced changes; with each change the smaller size changes less, the larger, more. Most metazoans reach a lower limit constrained by available space for nec- essary organ systems, but larger sizes, on the other hand, can accommodate new structural innovations and increase the complexity of existing morphologies. Although new taxa may originate from the smaller, more gener- alized forms of a morphospace size range, each subsequent origination will take place at a larger size and produce larger descendants. Phyletic size increase of this type may be de- scribed as stochastic and is not selectively directed towards larger sizes. The fossil record of the Polyplacophora is too incomplete to determine which family, the ischnochitonids or the chitonids, is the most ancient. Both may have been independently derived from the leptochitonids. It is more likely, however, in view of the similarities be- tween the two groups, that the chitonids have evolved from the ischnochitonids. This is the evolutionary Sequence assumed in this study. Leptochitonids occupy only a small part of the total size range for all chitons and are limited predominantly to the smallest forms, although a skewing towards larger sizes is evident (Fig. 12). Ischnochitonids may have been derived from a leptochitonid ancestor that radiated into the small size end of the ischnochiton potential size spectrum. Radia- tion by size alteration produced a strongly right-skewed distribution curve that includes most of the size range of all Recent species (Fig. 13). As a result of this increase, the mean size of ischnochitons exceeded that of the leptochitonids. Smaller ischnochitons became larger than small leptochitonids, and a subsequent evolutionary step from the ischnochitonids to the chitonids would result in yet another shift to larger sizes by the same mechanism. This created individuals large enough to start the intertidal colonization, from which even larger forms would be se- lected by the constraints of the habitat and the impetus of competition. The chitonids display a combination of subtidal and sheltered inter- tidal species comparable in size with ischno- chitonids, plus the intertidal component of the expected larger forms (Fig. 14). Species on the MT line exhibit a size in- crease without perturbations of onset age (b's = equal for the families of MT species) or accelerated growth rates (= equal k's). De- scendants increase their size past that of an- cestral adults but maintain a geometrically similar shape. Sakae (1968), in a similar study using nine species of chitons in three families, found that relative to valve |, all other valves became proportionately narrower and longer in large chitons. This would indicate a tendency for increasing vermiformity in large species, a conclusion that cannot be supported by this study. However, Sakae used a total width and length measurement of the valves that in- cluded the articulamentum; thus the total valve may become more elongate in larger chitons without increasing vermiformity. Mopaliidae The earliest known mopaliid is a Plaxiphora from the Lower Miocene or Upper Oligocene (A. G. Smith, 1960), a genus represented to- day by a small number of species that range in adult size from 10 to 140 mm. Like most mopaliids, they are associated with temperate or boreal intertidal habitats. In the northern hemisphere, the dominant genus is Mopalia; in the southern, it is Plaxiphora and related genera; and the two are bridged by the Jap- anese Hachijomopalia Taki, 1953. Although mopaliids cover a wide size range, most spe- cies are larger on the average than MT spe- cies. Their relatively recent emergence, large average size, and intertidal habitat all argue for a derivation from the MT line during the time of their Cenozoic radiation into the littoral zone. The physiological ability to withstand low temperatures separated these forms from the ecologically similar chitonids of more trop- ical regions. The mopaliids lack the gross morphological homogeneity of the MT species, particularly in the amount of tegmental coverage (Fig. 5). At a size of 50 mm, a species of Placiphorella has a 119% wider tegmentum than a MT spe- cies of comparable size; Amicula has the tegmentum of a comparable MT form with an 84% reduction. As in Stenoplax and Steno- chiton, generic changes have largely been effected by altering b between two genera (decreased in Amicula, increased in Placi- phorella). If we accept that mopaliids were de- scended from a MT ancestor, then it seems likely that at some point their scaling factors for valve coverage/width rates must have di- 234 WATTERS verged from the MT. It is unlikely that each genus arose through iteration from the ances- tral stock. It is more reasonable to postulate that a common ancestor diverged and then gave rise to additional genera by changes in b. Both Plaxiphora and Mopalia are very sim- ilar to the MT species for vermiformity and tegmental coverage (Figs. 4, 5). Either genus may actually be ancestral, given the imperfect fossil record of chitons. Other genera di- verged by additional changes in b, although all mopaliids are similar to MT species in the amount of vermiformity. Of all chiton genera, Placiphorella has de- viated the most from the MT line towards ovateness. It has deviated in several other related respects also. The anterior portion of the girdle has become greatly dilated, forming a flap-like extension. This structure is used to trap live prey, making these species the only known actively predaceous chitons (Barn- awell, 1959). Lacking cephalic eyes, the gir- dle must act not only as a trap but as a sen- sory device as well in order to capture prey. The girdle hairs of Placiphorella, as well as other mopaliids, have been shown to be in- nervated (Plate, 1902; Leise, 1983, 1988), al- though the role, if any, of these hairs in de- tecting prey is unknown. Widening the body (ovateness) increases the amount of area that is directed anteriorly, potentially maximiz- ing the sensory and mechanical potential to this region. This crude cephalization may be a prerequisite to a habitat or diet that demands more sophisticated sensory input and has developed in other chitons: Lorica H. & A. Adams, 1852; Loricella Pilsbry, 1893; and Componochiton Milne, 1963, of the Schizo- chitonidae, and Craspedochiton in the Cryp- toplacidae. There is no implication here that these genera are also carnivorous, but it is likely that their habitat (most are deep-water) necessitates a greater degree of sensory elaboration than that of other chitons. In- creasing ovateness seems to result in an an- teriorly directed increase of sensory potential. Amicula, and to a lesser degree Katharina, possess extreme tegmental reduction (with- out a concomitant decrease in the articula- mentum). Placiphorella has the greatest de- gree of tegmental coverage; this is correlated with the ovateness of that genus. Cryptoplacidae The cryptoplacids constitute the most aber- rant and misunderstood group of chitons; no other family encompasses such a wide vari- ety of forms, ranging from the ischnochiton- like Craspedochiton to the vermiform Crypto- plax. Many of the forms encountered in this family are reminiscent of mopaliids, and some taxa of the two families have often been con- fused and misplaced. Both families show a trend towards tegmental reduction that is more than a result of maintaining geometric similarity. The cryptoplacids also display a trend towards vermiformity that exceeds the requirements of geometric similarity. These two trends in conjunction have produced the entire range of cryptoplacid body forms. The most likely candidate for the ancestral cryptoplacid is a craspedochiton, particularly a Craspedochiton (Amblyplax)-like form. That group contains several important morpholog- ical features that relate it to MT species: divi- sion of the intermediate valves into discern- able lateral and central areas, multiple insertion plates in valve VIII, and an unde- fined jugum. Craspedochiton is the most isometric cryp- toplacid genus for vermiformity and tegmental coverage (Figs. 6, 7); thus, in addition to its otherwise MT-like features, it has also devi- ated the least of all cryptoplacid genera from the MT line. Other genera of cryptoplacids have diverged by alteration of b (and k, for vermiformity). The present genera, except for Craspedo- chiton, are allometrically related for vermifor- mity but differ in their values of b (Fig. 6). They have diverged from that genus by a change in k; Craspedochiton retains the an- cestral MT form dimensions. The most vermi- form of the cryptoplacids are Cryptoconchus, Cryptoplax, Choneplax, and Notoplax; an adult 50 mm Cryptoplax is 244% more vermi- form than a comparable MT species. At least some species of these four genera are often found in a habitat largely unused by other chitons, that being the narrow spaces and channels in corals and sponges (Ang, 1967; Ferreira, 1985; Lyons, 1988a). The sinuous shapes of these chitons may be an adaptation to this lifestyle. As previously mentioned, Cryptoconchus may be polyphyletic; its re- gression line therefore cannot be considered accurate (for vermiformity or tegmental cov- erage). Cryptoplacids also show a trend towards tegmental reduction (Fig. 7). Craspedochiton most resembles MT forms, and subsequent taxa may have been produced by changes in b. Cryptoplax has a k of 1.23, indicating that POLYPLACOPHORAN MORPHOSPACE 235 Increasing width Increasing length FIG. 15. Hypothetical diagram of evolutionary changes in form in the Polyplacophora. The majority of species occur along the isometric path MT. Differentiation of new families and genera involve a change in b, the scaling factor for vermiformity. A model for tegmental coverage would be similar. valve coverage increases with size; however, b is so low (0.16, the lowest of any group studied) that the line for valve coverage falls clearly on the reduction side of the isometric line. A 50 mm species of Cryptoplax has 66% ofthe coverage of a similar MT species. Cryp- toplax appears to be a parallel evolutionary line from Craspedochiton involving not only the reduction of the tegmentum but a reduc- tion of the articulamentum as well. CONCLUSIONS When simple body shape and amount of tegmental coverage are examined, chitons show trends that fit existing models of mor- phological change in time and space. The majority of chiton genera were eliminated dur- ing the Permo-Triassic extinction, apparently leaving small, offshore leptochitonids to act as the major source of future chiton evolution. Many Recent taxa have body shapes that evolved isometrically to those of their lepto- chitonid ancestors; larger forms are merely scaled-up models of smaller ones. This trend in size increase, often called Cope’s Rule, may be stochastic in nature (Stanley, 1973a). We may speculate that progressive size in- crease fortuitously enabled the chitons to col- onize newly forming intertidal regions. Large size minimized desiccation and supplied suf- ficient pedal force to withstand wave action, but larger size also increased food require- ments. Thus, competition for feeding territo- ries may have accelerated the evolution of intertidal species in the Cenozoic. Larger sizes also permitted new complexity in old an- atomical structures, and may have given rise to new structures. The esthete system reaches its zenith in the large intertidal forms. A very few forms diverged from this line towards vermiformity and new niches. Within vermiform MT genera, species are nearly iso- 236 WATTERS Increasing width Increasing length FIG. 16. Hypothetical diagram of evolutionary changes in vermiformity in the Cryptoplacidae. The original divergence involved a change in k, the growth rate, while subsequent originations transposed b. metrically related by body shape, and it is ev- ident that vermiformity arose within the an- cestral MT by a change of b. This is the same mechanism by which the Mopaliidae and Cryptoplacidae have diverged from the MT. In the Mopaliidae and the Cryptoplacidae, definite trends can be found in tegmental re- duction (relative to the MT group), and in cryptoplacids, also in vermiformity. These trends appear to have arisen by the alteration of b from a MT path. The resulting array of shapes included forms not encompassed by the MT species. Subsequent genera arose by further changes in b. These changes have ex- panded the morphospace of chitons beyond the boundaries of the ancestral MT taxa (Fig. 15). In the vermiformity of the Cryptoplacidae, the initial divergence from an presumed an- cestral Craspedochiton-like form involved a change in k, although subsequent genera are allometrically related and differentiated by transpositions of b (Fig. 16). The Crypto- placidae have diverged from the MT in both valve coverage and vermiformity. The results of this study suggest that the majority of living chitons are isometrically sim- ilar for overall shape and tegmental coverage. Although it is possible that some smaller taxa have been derived from larger ones, the main direction of chiton morphospace change has been size increase. Novel changes in overall shape and tegmental coverage have been brought about primarily by changes in b. — These changes have resulted in new higher — taxa, although the species of each new taxon are nearly isometrically related to each other (parallel to MT line); they “behave” within their new morphospace in the same manner as the ancestral MT species “behave” in | theirs. This suggests that new higher chiton | POLYPLACOPHORAN MORPHOSPACE 237 taxa may be the result of macroevolutionary processes: these taxa are “bumped” into a new region of the morphospace where the en- suing species continue to morphologically fol- low the trajectory of the ancestral group. The family Cryptoplacidae probably has as its most primitive living group the genus Craspedochiton. This genus has deviated lit- tle from the MT in either overall shape or tegmental coverage (as evidenced by its his- torically ambivalent placement in several fam- ilies). The remaining genera are allometrically similar but have deviated from isometry. Pre- sumably an as yet unidentified ancestor of Craspedochiton diverged by a change in k into a new morphospace and subsequently gave rise to new genera. This has generated the greatest array of body shape and tegmen- tal coverage in the Polyplacophora. ACKNOWLEDGEMENTS | would like to thank Dr. Ruth Turner, Mu- seum of Comparative Zoology, Harvard Uni- versity, and the late Dr. Joseph Rosewater, National Museum of Natural History, Wash- ington, D. C., for allowing me to examine the collections under their care. Ms. Kathy Borror, Dr. Abbott Gaunt, Dr. David Stansbery, and Dr. Barry Valentine, all Department of Zool- ogy, Ohio State University, kindly commented on the manuscript. Mr. C. John Finlay, Miami, Florida, kindly supplied specimens for study. The manuscript was significantly improved by comments by Dr. George Davis, Dr. Steven J. Gould, and an anonymous reviewer. LITERATURE CITED ABBOTT, R. T., 1974, American Seashells, 2nd ed. Van Nostrand Reinhold, New York. 663 pp. ALBERCH, P., S. J. GOULD, G. F. OSTER & D. B. WAKE, 1979, Size and shape in ontogeny and phylogeny. Paleobiology, 5: 296-317. ANG, E. A., 1967, Loricates of the Philippines. Nat- ural and Applied Science Bulletin, 20: 383-464. ASHBY, E., 1926, Victorian chitons. The Victorian Naturalists, 43: 7-16. ASHBY, E., 1928, South African chitons, being a description of the Polyplacophora represented in the Turton collection. Proceedings of the Mala- cological Society of London, 18: 76-92. ASHBY, E., 1931, Monograph of the South African Polyplacophora (chitons). Annals of the South Af- rican Museum, 30: 1-57. BARASH, A. & Z. DANIN, 1977, Polyplacophora (Mollusca) from the eastern Mediterranean. Jour- nal de Conchyliologie, 114: 3-28. BARNAWELL, E. B., 1959, The carnivorous habit among the Polyplacophora. Veliger, 2: 85-88. BERRY, S. S., 1917, Chitons taken by the United States Fisheries steamer “Albatross” in the northwest Pacific in 1906. Proceedings of the United States National Museum, 54: 1-17. BULLOCK, R. C., 1972a, On the taxonomy of Cha- etopleura fulva (Wood, 1815) (Mollusca: Polypla- cophora). Occasional Papers on Molluscs, 3: 177-192. BULLOCK, R. C., 1972b, Notes on the genus Chi- ton in the western Indian Ocean (Mollusca: Poly- placophora). Occasional Papers on Molluscs, 3: 237-252. BURGHARDT, G. 8 L. BURGHARDT, 1969, A col- lector's guide to west coast chitons. San Fran- cisco Aquarium Society Special Publication, (4): 1-45. CASTELLANOS, Z. J. A. DE, 1948, Sobre seis es- pecies de poliplacoforos Argentinos. Notas Zoo- logía, Museo de La Plata, 13: 191-202. CASTELLANOS, Z. J. A. DE, 1956, Catálogo de los poliplacoforos Argentinos y de aguas vecinas al estrecho de Magellanes. Revista Sección Zoolo- gía, Museo de La Plata, new series, 6: 465-486. COTTON, B. C., 1964, Chitons. South Australian Mollusca. South Australian Branch of the British Science Guild, Adelaide. 151 pp. CONNOR, M. S., 1975, Niche apportionment among the chitons Cyanoplax hartwegii and Mo- palia muscosa and the limpets Collisella limatula and Collisella pelta under the brown alga Pelvetia fastigiata. Veliger, Supplement, 18: 9-17. DALL, W. H., 1909, Report on a collection of shells from Peru, with a summary of the littoral marine Mollusca of the Peruvian zoological province. Proceedings of the United States National Mu- seum, 37: 147-294. DAWKINS, R. 8 J. R. KREBS, 1979, Arms races between and within species. Proceedings of the Royal Society of London, series В, 205: 489— SE DEMOPOULOS, P. A., 1975, Diet, activity and feeding in Tonicella lineata (Wood, 1815). Veliger, Supplement, 18: 42—46. FERREIRA, A. J., 1974, The genus Lepidozona in the Panamic province, with a description of two new species. Veliger, 17: 162-180. FERREIRA, A. J., 1976, A new species of Callisto- chiton in the Caribbean. Nautilus, 90: 46-49. FERREIRA, A. J., 1978a, The chiton species de- scribed by C. B. Adams, 1845, from Jamaica. Bulletin of Marine Science, 28: 81-91. FERREIRA, A. J., 1978b, The genus Lepidozona (Mollusca: Polyplacophora) in the temperate eastern Pacific, Baja California to Alaska, with descriptions of a new species. Veliger, 21: 19— 44. FERREIRA, A. J., 1978c, The species of chitons (Neoloricata: Ischnochitonidae) from the Gal- 238 WATTERS apagos Islands. Bulletin of the Southern Califor- nia Academy of Sciences, 77: 36-39. FERREIRA, A. J., 1978d, The genus Calloplax Thiele, 1909 (Mollusca: Polyplacophora) in the Americas. Bulletin of the Southern California Academy of Sciences, 77: 56-64. FERREIRA, A. J., 1979a, The genus Callistochiton Dall, 1879 (Mollusca: Polyplacophora) in the eastern Pacific, with a description of a new spe- cies. Veliger, 21: 444—446. FERREIRA, A. J., 1979b, The family Lepidopleu- ridae (Mollusca: Polyplacophora) in the eastern Pacific. Veliger, 22: 145-165. FERREIRA, A. J., 1980, A new species of Lepi- dopleurus Risso, 1826, (Mollusca: Polyplacoph- ora) in the deep waters of the eastern Pacific. Veliger, 23: 55-61. FERREIRA, A. J., 1985, Chiton (Mollusca: Polypla- cophora) fauna of Barbados, West Indies, with the description of a new species. Bulletin of Ma- rine Science, 36: 189-219. FERREIRA, A. J. & H. BERTSCH, 1979, A new species of chiton (Mollusca: Polyplacophora) from the Hawaiian Islands and Tahiti. Transac- tions of the San Diego Society of Natural History, 19: 75-84. FISCHER, F. P., 1988, The ultrastructure of the aesthetes in Lepidopleurus cajetanus (Polypla- cophora: Lepidopleurina). American Malacologi- cal Bulletin, 6: 153-159. FISCHER-PIETTE, E. & A. FRANC, 1960, Classe des polyplacophores. In GRASSE, P. P., ed., Traité de Zoologie, 5: 1701-1785. Masson et Cie, Paris. GLYNN, P. W., 1970, On the ecology of the Carib- bean chitons Acanthopleura granulata Gmelin and Chiton tuberculatus Linne. Smithsonian Con- tributions to Zoology, 66: 1-21. GOMEZ, R. L., 1975, An association between Nut- tallina californica and Cyanoplax hartwegii, two west coast polyplacophorans. Veliger, Supple- ment, 18: 28-29. GOULD, S. J., 1966, Allometry and size in ontogeny and phylogeny. Biological Review, 41: 587-640. GOULD, S. J., 1971, Geometric similarity in allo- metric growth: a contribution to the problem of scaling in the evolution of size. American Natu- ralist, 105: 113-1386. GOULD, S. J., 1972, Allometric fallacies and the evolution of Gryphaea: a new interpretation based upon White's criterion of geometric simi- larity. In: T. DOBZHANSKY, M. HECHT & W. STEERE, eds., Evolutionary biology, 6: 91-118. Appleton-Century-Crofts, New York. GOULD, S. J., 1974, The origin and function of “bizarre” structures: antler size and skull size in the “Irish Elk”, Megaloceros giganteus. Evolu- tion, 28: 191-220. GOULD, S. J., 1977, Ontogeny and phylogeny. Belknap Press, Cambridge, MA. 501 pp. HABE, Т. 4 К. ПО, 1965, Shells of the world т colour, 1. The northern Pacific. Hoikusha Pub- lishing, Osaka, 176 pp. HADDON, A C., 1886, Report on the Polyplacoph- ora collected by H. M. S. “Challenger” during the years 1873-1876. Zoology of the Challenger Ex- peditions, 43: 1-47. HALLAM, A., 1974, What controls rates of evolu- tion? Nature, 258: 493-496. HALLAM, A., 1975, Evolutionary size increase and longevity in Jurassic bivalves and ammonites. Nature, 258: 493-496. HALLAM, A., 1978, How rare is phyletic gradualism and what is its evolutionary significance? Paleo- biology, 4: 16-25. HOARE, В. D., М. T. STURGEON 4 Т. В. HOARE, 1972, Middle Pennsylvanian (Alleghany Group) Polyplacophora from Ohio. Journal of Paleontol- ogy, 46: 675-680. HULL, A. F. B. & J. RISBEC, 1930, The loricates of the New Caledonian region (Class Mollusca—Or- der Loricata). Australian Zoologist, 6: 372-386. HULL, A. F. B. & J. RISBEC, 1931, The loricates of the New Caledonian region (Class Mollusca— Order Loricata) Il. Australian Zoologist, 6: 277- 286, 372-386. HUXLEY, J. S., 1972 (originally published 1932), Problems of relative growth. Dover Press, New York. 312 pp. HYMAN, L. H., 1967, The invertebrates, Mollusca |, 6. McGraw-Hill, New York. vii + 792 pp. IREDALE, T. & A. F. B. HULL, 1927, A monograph of the Australian loricates. Royal Zoological So- ciety of New South Wales, Sydney. 168 pp. IREDALE, T. & A. F. B. HULL, 1929, The loricates of the Neozelanic region. Australian Zoologist, 5: 305-323; 6(1): 75-95; 6(3): 59-76, 119-164. KAAS, P., 1976, Notes on Loricata 8. Description of a new Acanthochitona species from the Bay of Bengal. Basteria, 40: 119-122. KAAS, P., 1979, The chitons (Mollusca: Polypla- cophora) of Mozambique. Annals of the Natal Museum, 23: 855-879. KAAS, P. & R. A. VAN BELLE, 1985a, Monograph of living chitons. 1. Order Neoloricata: Lepi- dopleurina. Brill/Backhuys, Leiden. 240 pp. KAAS, P. & R. A. VAN BELLE, 1985b, Monograph of living chitons. 2. Suborder Ischnochitonina. Ischnochitonidae. Schizoplacinae, Callochitoni- nae & Lepidochitoninae. Brill/Backhuys, Leiden. 198 pp. KAAS, P. & R. A. VAN BELLE, 1987, Monograph of living chitons. 3. Suborder Ischnochitonina. Ischnochitonidae. Chaetopleurinae & Ischnochi- toninae (pars). Additions to Vols. 1 & 2. Brill/ Backhuys, Leiden. 302 pp. KEEN, A. M., 1971, Sea shells of tropical west America, 2nd ed. Stanford University Press, Stanford. 1064 pp. KUES, B. S., 1974, A new subspecies of Cyanoplax dentiens (Polyplacophora) from San Diego, Cal- ifornia. Veliger, 16: 297-299. LEISE, E. M., 1983, Chiton integument: ultrastruc- ture and development of sensory ornaments. Doctoral Dissertation, University of Washington, Seattle, Washington. 196 pp. POLYPLACOPHORAN MORPHOSPACE 239 LEISE, E. M., 1988, Sensory organs in the hairy girdles of some mopaliid chitons. American Mal- acological Bulletin, 6: 141-151. LELOUP, E., 1940A, Les chitons du genre Crypto- plax Blainville, 1818. Bulletin du Musée Royal d'Histoire Naturelle de Belgique, 16(15): 1-32. LELOUP, E., 1940b, Deux nouvelles especes de polyplacophores, Cryptoconchus ashbyi et Cryp- toconchus roseus. Bulletin du Musée Royal d'Histoire Naturelle de Belgique, 16(33): 1-18. LELOUP, E., 1941, A propos de quelques acantho- chitons peu connus ou nouveaux, l.—Région Indo-Pacifique. Bulletin du Musée Royal d’His- toire Naturelle de Belgique, 17(24): 1-8. LELOUP, E., 1952, Polyplacophores de l'Ocean In- dien et des côtes de |’Indochine Française. Mé- moires de l'Institut Royal des Sciences Naturel- les de Belgique, (2)47: 1-69. LELOUP, E., 1956, Polyplacophora. Reports of the Lund University Chile Expedition 1948-49. Lunds Universitets Arsskrift, new series 2, 52(15) (Kungliga Fysiografiska Sällskapets Handlingar, new series, 67(15)): 3-93. LELOUP, E., 1961, Ischnochiton adamsii (Car- penter, 1863), /. dispar (Sowerby, 1832) et I. pseudostriolatus sp. nov. Bulletin du Musée Royal d'Histoire Naturelle de Belgique, 38(18): 1-8. LELOUP, E., 1965, Description d’une espéce nou- velle de chiton recueillie par la Calypso dans le Golfe de Guinée, Notoplax foresti sp. nov. Bulle- tin du Musée d'Histoire Naturelle, new series, 37(1): 155-161. LELOUP, E., 1966, Chitons de Simonstown et de l'Ocean Antarctique. Bulletin du Musée Royal d'Histoire Naturelle de Belgique, 42(27): 1-9. LINSENMEYER, T. A., 1975, The resistance of five species of polyplacophorans to removal from nat- ural and artificial surfaces. Veliger, Supplement, 18: 83-86. LYONS, W. G., 1988a, A review of Caribbean Acanthochitonidae (Mollusca: Polyplacophora) with descriptions of six new species of Acantho- chitona Gray, 1821. American Malacological Bul- letin, 6: 79-114. LYONS, W. G., 1988b, Class Polyplacophora. Common and scientific names of aquatic inver- tebrates from the United States and Canada: mollusks. American Fisheries Society Special Publication, (16): vii+ 277 pp. MACPHERSON, J. H. & С. J. GABRIEL, 1962, Ma- rine Mollusca of Victoria. Melbourne University Press, Melbourne. 475 pp. MARINCOVICH, L., 1973, Intertidal mollusks of Iquique, Chile. Science Bulletin, Natural History Museum, Los Angeles County, (16): 1-49. MILNE, K. L., 1963, A new deep-water chiton (Poly- placophora: Chitonidae) from eastern Australia. Journal of the Malacological Society of Australia, (7): 25-27. NIERSTRASZ, H. F., 1905, Die Chitonen der Si- boga-Expedition. Résultats des Explorations Zoologiques, Botaniques, Océanographiques et Geologiques а Бога du “Siboga, ” Monograph 48. 112 pp. PLATE, L. H., 1902, Die Anatomie und Phylogenie der Chitonen. Zoologische Jahrbücher, Supple- ment, 5: 15-216. PUTMAN, B. F., 1980, Taxonomic identification key to the described species of polyplacophoran mol- lusks of the west coast of North America. Pacific Gas and Electric Company Department of Engi- neering Research Report, 411-79.342: 1-165. REEVE, L. A., 1847-1848, Conchologia iconica, 4. Monograph of the genus Chitonellus, pl. 1; Mono- graph of the genus Chiton, pls. 1-27 (1847); pl. 28 (1848). London. RIEDL, R., 1963, Fauna und Flora der Adria. Ein systematischer Meeresfuhrer für Biologen und Naturfreunde. Paul Parey, Hamburg. 640 pp. SAKAE, A., 1969, Species-form allomorphosis in the valves of Japanese chitons. Proceedings of the Japanese Society of Systematic Zoology, (4): 15-22. SIRENKO, В. I., 1976, Systematics of chitons of the genus Lepidozona Pilsbry. Soviet Journal of Ma- rine Biology, 1: 172-186. SMITH, A. G., 1960, Amphineura. In MOORE, R. C., ed., Treatise on invertebrate paleontology, part I, Mollusca, 1: 41-76. Geological Society of America and University of Kansas Press, Lawrence, Kansas. SMITH, A. G., 1961, Four species of chitons from the Panamic province (Mollusca: Polyplacoph- ora). Proceedings of the California Academy of Sciences, 30: 81-90. SMITH, A. G. & I. COWAN, 1966, A new deep- water chiton from the northeastern Pacific. Oc- casional Papers of the California Academy of Sciences, 56: 1-15. SMITH, A. G. & A. J. FERREIRA, 1977, Chiton fauna of the Galapagos Islands. Veliger, 20: 82- 97. SMITH, A. G. & G. D. HANNA, 1952, A rare species of chiton from Pioneer Seamount off central Cal- ifornia. Proceedings of the California Academy of Sciences, 27: 389-392. SMITH, S. Y., 1975, Temporal and spatial activity patterns of the intertidal chiton Mopalia muscosa. Veliger, Supplement, 18: 57-62. STANLEY, S. M., 1973a, An explanation for Cope's Rule. Evolution, 27: 1-26. STANLEY, S. M., 1973b, Effects of competition on rates of evolution, with special reference to bi- valve mollusks and mammals. Systematic Zool- ogy, 22: 486-506. TAKI, IS., 1953, A peculiar chiton from Hachijo-jima with special reference to its distribution in the Pa- cific. Records of the Oceanographic Works of Ja- pan, 1: 60-65. THIELE, J., 1909-10, Revisions des Systems der Chitonen. Zoologica, Stuttgart, 22(56), part 1: 1- 70 (1909); part 2: 71-132 (1910). UTUNOMI, H., 1968, Colored illustrations of sea- shore animals of Japan. Hoikusha Publishing, Osaka. 168 pp. 240 WATTERS VALENTINE, J. W., 1973, Phanerozoic taxonomic diversity: a test of alternate models. Science, 181: 1078-79. VAN BELLE, R. A., 1978, Sur la classification des Polyplacophora: VII. Errata, additions et rectifica- tions. Resume de la classification. Index alpha- betique. Informations de la Societe Belge de Ma- lacologie, (6)(3): 65-82. WATTERS, С. Т., 1981, Two new species of Acan- thochitona (Polyplacophora: Cryptoplacidae) from the New World. Nautilus, 95: 171-177. WATTERS, G. T., 1990, A review of the Recent eastern Pacific Acanthochitoninae (Mollusca: Polyplacophora: Cryptoplacidae) with the de- scription of a new genus, Americhiton. Veliger, 33: 249-279. WHITE, J. F. 8 S. J. GOULD, 1965, Interpretation of the allometric coefficient in the allometric equation. American Naturalist, 99: 5-18. YAKOVLEVA, A. M., 1952, Pantsyrnye mollyuski morei SSSR [Shell-bearing mollusks (Loricata) of the seas of the USSR]. In: E. М. PAVLOVSKII, ed., Atlas Bespozyonochnykh Dal'novostochnykh Morei SSR, (45): 1-107. (English translation, 1965). Revised Ms. accepted 29 October 1990 MALACOLOGIA, 1991, 33(1-2): 241-253 MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS: REASSESSING THE ASSUMPTION OF PRESERVATION-INDUCED DEFORMATION JANET В. VOIGHT" Department of Ecology and Evolutionary Biology, The University of Arizona, Tucson, Arizona 85721, U.S.A ABSTRACT Octopod taxonomists have made little use of external morphology; individual variation among specimens, investigator bias, and especially preservation artifacts are thought to limit the reli- ability of data. However, within each of ten species, logarithmically transformed linear measure- ments correlate highly with In mantle length (a common measure of body size), consistent with the hypothesis of minimal intraspecific shape variation. Octopod specimens exhibit compara- tively little non-allometric variation in shape; allometric growth coefficients derived from regres- sions reliably estimate specific growth patterns. Reports of variation are probably at least par- tially due to the use of proportions in taxonomic studies. Key words: Octopus, morphology, morphological variation, Octopodinae, allometric growth. INTRODUCTION Octopods are soft-bodied, lacking any but the most rudimentary shell remnants and a small section of hyaline cartilage over the brain. Formalin fixation produces variable amounts of shrinkage (Nixon, 1971), and for- malin can induce deformation of alcohol-pre- served specimens after as many as 70 years in preservation (Pickford, 1964). These phys- ical characteristics, the lack of well-defined anatomical points on which to base a system of measurements, and reported extensive in- traspecific variation (Robson, 1929) have lim- ited the use of external morphology in octo- pod taxonomy. Increasingly, such other characters as behaviors and chromatophore patterning (Hanlon, 1988; Roper & Hochberg, 1988) are being sought to help resolve taxo- nomic dilemmas. External morphology of octopods has con- ventionally been described by a set of indices defined by Robson (1929) in which measure- ments of one body character are expressed as proportions of another. Ratios seem intu- itively to correct for size differences and are used in octopod taxonomy to describe growth (e.g. Pickford, 1945; Mangold & Portmann, 1964; Burgess, 1966). However, unsus- pected biases have been shown in ratios in other groups, and may especially influence taxonomic accounts (Pearson, 1897; Pickford & McConnaughey, 1949; Christiansen, 1954; Minkoff, 1965; Blackith & Reyment, 1971; Atchley et al., 1976; Strauss, 1985). Biases can be created by, among other things, differ- ential growth rates (i.e. allometries) of the body sections of interest and by inevitable measurement error that is compounded by the calculation of ratios. Studies of external morphology of octo- puses have been limited, as they have been for most soft-bodied organisms. How speci- mens react to fixation and preservation has been repeatedly documented in such groups as fishes, medusae and worms (Hile, 1936; Farris, 1963; Howmiller, 1972; Theilacker, 1980; Mills et al., 1982; Fowler & Smith, 1983; LaFontaine & Leggett, 1989). Specimens typ- ically contract on fixation, eventually their weight and morphology stabilize. Whether such equilibrated specimens contain biologi- cally informative data concerning size and shape has been rarely addressed. Although variation in octopod morphology is widely bemoaned, and the potential for preservation-induced variation seems obvi- ous, the extent of morphological variation in preserved specimens has not been quantita- tively documented. This paper evaluates morphological variation in ten species of shal- low-water octopods and tests whether in- traspecific growth patterns can be traced us- ing preserved specimens. "Present Address: Department of Zoology, Field Museum of Natural History, Roosevelt Road at Lake Shore Drive, Chicago, Illinois 60605, U.S.A. 242 VOIGHT TABLE 1. Exponential growth rates of Octopus species reported in laboratory growth experiments. Measurements made on live, narcotized individuals. w = wet weight t = time. Species Growth Rate = (225 х 076383 OSOS = (1.278 105 Octopus briareus === м: = (9.48 x 10 rt! у (В 2810-Е Е x ыы Octopus joubini Octopus bimaculoides W = (2:59 10-92 = м = (5.65 X 10 2° Octopus digueti w= (6.78 x 1093973 MATERIALS AND METHODS Octopods grow at multiplicative rates; lab- oratory rearing experiments have docu- mented nearly uniform growth rates that are described by power functions during most of the life cycle for the species studied (Table 1). Because this growth pattern is documented, all data in this study were transformed to nat- ural logarithms prior to analysis. Such trans- formation both linearizes data without affect- ing allometric relationships and equalizes error over wide size ranges (Strauss, 1985), thus facilitating comparisons over the 30-fold size differences considered here. The innate relationship between any two body components is expected to show a high correlation, defined by a line on a In-In plot; because body components are documented to grow at constant rates, they will show allo- metric growth relative to one another (Lande, 1985). However, if either character varies in- dependently of the other, the expected corre- lation will be reduced. Low correlations would demonstrate that unique variation in the char- acters eradicates the expected pattern of uni- form growth, and would show that external morphology has little value in octopod taxon- omy. High correlations would demonstrate that such variation is minimal and that in- creases in overall body size are partitioned in an orderly manner over the body measure- ments considered. High correlations would support the use of external morphology in tax- onomy. If intraspecific variation is minimal (i.e. if correlations are high) specimens of a wide size range can be used to model the average growth of the species over those sizes. Individual specimens at given sizes г2 Reference 0.999 Hanlon (1983a) 0.986 Hanlon (1983a) 0.980 Hanlon & Wolterding (1989) 0.991 Forsythe (1984) 0.980* Opresko & Thomas (1975) 0.991 Forsythe & Hanlon (1981) 0.994 (@18°, Forsythe & Hanlon (1988) 0.975 (@23°, Forsythe & Hanlon (1988) 0.989 DeRusha et al. (1987) *r? calculated by Hanlon (1983b) would act as proxies for the entire species at that size. Information on species-specific growth can then perhaps distinguish species. The slope of the regression line on a In-In plot is the growth coefficient of the characters relative to each other. A slope of 1.00 repre- sents isometric growth, showing that the growth rates of the two characters are equal. Slopes greater than 1.0 indicate positive al- lometry, that the character plotted on the Y- axis grows more rapidly than the one on the X-axis. In contrast, slopes less than 1.00 in- dicate negative allometry. Standard octopod measurements defined by Robson (1929) were the basis for this study. The measurements are: Mantle Length (ML), the distance from the posterior tip of the mantle to the midpoint between the eyes; Mantle Width (MW), the greatest width across the dorsal mantle; Head Width (HW), the greatest width of the head including the eyes; Arm Length (AL), the length of the longest arm from the beak to the arm tip; Arm Width (AW), the width of the stoutest (right) arm at its midpoint; Sucker Diameter (SD), the outer diameter of the largest normal sucker (ex- cluding any specially enlarged suckers); and Web Depth (WD), the greatest depth of any web sector measured from the beak to the midpoint between two arms. These measure- ments are not necessarily homologous (some, for instance AL and AW, do not nec- essarily refer to the same limb), but they are extremal distances as defined by Bookstein (1978). Data were primarily taken from published accounts of the best known American octo- pod fauna, that of the western Atlantic and the Caribbean (Table 2). To supplement the in- MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS 243 TABLE 2. Reported are the species of Octopodinae included, the size range (in mm ML), “age” (the number of years between the date of capture and publication), number of specimens, and literature sources for each. “*” indicates data were published as ratios and ML; back-calculation of raw data was required. Size Range Age 6.3-92.9 1-86 Species Octopus briareus Octopus burryi 21-65 0-79 Octopus defilippi 2.7-85.6 2-112 Octopus filosus 6-72 2-26 Octopus joubini 11.5-54 6-7 Octopus macropus 3.3-173 2-83 Octopus maya 12.8-119 1-5 Octopus vulgaris 13-185 2-104 Octopus zonatus 14-30 2 Euaxoctopus pillsburyae 17-24 7-9 sufficient literature data available for Octopus maya and Octopus briareus, | measured lab- oratory-reared specimens with digital cali- pers. Data for trans-Atlantic species include both New and Old World specimens (e.g. Oc- topus burryi and O. vulgaris). A single speci- men of the 350 treated was deleted from this analysis. It was apparently misidentified; both its morphology and collection depth were in- consistent with those of the species to which it had been assigned. Despite the recent dis- covery that, as currently recognized, O. jou- bini contains a cryptic species, the data Pick- ford (1945) reported were treated as if from a single species. Use of published data avoids any personal bias implicit in my own data and acts conser- vatively against the colinearity hypothesis be- ing tested. Pooling data from several authors for a single species, as in O. burryi, O. defil- ippi and O. macropus (Table 2), includes po- tential investigator biases, increasing the amount of variation expected in the results. In some cases (Table 2), data were reported n References 41 Pickford (1945), n = 11*; Rees (1950), n = 1; Lab, n = 29 16 Voss (1951), п = 6; Pickford (1955), n = 1*; Adam (1960;1961), n = 8%; Palacio (1977), п = 1* 42 Voss (1964), п = 7*; Palacio (1977), n = 4*; Arocha & Robaina (1984), n = 4*; Rees (1954), n = 27 и 106 Alvina (1965), п = 106 32 Pickford (1945), п = 32* 26 — Pickford (1945), п =8*; Rees (1950, 1955), п = 12; Voss & Phillips (1957), п = 2; Palacio (1977), n = 3*; Adam (1951), п = 1 46 Voss & Solis (1966), п = 9; Lab, п = 37 48 Pickford (1945, 1955), п = 48)* 5 Voss (1968), п = 5 II 4 Voss (1975), n = 4 only as ratios and mantle lengths, requiring back-calculation of raw measurements. Back- calculation also increases the variation to be expected in this study due to the larger, com- pounded sampling error of the ratio with re- spect to the original measurements. Because the data are so varied in origin (collection dates of the specimens range from 1862 for O. macropus (Pickford, 1945) to 1986 for laboratory-reared specimens of O. maya), the techniques used to kill and fix the specimens can be assumed to be highly var- ied, but are unrecorded. Pickford (1945) noted the condition of each specimen she studied; her data reflect a wide range of spec- imen textures and consistencies. The time a specimen spent in preservation fluid prior to measurement may affect its mor- phology. | estimated this interval by the differ- ence between the year of specimen collection and the year of publication, although publica- tion date is admittedly an imprecise estimator. Specimens included in this study spent from O to 112 years in preservation fluid (Table 2). If 244 VOIGHT correlations from species with the greatest temporal variation (O. defilippi, O. vulgaris) are lower than from species with lower tem- poral variation (O. zonatus, Euaxoctopus pillsburyae), the duration of preservation will be shown to increase morphological deforma- tion. Correlations between In mantle length (ML) and the 6 In body measurements were calcu- lated with NCSS (Number Crunching Statisti- cal System) on an IBM-compatible personal computer. The regression equations reported here are functional regressions in which the variation is equally partitioned on the X and Y axes (Ricker, 1973). ML is the proxy for body size due to (1) its demonstrated high cor- relation with body weight in narcotized O. vulgaris, O. briareus and O. digueti (Nixon, 1971; Smale & Buchan, 1981; Aronson, 1982; Voight, unpub. data), and (2) the fact that ML is the standard size measure and its raw value is reported in virtually all octopod stud- ies, eliminating error that might result from us- ing a standard measure of size back-calcu- lated from reported ratios. Whether the regression slopes indicate positive or negative allometry, or instead can- not reject the null hypothesis of isometry, was determined by the 95% confidence interval of the slope. If the confidence interval included 1.0, growth was assumed to be isometric rel- ative to mantle lengthening. If the confidence interval included a value between 0.9 and 1.1, the hypothesis of isometry could not be fully rejected, and slight positive or negative allom- etry was indicated. A statistical analysis of in- vestigator bias (i.e. whether intraspecific vari- ation increases with the number of different investigators providing data) was prevented by the few specimens included in most re- ports and the limited size overlap between re- ports. RESULTS AND DISCUSSION The variation long attributed to octopod specimens has been exaggerated. The linear correlations between In MW, In HW, In AL (Fig. 1), In AW (Fig. 2), In SD (Fig. 3), In WD (Fig. 4) and In ML in nine of the ten species considered are highly significant (Table 3). The figures show the correlations are not sta- tistical artifacts created by clustered data with singular outliers; similar patterns hold for MW and HW (Voight, 1990). Only Euaxoctopus pillsburyae (n = 4; Table 2) fails to consis- tently show significant correlations (Table 3). The narrow size range available for this spe- cies prevents a valid test of the morphological relationships. With a slightly wider size range, the five specimens of O. zonatus generally show significant correlations (Table 3). Although investigator bias could not be quantified, the number of investigators provid- ing data for a species did not affect the mor- phological relationships. Both O. macropus with five investigators and O. defilippi with four show high correlations. The correlation coefficients for O. burryi, with data from four investigators, though still highly significant, are lower. These slightly lower correlations are more likely due to the few specimens of a narrower size range available (Table 2) than to investigator-induced variation. The time interval a specimen spent in pres- ervation fluid did not contribute to morpholog- ical deformation. Correlation coefficients from species with the greatest temporal variation and from those with lesser variation are not different (Table 3). The high correlations of each of six mea- surements with ML (Table 3) demonstrate that these body components are tightly regu- lated throughout post-settlement life in west- ern Atlantic octopods. The correlations are so great that comparatively little intraspecific variation can be documented among pre- served specimens. Apparently, the treatment history of the specimen, the state and dura- tion of preservation, and whatever abuses or dissections to which the specimen had been subjected, as well as potential biases of indi- vidual investigators, contribute little to the re- lationships among body components within a species. Considering the poor anatomical definition of the measurements, and that the measurements do not necessarily compare homologous points (MW, HW), or even ho- mologous limbs (AW, AL, WD, SD), their con- sistency is remarkable. A single condition is seen here to deform specimens. Senescent (post-egg laying) oc- topuses, seen in O. joubini (Pickford, 1945, n = 12), О. macropus (Voss 4 Phillips, 1957, п = 1), and O. briareus (unpub. data, n = 2), show abnormally long MLs relative to other body parameters. Although senescent speci- mens of O. briareus and O. macropus are plotted with conspecifics (Figs. 1—4), they were not included in the analyses reported in Table 3. Inclusion of senescent animals did not alter the significance of the correlations, but slightly changed the regression equa- MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS 245 TABLE 3. Reported are the correlations of the six In external measurements with In ML, slopes defined by functional regression, their standard errors, and allometries defined by the confidence interval of the slope (see text for details). An asterisk after the species name indicates senescent female(s) removed from data before correlations calculated; asterisks following the correlation coefficient indicate level of significance (*** p<.001; ** p<.01 * p<.05). “?” indicates marginally non-isometric growth. м Species n In MW versus In ML briareus* 39 burryi 16 defilippi 40 filosus 106 joubini 20 macropus 24 maya 46 vulgaris 45 zonatus 5 Euaxoctopus 4 In HW versus In ML briareus* 39 burryi 16 defilippi 42 filosus 106 joubini 20 macropus* 24 maya 46 vulgaris 47 zonatus 5 Euaxoctopus 4 In AL versus In ML briareus* 39 burryi 15 defilippi 41 filosus 106 joubini 19 macropus* 23 maya 42 vulgaris 42 zonatus 5 Euaxoctopus 2 In SD versus In ML briareus* 39 burryi 16 defilippi 31 filosus 106 joubini 19 macropus* 25 maya 45 vulgaris 44 zonatus 4 Euaxoctopus 4 In WD versus In ML briareus 37 burryi 15 defilippi 13 filosus 106 r 0.969*** 0.889*** 0.970*** 0.960*** 0.963*** 0.984*** 0.959”** 0.964*** 0.895* 0.982* 0.964*** 0.914*** 0.954*** 0.857*** 0.9707 0.982*** 0.954*** 0.925*** 0.9737 0.826 NS 0.974*** 0.928*** 9.9797 0.962*** 0.940*** 0.984*** (USAR 0.940*** ОЭ = 0.980*** 0.888*** 0.934*** 0.965777 0.968*** 0.988*** 0.937977 0.941*** 0.892NS 0.604 NS 0.9597 OS 0.723** 0:956""" Allom. Slope St. Err. ISO. 1.062 + 0.043 ISO. 1.060 + 0.130 neg? 0.861 + 0.034 iSO. 0.986 + 0.024 iso. 0.963 + 0.061 iso 0.959 + 0.038 iSO. 0.969 + 0.042 iso. 0.988 + 0.040 iso. 0.797 + 0.206 neg. 0.544 + 0.073 neg? 0.918 + 0.040 neg? 0.788 + 0.086 neg. 0.750 + 0.036 neg? 0.871 + 0.044 neg. 0.696 + 0.040 neg? 0.863 + 0.035 neg. 0.727 + 0.033 neg. 0.765 + 0.043 neg. 0.620 + 0.083 pos. 1.363 + 0.051 pos? 1.262 + 0.130 pos. 1.490 + 0.049 pos. 1.336 + 0.036 iso. 1.134 + 0.094 pos. 1.595 + 0.062 pos. 1.197 + 0.040 iso. 1.119 + 0.060 150. 2153 == 0:14 pos. 1.225 + 0.040 iso. 1.284 + 0.158 iso. 1.007 + 0.067 pos.? 1.136 + 0.029 iSO. 0.933 + 0.056 pos? 1.089 + 0.036 pos? 1113=210035 iso. 1.025 + 0.054 pos. 1.267 + 0.061 pos. 1.290 + 0.102 ISO. 0.767 + 0.160 pos. 1.314 + 0.038 (continued) 246 VOIGHT TABLE 3. (Continued) Species n r Allom. Slope+ St. Err. joubini 19 0.846*** iso. 1.105 + 0.143 macropus 11 0.979*** ISO. 1.006 + 0.064 maya 44 0.974*** pos? 1.136 + 0.040 vulgaris 40 0.964*** iso. 1.066 + 0.046 zonatus 4 0.994*** neg. 0.717 + 0.056 Euaxoctopus 4 —0.573NS In AW versus In ML briareus* 35 0.941*** pos. 1.304 + 0.077 burryi 7 0.987*** iso. 1.136 + 0.081 defilippi 13 0.881 *** neg. 0.500 + 0.071 filosus 106 0.940*** pos. 1.202 + 0.040 Joubini 16 0.982*** neg? 0.889 + 0.045 macropus* 11 0.977*** iso. 0.919 + 0.062 maya 46 0.968*** pos? 1.184 + 0.045 vulgaris 23 09532 iso. 0.995 + 0.066 zonatus 5 0.930* 150. 0.720 + 0.153 Euaxoctopus 4 0.966* ISO. 1.299 + 0.238 tions. Their deletion assures the reliability of the growth coefficients. Other specimens of O. defilippi and O. macropus also appear to display abnormally long MLs. However, the lack of independent data indicating that these specimens are indeed senescent prohibits their exclusion. Apparently the muscles of senescent and normal octopuses react differently to preser- vation and therefore create intraspecific dif- ferences. Radio-labelled leucine injected into the blood stream of reproductively maturing octopuses is not invested in muscle tissue as it is very quickly in immature octopuses (O’Dor & Wells, 1978). If reproductive matu- rity and subsequent senescence reduce pro- tein synthesis in muscles, deterioration of the somatic musculature results (O’Dor & Wells, 1978). When preserved, senescent octo- puses do not respond as do immature octo- puses and deformation results. Literature data on O. defilippi by Robson (1929), O. ornatus by Voss (1981) and O. ra- panui by Voss (1979) demonstrate that man- tle and arm elongation may also occur in less mature animals. No causal mechanism of de- formation has been suggested to my knowl- edge, but collection technique is implicated. Most deformed specimens of O. rapanui were collected with the fish poison Rotenone (Voss, 1979); how the deformed specimens of O. ornatus were collected was unreported (Voss, 1981). Some, but not all, specimens of eastern Pacific octopuses collected with fish poison show similar mantle and arm deforma- tion (unpub. data). Whether chemicals used in collection or the treatment and preservation of poisoned octopuses create mantle and arm deformation requires experimental testing. The highly significant correlations between body components allow the slopes of the re- gression lines to be interpreted as growth co- efficients. Table 3 shows general growth pat- terns. The mantle widens at the same rate that it lengthens; that is, the two traits are iso- metric with respect to one another. The head generally widens more slowly than the mantle lengthens; growth is negatively allometric. Arm growth, in contrast, is generally positively allometric relative to mantle lengthening. Arm lengthening, and growth rates of SD, WD and AW, the characters which diverge most strongly between species, may be the best traits to distinguish species. The growth patterns documented by pains- takingly narcotizing and handling a small group of laboratory-reared animals (e.g. For- sythe, 1984; DeRusha et al., 1987) compare well with the patterns derived from preserved specimens. Positive allometric growth of the arms relative to the mantle in O. joubini (For- sythe, 1984) is also seen in preserved spec- imens. The rapid arm lengthening and thick- ening in O. briareus are consistent with lab growth data (Hanlon, 1977), although stage- specific changes in growth cannot be ad- dressed with specimens of unknown ages. Unfortunately, most lab-rearing experiments report growth as a function of size versus time (Boyle, 1983; Hanlon & Forsythe, 1985; For- MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS 1.1 3.1 O. joubini 1.1 3.1 5.1 O. macropus 247 3.1 5.1 O. defilippi ott wat ЦИ 6.9 4.9 29 0.9 1.1 3.1 5.1 О. vulgaris FIG. 1. Plots of In arm length (AL) versus In ML for each species with n>6. See Table 3 for correlation and regression statistics. Specimens known to be senescent indicated by a large dot. sythe & Van Heukelem, 1987), prohibiting di- rect comparison of lab-derived growth rates to those in this study. The growth allometries reported here pre- dict species morphology; these predictions can be compared to reports in the taxonomic literature. For example, O. defilippi diverges strongly from the other species considered. Its mantle width is slightly negatively allomet- ric with respect to mantle length, whereas its head width, arm width have decidedly nega- tive allometries. Sucker growth and web deepening are isometric. Only AL shows dis- tinctly positive growth. These growth patterns 248 VOIGHT 2.8 1.4 0.0 1.1 3.1 5.1 1.1 3.1 5.1 O. burryi O. filosus 11 3.1 5.1 1.1 31 5.1 O. macropus O. vulgaris FIG. 2. Plots of In arm width (AW) versus In ML for each species with n > 6. (See Figure 1 caption for details.) produce a slender-bodied octopus with a sus (= O. hummelincki) each of the six char- shallow web and many small suckers on long, acters shows positive allometry, predicting a thin arms, consistent with the morphology re- robust species with a broad mantle and head, ported by Voss (1964). In contrast, in O. filo- a deep web and long, thick arms, bearing MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS 249 1.1 3.1 51 11 3.1 5.1 O. briareus O. defilippi 1.1 3.1 5.1 1.1 3.1 5.1 O. macropus O. vulgaris FIG. 3. Plots of In sucker diameter (SD) versus In ML for each species with n>6. Male specimens of Octopus burryi are indicated by large dots. larger (and therefore probably fewer) suckers. In the older literature (Mangold & Port- This matches the species’ overall morphology mann, 1964; Burgess, 1966) ratios were used reported by Roper et al. (1984). to analyze octopod growth. However, ratios 250 LA 3.1 5.1 VOIGHT 1.1 3.1 5.1 O. filosus 1.1 3.1 5.1 O. macropus 1.1 3.1 5.1 O. vulgaris FIG. 4. Plots of In web depth (WD) versus In ML for each species with n>6. (See Figure 1 caption for details.) can obscure relationships between charac- ters. Ratios usually change with increasing size, due not to biological reality but rather to Statistical artifacts associated with ratios (Strauss, 1985). As a result, statements de- scribing growth derived from this type of anal- ysis are confounded. This study found two types of ratio-induced error. First, ratios have obscured morphologic patterns. The sexual difference in sucker -di- | | | MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS ameter of O. burryi has been undetected, de- spite the fact that males have suckers up to twice as large as do females, after a certain body size (Fig. 3; Adam, 1961). Typically in Octopus sexually dimorphic suckers are de- fined as the two or three suckers that are con- spicuously enlarged relative to adjacent suck- ers. However, there are no conspicuous differences between adjacent sucker sizes in males of O. burryi; all suckers may enlarge simultaneously. The second type of error attributable to ra- tios is misidentification. Bivariate plots of SD versus ML clearly revealed a specimen iden- tified as O. defilippi by Palacio (1977) based on the analysis of ratios, had exceptionally small suckers relative to other members of that species. The morphological inconsis- tency, coupled with the 800-m collection depth (Palacio, 1977), suggests that this specimen was misidentified, and is probably referable to Benthoctopus (R. B. Toll, pers. comm.). Octopus defilippi and Euaxoctopus spp., although overtly similar (Voss, 1975), are also clearly separable by bivariate plots of sucker size versus mantle length, although Sucker Diameter Index (the ratio of sucker diameter to mantle length) values do not sep- arate the taxa. CONCLUSIONS Ratios in octopod taxonomy have led to misidentifications, obscured morphologic pat- terns, and contributed to the erroneous as- sumption of excessive morphological varia- tion. Although intuitively a means of escaping size-bias, ratios are only reliable in rare cases where the regression line passes through the Origin (that is, when the intercept is zero on a In-In plot). Only two of these 60 regression lines here approach this strict criterion (Voight, 1990). Ratios calculated from data that violate this criterion are correlated, often strongly, with body size, as has been noted in the squids Illex (Mangold et al., 1969) and Taonius (Dilly & Nixon, 1976) and in Octopus (Pickford & McConnaughey, 1949). Because ratios are size-linked, they usually fail to distinguish spe- cies, but even when diagnostic they may re- flect overall size differences rather than real biological differences. That morphological traits are highly corre- lated with body size in preserved cephalo- pods has been previously reported—Octopus bimaculoides and O. bimaculatus (Pickford & 251 McConnaughey, 1949), Macrotritopus spp. (= O. defilippi) (Rees, 1954), and O. filosus (Burgess, 1966); in the cranchiid squids 7. megalops (Dilly & Nixon, 1976) and Gali- teuthis glacialis (McSweeny, 1978); the sepi- olid Euprymna (Okutani & Horita, 1987); and the bathyteuthid Bathyteuthis (Roper, 1969)— as it has been in at least some specimens of heteropod gastropods (Seapy, 1985). The concept that these “naked” mollusks are ex- tremely variable, reinforced by experience with conspecific specimens of very different textures and consistencies, may have pre- vented these investigators from realizing the significance of the correlations they reported. Voss (1977) recommended re-evaluation of the basic tenets of cephalopod systematics. This study demonstrates the value of such research. External morphology of soft-bodied organisms such as octopods can be suffi- ciently robust to provide a basis for taxonomy. ACKNOWLEDGMENTS The assistance of R. E. Strauss was vital to this project. M. J. Brooks made valuable com- ments on the manuscript as did two anony- mous reviewers. R. T. Hanlon of the Univer- sity of Texas Marine Biomedical Institute (UTMBI) loaned the laboratory-reared speci- mens. The Hawaiian Malacological Society kindly provided financial support. LITERATURE CITED ADAM, W., 1951, Les céphalopodes de l'Institut Francais de l'Afrique Noire. Bulletin de l'Institut Frangais d’Afrique Noire, 13: 771-787. ADAM, W., 1960, Les cephalopodes de l'Institut Francais de l'Afrique Noire. Il. Bulletin de l'Institut Frangais d’Afrique Noire, 22: 465-511. ADAM, W., 1961, Céphalopodes. Annales de l'In- stitut Océanographie Paris, 39: 227-235. ALVINA, L. H., 1965, A study of the morphology and biology of Octopus hummelincki Adam, 1936. Ms. Thesis, University of Miami, 146 pp. AROCHA, F. & G. ROBAINA., 1984, First record of Octopus defilippi Verany, 1851 in Venezuelan coastal waters. Nautilus, 98: 25-28. ARONSON, R. B., 1982, An underwater measure of Octopus size. Veliger, 24: 375-377. ATCHLEY, W. R., C. T. GASKINS, & D. ANDER- SON, 1976, Statistical properties of ratios. |. Em- pirical results. Systematic Zoology, 25: 137-148. BLACKITH, R. E. & R. A. REYMENT, 1971, Multi- variate morphometrics. Academic Press, Lon- don, 412 pp. 252 VOIGHT BOOKSTEIN, F. L., 1978, The measurement of bi- ological shape and shape change. Lecture notes in Biomathematics, Vol. 24, Springer-Verlag, New York, 191 pp. BOYLE, P. R. 1983, Cephalopod life cycles Vol. 1. Academic Press, London, 475 pp. BURGESS L. A, 1966, A study of the morphology and biology of Octopus hummelincki Adam 1936 (Mollusca: Cephalopoda). Bulletin of Marine Sci- ence, 16: 762-813. CHRISTIANSEN, K., 1954, Ratios as a means of specific differentiation in Collembola. Entomol- ogy News, 65: 176-177. DERUSHA, В. H., J. W. FORSYTHE & В. T. HAN- LON, 1987, Laboratory growth, reproduction and life span of the Pacific pygmy octopus, Octopus digueti. Pacific Science, 41: 104-121. DILLY, P. N. & M. NIXON, 1976, Growth and de- velopment of Taonius megalops (Mollusca: Cephalopoda), and some phases of its life cycle. Journal of Zoology, London, 179: 19-83. FARRIS, D. A., 1963, Shrinkage of sardine (Sardi- nops caerulea) larvae upon preservation in buff- ered formalin. Copeia, 1963: 185-186. FORSYTHE, J. W., 1984, Octopus joubini (Mol- lusca: Cephalopoda): a detailed study of growth through the full life cycle in a closed seawater system. Journal of Zoology, London, 202: 393— 417. FORSYTHE, J. W. & R. T. HANLON, 1981, First rearing of Octopus joubini Robson 1929 on mysi- dacean and caridean shrimps. Bulletin of the AMU for 1980: 42-45. FORSYTHE, J. W. & R. T. HANLON, 1988, Effect of temperature on laboratory growth, reproduc- tion and life span of Octopus bimaculoides. Ma- rine Biology, 98: 369-379. FORSYTHE, J. W. & W. F. VAN HEUKELEM, 1987, Growth. Pp. 135-156, in: P. R. BOYLE, ed., Cephalopod life cycles Vol. 2, Academic Press, London. FOWLER, G. M. & S. J. SMITH, 1983, Length changes in silver hake (Merluccius bilinearis) lar- vae: effects of formalin, ethanol and freezing. Ca- nadian Journal of Fisheries and Aquatic Sci- ences, 40: 866-870. HANLON, R. T., 1977, Laboratory rearing of the Atlantic reef octopus, Octopus briareus Robson, and its potential for mariculture. World Maricul- ture Society, 8: 471--482. HANLON, В. Т., 1983a, Octopus joubini. Pp. 293— 310, in: P. R. BOYLE, ed., Cephalopod life cycles Vol. 1, Academic Press, London. HANLON, В. T., 1983b, Octopus briareus. Pp. 251— 266, in: Р. В. BOYLE, ed., Cephalopod life cycles Vol. 1, Academic Press, London. HANLON, R. T., 1988, Behavioral and body pat- terning characters useful in taxonomy and field identification of cephalopods. Malacologia, 29: 247-265. HANLON, R. T. & J. W. FORSYTHE, 1985, Ad- vances in the laboratory culture of octopuses for biomedical research. Laboratory Animal Science, 35: 33-40. HANLON, R. T. & M. R. WOLTERDING, 1989, Be- havior, body patterning, growth and life history of Octopus briareus cultured in the laboratory. American Malacological Bulletin, 7: 21-45. HILE, R., 1936, Age and growth of the cisco, Leucichthys artedi (LeSueur), in the lakes of the northeastern Highlands, Wisconsin. Bulletin of the Bureau of Fisheries, 44 (19): 211-317. HOWMILLER, R. P., 1972, Effects of preservatives on weights of some common macrobenthic inver- tebrates. Transactions of the American Fisheries Society, 101: 743-746. LAFONTAINE, Y. D. & W. C. LEGGETT, 1989, Changes in size and weight of hydromedusae during formalin preservation. Bulletin of Marine Science, 44: 1129-1137. LANDE, R., 1985, Genetic and evolutionary as- pects of allometry. Pp. 21-32, in: W. L. JUNGERS, ed., Size and scaling in primate Bi- ology, Plenum Press, New York. MANGOLD, K., С. С. LU & F. A. ALDRICH, 1969, A reconsideration of forms of squid of the genus Illex (стае, Ommastrephidae). Il. Sexual di- morphism. Canadian Journal of Zoology, 47: 1153-1156. MANGOLD, K. & A. PORTMANN, 1964, Dimen- sions et croissance relatives des Octopodidés Méditerranéens. Vie et Milieu, (Supp. 17): 213— 233. MCSWEENY, E. S., 1978, Systematics and mor- phology of the Antarctic cranchiid squid Gali- teuthis glacialis (Chun). Biology of the Antarctic Seas VII. Antarctic Research Series, 27: 1-39. MILLS, E. L., K. PITTMAN & B. MUNROE, 1982, Effect of preservation on the weight of marine benthic invertebrates. Canadian Journal of Fish- eries and Aquatic Sciences, 39: 221-224. MINKOFF, E. C., 1965, The effects on classification of slight alterations in numerical technique. Sys- tematic Zoology, 14:196-213. NIXON, M., 1971, Some parameters of growth in Octopus vulgaris. Journal of Zoology, London, 163: 277-284. O'DOR, В. К. & M. J. WELLS, 1978, Reproduction versus somatic growth: Hormonal control in Oc- topus vulgaris. Journal of Experimental Biology, 77: 15-31. OKUTANI, T. & E. HORITA, 1987, Identity of Eu- prymna berryi Sasaki, 1929 (Cephalopoda: Se- piolidae). Venus, Japanese Journal of Malacol- ogy, 46: 91-107. OPRESKO, L. & R. THOMAS, 1975, Observations on Octopus joubini: some aspects of reproduc- tive biology and growth. Marine Biology, 31: 51— 61. PALACIO, F. J., 1977, A study of coastal cephalo- pods from Brazil with a review of Brazilian zoo- geography. Ph.D. Dissertation, University of Mi- ami, 311 pp. PEARSON, K., 1897, On a form of spurious corre- lation which may arise when indices are used in MORPHOLOGICAL VARIATION IN OCTOPOD SPECIMENS 253 the measurement of organs. Proceedings of the Royal Society of London, 60: 489-502. PICKFORD, G. E., 1945, Le poulpe Americain: a study of the littoral Octopoda of the western At- lantic. Transactions of the Connecticut Academy of Arts and Sciences, 36: 701-811. PICKFORD, G. E., 1955, A revision of the Octopo- dinae in the collections of the British Museum. Part I. Further light on the Octopus rugosus prob- lem. Bulletin of the British Museum of Natural History (Zoology), 3: 151-162. PICKFORD, G. E., 1964, Octopus dofleini. Bulletin of the Bingham Oceanographic Collection, 19: 5-70. PICKFORD, G. E. & B. H. MCCONNAUGHEY, 1949, The Octopus bimaculatus problem: a study in sibling species. Bulletin of the Bingham Oceanographic Collection, 12: 1-66. REES, W. J., 1950, Notes on Cephalopoda from the Caribbean. Proceedings of the Malacological Society of London, 28: 107-114. REES, W. J., 1954, The Macrotritopus problem. Bulletin of the British Museum of Natural History (Zoology), 2: 69-99. REES, W. J., 1955, The larvae and late-larval stages of Octopus macropus Risso. Proceedings of the Malacological Society of London, 31: 185— 189. RICKER, W. E., 1973, Linear regressions in fisher- ies research. Journal of the Fisheries Research Board of Canada, 30: 409-434. ROBSON, С. C., 1929, A monograph of the Recent Cephalopoda. |. Octopodinae. British Museum (Natural History), London, 236 pp. ROPER, С. Е. E., 1969, Systematics and zoogeog- raphy of the worldwide bathypelagic squid Bathy- teuthis (Cephalopoda: Oegopsida). United States National Museum Bulletin, 291: 1-210. ROPER, C. F. E. & F. G. HOCHBERG, 1988, Be- havior and systematics of cephalopods from Liz- ard Island, Australia, based on color and body patterns. Malacologia, 29: 153-193. ROPER, С. Е. E., М. J. SWEENEY & С.Е. NAUEN, 1984, FAO species catalogue. Vol. 3. Cephalo- pods of the world. An annotated and illustrated catalogue of species of interest to fisheries. FAO Fisheries Synopsis, 125: 277 pp. SEAPY, R. R., 1985, The pelagic genus Pterotra- chea (Gastropoda: Heteropoda) from Hawaiian waters: a taxonomic review. Malacologia, 26: 125-135. SMALE, M. J. & P. R. BUCHAN, 1981, Biology of Octopus vulgaris off the east coast of South Af- rica. Marine Biology, 65: 1-12. STRAUSS, В. E., 1985, Evolutionary allometry and variation in body form in the South American cat- fish genus Corydoras (Callichthyidae). System- atic Zoology, 34: 381-396. THEILACKER, G. H., 1980, Changes in body mea- surements of larval northern anchovy, Engraulis mordax and other fishes due to handling and preservation. Fisheries Bulletin, 78: 685-692. VOIGHT, J. R., 1990, Population biology of Octo- pus digueti and the morphology of American trop- ical octopods. Ph.D. Dissertation, University of Arizona, Tucson, 196 pp. VOSS, G. L., 1951, Further description of Octopus burryi Voss with a note on its distribution. Bulletin of Marine Science, 1: 231-240. VOSS, G. L., 1964, Octopus defilippi Verany 1851, an addition to the cephalopod fauna of the west- ern Atlantic. Bulletin of Marine Science, 14: 554— 560. VOSS, G. L., 1968, Octopods from the R/V Pills- bury southwestern Caribbean Cruise, 1966, with a description of a new species, Octopus zonatus. Bulletin of Marine Science, 18: 645-659. VOSS, G. L., 1975, Biological results of the Univer- sity of Miami deep-sea expeditions 115. Euaxoc- topus pillsburyae, new species (Mollusca: Ceph- alopoda) from the southern Caribbean and Surinam. Bulletin of Marine Science, 25: 346-— 352. VOSS, G. L., 1977, Present status and new trends in cephalopod systematics. Symposia of the Zoo- logical Society of London, 38: 49-60. VOSS, G. L., 1979, Octopus rapanui, new species, from Easter Island (Cephalopoda: Octopoda). Proceedings of the Biological Society of Wash- ington, 92: 360-367. VOSS, С. L., 1981, A redescription of Octopus or- natus Gould, 1852 (Octopoda: Cephalopoda) and the status of Callistoctopus Taki, 1964. Pro- ceedings of the Biological Society of Washington, 94: 525-534. VOSS, G. L. & C. PHILLIPS, 1957, A first record of Octopus macropus Risso from the United States with notes on its behavior, color, feeding and go- nads. Journal of the Florida Academy of Science, 20: 223-232. VOSS, G. L. & M. SOLIS, 1966, Octopus maya, a new species from the Bay of Campeche, Mexico. Bulletin of Marine Science, 16: 615-625. Revised Ms. accepted 22 January 1991 MALACOLOGIA, 1991, 33(1-2): 255-272 THE IMPACT OF MOLLUSCICIDES ON ENZYME ACTIVITIES IN THE HEPATOPANCREAS OF DEROCERAS RETICULATUM (MULLER) Rita Triebskorn Zoologisches Institut I, Abt. Okologie/Morphologie, Im Neuenheimer Feld 230, 6900 Heidelberg, Germany ABSTRACT The influence of three commercial molluscicide pellets, Cloethocarb, Mesurol, and Spiess Urania 2000, on the activities of six enzymes in the hepatopancreas of Deroceras reticulatum were investigated by light and electron microscope histochemisty as well as by photometric studies. In the digestive cells, enzymes catalyzing energy-producing digestive processes (non- specific esterases and acid phosphatases) are induced, while, in the crypt cells, enzymes related to energy-consuming metabolic pathways often involved in detoxification (alkaline phos- phatase, and NADPH-neotetrazoliumreductase) are activated. Key words: Deroceras, molluscicides, enzymes, hepatopancreas. INTRODUCTION Xenobiotics are known to interact directly with enzymes (Wilkinson, 1976; Dauterman & Hodgson, 1990; Eldefrawi & Eldefrawi, 1990). Carbamates, for example, bind to esterases (Gordon & Eldefrawi, 1960; Tegelsstróm & Wahren, 1972) or, more specifically to cho- linesterases as competitive inhibitors (Metcalf & Fukuto, 1965; Young & Wilkins, 1989; El- defrawi & Eldefrawi, 1990). Furthermore, toxic substances can influence the homeostasis of the cell by destruction of such cellular com- ponents as mitochondria (Triebskorn, 1988). In general, the destruction of cellular struc- tures may result from an interaction between lipophilic chemicals and membranes (Sparks, 1972; Trump et al., 1981; Cascorbi & Pauli, 1990). The cell is able to react to alterations in cell homeostasis either by intensification of special metabolic pathways or by induction of particular enzymes. For example, enzymes normally catalyzing oxidative digestive pro- cesses or being involved in steroid metabolism might be used in detoxification through oxida- tion (den Besten et al., 1990). The activation of such oxidative processes by xenobiotics has often been described for vertebrates (Stohs et al., 1976; Hinton et al., 1978; Kagan, 1988), but is also mentioned for invertebrates (Lee, 1981; Widdows et al., 1981; Stegeman, 1985; Livingstone, 1988). In the present study, enzyme activities in 255 the hepatopancreas of Deroceras reticulatum were studied. In this organ, one cell type in particular, the crypt cell, which is also called basophil cell, seems to possess a special function in detoxification of pollutants (Sim- kiss & Mason, 1984). This assumption is sup- ported by my own investigations in which it has been shown that radioactive material is present in the crypt cells shortly after feeding the animals with **C-labeled Cloethocarb (Triebskorn et al., 1990). Furthermore, Ca- jaraville et al. (1990) observed an increase in the number of basophil cells after exposing animals to petrol hydrocarbons. In the present study, first, molluscicide-in- duced alterations in the activity of lysosomal enzymes involved in intracellular digestion (esterases and acid phosphatases) have been investigated. Alterations in lysosomal hydrolases induced by different kinds of stress have been described by Moore (1976), Moore & Halton (1977), Moore (1982), Moore et al. (1982), Cajaraville et al. (1989). Second, enzymes related to transport pro- cesses (alkaline phosphatases and ATPases) were studied. Banna (1980) demonstrated that after application of the molluscicide Fres- con the activity of the alkaline phosphatase in the hepatopancreas of Bulinus truncatus is higher than in control animals. Finally, two enzymes known to be involved in oxidative detoxification (NADPH-neotetra- zolium reductase and arylhydrocarbon hy- droxylase) were analyzed. 256 TRIEBSKORN TABLE 1. Time spans between the first ingestion of the different molluscicides and dissection. L: Light microscopy; E: Electron microscopy; P: Photometric measurment Cloethocarb Mesurol Metaldehyde 2% 0.1% 0.01% 0.001% 4% 4% Non-spec. esterase 5hL 6hL 6hL — 5hL 5hL 6hE 6hE 6hE — — — Acid phosphatase 5hL 5hL 5hL — 5hL 5hL 1hE — — — — 14hE — — 14hE — 5hP Alk. phosphatase 5hL 5hL 5hL — 5hL 5hL 1hE — — — — 14hE — — 14hE — 5hP ATPase 1.5hL — — 1.5hL 4hL 5hL — — 5hL 5hL NADPH-neotetra- 5hL 5hL 5hL 3wL — 5hL zoliumreductase Arylhydrocarbon- 5hP 5hP 5hP hydroxylase MATERIALS AND METHODS Laboratory-reared Deroceras reticulatum were fed molluscicides, which were applied either as commercial pellets or as a self-made wheat-bran agar formulation containing differ- ent concentrations of the molluscicide agents. Control animals were fed wheat-bran agar. The following substances were used: Mesurol: commercial pellets containing 4% of the effective substance 4-(methyl-thio)- 3,5-xylyl-methyl-carbamate. Spiess-Urania 2000: commercial containing 4% metaldehyde. Cloethocarb: wheat-bran agar containing 2%, 0.1%, 0.01% or 0.001% of 2-(2-chloro-1- methoxy-ethoxy-)phenyl-N-methylcarbamate. The animals were dissected at different times after the first ingestion of the mollusci- cides. The duration of exposure is illustrated in Table 1. For light microscope, histochemical en- zyme tests, the digestive system was isolated and the hepatopancreas was either frozen in isopentan, which was cooled in liquid nitro- gen, or was fixed for 1 h in 2% glutaraldehyde (dissolved in 0.01 M cacodylic buffer, pH 7.4). Fixed material was embedded without being dehydrated in HistoResin. For enzyme tests on the electron micro- scope level, vibratome-cut sections of the he- patopancreas (of about 100 um) were fixed in pellet a 2% glutaraldehyde solution in cacodylic buffer (0.01 M, pH 7.4) for 45 min (at 4°C). Afterwards, they were rinsed in the same buffer (in the test for esterases 10% dimethyl- sulfoxide was added), and then they were transfered into the incubation medium. After incubation, the tissues were postfixed in re- duced osmium (Karnovsky, 1971), stained en bloc with uranylacetate, dehydrated, and em- bedded in Spurr’s medium (Spurr, 1969). Ultrathin sections of 100-200 nm were cut on a Reichert-OM-U2 ultramicrotome and ex- amined in a Zeiss EM 9 without further stain- ing. For photometric measurements of enzyme activities, the animals were dissected, the di- gestive tract and the hepatopancreas were isolated, the gut content was removed, and finally the tissues were frozen at —80°C ina deep-freezer. For arylhydrocarbonhydroxy- lase, fresh tissues (the hepatopancreas not removed from the digestive tract) were used. The tissues were put in a buffer specific for the enzyme tested, homogenized either in a Poly- tron-Kinematica potter or in a hand-potter (for arylhydrocarbonhydroxylase), and centri- fuged in a Sorvall RC 2-B centrifuge. For the different enzyme tests, saturation curves were obtained in a LKB Biochrom Ultraspec Plus or in a Beckman DU-6 spectrophotometer. Finally, the enzyme activity could be calcu- lated using the following formula: IMPACT OF MOLLUSCICIDES ON DEROCERAS 257 AE/min x v р = ——— [muUnits/ml] exdxv c: enzyme activity AE: difference of extinction V: test volume (content of the cuvette, 1ml) v: sample volume (0.02 ml) d: diameter of the cuvette (1 cm) e: coeffizient of extinction (cm°/umol) The following methods were used Non-specific esterases (NE) Light microscope-HistoResin sections Method: Lojda et al. (1976) Substrate: natriumdisphosphate Electron microscope Method: Bell & Barnett (1965) Substrate: thioacetoacid Acid phosphatases (AcP) Light microscope-HistoResin sections Method: Werner (1986) Substrate: naphthol-AS-biphosphate Electron microscope Method: Robinson & Karnovsky (1983) Substrate: Na-B-glycerophosphate Photometric measurement Method: Bergmeyer (1970) Substrate: p-nitrophenylphosphate Alkaline Phosphatases (AIP) Light microscope-HistoResin sections Method: Werner (1986) Substrate: naphthol-AS-MX-phosphate Electron microscope Method: Robinson & Karnovdky (1983) Substrate: Na-B-Glycerophosphate Photometric measurement Method: Bergmeyer (1970) Substrate: p-nitrophenylphosphate Na*-Ka* -АТРазе (ATP) Light microscpoe-cryosections Method: Wachstein & Meisel (1957) Substrate: ATP -Na ‘ -salt NADPH-neotetrazoliumreductase (NTR) Light microscope-cryosections Method: Lojda et al. (1976), Bayne et al. (1985) Substrate: NADPH Arylhydrocarbonhydroxylase (AHH) Photometric measurement Method: Collodi et al. (1984) Substrate: diphenyloxazole RESULTS In the following, molluscicide-induced alter- ations of enzyme activity in the hepato- pancreas of Derocras reticulatum will be described. The results from the enzyme his- tochemical studies are confined to reactions in the three cell types of the hepatopancreas: the digestive cells, the crypt cells, and the ex- cretory cells. For the photometric measure- ments, the enzyme activity in the hepatopan- creas will be compared with that in the alimentary tract, except for the arylhydrocar- bonhydroxylase, because for this test, the he- patopancreas had not been removed from the digestive tract. Non-specific Esterases Light microscopy: Control. Most of the reac- tion product is localized in small vesicles in the apical half of the digestive cells (Fig. 1). Less intense reaction can be observed in the cytoplasm. Cloethocarb. With all Cloethocarb concen- trations, but especially with the 0.01% oral dose (OD), the reaction of esterases in the cytoplasm of digestive cells is stronger than in the controls. Large amounts of reaction prod- uct can be found in the tubular lumen (Fig. 2). Mesurol. After ingestion of Mesurol, the re- action in the digestive cells is more intense than in the controls. The reaction products are dispersed throughout the cytoplasm of the cells. Metaldehyde. The cytoplasm of the diges- tive cells stained more intensely than in the controls. Electron microscopy: Control. Reaction prod- uct is localized in the cytoplasm and on the microvilli of the digestive cells (Fig. 13), and in the endocytotic channels and vesicles origi- nating between them (Fig. 14). Vesicles in the apical half of the cell show enzyme activity, as well. Small amounts of precipitate can be found dispersed in the cytoplasm of digestive and crypt cells. Furthermore, small vesicles in the crypt cell stained intensely. Cloethocarb. After application of 2% or 0.1% OD, the reaction of non-specific ester- TRIEBSKORN IMPACT OF MOLLUSCICIDES ON DEROCERAS 259 ases in the cytoplasm as well as on the mi- crovilli and cell bases of digestive and crypt cells is stronger than in controls (the last two reactions can also be observed with 0.01% OD) (compare Figs. 13, 14 with 15, 16). Ex- cept for 0.01% OD, precipitate in the endocy- totic channels and vesicles of the digestive cells is not observable. Acid phosphatase Light microscopy: Control. Reaction product is localized distinctly in small vesicles in the most apical parts of the digestive cells (Fig. 3). The enzyme activity in the cytoplasm of the digestive, crypt, and excretory cells is less intense. Cloethocarb. After poisoning the animals with Cloethocarb in concentrations higher than 0.01%, only few reaction of the enzyme can be observed in the digestive cells (Fig. 5). After 0.01% OD, only few vesicles in the mid- dle of the digestive cells are stained. Mesurol. The reaction after uptake of Mesurol is less intense than in controls, but more intense than after Cloethocarb inges- tion. In some cells, precipitate can be ob- served in vesicles in the middle of the diges- tive cells as well as dispersed over the cytoplasm (Fig. 4). Metaldehyde. Phosphatase activity in the digestive cells was reduced after poisoning the animals with metaldehyde. However, the reaction is more intense than after ingestion of the two carbamates. In contrast to the con- trols, reaction product can be found in the basal parts of the digestive cells. Electron microscopy: Control. In the diges- tive cells, reaction product is localized in small vesicles that are often attached to large vac- uoles or lipid droplets in the apical half of the cells (Fig. 17). Little enzyme activity can be found on the microvilli of digestive, crypt, and excretory cells. Cloethocarb: As soon as 1 h after poisoning the animals with 2% Cloethocarb, but espe- Cially after 14 h, the most striking enzyme ac- tivity can be observed in the endoplasmic re- ticulum, the Golgi apparatus (Figs. 18, 19, 20), and in the basal labyrinth of the crypt cells. Large amounts of precipitate can also be found in the cytoplasm and in small vesi- cles in close contact to dictyosomes and en- doplasmic reticulum. The enzyme reaction in this cell type is more intense than in the con- trols. In the digestive cells, enzymatically ac- tive vesicles can be found in the basal half of the cells. The microvilli of digestive and ex- cretory cells show intense enzyme activity. After 1 h, enzyme activity is also present in the lumen of the gland tubules, especially in regions where cells are extruded from the ep- ithelia. Furthermore, especially after 14 h, the hemolyph space is distinctly stained. Mesurol: 14 h after the first ingestion of Mesurol, large numbers of heavily destroyed digestive and crypt cells can be found, with precipitate dispersed throughout the cyto- plasm. Phosphatase activity in the endoplas- mic reticulum, the Golgi apparatus, and the vesicles in the crypt cells is as intense as after Cloethocarb poisoning. In the digestive cells that are still intact, reaction product is local- ized on the microvilli and in small vesicles be- tween the microvilli. The hemolymph space is heavily stained. Photometric measurement: Figure 25 clearly demonstrates that the activity of acid phos- phatases in the hepatopancreas (9.43) is nearly twice as high as in the alimentary tract (4.71). Whereas the enzyme activity in the al- imentary tract was not influenced by the mol- luscicides, all of the molluscides tested in- duced a similar reduction in acid phosphatase activity in the hepatopancreas by approxi- mately 35%. Alkaline Phosphatase Light microscopy: Control. In the most basal parts of the epithelial cells, a small layer showing enzyme activity can be observed. FIGS. 1-7. Light microscope staining for non-specific esterases (NE), acid (AcP) and alkaline (AIP) Phos- patases). 1. NE, control. Enzyme reaction in vesicles in the apical half of the digestive cells (arrows), x 380. 2. NE, 2% Cloethocarb, 5 h. Increased enzyme activity in the lumen of the tubules (arrows), x 130. 3. AcP, control. Positive reaction in vesicles in the most apical parts of the digestive cells (arrows), x 240. 4. AcP, Mesurol, 5 h. Minimal precipitate in the middle of the digestive cells (arrows), x 200. 5. AcP, 2% Cloethocarb, 5 h. weak enzyme reaction in the digestive cells (arrows), x 250. 6. AIP, 0.01% Cloethocarb, 5 h. Reaction product on the cell bases, in apical lying vesicles of the digestive cells and in excretory granules (arrows), x 250. 7. AIP, 2% Cloethocarb, 5 h. Strong reaction on the cell bases and in excretory granules (arrows), х 250. 260 TRIEBSKORN Furthermore, reaction product is localized in small vesicles in the apical half of the diges- tive cells and in the excretory cells. Cloethocarb. After uptake of 2% or 0.1% Cloethocarb, the reaction in the most basal parts of the cells is more intense than in con- trols or after 0.01% Cloethocarb, which is sim- ilar to that in controls (compare Figs. 6 and 7). Vesicles in the digestive cells stain less and reaction in the excretory cell is more distinct than in controls. Mesurol: The reaction can be compared with that occuring after Cloethocarb intoxica- tion. The most striking reaction is to be found in the excretory cells. Metaldehyde. After ingestion of the metal- dehyde-containing pellets, reaction in the basal parts of the cells, in the excretory cells and in the lumen of the tubules is more in- tense than in the controls. Electron microscopy: Control. While on the microvilli of digestive and crypt cells only few reaction products could be found, the apical surfaces of the excretory cells reveal intense phosphatase activity. In the digestive and crypt cells, the cisternae of the endoplasmic reticulum located close to the cell surface are distinctly stained. Furthermore, small phos- phatase-positive vesicles are present in the basal parts of the crypt cells. Precipitate can also be found in the excretory vacuoles. Rel- atively little reaction product can be observed dispersed throughout the hemolymph. Cloethocarb. After 1 h, but more intensely after 14 h, the most striking reaction of alka- line phosphatase can be found in the endo- plasmic reticulum of both the digestive and the crypt cells (Figs. 21, 22). Precipitate is also present in cis-face cisternae of the Golgi apparatus and in basally lying vesicles of the crypt cells (Figs. 23, 24). In the digestive cells, reaction product often surrounds lipid droplets. Large amounts of precipitate can be found in the hemolymph space and in the tu- bular lumen. Mesurol. Similar to ingestion of Cloetho- carb, the enzyme activity in the endoplasmic reticulum of digestive and crypt cells is in- tensified. The most distinct reaction can be observed in cisternae lying next to the cell surface. Furthermore, large amounts of pre- cipitate are present in small vesicles in the basal parts of the crypt cells and in the he- molymph space. Photometric measurement: Figure 26 shows that in control animals the activity of alkaline phosphatase in the hepatopancreas (2.35) is lower than in the alimentary tract (3.66). In both parts of the digestive system, the phos- phatase activity is activated by molluscicides. After poisoning the animals with metalde- hyde, the enzyme activity in the hepatopan- creas is more than twice as high as in the controls (6.53). After application of the two carbamates, it is increased to 3.97 (Cloetho- carb) or 4.33 (Mesurol). ATPase Light microscopy: Control. Enzyme activity can especially be found in the basal parts of the epithelial cells, as well as in the muscle and nerve tissue underlying the epithelia. Positively reacting vesicles are also present in the apical parts of the digestive cells and in vesicles in the crypt cells (Fig. 8). Cloethocarb. The reaction in the basal parts of the epithelial cells and in the muscle and nerve tissue is slightly reduced. Large amounts of reaction product can be found in the apical parts of digestive cells, either local- ized in vesicles or dispersed throughout the cytoplasm. Mesurol. The reactions are similar to those after Cloethocarb ingestion (Fig. 9). Metaldehyde. Enzyme activity in the apical parts of the digestive cells is more intense than in controls. Reaction product is distinctly localized in vesicles and on the microvilli bor- der (Fig. 10). NADPH-Neotetrazoliumreductase Light microscopy: Control. In the control an- imals, enzyme activity can be found in the crypt cells, especially on the cell apices (Fig. 11). The apices of the digestive cells are free of precipitate. Cloethocarb. After uptake of 2% or 0.1% FIGS. 8-10. Light microscope staining for ATPase (ATP). 8. ATP, control. Strong reaction on the cell bases, in vesicles of crypt (CC) and digestive cells (DC) as well as in muscle tissue, х 260. 9. ATP, Mesurol, 5 h. Strong reaction on the cell apices and in muscle tissue, x 400. 10. ATP, Metaldehyde, 4 h. Strong enzyme activity on the microvillous border and in apical lying vesicles of the digestive cells as well as in muscle tissue, X 260. IMPACT OF MOLLUSCICIDES ON DEROCERAS 261 TRIEBSKORN IMPACT OF MOLLUSCICIDES ON DEROCERAS 263 Cloethocarb, the enzyme activity in the he- patopancreas is stronger than in the controls. Especially in the crypt cells, reaction product is dispersed throughout the cytoplasm (Fig. 12). After ingestion of sublethal concentra- tions (0.001%), the reaction in the crypt cells is more intense than after 5% or 0.1%. Metaldehyde: The reaction can be com- pared with that after ingestion of sublethal Cloethocarb concentrations. Arylhydrocarbonhydroxylase Photometric measurement: Figure 27 dem- onstrates that the enzyme is clearly activated by metaldehyde (from 5.19 to 10.39). Mesurol induces a slight increase (to 7.7), whereas the enzyme activity is not influenced by Cloe- thocarb (5.3). DISCUSSION In the present study, it is shown that poi- soning slugs with molluscicides induces mod- ifications in the activity of several enzymes. Alterations in enzyme activity can be corre- lated with molluscicide-induced changes in ul- trastructure of the cells in the hepatopancreas of Deroceras reticulatum described previ- ously (Triebskorn, 1989; Triebskorn & Kü- nast, 1990). In a first step, alterations of hydrolases were investigated. The impact of mollusci- cides on these intracellular digestive en- zymes is very important, as they are involved in essential metabolic pathways during intra- cellular digestion, breaking down ester com- pounds by hydrolysis. Because it has already been reported by Tegelstóm & Wahren (1972) that carbamates not only inhibit cho- linesterases but influence esterases in gen- eral, the affect of Cloethocarb on non-specific esterase activity was examined. Non-specific esterases is the term used for a group of enzymes that includes carboxylesterases, arylesterases and acetylesterases. An inhibi- tion of these non-specific esterases could be observed only in the cells of the alimentary tract (e.g. in the oesophagus, crop stomach and gut) (Triebskorn, unpublished). In the he- patopancreas, the enzyme activity in the cy- toplasm of the cells was stronger after appli- cation of the molluscicide than in controls. This might be related to the fact that several types of esterases exist in the different parts of the digestive system with different sensitiv- ities to carbamates. After carbamate poisoning, the high activity of esterases in the cytoplasm of digestive cells and the lack of reaction product in the vesicles might be due to the destruction of vesicle membranes, as described by Trieb- skorn & Künast (1990). Esterases normally found in membrane-bound compartments are set free into the cytoplasm and catalyze reac- tions leading to autolysis. Such autolytic pro- cesses are often discussed for terrestrial and marine invertebrates as a result of stress- induced instability of lysosome membranes (Moore, 1976; Oxford & Fish, 1979; Ca- jaraville et al., 1989). However, after metalde- hyde ingestion, large amounts of non-specific esterases can be found in the digestive cells, even though the lysosomal system is not as damaged as after carbamate poisoning (Trieb- skorn, 1989). In general, the results of the electron micro- scope test in control animals make evident that some of the non-specific esterases local- ized in vesicles in the apical parts of the di- gestive cells of the hepatopancreas are not produced by these cells, but are resorbed by pinocytotic processes. This can be assumed from the fact that no Golgi apparatus could be found producing these esterase-positive ves- icles and that high enzyme activity is present in the lumen of the tubules as well as in pi- nocytotic vesicles and channels. The presence of non-specific esterases in vesicles of the crypt cells might also be due to intracellular digestive processes. Especially because crypt cells are thought to be involved in detoxification of such foreign compounds as heavy metals (Simkiss & Mason, 1984), a functional role of esterases in detoxification might be possible. The role of esterases in detoxification of carbamates is mentioned by Gordon & Eldefrawi (1960), who speak of special “carbamate-esterases.” As asecond intracellular digestive enzyme- system, the acid phosphatases, known as key enzymes in primary lysosomes (Goldfischer et al., 1964), were examined. Enzymes of this group are able to catalyze the breakdown of FIGS. 11, 12. Light microscope staining for NADPH-neotetrazolium reductase (NTR). 11. NTR, control. Reaction on the apices of crypt cells, x 480. 12. NTR, 0.1% Cloethocarb 5 h. Increased enzyme activity in the crypt cells, x 300. TRIEBSKORN % XL de 4 < OS IMPACT OF MOLLUSCICIDES ON DEROCERAS 265 esterbonds in orthophosphate esters under acid conditions and are involved in the attack of pyrophospate bonds. They act additionally as transphosphorylases (Lojda et al., 1976). The distribution of this enzyme activity in the three cell types of the hepatopancreas agrees well with that described by Sumner (1969) for the midgut gland cells of Mytilus edulis and Helix pomatia and by Bowen (1970) for those of Arion ater. In the crypt cells, Golgi cisterns, which stained strongly for the enzymes, are similar to those described as rigid lamellae, and large phosphatase-active vesicles seem to be con- densing vacuoles (Novikoff et al., 1977; Hand & Olivier, 1977). In the vacuolar system of the digestive cells, the positive-reacting vesicles might be primary lysosomes. It had been expected that molluscicide ingestion, leading to lysosomal instability, would result—as in the case of es- terases—in increased phosphatase activity in the cytoplasm and reduced activity in the ves- icles. This turned out not to be true. The light microscope tests clearly demonstrate that af- ter Cloethocarb poisoning, phosphatase ac- tivity is totally diminished, where as after Mesurol and metaldehyde a little activity can still be observed in basally located vesicles. This speaks in favour of two phenomena: first, the membranes of primary lysosomes seem to be less unstable than those of secondary lysosomes, which contain non-specific es- terases and break down under stress condi- tions; second, acid phosphatases seem to catalyze reactions at the beginning of intra- cellular digestion shortly after the fusion of pri- mary lysosomes with digestive vacuoles or secondary lysosomes. Molluscicides induce the fusion activity (eventually due to an im- pact on the cytoskeleton) and therefore ac- celerate phosphatase-catalyzed primary di- gestive processes. This hypothesis is also supported by the present photometric mea- surements, which demonstrate the impor- tance of the hepatopancreas in digestive processes catalyzed by acid phosphatase. The results of electron microscope studies, however, do not correspond well with the mol- luscicide-induced reactions observed in light microscope studies. After taking up Cloetho- carb, the digestive cells show small vesicles with distinct phosphatase activity in the basal parts of the cell. Furthermore, the strong re- action on the microvilli and in the tubular lu- men could not be found in light microscope studies. The differing results, obtained by dif- ferent methods, might be due to the fact that in all these tests different substrates for acid phosphatases were used, and more or less different types of acid phosphatases may have been stained. The substrate specifity of enzyme groups and the problems that may arise if results obtained by different methods are compared are also emphasized by Newell (1977), Oxford & Fish (1979), and Bowen (1981). As enzymes possibly involved in transport processes (Lojda et al., 1976), alkaline phos- phatases and ATPases were investigated. Al- kaline phosphatases break down ester com- pounds of orthophosphate acids under alkaline conditions between pH 9.2 and 9.8 (Lojda et al., 1976). A special type of alkaline phosphatase with a pH optimum of 7.5 is ca- pable of catalyzing the hydrolytic break down of ATP. This is the cell membrane ATPase, the activity of which depends on the presence of Ма ' - and K* - ions. All molluscicides tested induced an in- crease in the activity of the alkaline phos- phatases in the cells of the hepatopancreas. This was evident in the light and electron mi- croscope tests, as well as in the photometric measurements. The increased activity of the enzyme, especially on the basal cell surfaces, might be correlated with an intensified trans- port of the molluscicides from the hemolymph space into the cells of the hepatopancreas. The presence of ATPase in these regions of FIGS. 13-18. Electron microscope investigations for non-specific esterases (NE) and acid phosphatases (AcP). 13. NE, control. Enzyme reaction in the cytoplasm and on the microvilli (MV) of a digestive cell, x 14200. 14. NE, control. Reaction product in pinocytotic channels and vesicles of a digestive cell, x 42000. 15. NE, 2% Cloethocarb, 6 h. Strong enzyme activity in the cytoplasm of a digestive cell, x 14200. 16. NE, 0.01% Cloethocarb, 6 h. Enzyme activity in pinocytotic channels of a digestive cell, x 14200. 17. AcP, control. Vesicles with enzyme activity in a digestive cell, x 31300. 18. AcP, 2% Cloethocarb, 14 h. Endoplasmic reticulum (ER) with enzyme activity surrounding lipid droplet (L), x 19300. TRIEBSKORN IMPACT OF MOLLUSCICIDES ON DEROCERAS 267 ACI Vos phatase SS 7 6 57 Y +1.0 fr = / 4.46 +0.6 / Co Bas Mes Met n = 4 FIG. 25. Photometric measurement of acid phosphatases in the hepatopancreas (HP) and the digestive tract (ОТВ) of Deroceras reticulatum in controls (Co) and 5 h after ingestion of Cloethocarb (Bas), Mesurol (Mes) or the metaldehyde containing molluscicide Spiess Urania 2000 (Met). For each treatment, four animals were measured three times each. the cells also supports this hypothesis. It is further supported by the results of autoradio- graphic studies (Triebskorn et al., 1990), in which it was demonstrated that radio-labeled Cloethocarb can be found in the crypt cells only 1 h after the first ingestion of the mollus- cicide. The enhanced activity of alkaline phos- phatase in the cisternae of the granular en- doplasmic reticulum in the crypt cells might be due to an activation of intracellular, energy- consuming processes. It is possible either that the synthesis of special enzymes even- tually involved in detoxification is activated or that intracellular transport of essential metab- olites is reinforced due to an increased en- ergy demand. The latter is supported by the fact that immediately after poisoning with met- aldehyde or Cloethocarb, storage products in the cells, especially in the crop, are reduced (Triebskorn, 1989; Triebskorn & Künast, 1990), and that this reduction is proceeded by FIGS. 19-24. Electron microscope staining of acid (AcP) and alkaline (AIP) phosphatases. 19. AcP, 2% Cloethocarb, 1 В. Enzyme-active vesicles (V) in a crypt cell arising from a Golgi apparatus (GA), x 20000. _ 20. AcP, 2% Cloethocarb, 14 h. Phosphatase activity in a cisterna of the Golgi apparatus (GA) and in Golgi vesicles (V), x 42000. 21. AIP, 2% Cloethocarb, 1 h. Phosphatase activity in the endoplasmic reticulum (ER) of a crypt cell, x 10900. 22. AIP, 2% Cloethocarb, 1 h. Endoplasmic reticulum (ER) in the basal part of a crypt cell (B) with enzyme activity, x 14200. | 23. AIP, 2% Cloethocarb, 1 h. Crypt cell with reaction in the endoplasmic reticulum (ER) and in a cisterna | of the Golgi apparatus (GA), x 19900. | 24. AIP, 2% Cloethocarb, 1 h. Vesicle (V) in the basal part of a crypt cell with enzyme activity, x 15600. 268 TRIEBSKORN Alkaline Phosphatase | FIG. 26. Photometric measurement of alkaline phosphatases in the hepatopancreas (HP) and the digestive tract (DTR) of Deroceras reticulatum in controls (Co) and 5 h after ingestion of Cloethocarb (Bas), Mesurol (Mes) or the metaldehyde containing molluscicide Spiess Urania 2000 (Met). For each treatment, four animals were measured three times each. a proliferation of the endoplasmic reticulum, which shows high alkaline phosphatase activ- ity. As enzymes that might be active in oxida- tive detoxification, NADPH-neotetrazolium- reductase (NTR) and arylhydrocarbonhy- droxylase (AHH) were tested. NTR, a NADPH-cytochrome P-450-reductase with tetrazolium-salt as an artificial substrate, was used by Widdows et al. (1981), Moore & Lowe (1985), and Nott et al. (1985) as an indicator enzyme for environmental stress. Both the NTR and the AHH belong to the MFO- system (mixed function oxygenases), which is normally involved in the metabolism of steroid hormones (Lee, 1981). Enzymes of this system are characterized by a low sub- strate specificity (Netter, 1980). They are therefore able to bind several kinds of toxi- cants or xenobiotics. Furthermore, in mam- mals, enzymes of the MFO-system are known to be induced within a few hours (Kagan, 1988). NTR was activated by both molluscicides tested, especially in the crypt cells. The acti- vation after sublethal concentrations of Cloe- thocarb and after metaldehyde application was stronger than after application of highly concentrated Cloethocarb pellets. Because metaldehyde and low concentrations of Cloe- thocarb induced less degenerative effects in the cells of the hepatopancreas, but conspi- cious alterations of the endoplasmic reticulum (Triebskorn, 1989), two conclusions can be drawn: First, because after metaldehyde as well as after low Cloethocarb concentrations only moderate degenerative effects occur in the cells, reactive processes may be switched on, allowing the animals to react to the poisoning. This partly agrees with the quantitation of.the AHH, which is most intensely activated by met- IMPACT OF MOLLUSCICIDES ON DEROCERAS 269 MFO Y son % Co Bas Mes ni = 7 FIG. 27. Photometric measurement of the MFO enzyme arylhydrocarbonhydroxylase in control animals (Co) of Deroceras reticulatum and 5 h after ingestion of Cloethocarb (Bas), Mesurol (Mes) or the metaldehyde containing molluscicide Spiess Urania 2000 (Met). For each treatment, seven animals were measured three times each. aldehyde. The reason why Cloethocarb does not induce this enzyme, whereas Mesurol, the molluscicide that leads to heavy ultrastructure damage does is not known. It is possibly due to the halogenisation of Cloethocarb. Second, ultrastructural alterations in the endoplasmicreticuluminthecryptcells (Triebs- korn, 1989; Triebskorn & Künast, 1990) may possibly be related to the activation of the NTR, which is known to be localized on the smooth endoplasmic reticulum (Smuckler & Arcasoy, 1969; Netter, 1980; Lee, 1981). The relation between changes in the endoplasmic reticulum and induction of MFO enzymes has already been described for vertebrates (Klaunig et al., 1979). Comparing these results to those obtained in experiments with radio-labeled material (Triebskorn et al., 1990), another conclusion is that Cloethocarb is transported from the he- molymph into the crypt cells of the hepato- pancreas, where metabolic processes may take place. In contrast to heavy metals, which are detoxified in the crypt cells by binding to spherites and are finally stored in these cells (Simkiss & Mason, 1984), the molluscicide is retransported after a possible metabolisation into the hemolymph. Radioactive material can be detected in hemolymph cells (Triebskorn et al., 1990). Finally, it should be emphasized that en- zymes of the MFO system are involved in the metabolism of steroids in untreated animals (den Besten et al., 1990). Only in case of poi- soning, they have a function in detoxification. Summarizing these results, it seems evi- dent that in the digestive cells, intravacuolar digestive processes are intensified, leading to large secondary lysosomes with fragile membranes. The cells activate intracellular digestion, winning energy for reactive, energy consuming, and possibly detoxification pro- cesses that probably take place in the crypt cells. Reactions in the excretory cells are dif- 270 TRIEBSKORN ficult to interpret because there is only little information about their function in the hepato- pancreas of untreated animals. In any case, the high activity of alkaline phosphatase on the microvilli and in the large central vacuole suggests that active transport processes also take place in these cells. In general, the question arises about the way in which enzyme activities are altered in a time span of only a few hours. Inhibitory effects might be related to the fact that the effective substances may interact di- rectly with enzymes as competitive inhibitors. This is often described for carbamates and cholinesterases (Young & Wilkins, 1989). Furthermore, the production of metabolites acting as inhibitors may be intensified by poi- soning. Finally, the destruction of the or- ganelles as sites of the localization of en- zymes may lead to a reduction in the activity of those enzymes. The reason for enzyme activation might be due to an interaction of the effective sub- stance or of its metabolites with repressors, leading to the transcription of normally re- pressed genes. For this, the presence of in- ductive enzymes must be postulated. Further- more, it is also possible that the direct interaction of the chemical or its metabolite with the respective enzyme leads to an alter- ation in the enzyme configuration. As a result, an enzyme with a moderate activity on the substrate used could become less specific and its activity might be increased. ACKNOWLEDGEMENTS | thank Prof. W. Wieser for allowing the en- zymatic studies at the Institute of Zoophysiol- ogy, Innsbruck, and Reinhard Dallinger, Rein- hard Lackner, and Sylvia Hinterleitner for their friendly help during these experiments. | am very grateful to BASF, Limburgerhof, especially to Prof. Dr. C. Kunast, for financial support of this study. Many thanks are also due to David Russell for the revision of the English version. LITERATURE CITED BANNA, H. B. M., 1980, Histochemical studies of some enzymes in the tissues of the schistosome vector snail Bulinus truncatus (Andouin) with special reference to the effects of a molluscicide. |. Hydrolases. Histochemical Journal, 12: 145— 152. BAYNE, B. L., D. A. BROWN, K. BURNS, D. R. DIXON, A. IVANOVICI, D. R. LIVINGSTONE, D. M. LOWE, M. N. MOORE, A. R. D. STEBBING & J. WIDDOWS, 1985, The effects of stress and pollution on marine animals. Praeger Scientific, New York. BELL, M. & J. BARNETT, 1965, The use of thio- substituted carboxylic acids as histochemical substrates. Journal of Histochemistry and Cy- tochemistry, 13: 611-628. BERGMEYER, H. U., 1970, Methoden der enzyma- tischen Analyse. Verlag Chemie, Weinheim. BOWEN, I. D., 1970, The fine structural localization of acid phosphatase in the gut epithelial cells of the slug Arion ater (L.). Protoplasma. 70: 247- 260. BOWEN, I. D., 1981. Techniques for demonstrating cell death. Pp. 379-444, In: J. D. BOWEN & R. A. LOCKSHIN, 1981, Cell death in biology and pathology. Chapman and Hall, London, New York. CASCORBI, |. & W. PAULI, 1990. Phenols alter physico-chemical properties of plasma mem- branes: |: Effects on membrane fluidity. Abstracts of the 12th Annual Conference of Physiological and Biochemical Approaches to the Toxicological Assessment of Environmental Pollution. P 2-42. CAJARAVILLE, M. P., J. A. MARIGOMEZ & E. AN- GULO, 1989, A stereological survey of lysosomal structure alterations in Littorina littorea exposed to 1-Naphthol. Comparative Biochemistry and Physiology 93c (2): 231-237. COLLODI, P., M. S. STEKOLL, D. R. STANLEY & D. RICE, 1984, Hepatic aryl hydrocarbon hydrox- ylase activities in Coho salmon (Oncorhynchus kisutch) exposed to petroleum hydrocarbons. Comparative Biochemistry and Physiology, 79c (2): 337-341. DAUTERMAN, W. C. & E. HODGSON, 1990. Me- tabolism of xenobiotiocs. Pp: 19-55, In: E. HODGSON & KUHR, R. J., eds., 1990, Safer insecticides. Development and use. Marcel Dek- ker, Inc., New York, Basel. DEN BESTEN, P. J., J. В. MAAS, J. M. L. ELEN- BAS, S. J. DIELEMAN & D. R. LIVINGSTONE, 1990. Mixed-Function oxygenase system and the metabolism of steroids and benzo[a]pyrene in the sea star Asterias rubens. Abstracts of the 12th Annual Conference of Physiological and Bio- chemical Approaches to the Toxicological As- sessment of Environmental Pollution. P 2-30. ELDEFRAWI, M. E. & A. Т. ELDERFRAWI, 1990. Nervous-system-based insecticides. Pp. 155- 209, In: E. HODGSON & R. J. KUHR, eds., 1990, Safer insecticides. Development and use. Marcel Dekker, Inc., New York, Basel. GOLDFISCHER, R., E. ESSNER & A. B. NO- VIKOFF, 1964, The localization of phosphatase activities in the level of ultrastructure. Journal of Histochemistry and Cytochemistry, 12: 72-93. GORDON, H. T. & M. E. ELDEFRAWI, 1960, Ana- IMPACT OF MOLLUSCICIDES ON DEROCERAS 271 log-synergism of several carbamate insecticides. Journal of Economic Entomology, 53: 1004— 1009. HAND, A. R. & C. OLIVIER, 1977. Cytochemical studies of GERL and its role in secretory granule formation in exocrine cells. Histochemical Jour- nal, 9: 375-392. HINTON, D. E., J. E. KLAUNIG & M. M. LIPSKY, 1978, PCB-induced alterations in teleost liver: A model for environmental disease in fish. Marine Fisheries Review, 40 (10): 47—50. KAGAN, J. S., 1988, Effect of monooxygenase en- zyme system on xenobiotics. Monatshefte für Ve- terenärmedizin, 43 (22): 791-793. KARNOVSKY, M. J., 1971, Use of ferrocyanide- reduced osmium tetroxide in electron micros- copy. Journal of Cell Biology, 51: Abstr. 284. KLAUNIG, J. E., M. M. LIPSKY, В. Е. TRUMP & D. E. HINTON, 1979, Biochemical and ultrastruc- tural changes in teleost liver following subacute exposure to PCB. Journal of Environmental Tox- icology, 2: 953-963. LEE, R. F., 1981, Mixed-function oxidases (MFO) in marine invertebrates. Marine Biological Letters, 2: 87-105. LIVINGSTONE, D. R., 1988, Responses of mi- crosomal NADPH-cytochrom-c reductase activity and cytochrom P450 in digestive glands of Mytilus edulis and Littorina littorea to environ- mental and experimental exposure to pollutants. Marine Ecology-Progress Series, 46: 37-43. LOJDA, Z., R. GOSSRAU & T. H. SCHUBLER, 1976, Enzymhistochemische Methoden. Springer Verlag, Berlin, Heidelberg, New York. METCALF, R. L. & T. R. FUKUTO, 1965, Effects of chemical structure on intoxication. Journal of Ag- riculture and Food Chemistry, 13:220-231. MOORE, М. N., 1976, Cytochemical demonstration of latency of lysosomal hydrolases in digestive cells of the common mussel Mytilus edulis, and changes induced by thermal stress. Cell and Tis- sue Research, 175: 279-287. MOORE, M. N., 1982, Lysosomes and environ- mental stress. Marine Pollution Bulletin, 13: 42— 43. MOORE, M. N. & D. W. HALTON, 1977, The cy- tochemical localization of lysosomal hydrolases in the digestive cells of littorinids and changes induced by larval termatode infection. Zeitschrift fur Parasitenkunde, 53: 115-122. MOORE, M.N., R. K. PIPE & S. V. FARRAR, 1982, Lysosomal and microsomal responses to envi- ronmental factors in Littorina littorea from Sullom Voe. Marine Pollution Bulletin, 13: 340-345. MOORE, M. N. & D. M. LOWE, 1985, Cytological and cytochemical measurements. Pp. 46-74, in: В. L. BAYNE et al., eds., The effects of stress and pollution on marine animals, Praeger Scientific, New York. NETTER, K. J., 1980, Inhibition of oxidative drug metabolism in microsomes. Pharmacology and Therapeutics, 10: 515-535. NEWELL, P. F., 1977, The structure and enzyme histochemistry of slug skin. Malacologia, 16(1): 183-195. NOTT, J. A., M. N. MOORE, L. J. МАММ & К. P. RYAN, 1985, The fine structure of lysosomal membranes and endoplasmic reticulum in the di- gestive cells of Mytilus edulis exposed to An- thracene and Phenanthrene. Marine Environ- mental Research, 17: 226-229. NOVIKOFF, A. B., M. MORI, N. QUINTANA & A. YAM, 1977, Studies of the secretory process in the mammalian exocrine pancreas. Journal of Cell Biology, 75: 148-165. OXFORD, С. $. 8 L. J. FISH, 1979. Ultrastructural localization of esterase and acid phosphatase in digestive gland cells of fed and starved Cepaea nemoralis (L.) (Mollusca, Helicidae). Proto- plasma, 101: 181-196. ROBINSON, J. M. & M. J. KARNOVSKY, 1983, Ultrastructural localization of several phos- phatases with cerium. Journal of Histochemistry and Cytochemistry, 31(10): 1197-1208. SIMKISS, K & A. Z. MASON, 1984, Cellular re- sponses of molluscan tissues to environmental metals. Marine Environmental Research, 14: 103-118. SMUCKLER, E. A. & M. ARCASOY, 1969, Struc- tural and functional changes of the endoplasmic reticulum of hepatic parenchymal cells. /nterna- tional Reviews of Experimental Pathology, 7: 305-418. SPARKS, A.K., 1972, Invertebrate pathology, non- communicable diseases. Academic Press, New York, London. SPURR, A. R., 1969, A low viscosity embedding medium for electron microscopy. Journal of Ul- trastructural Research, 26: 31-43. STEGEMAN, J. J., 1985, Benzo-[a]-pyrene oxida- tion and microsomal enzyme activity in the mus- sel Mytilus edulis and other bivalve mollusc spe- cies from the western atlantic. Marine Biology, 89: 21-30. STOHS, S. J., R. C. GRAFSTROM, M. D. BURKE & S. ORRENIUS, 1976, Xenobiotic metabolism and enzyme induction in isolated rat intestinal mi- crosomes. Drug Metabolism Disposition, 4: 517- Sie SUMNER, A. T., 1969, The distribution of some hydrolytic enzymes in the cells of the digestive gland of certain lamellibranchs and gastropods. Journal of Zoology, London, 158: 277-291. TEGELSSTROM, H. & H. WAHREN, 1972, The effect of an N-methyl carbamate on esterases from snail, mouse and man studied by starch-gel electrophoresis. Comparative Biochemistry and Physiology, 43b: 339-343. TRIEBSKORN, R., 1988, Molluskizid-induzierte Reaktionen im Verdauungstrakt von Deroceras reticulatum (Múller). Verhandlungen der Deut- schen Zoologischen Gesellschaft, 81: 332. TRIEBSKORN, R., 1989, Ultrastructural changes in the digestive tract of Deroceras reticulatum (Muller) induced by a carbamate molluscicide 272 TRIEBSKORN and by metaldehyde. Malacologia, 31(1): 141— 156. TRIEBSKORN, R. & C. KUNAST, 1990, Ultrastruc- tural changes in the digestive system of Dero- ceras reticulatum (Mollusca, Gastropoda) in- duced by lethal and sublethal concentrations of the carbamate molluscicide Cloethocarb. Mala- cologia, 32(1): 87-104. TRIEBSKORN, R., C. KUNAST, R. HUBER & G. BREM, 1990, The tracing of a '*C-labeled car- bamate molluscicide through the digestive tract of Deroceras reticulatum. Pesticide Science, 28: 321-330. TRUMP, В. F., I. К. BEREZESKY & A. В. OS- ORNIO-VARGAS, 1981, Cell death and the dis- ease process. The role of calcium. In: J. D. BO- WEN & R. A. LOCKSHIN, eds., Cell death in biology and pathology. Chapman and Hall, Lon- don, New York. WACHSTEIN, M. & E. MEISEL, 1985, Histochem- istry of hepatic phosphatases at a physiologic PH with special reference to a demonstration of bile canaliculi. American Journal of Clinical Pathol- ogy, 27: 13-23. WERNER, J., 1986, Einbettungs- und Färbe- methoden für das RWL-Medium. RWL Histotech- nologie, Bruchmünhl. WIDDOWS, J., B. L. BAYNE, P. DONKIN, D. R. LIVINGSTONE, D. M. LOWE, M. N. MOORE & P. N. SALKELD, 1981, Measurement of the re- sponses of mussels to environmental stress and pollution in Sullom voe: a baseline study. Pro- ceedings of the Royal Society of Edinburgh, 80B: 332-338. WILKINSON, C. F., ed., 1976, Insecticide biochem- istry and physiology. Heyden publishing, London, New York, Rheine. YOUNG, A. G. & R. M. WILKINS, 1989, The re- sponse of invertebrate acetylcholin-esterase to molluscicides. Proceedings of the British Crop Protection Council., 41: 121-128. Revised Ms. accepted 23 January 1991 MALACOLOGIA, 1991, 33(1-2): 273-280 MORPHOLOGY OF THE MATURE SPERMATOZOON OF BULINUS TROPICUS (KRAUSS, 1848) (GASTROPODA : PLANORBIDAE) Therese D. Brackenbury & С. С. Appleton’ Department of Zoology and Entomology, University of Natal, P. O. Box 375, Pietermaritzburg, 3200, South Africa ABSTRACT Using the TEM, the morphology of the spermatozoon of Bulinus tropicus was shown to be of the modified type, with features in common with other basommatophoran species. A comparison with another planorbid, Biomphalaria glabrata, suggests that differences do however occur at the subfamily level. For example, the former species differed from the latter in that its sperma- tozoon was shorter, the nucleus was bulbous, a glycogen ring was present at the base of the nucleus, and the maximum number of glycogen helices was four. If spermatozoon morphology can be shown to be unique for individual species within the family Planorbidae, it may provide a useful taxonomic tool in clarifying uncertainties in the systematics of certain species groups. Key words: spermatozoon, ultrastructure, Bulinus tropicus, Planorbidae. INTRODUCTION Spermatozoon morphology is proving to be a useful tool in determining systematic and phylogentic relationships among proso- branchs and opisthobranchs (Healy, 1988a,b; Hodgson & Bernard, 1988). Within the Pul- monata, many authors, for example, Kitajima & Paraense (1976), Maxwell (1975), Healy (1983) and Healy & Jamieson (1989), have examined sperm morphology, but differences appear to be slight. The aim of the present study is to extend this knowledge to the African Basommatophora by examining the mature spermatozoon of Bulinus tropicus (Krauss, 1848) (Planorbidae), a species widespread in southern Africa and of economic importance because it serves as the intermediate host for paramphistome cattle flukes. Brown (1980) noted that bulinid species have been defined principally on shell, anatomical and radular characters but that this is rarely completely satisfactory. The B. tropicus/natalensis com- plex (Brown et al., 1971) is one such case, and itis hoped that the present study in which the morphology of the B. tropicus spermatozoon is described in detail will be of value in helping to define these closely related species, both of which occur in South Africa. MATERIALS AND METHODS Adult B. tropicus were obtained from the Umsindusi River, Pietermaritzburg, South Af- 'To whom correspondence should be addressed. 273 пса (29°37’S, 30°26’30”Е). This site is 28 km northwest of locality number 52 of Brown et al. (1971a,b). The hermaphrodite ducts of five individuals were dissected out and teased apart in 0.7% sodium chloride solution on a glass slide, and the general morphology of living spermatozoa examined under phase contrast with an Olympus BH-2 light micro- scope. A second group of three snails was relaxed in a mixture of menthol and chloral- hydrate crystals (van Eeden, 1958), their shells carefully crushed between two glass slides, and the hermaphrodite ducts removed. Pieces of the latter were fixed in 3% glutaral- dehyde at 5°C in 0.05 M phosphate buffer (pH 7.2) for two hours. They were then rinsed twice in 0.05M phosphate buffer for 30 min- utes each, post-fixed in 1% osmium tetroxide for two hours, rinsed again, and dehydrated through an ethanol series (30%-100%). For SEM examination, the tissue was critical point dried by using a Hitachi HCP Critical Point Dryer, sputter coated with gold paladium by a Polaron E5100 Sputter Coater, and viewed with a Hitachi S-570 Scanning Electron Micro- scope. The same procedure was followed for TEM preparation, except that after dehydra- tion the tissue was gradually embedded in Spurr's epoxy resin. Thin sections were cut using a LKB Ultratome 3 and placed on cop- per viewing grids. Sections were then stained with uranyl acetate and lead citrate, and ex- amined with a JEOL 100 CX Transmission Electron Microscope. 274 BRACKENBURY & APPLETON RESULTS Different proposals for sub-dividing the gastropod spermatozoon have been made by Kitajima & Paraense (1976), Healy (1983, 1988a), and Sumikawa & Funakoshi (1984). In the present study, we have combined these schemes so as to best describe the sperma- tozoon of B. tropicus: the acrosome (A), the nucleus (N)—the midpiece: (i) the neck region (NK), (ii) the glycogen helix regions (4H-1H), (iii) the mictochondrial derivative region, (M)—the glycogen piece (G), and the end piece (E) (Fig. 1). Light Microscope Observations Under phase contrast, the spermatozoon of B. tropicus was visible as a thin filament 560 um long with a slightly enlarged anterior head. Under oil immersion, this head was seen to be conical in shape. For a short dis- tance along the length of the midpiece, just posterior to the head, helical coiling of the tail was pronounced (Fig. 2). In some cases, Cy- toplasmic swellings were seen along the length of the tail. Scanning Electron Microscope (SEM) Observations The head (acrosome and nucleus) mea- sured 4.8 jm, was corkscrew-shaped and an- gled at the point where the acrosome (0.62 x 0.16 рт) capped the nucleus anteriorly (Fig. 3). At the base of the head, there was a pro- nounced constriction forming the neck region of the midpiece. The tail was helically coiled, and the end piece narrowed to a short, ta- pered tip measuring 2.3 x 0.42 um (Fig. 4). Transmission Electron Microscope (TEM) Observations Acrosome: The acrosome sheathed the nu- cleus anteriorly, measured 2.2 x 0.1 рт, and fitted in a groove at the anterior end of the nucleus. It was constructed of semi-electron dense material (Fig. 5). Nucleus: The nucleus of the head was ex- tremely electron dense and measured 3.5 um in length. It was bulb-shaped, with its anterior section angled to one side, then re-aligned, terminating in atapered, rounded point. Three distinct lateral projections, the helical keels, were present. The nucleus had a basal in- vagination, the implantation fossa, and was 2H 1H G 1 E FIG. 1. Diagrammatical representation of a whole mature spermatozoon of B. tropicus (total length 560 рт) showing the morphological divisions used in this study. A = acrosome; E = end piece; G = glycogen piece; 4H = quadruple helix region of midpiece; 3H = triple helix region of midpiece; 2H = double helix region of midpiece; 1H = single helix region of midpiece; M = mitochondrial deriv- ative region of midpiece; N = nucleus; NK = neck region of midpiece. SPERM OF BULINUS TROPICUS 275 FIG. 2. Anterior region of a spermatozoon of B. tropicus under oil emersion showing the head and pro- nounced helical coiling of the tail. h = head; t = tail. covered by a loosely fitting plasma membrane (Fig: 5). The Midpiece: (i) The neck region: The mid- piece was inserted into the implantation fossa via the basal body of the axoneme (Fig. 6). This basal body was the insertion point of nine coarse fibres, which were enclosed by a ring of glycogen. These fibres were thick, pe- riodically banded, and in longitudinal section showed a crescent-shaped arrangement measuring 0.09 um at the widest part (Fig. 6). The cross-striations occurred at 0.01 um in- tervals along the fibres’ length. In cross sec- tion, the nine individual fibres each had a tri- angular form in a radially symmetrical pattern (Fig. 7). The coarse fibres became more slen- der posteriorly and were tightly apposed to the outer surface of the axial complex in such a way that each fibre practically merged with the axoneme’s peripheral doublets (Fig. 6). The plasma membrane was loosely applied to the mitochondrial derivative complex, which consisted of two derivatives: the paracrystal- line and the mitochondrial matrix layers. In longitudinal section, 11 outer paracrystalline layers were observed (total thickness 0.12 um), ensheathing two mitochondrial matrix layers, which together measured 0.08 um. The glycogen helices were incorporated be- tween these two derivatives. Underlying the mitochondrial matrix layers were three inner paracrystalline layers surrounding the axial complex (Fig. 6). (ii) Glycogen helix region: Helical ridges containing glycogen were situated immedi- ately distal to the neck region. In transverse section, the maximum number of helices seen was four, each measuring approximately 0.19 x 0.09 um (Fig. 8). These glycogen helices were enclosed by the outer paracrystalline and mitochondrial matrix layers. The axial complex had a 9 + 2 microtubule arrange- ment. Helices were lost consecutively along the length of the midpiece and transverse sec- tions showed that, posterior to the termination of the first helix, the remaining three became equidistant from each other (Fig. 9). Termina- tion of these helices continued down the mid- piece until all had disappeared (Figs. 10-14). The single helix region also saw the termina- tion of one of the mitochondrial matrix layers. (iii) Mitochondrial derivative region: This was characterized by the absence of glyco- gen helices, the presence of a loosely fitting plasma membrane, outer and inner paracrys- FIGS. 3, 4. SEM. Mature spermatozoon of B. tropi- cus. 3: Head region. h = head. 4: Tail tip. e = end piece. talline layers, and a single mitochondrial ma- trix layer (Fig. 14). Transverse sections ob- tained here showed fine detail of the axoneme, i.e. the peripheral doublets that consisted of an A-subtubule connected to a BRACKENBURY & APPLETON FIGS. 5-7. TEM. Mature spermatozoa of B. tropi- cus. 5: Longitudinal section through the head. ac = acrosome; k = helical keel; n = nucleus; p = plasma membrane. 6: Longitudinal section through the neck region. c = paracrystalline layer; cf = coarse fibres; d = mitochondrial matrix layer; gr = glycogen ring. 7: Transverse section through the neck region. a = axoneme; c = paracrystalline layer; cf = coarse fibres; gr = glycogen ring; п = nucleus. SPERM OF BULINUS TROPICUS CT. FIGS. 8-11. Transverse sections through the glycogen helix regions of the midpiece of a mature sperma- tozoon of В. tropicus. 8: Quadruple helix region. а = axoneme; с = paracrystalline layer; d = mitochondrial layer; g = glycogen helix; p = plasma membrane. 9: Triple helix region. a = axoneme; g = glycogen helix; р = plasma membrane. 10: Double helix region. а = axoneme; с = paracrystalline layer; d = mitochondrial matrix layer; g = glycogen helix; p = plasma membrane. 11: Single helix region. g = glycogen helix; 1h = single helix region. B-subtubule (the latter incomplete) and a pair of separate centriole microtubules. The Glycogen Piece: The change from the mitochondrial derivative region of the mid- piece and the glycogen piece was abrupt, de- marcated by the annulus (Fig. 15). The diam- eter of the tail decreased from 0.28 um to 0.22 um, and the mitochondrial derivative was replaced by a glycogen sheath around the axoneme. The plasma membrane was closely applied to this glycogen layer, and the 9 + 2 microtubules were clearly distinguish- able in the axoneme (Fig. 16). The End Piece: This last division of the sper- matozoon had a maximum diameter of 0.14 um (Fig. 17) and included the termination of the glycogen sheath, so that only the plasma membrane and axoneme remained. The A- subtubule of each doublet was seen to have two dynein arms that pointed toward the B- 278 BRACKENBURY & APPLETON a | FIGS. 12,13. Longitudinal sections through the gly- cogen helix region of the midpiece of a mature spermatozoon of B. tropicus. 12: Triple and single helix regions. g -= glycogen helix; 1h = single helix region; 3h = triple helix region. 13: Double helix region. g = glycogen helix; 2h = double helix region. subtubule of the next doublet in a clockwise direction. A series of radii were observed to extend between the doublets and the central pair of microtubules. DISCUSSION The mature spermatozoon of the planorbid gastropod described in this study, B. tropicus, FIG. 14. Transverse section through the mitochon- drial derivative region of the midpiece of a mature spermatozoon of В. tropicus. a = axoneme; с = paracrystalline layer; d = mitochondrial matrix layer; p = plasma membrane. FIG. 15. Longitudinal section through the transition zone between the midpiece and the glycogen piece of a mature spermatozoon of B. tropicus. an = an- nulus; G = glycogen piece; M = midpiece. was of the modified type and conformed to the general pattern recorded for other pulmo- nate species by Anderson & Personne (1967, 1970), Maxwell (1975), Kitajima & Paraense (1976), Dan & Takaich (1979), Healy (1983), Sumikawa & Funakoshi (1984), and Healy (1988a). This may be summarized as com- prising: 1. an acrosome, 2. anucleus that is helically keeled, with a SPERM OF BULINUS TROPICUS 219 FIG. 16. Transverse section through the glycogen piece of a mature spermatozoon of B. tropicus. cm = centriole microtubules; gs = glycogen sheath; p = plasma membrane; pm = peripheral microtu- bules. basal invagination containing the basal body, 3. an axoneme that is associated with pe- riodically banded coarse fibres, 4. a midpiece, composed of prominent he- lically wound paracrystalline and mito- chondrial matrix layers, 5. mitochondrial derivatives incorporating a variable number of glycogen—filled helices, 6. a glycogen piece and an annulus, and 7. an end piece. Variations within this theme have been ге- ported for several pulmonate families by the authors mentioned above, and we have sim- ilarly noted several such differences in the B. tropicus spermatozoon. It is however most useful here to compare the latter with the only other planorbid snail whose spermatozoan ul- trastructure has, as far as we are aware, been described in detail, viz. Biomphalaria glabrata by Kitajima & Paraense (1976). A number of discrepancies were noted be- tween the spermatozoa of these two planor- bids. For example, the overall length of the spermatozoon of В. tropicus (560 jm) was shorter than that of В. glabrata (650 um), i.e. by about 14%. The acrosome of the latter species was narrower and longer (1 wm in length) than that of В. tropicus (0.62 jm). The head of the spermatozoon of B. tropicus (4.8 x 0.63 um) was slightly smaller than that of В. glabrata (5 x 1 рт). In addition, the nu- cleus of the latter species was more elongate FIG. 17. Transverse section through the end piece of a mature spermatozoon of B. tropicus. cm = centriole microtubules; p = plasma membrane; pm = peripheral microtubules. and torpedo-shaped than the nucleus of B. tropicus, which is described here as bulbous. For both snails, the nuclear material was extremely electron dense and helically coiled. However, B. glabrata possessed four, slightly projecting, helical keels, one more than B. tropicus. Another notable difference was the absence of the glycogen ring at the base of the nucleus in B. glabrata. Although the neck regions and midpieces of both species were similar, Kitajima & Paraense (1976) noted that for B. glabrata the maximum number of glycogen helices was two, whereas we have found four in B. tropicus. The only other dif- ference we recorded was that in B. tropicus the abrupt transition between the midpiece and the glycogen piece occurred with a greater change in the diameter of the sper- matozoon than it did in B. glabrata. Kitajima & Paraense (1976) do not describe the shape of the tail tip, and therefore we cannot make a comparison between the two species. This study has shown that there are mor- phological differences between mature sper- matozoa within the family Planorbidae at the subfamily level, i.e. between Bulinus tropicus (Bulininae) and Biomphalaria glabrata (Plan- orbinae). Whether this will prove a useful character at the generic or species level within these subfamilies can only be fully evaluated by further study. 280 BRACKENBURY & APPLETON ACKNOWLEDGEMENTS We wish to acknowledge the help of the staff of the Electron Microscope Unit of the University of Natal, Pietermaritzburg, and the Foundation for Research Development (С.5.1.В.) for financial support. LITERATURE CITED ANDERSON, W. A. & P. PERSONNE, 1967, The fine structure of the neck region of spermatozoon of Helix aspersa. Journal of Microscopy, 6: 1033-1042. ANDERSON, W. A. & P. PERSONNE, 1970, The localization of glycogen in the spermatozoa of various invertebrate and vertebrate species. Journal of Cell Biology, 44: 29-51. BROWN, D. S., 1980, Freshwater snails of Africa and their medical importance. Taylor & Francis Ltd., London. BROWN, D. S., G. OBERHOLZER & J. A. VAN EEDEN, 1971, The Bulinus natalensis/tropicus complex (Basommatophora: Planorbidae) in southeastern Africa. 1. Shell, mantle, copulatory organ and chromosome number. Malacologia, 11: 131-170. DAN, D. C. & S. TAKAICHI, 1979, Spermatogene- sis in the pulmonate snail Euhadra hickonis. 3: Flagellum formation. Development, Growth and Differentiation, 21: 71-86. HEALY, J. M., 1983, An ultrastructural study of ba- sommatophoran spermatozoa (Mollusca : Gas- tropoda). Zoologica Scripta, 12: 57-66. HEALY, J. M., 1988a, Sperm morphology and its systemtic importance in the Gastropoda. Malaco- logical Review, 4: 251-266. HEALY, J. M., 1988b, Sperm morphology in Serpu- lorbis and Dendropoma and its relevance to the systematic position of the Vermetidae (Gas- tropoda). Journal of Molluscan Studies, 54: 295— 308. HEALY, J. M. & B. G. M. JAMIESON, 1989, An ultrastructural study of spermatozoa of Helix as- persa and Helix pomatia (Gastropoda, Pulmo- nata). Journal of Molluscan Studies, 55: 389— 404. HODGSON, А. N. & В. T. Е. BERNARD, 1988, A comparison of the structure of the spermatozoa and spermatogensis of 17 species of patellid lim- pets (Mollusca: Gastropoda: Archaeogas- tropoda). Journal of Morphology, 195: 205-223. KITAJIMA, E. W. & W. L. PARAENSE, 1976, The ultrastructure of the mature sperm of the fresh- water snail Biomphalaria glabrata (Mollusca : Gastropoda). Transactions of the American Mi- croscopic Society, 95: 1-10. MAXWELL, W. L., 1975, Scanning electron micro- scope studies of pulmonate spermatozoa. The Veliger, 18: 31-33. SUMIKAWA, S. & C. FUNAKOSHI, 1984, The fine structure of mature spermatozoa in two species of the Siphonariidae (Pulmonata: Basommato- phora). Kairuigaku Zasshi (Venus), 42: 143-155. VAN EEDEN, J. A., 1958, Two useful techniques in freshwater malacology. Proceedings of the Mal- acological Society of London, 33 (2): 64—66. Revised Ms. accepted 28 January 1991 MALACOLOGIA, 1991, 33(1-2): 281-288 TROPICAL “PHYSIOGNOMY” OF A LAND SNAIL FAUNULE FROM THE EOCENE OF SOUTHERN CALIFORNIA Barry Roth Museum of Paleontology, University of California, Berkeley, California 94720, U.S.A. ABSTRACT Land snail fossils from the Santiago Formation, Eocene of San Diego County, southern California, have shell height (h) and diameter (d) relationships differing from any found in the Recent land mollusk fauna of California or most other analyzed northern hemisphere faunas. Similar h,d values are found in the Recent Helicostylinae of the Philippine Islands and Papuin- inae of northern Australasia. Both Helicostylinae and Papuininae represent endemic radiations on island groups in wet tropical forest. The Santiago Formation fossils may represent such a radiation. Around 43 million years ago, southwestern California may have been similar in climate and isolation to the present-day Philippines. INTRODUCTION Because of their evolutionary conserva- tism, fossil land mollusks are useful in inter- preting ancient environments (Roth, 1984, 1988). Late Mesozoic or early Tertiary speci- mens often show shell characters that allow them to be assigned to modern families (Solem, 1978) or genera (e.g. Roth, 1984, 1986, 1988). In other cases, critical charac- ters for generic assignment are lacking, but conclusions can be drawn from a combination of taxonomic and non-clade-specific charac- ters (e.g. Roth & Pearce, 1988, Helicinidae greater than 10 mm in diameter). However, because much land snail classification is based on soft anatomy, and convergence in shell morphology is common, it is not always possible to assign land snail fossils to specific taxonomic groups or to base paleoenviron- mental analysis on convincing modern homo- logues. In cases where taxonomic information is unavailable (or where the fossils belong to groups without living representatives), it would be useful to have a style of analysis that proceeds on physical appearance, with- out reliance on taxonomy. For example, pa- leobotanists have long known that there is a correlation between vegetation type and the proportion of entire-margined leaves in an as- semblage (Baily & Sinnott, 1915, 1916) and that this correlation is independent of the flo- ristic (that is, taxonomic) composition of the vegetation. By analogy, the percentage of en- tire-margined leaves in a fossil assemblage becomes an index of the vegetation type and, 281 by extension, of the paleoclimate. Wolfe (1969) refined this approach by adding an- other non-taxonomic but environmentally sensitive criterion, leaf size, and applied the method extensively to Tertiary floras (e.g. Wolfe, 1971, 1978, 1981). This paper is an attempt to use one set of non-taxonomic physical criteria for the inter- pretation of an early Tertiary land snail faunule. Shell height (h) and diameter (a) have long been used in the description and identification of gastropod species. Cain (1977, 1978a, 1978b, 1981) recognized an important regularity in the distribution of h and din modern land snail faunas: for fully terres- trial, freely crawling snails that can retract completely into their shells (i.e. excluding slugs and semi-slugs), h versus d is distrib- uted bimodally, with the vast majority of shells being either high-spired (h markedly greater than d) or equidimensional to low-spired (h equal to, to markedly lower than, d). This bi- modal distribution holds true through all size classes and applies to both stylommatopho- ran (Cain, 1977, 1981) and operculate (Cain, 1978b) land snails, with a small number of interesting exceptions. Normally, regional snail faunas include points in both the upper and the lower fields (referred to here as the U group and L group respectively, following Cain [1980]). Bradybaenidae of the subfamily Helicostyli- nae in the Philippine Islands and Camaenidae of the subfamily Papuininae in northern Aus- tralasia are anomalous, falling between the U group and the L group and overlapping both (Cain 1978a). Both Helicostylinae and Pap- 282 ROTH uininae are the only endemic subfamilies of their respective island groups, and both ap- parently represent local radiations in settings of wet tropical forest. The total snail faunas of the Philippines and of New Guinea and the related island groups include points in U and L groups as usual, in addition to the anoma- lous scatter. A fossil faunule showing the usual bimodal distribution of h,d would not be particularly in- formative (except as an additional instance of the rule), because the pattern is so nearly general; but faunules with an anomalous dis- tribution are informative to the extent that we can understand the conditions associated with the anomaly. To this extent as well, the distribution of h,d can be used as a physiog- nomic index, similar to the percentage of en- tire-margined leaves in a fossil florule. In 1987, members of the Department of Pa- leontology of the San Diego Natural History Museum collected fossil vertebrates and in- vertebrates (including land snails) from the Santiago Formation (Eocene) in northwest- ern San Diego County, California. The poor preservation (nearly all are internal molds lacking shell or an impression of the external detail) precludes their identification to family or genus. The range of h,d falls mostly out- side the h,d field of modern Californian land mollusks but within the fields occupied by liv- ing Helicostylinae and Papuininae. MATERIAL AND METHODS Geologic Setting The Santiago Formation consists of ma- rine, lagoonal-estuarine, and nearshore non- marine-fluviatile sedimentary rocks cropping out in Orange County and northwestern San Diego County, California (Golz & Lillegraven, 1979). Over this large area of outcrop, the formation varies in thickness from about 823 m in the type area to over 1,800 m on Camp Pendleton, thinning to about 200 m near Oceanside. Several workers (Morton, 1974; Schoellhamer et al., 1981) have proposed in- formal divisions of the formation, recognizing a lower crossbedded sandstone unit, a middle pebbly sandstone unit, and an upper siltstone unit. The formation apparently was deposited over a considerable span of middle Eocene time. Marine mollusks of the “Domengine” Stage (early middle Eocene) have been coi- lected from the type area of the formation in the Santa Ana Mountains, Orange County (Woodring & Popenoe, 1945). Vertebrate fos- sils can be divided into at least two biostrati- graphic assemblages, one containing early Uintan land mammals and another containing late Uintan mammals (Golz & Lillegraven, 1977; Lillegraven, 1979, 1980; S. L. Walsh & T. A. Demere, work in progress). No radio- metric dates have been obtained in associa- tion with the Santiago Formation faunas, but mammalian faunal correlations suggest an approximate age range for the land mammals of 48 to 43 Ma (million years), early middle to late middle Eocene (Lillegraven, 1980; Berg- gren et al., 1985). Material Examined The fossils are predominantly internal molds without shell material. The matrix is a light gray, silty, arkosic sandstone. The few partial external molds present are too coarse to show any sculptural detail. Many of the specimens are distorted by compression, ob- lique to the axis of coiling. Only one specimen retains any shell material. It has crude, colla- bral growth lines (Fig. 1); the crystalline struc- ture is three-layered, with outer and inner ra- dial crossed-lamellar layers (each about 25% of the shell thickness) enclosing a concentric crossed-lamellar layer. This type of structure is not taxonomically diagnostic. The speci- mens more likely represent pulmonates than prosobranchs; internal whorl partitions are present, not resorbed as in Helicinidae (Ar- chaeogastropoda), and there is no evidence of complete peritremes, such as would be found in Cyclophoridae (Mesogastropoda). Because of the poor state of preservation, it is not possible to state exactly how many taxa are present in the sample, but | estimate that the internal molds represent a minimum of five (and perhaps a maximum of ten) land snail species. Figures 1—5 illustrate represen- tative specimens. In addition, opercula up to 7.1 mm in length, assignable to family Helicinidae, are present. They have lateral nuclei with concentric growth lines as in the genus Helicina La- marck, 1799, and are comparable to opercula of Recent species with shells approximately 12 mm in diameter. The opercula cannot be associated with any shell fossils in the present collection. Also present are internal molds that resem- ble high-spired pleurocerid or thiarid fresh- water gastropods. PHYSIOGNOMY OF A LAND SNAIL FAUNULE 283 FIGS. 1-5. Representative land snail fossils from SDSNH Locality 3276; Oceanside, San Diego County, California. Santiago Formation, middle Eocene. x 1.1 All fossils were collected from a 0.6-1.2 m- thick, pinkish green, locally cemented, locally laminated, moderately well sorted, fine- to very fine-grained, silty, arkosic sandstone (SDSNH Locality 3276; Oceanside, San Diego County, California; coll. by B. O. Riney and R. A. Cerutti, 3 March 1987). This fossil- iferous sandstone occurs near the top of a >15 m-thick sequence of interbedded white, coarse-grained, arkosic sandstones and green, laminated, sandy claystones. The in- vertebrate fossils were collected in associa- tion with a diverse assemblage of well pre- served late Uintan terrestrial mammals. Specimens were measured with a hand- held vernier caliper, to the nearest 0.1 mm. Height (h) was measured parallel to the axis of coiling and diameter (d) across the greatest dimension perpendicular to the axis. Addition- ally, for distorted or incomplete specimens, the approximate original h and d were esti- mated by eye. The h,d values of Santiago Formation specimens were plotted for comparison on the same base as the h,d fields of Recent native Californian land mollusks and other selected northern hemisphere faunas, non- helicostyline land stylommatophorans of the Philippine Islands, Helicostylinae, and Pa- puininae, using data presented diagrammati- cally by Cain (1977, 1978a). The thermal classification of climates follows Wolfe (1979, fig. 3). 50 - mm ae 40> er 30F+ BA me 20+ 10> | NE 0 y = ee | = =e ES = 0 10 20 30 40 50mm 6 d FIG. 6. Distribution of shell height (h) and diameter (d) for land snail fossils from Santiago Formation, San Diego County, California. The bisector (h = d) is given in this and following figures as a reference. RESULTS Figure 6 shows the h,d scatter for the San- tiago Formation specimens examined. In most cases, estimated reconstruction of the original shape of distorted specimens (Fig. 7) moves the points closer to the bisector and 284 ROTH h 507 E mm whe / Y Y 40> WA AA р A 30> Le Y A 20+ es 11 10+ 0 — 0 10 20 30 40 50mm 7 d FIG. 7. Distribution of reconstructed height (h) and diameter (d) of Santiago Formation land snails. Tails of arrows represent observed values; points, the estimated values for reconstructed shells. farther away from the usual L group field of modern land snail faunas. The following com- parisons are based on non-adjusted values and therefore represent the minimum proba- ble departures from “normal” L group values. The upper two-thirds of the Santiago For- mation h,d field, representing large, relatively equidimensional shells, falls outside the Re- cent Californian field (Fig. 8). Approximately the same relation holds true between the Santiago Formation field and the L group fields of such diverse faunas as those of Puerto Rico, North and South Carolina, Texas (with a few species of Humboldtiana falling just below the bisector in the h = 22 to 28, d = 25 to 33 range), Arizona, western Europe (with a few exceptions as discussed below), and the Palearctic realm. Similarly, the upper part of the Santiago Formation field falls out- side the L group field of Recent non-helico- styline Philippine stylommatophorans (Fig. 9). However, the Santiago Formation field falls entirely within the helicostyline field (Fig. 10) and emerges only at one end of the papuinine field (Fig. 11). Values of h,d are not uniformly distributed within the helicostyline field. The data of Cain (1978a) show a bimodal scatter within the He- licostylinae itself, with only two genera really well represented both above and below the bisector. The Santiago Formation points are almost entirely within the field of a single ge- nus, Chloraea Albers, 1850 (based on 38 pairs of values plotted by Cain, 1978a, fig. 2). A few other Recent snail groups show a similar type of anomalous h,d scatter. Values for the genus Helix Linnaeus, 1758, run along (and, on average, slightly above) the bisector, in the 15-50 mm range (Cain, 1981, fig. 9). The remainder of the subfamily Helicinae falls within the usual L group field. The larger shells from the Santiago Formation, with d = 30 mm, fall in the gap between the scatters for Helix and the rest of the Helicinae. Values for the Mexican and southwestern United States genus Humboldtiana von Ihering, 1892 (family Humboldtianidae), fall slightly below the bisector, with d ~ 20-50 mm (Evanoff & Roth, in preparation). (Two aberrantly flat- shelled species of Humboldtiana, H. plana Metcalf & Riskind, 1976 [holotype hd = 19.8,43.6 mm], and H. eulaliae Metcalf, 1984 [holotype h,d = 17.2,28.1 mm] are excluded.) Only the most nearly equidimensional San- tiago Formation specimens of d = 30 mm overlap the Humboldtiana field slightly; most fall below it (Fig. 12). Certain large Cyclophoridae (Mesogas- tropoda), especially some from the Indian subcontinent (including Ceylon and Burma) and southeast Asia (Cain 1978b), occupy the same field as Santiago Formation snails (Fig. 13). DISCUSSION The fact that numerous land mollusk fau- nas show approximately the same dual scat- ter of h,d values irrespective of their taxo- nomic composition strongly suggests that the pattern is dictated by functional relations, as yet not well understood (see Goodfriend, 1986). The Recent native land mollusk fauna of California consists of about 300 species (Roth, in preparation). Perhaps 40% of these are short-ranging endemics, mainly in the Helminthoglyptidae, that replace each other over short geographic distances. Diversity in sympatry is low (maximum observed sympat- ric diversity, nine species; six is more usual, and two to three frequent). Above h ~ 7 mm, the U group is lacking. In a sense, it is mis- leading to speak of a “Californian” fauna, because the state is a composite of several zoogeographic regions. However, even re- stricting consideration to southern California PHYSIOGNOMY OF A LAND SNAIL FAUNULE 285 о 10 20 30 40 50 60 70 80mm O d 11 d FIG. 8. The h,d field of Santiago Formation land snails (shaded) compared to the L group field of Recent native land snails of California (data from Cain, 1977). Crosses indicate three outlying values for Recent species of Californian Helminthoglyptidae. FIG. 9. The h,d field of Santiago Formation land snails (shaded) compared to the field of Recent non- helicostyline stylommatophorans of the Philippine Islands (data from Cain, 1978a). FIG. 10. The h,d field of Santiago Formation land snails (shaded) compared to the field of the Recent Helicostylinae (data from Cain, 1978a). FIG. 11. The h,d field of Santiago Formation land snails (shaded) compared to the field of the Recent Papuininae (data from Cain, 1978a). south of the Transverse Ranges does not greatly alter the limits of the h,d field. The present-day climate of San Diego County is warm subtropical, equable, and summer-dry. At San Diego the mean annual temperature is 17.3°C, and the mean annual range is 9.3°C; mean annual precipitation is 26.4 cm, less than 2.4 cm falling between May and October (Elford, 1970). The land mollusk faunas of New Guinea and the Philippines are highly diverse, al- though available literature does not allow a 286 ROTH 30 F 20: 0 - i 4 =! 10 20 30 40 50mm 12 ° р eo oF FIG. 12. The h,d field of Santiago Formation land snails (shaded) compared to the field of Recent species of Humboldtiana (data from various sources). FIG. 13. The h,d field of Santiago Formation land snails (shaded) compared to the field of Recent Cyclo- phoracea of the Indian subcontinent, including Ceylon and Burma (data from Cain, 1978b). precise estimate of actual numbers. Cain's (1978a, figs. 1, 2) h,d diagrams for the Phil- ippines, which exclude slugs and semi-slugs, are based on approximately 415 taxa. In gen- eral, Recent land snail diversity is high on is- lands and island groups, particularly those that are sufficiently high and topographically diverse to have some complexity of moisture regimes (Solem, 1984). High insular diversity generally is the result of local radiation by one or a few taxonomic groups. In the Philippines and New Guinea, part of this diversification has involved the invasion by the Helicostyli- nae and Papuininae of new h,d space, pre- sumably with the development of new func- tional/ecological relationships. Both of these radiations occurred in settings of wet tropical forest. The climate of the Phil- ippines is tropical, with mean annual temper- atures generally between 25.5°C and 27.5°C, mean annual range of temperature 2.6°C (av- erage of 42 stations), and normal annual rain- fall of 253 cm (Flores & Balagat, 1969). The distribution of precipitation through the year varies from place to place. In many areas, particularly in the southern and eastern parts of the archipelago, there is no dry season, and at least 6 cm of rainfall in the driest month. Even in areas with less than 6 cm precipitation in the driest month, the ground remains suffi- ciently wet throughout the year to support rain- forest. The climate is influenced by monsoons and tropical cyclones; thunderstorms, occur- ring in most months of the year, add signifi- cantly to the total precipitation. The climate of eastern Indonesia, in the range of Papuininae, is similar (Sukanto, 1969), tropical and wet at low elevations, grading to warm subtropical and wet at higher elevations. Large cyclophorids of India and southeast Asia support approximately the same infer- ences as the Helicostylinae and Papuininae: radiation in wet tropical and subtropical for- ests, with new adaptive relations accompa- nied by expansion of the h,d field. The fact that operculate and stylommatophoran land snails fall into much the same h,d scatters worldwide is a further indication that h and d respond to pervasive ecological conditions in the terrestrial environment (Cain, 1978b). Ap- parently, in a given area, available lineages evolve to fill out the prospective h,d range. Several other lines of evidence support the model of geographically isolated and wet tropical to subtropical conditions in southern California during the late middle Eocene. Ver- tebrate faunas of the Santiago Formation show greater endemism and fewer similarities with the faunas of the western interior of North America than earlier middle Eocene faunas (e.g. those from the Friars and Mission Valley formations) (Lillegraven, 1979, 1980). This is true even of large, cursorial mammals such as carnivores and artiodactyls. Lille- graven (1979) interpreted this difference as representing the closing of a dispersal corri- dor between the coast and the interior after early Uintan time. In his paleogeographic re- PHYSIOGNOMY OF A LAND SNAIL FAUNULE 287 construction, at the time of deposition of the Friars and Mission Valley formations, a low- land corridor allowed dispersal from the coastal plain of California, via what is now the Mojave Desert and Great Basin, to and from the central Rocky Mountains. By later Uintan time, faunal exchange between coast and in- terior was limited. Conceivably, topographic or climatic barriers affecting the vertebrate fauna might also have affected snails. Nilsen & McKee’s (1979) early and middle Eocene paleogeographic map of the western United States shows southern California iso- lated from the western interior by an ancestral cordillera; the continental borderland was complex, with numerous embayments, ba- sins, and islands. (The present Channel Is- lands of California, while small in area [904 km?] and close to the mainland [21-98 km off- shore], as a group have 67% species ende- mism.) In the Western Hemisphere at present, He- licinidae with shell diameter greater than about 10 mm occur only in tropical to subtrop- ical regions (Roth & Pearce, 1988). The pres- ence in the Santiago Formation of helicinid opercula indicative of shells of this size is con- sistent with the tropical to subtropical condi- tions inferred from the h,d relations. Identifiable fossil land snails from other Eocene formations of southern California be- long to genera (or quasi-taxonomic groups such as “helicinids over 10 mm in diameter’) that are tropical to subtropical at present (Roth, 1988; Roth & Pearce, 1988). The analogy with Helicostylinae and Pa- puininae is weakened by the absence of shells with h > d. If the Santiago Formation fossils represented a random sample of a ra- diation with the h,d scatter of the Helicostyli- nae, one would expect specimens with h > d to be present. In the absence of any studies on the taphonomy of large land snail shells in a forest environment, it is difficult to comment meaningfully on this discrepancy. However, a single deposit presumably would not receive shells representing an entire fauna, but rather a local subset, in which not all h,d combina- tions might be present. For example, a faunule from the Leron Formation, fluvial, la- goonal, and nearshore marine deposits of Pleistocene age in Papua New Guinea, con- tains probable papuinine camaenid snails of h < d only (L. D. Abbott, personal communica- tion). In conclusion, this study provides at least preliminary support for the use of shell height and diameter relationships in interpreting fos- sil land snail faunules, especially in combina- tion with other forms of paleoclimatic evi- dence. Where other evidence is lacking, the h,d scatter of a faunule may give at least a first approximation of ancient climate. Possi- ble improvements might include use of some form of dispersion analysis rather than the minimum convex polygon method used here (which may give too much emphasis to the outliers of a scatter). The utility of this method of analysis will improve as the relationship be- tween climate and the physiognomies of mod- ern land snail faunas is understood in more detail. ACKNOWLEDGMENTS | am grateful to Thomas A. Demere for call- ing this material to my attention, supply strati- graphic information, and commenting on the manuscript. Arthur J. Cain provided a helpful critique. Emmett Evanoff discussed with me the size and shape relations of Recent and fossil Humboldtianidae. Lon D. Abbott al- lowed citation of fossils found during his geo- logical work in Papua New Guinea. LITERATURE CITED BAILEY, |. W. & E. W. SINNOTT, 1915, A botanical index of Cretaceous and Tertiary climates. Sci- ence, 41: 831-834. BAILEY, 1. W. 8 E. W. SINNOTT, 1916, The climatic distribution of certain types of angiosperm leaves. American Journal of Botany, 3: 24-39. BERGGREN, W. A., D. V. KENT, J. J. FLYNN & J. A. VAN COUVERING, 1985, Cenozoic geochro- nology. Geological Society of America Bulletin, 96: 1407-1418. CAIN, A. J., 1977, Variation in the spire index of some coiled gastropod shells, and its evolution- ary significance. Philosophical Transactions of the Royal Society of London, ser. B, Biological Sciences, 277: 377-428. CAIN, A. J., 1978a, Variation of terrestrial gastro- pods in the Philippines in relation to shell shape and size. Journal of Conchology, 29: 239-245. CAIN, A. J., 1978b, The deployment of operculate land snails in relation to shape and size of shell. Malacologia, 17: 207-221. CAIN, A. J., 1980, Whorl number, shape, and size of shell in some pulmonate faunas. Journal of Conchology, 30: 209-221. CAIN, A. J., 1981, Variation in shell shape and size of helicid snails in relation to other pulmonates in 288 ROTH faunas of the Palaearctic region. Malacologia, 21: 149-176. ELFORD, C. R., 1970, Climate of California. Cli- matography of the United States, no. 60-4. U. S. Department of Commerce. EVANOFF, E. & B. ROTH, in preparation, Mollus- can paleontology of the White River Formation (Eocene and Oligocene) in the Douglas area, south-central Wyoming. FLORES, J. F. & V. F. BALAGAT, 1969, Climate of the Philippines. Pp. 159-213 in Climates of northern and eastern Asia (World Survey of Cli- matology, vol. 8). H. ARAKAWA, ed., Elsevier, Amsterdam. GOLZ, D. J. & J. A. LILLEGRAVEN, 1977, Summary of known occurrences of terrestrial vertebrates from Eocene strata of southern California. Con- tributions to Geology, University of Wyoming, 15: 43-65. GOODFRIEND, G., 1986, Variation in land-snail shell form and size and its causes: a review. Sys- tematic Zoology, 35: 204-223. LILLEGRAVEN, J. A., 1979, A biogeographical problem involving comparisons of later Eocene terrestrial vertebrate faunas of western North America. Pp. 333-347 in Historical biogeogra- phy, plate tectonics, and the changing environ- ment. J. GRAY & A. J. BOUCOT, eds., Oregon State University Press, Corvallis. LILLEGRAVEN, J. A., 1980, Primates from the later Eocene rocks of southern California. Journal of Mammalogy, 61: 181-204. METCALF, A. L., 1984, A new Humboldtiana (Pul- monata: Helminthoglyptidae) from extreme east- ern Chihuahua, Mexico. The Nautilus, 98: 145— 147. METCALF, A. L. & D. H. RISKIND, 1976, A new Humboldtiana (Pulmonata: Helminthoglyptidae) from Coahuila, Mexico. The Nautilus, 90: 99— 100. MORTON, P. K., 1974, Geology and engineering geologic aspects of the south half of the Canada Gobernadora Quadrangle, Orange County, Cali- fornia. California Division of Mines and Geology Special Report 111: 1-30. NILSEN, T. H. 8 E. H. McKEE, 1979, Paleogene paleogeography of the western United States. Pp. 257-276 in Cenozoic paleogeography of the western United States. J. M. ARMENTROUT, M. R. COLES & H. TERBEST, eds., Pacific Coast Paleogeography Symposium 3. Pacific Section, Society of Economic Paleontologists and Miner- alogists. ROTH, B., 1984, Lysinoe (Gastropoda: Pulmonata) and other land snails from Eocene-Oligocene of Trans-Pecos Texas, and their paleoclimatic sig- nificance. The Veliger, 27: 200-218. ROTH, B., 1986, Land mollusks (Gastropoda: Pul- monata) from early Tertiary Bozeman Group, Montana. Proceedings of the California Academy of Sciences, 44: 237-267. ROTH, B., 1988, Camaenid land snails (Gas- tropoda: Pulmonata) from the Eocene of south- ern California and their bearing on the history of the American Camaenidae. Transactions of the San Diego Society of Natural History, 21: 203— 220. ROTH, B., in preparation, A manual of the land snails and slugs of California. California Depart- ment of Food & Agriculture. ROTH, B. & T. A. PEARCE, 1988, “Micrarionta” dallasi, a helicinid (prosobranch), not a helm- inthoglyptid (pulmonate), land snail: paleocli- matic implications. The Southwestern Naturalist, 33: 117-119. SCHOELLHAMER, J. E., J. G. VEDDER & В. Е. YERKES, 1981, Geology of the northern Santa Ana Mountains, California. U. S. Geological Sur- vey Professional Paper 420-D: 1-109. SOLEM, A., 1978, Cretaceous and early Tertiary camaenid land snails from western North Amer- ica (Mollusca: Pulmonata). Journal of Paleontol- ogy, 52: 581-589. SOLEM, A., 1984, A world model of land snail di- versity and abundance. Pp. 6-22 in World-wide snails. A. SOLEM & A. C. VAN BRUGGEN, eds., E. J. Brill/W. Backhuys, Amsterdam. SUKANTO, M., 1969, Climate of Indonesia. Pp. 215-229 in Climates of northern and eastern Asia (World Survey of Climatology, vol. 8), H. ARAKAWA, ed., Elsevier, Amsterdam. WOLFE, J. A., 1969, Neogene floristic and vegeta- tional history of the Pacific Northwest. Madrono, 20: 83-110. WOLFE, J. A., 1971, Tertiary climatic fluctuations and methods of analysis of Tertiary floras. Palae- ogeography, Palaeoclimatology, Palaeoecology, 9: 27-57. WOLFE, J. A., 1978, A paleobotanical interpreta- tion of Tertiary climates in the northern hemi- sphere. American Scientist, 66: 694—703. WOLFE, J. A., 1979, Temperature parameters of humid to mesic forests of eastern Asia and rela- tion to forests of other regions of the northern hemisphere and Australasia. U. S. Geological Survey Professional Paper 1106: 1-37, pls. 1-3. WOLFE, J. A., 1981, Paleoclimatic significance of the Oligocene and Neogene floras of the north- western United States. Pp. 79-101 in Paleobot- any, Paleoecology, and Evolution, vol. 2, K. J. NIKLAS, ed., Praeger Publishers, New York. WOODRING, W. P. & W. P. POPENOE, 1945, Pa- leocene and Eocene stratigraphy of northwest- ern Santa Ana Mountains, Orange County, Cali- fornia. U. S. Geological Survey Oil and Gas Investigations Preliminary Chart 12. Revised Ms. accepted 7 February 1991 MALACOLOGIA, 1991, 33(1-2): 289-338 SYSTEMATIC REVIEW AND FUNCTIONAL MORPHOLOGY OF THE MANGROVE SNAILS TEREBRALIA AND TELESCOPIUM (POTAMIDIDAE; PROSOBRANCHIA) Richard S. Houbrick Department of Invertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560 U.S.A. ABSTRACT Diagnoses of the genera Telescopium Montfort, 1810, and Terebralia Swainson, 1840, are made, based on conchological and anatomical characters. Telescopium telescopium (Linné, 1758) is the only living member of the genus. Three Recent Terebralia species are recognized: Terebralia palustris (Linné, 1758), Terebralia sulcata (Born, 1778) and Terebralia semistriata (Mörch, 1852). Complete synonymies and descriptions are presented for each genus-level and species-level taxon. Cerithium crassum Lamarck, 1822, and Strombus semistriatus Roding, 1798, are determined to be nomina dubia. Shells, radulae and anatomical features are described and compared, and intraspecific variability is noted. Both genera date from the mid-Tertiary, and most living species have fossil records extending back to the Miocene. Both genera are confined today to the Indo-West Pacific. The ecology, life history, fossil record and geographical distri- bution of each species are presented. Species of both genera are amphibious surface-dwellers of muddy substrates in mangrove forests, and have reduced ctenidia. Most are algal-detritivores and have a taenioglossate radula, complex stomach and long crystalline style. The radula and buccal mass of Telescopium are very small, but the snout is extremely long. The radula and buccal mass of Terebralia species are robust and large. Terebralia palustris is unusual in undergoing a radular transformation between juvenile and adult stages that corresponds to the trophic dimorphism between the two stages: juveniles eat detritus, whereas adults eat fallen mangrove leaves. Species of both genera are gonochoristic with aphallate males and have open pallial gonoducts. The pallial oviduct of Telescopium is highly complex, comprising four laminae and very large, transverse glandular ridges adjacent to the oviductal groove. The pallial oviducts of Terebralia species are unusual among those of cerithioideans in having medial fusions. Terebralia males produce elaborate, crescentic spermatophores sculptured with many trans- verse keels. Spawn masses of both genera are deposited with the aid of a bulbous ovipositor located on the right side of the foot and comprise long, irregular gelatinous strings filled with very small egg capsules. The ovipositor of Terebralia species extends deep into the foot, forming an inner glandular chamber of ectodermal origin, which is thought to secrete the outer coat of the gelatinous string during spawn deposition. Species of both genera have free-swimming larval stages. The subfamily Batillariinae is excluded from the Potamididae and raised to familial rank. The genus Cerithidea Swainson, 1840, is considered closely related to Telescopium and Tere- bralia. Key Words: Potamididae, Telescopium, Terebralia, systematics, anatomy, reproductive biol- ogy, ecology, mangroves. INTRODUCTION One of the most conspicuous elements of the malacofauna of the vast mangrove swamps of the Indo-West Pacific province are the ubiquitous potamidid snails of the genera Telescopium Montfort, 1810, and Terebralia Swainson, 1840. Terebralia comprises three Recent species, Terebralia palustris (Linne, 1758), Terebralia sulcata (Born, 1778) and Terebralia semistriata (Mörch, 1852); Tele- 289 scopium telescopium (Linne, 1758) repre- sents a now monotypic genus. The taxonomy of Potamididae H. & A. Ad- ams, 1854, is in disarray because earlier au- thors relied chiefly on shell characters in their classifications and did not set clear limits to higher taxa. They were also unaware of the significant anatomical differences between the Cerithiidae Ferussac, 1819, and the Pot- amididae, and commonly referred many pot- amidid taxa to the Cerithiidae; moreover, ear- 290 HOUBRICK lier workers did not understand the anatom- ical differences between the two subfamilies thought to comprise Potamididae (Potamid- inae H. & A. Adams, 1854, and Batillariinae Thiele, 1929). Genus-level taxa from each pot- amidid subfamily were frequently mixed with one another and with other cerithiid taxa in varying combinations that included nomina of extinct higher taxa. The resultant entangled higher classification is misleading and engen- ders frustrating searches for taxa in the older literature. Terebralia and Telescopium species are large prosobranchs that occur in great num- bers and frequently dominate the surface of the muddy substrates of mangrove forests. It is therefore surprising that so little is known of their anatomy, reproductive biology or life his- tories. Although a number of ecological stud- ies on migratory movements and vertical dis- tribution in the mangrove habitat have been published, there has been no attempt com- prehensively to integrate these observations and other anecdotal information. So little is known about the soft anatomy and reproduc- tive biology of these large potamidids that it has been impossible to establish a good data base for character analysis; thus, the phylog- eny and systematic standing of these taxa in relation both to other potamidids and to other cerithioidean families has been unclear. It is the goal herein to rectify this situation. MATERIALS AND METHODS All three living species of Terebralia were studied in the field and in the laboratory. Pop- ulations of Terebralia sulcata were studied for ten days in the New Territories of Hong Kong and at the University of Hong Kong. Work on Terebralia semistriata and Terebralia palus- tris was conducted for five days at Magnetic Island and Townsville, Queensland, Australia. The anatomy of both Terebralia palustris, the type species of Terebralia, and of Terebralia sulcata is described in some detail, followed by a brief description of Terebralia semistri- ata. Observations and dissections of live Telescopium telescopium, the type of the ge- nus by tautonomy, were conducted at Mag- netic Island, Queensland, Australia. Living adult specimens of all of these species were removed from their shells by cracking them in a large vise. Animals were extracted from their shells, and care was taken to remove the numerous whorls of the visceral mass without damaging the thin mantle tissue, especially around the pallial gonoducts. Extracted snails were kept alive in aquaria and finger bowls. No adult males of Terebralia palustris or Tele- scopium telescopium were found. It is very difficult to remove these snails from their strong shells successfully without injuring and destroying tissues; moreover, damaged or injured snails secrete great quantities of mucus, which hinders dissection and fixation. Thus, it was not possible to pre- pare and examine numerous individuals in the short time spent in the field, but sufficient material was examined for comparative dis- sections. Animals were relaxed in 7.5% MgCl. solution for anatomical study and dis- section under a Wild M-8 dissecting micro- scope. A solution of methylene blue/basic- fuchsin was used to enhance anatomical features. Tissues were fixed in Bouin's solu- tion in seawater or in 10% seawater formalin and were sectioned with a razor blade. Whole animals were too large for histological sec- tioning, but selected organs and tissues were embedded in paraffin, sectioned at 7 um, and stained with Harris’ hematoxylin and eosin or Mallory’s triple stain for histological exam- ination. Photomicrographs of sections were made with a Zeiss Photomicroscope Ill. Tis- sues were Critical point dried and these, along with radulae and spermatophores, were ex- amined with a Zeiss Novascan-30 scanning electron microscope. Voucher specimens have been deposited in the USNM. Preserved museum specimens of species from other geographic regions were also used for study, but most were too poorly fixed for careful dis- section. The discussion of anatomical characters of each genus and the final discussion of pot- amidid anatomy should be understood with reference to my paper on cerithioidean phy- logeny (Houbrick, 1988). Abbreviations: AMNH—American Museum of Natural History; AMS—The Australian Mu- seum, Sydney; ANSP—Academy of Natural Sciences of Philadelphia; BMNH—British Mu- seum (Natural History); CNHM—Field Mu- seum of Natural History; DMNH—Delaware Museum of Natural History; FSM—Florida State Museum, Gainesville; MCZ—Museum of Comparative Zoology; MNHNP—Museum National d‘Histoire Naturelle, Paris; USNM— United States National Museum; WAM— Western Australian Museum; ZMA—Zoo- logisch Museum, Amsterdam. REVIEW OF TEREBRALIA AND TELESCOPIUM 291 SYSTEMATICS POTAMIDIDAE H. & A. ADAMS, 1854 TELESCOPIUM MONTFORT, 1810 Telescopium Montfort, 1810:438 (Type spe- cies, by od: Telescopium indicator Mont- fort, 1810 [= Trochus telescopium Linné, 1758, type species by tautonomy)). Tryon, 1882:250; Cossmann, 1906:124— 125; Wenz, 1940:743. Diagnosis: Shell very large, thick, conical, with numerous flat-sided whorls sculptured with spiral grooves; shell base with concentric cords and deep channel around columellar pillar. Aperture narrowly ovate, tangential (perpendicular) to shell axis with twisted, channeled columella, and outer lip curved to- ward centrally placed, short siphonal canal. Operculum corneous, circular and multispiral with central nucleus. Lateral tooth with broad lateral lamella. Snout very long with small buccal mass and very small taenioglossate radula. Rachidian tooth ovoid with broad cen- tral cusp. Mantle with siphonal light-sensory organ (pallial eye). Ovipositor on right side of foot in females. Pallial gonoducts completely open, highly complex. Zygoneurous nervous system. Egg capsules deposited in gelatinous strings. Remarks: Telescopium was thought to date from the Late Cretaceous (Turonian) by Coss- mann (1906:123), but most of the fossils at- tributed to this genus are campanilids. The earliest species that appears to belong to Telescopium as now understood is a Miocene fossil, Telescopium pseudobeliscus (Grate- loupe, 1832). Recent Telescopium is a mono- typic genus. The genus appears never to have been species-rich, but some fossil spe- cies have been described. Telescopium telescopium and Campanile symbolicum lredale, 1917, have convergent shell morphologies and were placed in the same group in some of the older monographic literature. The shells of many extinct Campa- nile species also resemble Telescopium, but recent studies have shown that the campa- nilid group constitutes a superfamily of its own (Houbrick, 1989). Telescopioidea Sacco, 1895:56 (Type species, by o.d.: Cerithium charpentieri Basterot, 1825) and Campanilop- sis Chavan, 1948 (Type species, by o.d.: Cer- ithium ceres Orbigny, 1847) were proposed as subgenera of Telescopium, but both should be excluded from it because their re- spective type species are members of Cam- paniloidea. A number of Italian Tertiary fossils lacking apertures were allocated to Tele- scopium by Sacco (1895), who noted that they differed sufficiently from Recent Tele- scopium to warrant establishment of a subge- nus, Telescopioidea Sacco, 1895, to accom- modate them. Sacco (1895) noted that they more closely resembled living Campanile symbolicum than they did Telescopium; ex- amination of these taxa confirms that they should be regarded as members of the Cam- panilidae. Authors such as Adams & Adams (1858:291) frequently included Campanile symbolicum lredale, 1917 (cited as Cerithium laeve Quoy & Gaimard, 1834) in the genus Telescopium. Fragmentary fossils of species of Campa- nile (Campanilidae) and the extinct potamidid, Vicarya Archiac & Haime, 1854, are com- monly assigned to Telescopium by authors. Shuto (1978:116, pl. 15, fig. 4) referred Pot- amides jogjacartense (Martin, 1914) to Tele- scopium, but judging from his illustrations of the specimen, it is doubtful that it should be included in the genus. Although many reports of fossil Tele- scopium in the literature are erroneous and based on fragments of other taxa, numerous fossils of Telescopium from Cenozoic depos- its in East Africa, Indonesia and the Philip- pines are either conchologically conspecific with Recent Telescopium telescopium, or very closely related to it. Allan (1950:86) included Telescopium in the family Telescopiidae, a name he substi- tuted for Potamididae without giving any rea- sons or comment, but there is no justification for this name change, and Telescopiidae must be considered a synonym of Potamid- idae. In the older literature, Telescopium is sometimes cited as a subgenus of Potamides Brongniart, 1810 (see Tryon, 1882:250-251) or as a synonym of Pyrazus Montfort, 1810 (see Adams & Adams, 1858:291), which is a batillariid. Telescopium telescopium (Linne) (Figs. 1-6) Trochus telescopium Linné, 1758:760 (Holo- type: Linnaean Collection, London; Type locality herein selected: Ambon); Linne, 1767:1231; Dodge, 1958:204-205. Cerithium telescopium Bruguiere, 1792:483; Kiener, 1841:88-89, pl. 28, fig. 4; Kobelt, 1898:57, plate 12, fig. 1. Telescopium indicator Montfort, 1810:438, 292 HOUBRICK figure (in part; Type not found; Montfort's figure, a Pyrazus species, here selected to represent lectotype). Telescopium fuscum Schumacher, 1817:233 (Type not found; Schumacher's figure reference to Buonanni, 1709, here se- lected to represent lectotype); Boettger, 1890:167. Potamides (Telescopium) telescopium К. Martin, 1884:145, 328, 348; Martens, 1897:180-181; Koningsberger, 1915: 446; Lischke, 1914:259; Oostingh, 1923: 75; Benthem Jutting, 1929:86; Rensch, 1934:339. Telescopium telescopium K. Martin, 1899: 220, pl. 33, figs. 509, 509a; 1919:94, 128, 137; Regteren Altena, 1941:13-14; Wissema, 1947:46; Butot, 1954:10; Ben- them Jutting, 1956:439-441, figs. 100, 108; 1959:105-106; Cernohorsky, 1972: 61, pl. 13, fig. 6; Brandt, 1974:196, pl. 15, fig. 61. Telescopium mauritsi Butot, 1974:7-12, pls. 1, figs. 3, 4; 2, figs. 2, 3 (Holotype: ZMA; Type locality: Pulo Panaitan, Sunda Strait, Indonesia); Benthem Jutting, 1959:105. Synonymic Remarks. Specimens of this dis- tinctive species occur in both the Linnaean Collection in London and in the “Museum UI- ricae” of the Uppsala Collection (Dodge, 1958:204-205). According to Dodge (1958), who presented a lengthy commentary on the types and Linnaean figure references for this species, no question has ever been raised as to the identity of Telescopium telescopium, and its synonymy is “unimpeachable.” He be- lieved that the specimen in the Linnaean col- lection in London should be accepted as Linné's probable type specimen, and | con- cur. Although Вгидшеге (1792) cited Cerith- jum telescopium without attributing author- ship to Linne, he did not intend to describe it under his own authorship, for he clearly stated that he merely was transferring Linné's specimen (species) from Trochus to Cerith- ¡um (Bruguiere, 1792:484). Most of Montfort's (1810) figure references for Telescopium in- dicator agree with Linné’s type specimen and figure references for Trochus telescopium, but Montfort’s (1810: figure, p. 438) own fig- ure of Telescopium indicator definitely does not represent Telescopium, but appears to be a Pyrazus species; thus, his concept of Tele- scopium indicator appears to be a mixture based upon the Telescopium figures of other authors and a Pyrazus species. The figure references for Telescopium fuscum Schuma- cher are the same as Linné’s for Trochus tele- scopium. Telescopium mauritisi Butot, 1954, is based on a worn gerontic specimen of Tele- scopium telescopium (see discussion below). Description Shell: Shell (Fig. 1) very large, solid, conical- trochoid, reaching 130 mm in length, 50 mm in width, and comprising 12 to 16 flat-sided whorls with an apical angle of 30-36 de- grees. Protoconch unknown. Early whorls sculptured with beaded spiral cord at suture and with two smooth spiral cords (Fig. 1H). Adult whorls sculptured with three large flat spiral cords and one narrow one, and with microscopic, colabral, axial striae (Fig. 11,J). Spiral cords sometimes disappearing with age or erosion. Suture weakly defined. Body whorl carinate in juveniles but expanded and broad in adults. Shell base with concentric cords and deep channel around columellar pillar. Aperture small, narrow, nearly tangen- tial (perpendicular) to shell axis, with well- rounded outer and basal margins. Outer lip thin, concave; basal lip hook-like, curved to- wards siphonal canal. Columella convex, hav- ing callus with thick, channeled, twisted fold extending into shell aperture and up columel- lar axis for length of shell (Fig. 1L). Shell color uniform dark reddish-brown to black, with whitish to light brown columellar callus. Operculum (Fig. 1M,N) corneous, small, cir- cular, multispiral (9-17 whorls), with central nucleus, transparent and fringed at margins. External Anatomy: Animal black-gray with dirty-white sole. Snout large, long (Fig. 3B, sn), having pair of short broad cephalic ten- tacles, each with tiny black eye at peduncular base. Tentacles sharply constricted at tips. Snout dorso-ventrally compressed, lined with many deep transverse wrinkles (Fig. 2B,C), very supple, extensible for considerable dis- tance. Snout tip (Fig. 2B) with thick, whitish, fleshy pad having vertical slit-like mouth (Fig. 2C). Foot large with whitish sole having many branching, transverse furrows (Fig. 2D). Left lateral and ventrolateral sides of foot each with deep, rounded groove (Fig. 3, llg, vg); groove strongly developed on left side. Ante- rior mucous gland opening a narrow, deep, slit (Fig. 2D; 3B, amg) extending halfway around sides of sole and heavily ciliated within (Fig. 2E). Groove emerging from exhal- ant siphon long, deep, highly ciliated in both REVIEW OF TEREBRALIA AND TELESCOPIUM 293 FIG. 1. Telescopium telescopium, shells and operculum. A-D, apertural, dorsal, side and anterior views of adult shell from Cairns, Queensland, Australia; 98.9 x 47.3 mm (USNM 795231). E-G, apertural, dorsal and anterior views of badly eroded, gerontic shell from Halmahera, Moluccas; 103 x 51 mm (USNM 837107). H, juvenile shell from Cebu, Philippines; 40.5 x 20 mm (USNM 419345). I-K, apertural, dorsal and anterior views of uneroded, cleaned shell showing sculptural details, from Batangas Bay, Luzon, Philippines; 86.5 x 37.7 mm (USNM 846507). L, shell sectioned through axis showing columellar pillar with weak interior plait. M, N, attached (M) and free (N) sides of operculum; 11.9 mm diameter. sexes, running down right side of foot (Fig. 3A, cg) in males, ending at edge of sole; in females, ending at oval, sooty-colored, warty ovipositor (Fig. 3A, ovp) near sole margin. Ovary pale yellow; testis orange. Mantle dirty white. Dorsal half of mantle edge with small, weak, wart-like papillae (Fig. 3A, mp) and with median cleft (Fig. 3B, mc) about 5 cm deep, separating inhalant siphon area from rest of mantle edge; ventral half of mantle edge smooth. Edge of interior of right side of mantle with thin glandular line of unknown function. Inhalant siphon thick, and interior edge with one or two orange-pigmented spots, each surrounding black, pit-like light-sensitive or- gan (eye) containing lens. 294 HOUBRICK FIG. 2. Telescopium telescopium, SEM photographs of critical point dried tissues. A, transverse cut through right side of foot showing glandular ovipositor (ovp); bar = 0.8 mm. B, C, two views of snout tip showing pad-like tip, fleshy lips and hook-like structure at dorsal side of mouth (arrow); bar (B) = 0.9 mm, bar (C) = 0.8 mm. D, anterior edge of sole showing slit-like opening to anterior mucous gland (arrow); bar = 0.9 mm. E, dense cilia lining opening to anterior mucous gland; bar = 20 pm. F, left side of mantle roof showing ctenidium (ct) and osphradium (os), and large ridge (r) between them; bar = 0.9 mm. G, detail of narrow posterior ctenidial filaments (ct); bar = 0.9 mm. H, detail of osphradium (os) in deep pit bordered by wide ridges; bar = 0.2 mm. Mantle Cavity: Mantle cavity deep, spacious, occupying last two whorls. Osphradium (Fig. 2F-H, os) a short narrow ridge sunk into deep trench next to whitish, thick, crescentic ridge (Fig. 2F, r). Osphradium close to ante- rior part of ctenidium and about one-fourth ctenidial length. Ctenidium (Fig. 2F,G, ct) large, broad, very shallow, becoming posteri- orly weaker. Ctenidium beginning about 5 mm from mantle edge, extending posteriorly for length of mantle cavity (two whorls). Anterior part of ctenidium with narrow, equilateral triangular filaments rapidly becoming very broad and shallow on left (Fig. 3D), attaining ribbed appearance throughout much of medial and posterior parts; posterior cteni- dial filaments becoming elongated, forming shallow, longitudinal ridges; raised, triangu- lar part of each posterior filament shifts to right, adjacent to hypobranchial gland (Fig. REVIEW OF TEREBRALIA AND TELESCOPIUM 295 D FIG. 3. Telescopium telescopium, anatomical features. A, right side of female showing elongate snout, large foot, ciliated groove and ovipositor on side of foot. B, anterior view of head-foot showing mantle edge, head, snout and sole; note opening of anterior mucous gland (amg) along edge of sole. C, central nervous system showing relationship of ganglia and long supraesophageal connective (sec); visceral ganglion and visceral loop not shown. D, individual ctenidial filaments from anterior (left) to posterior (right) mantle cavity. amg, opening to anterior pedal mucous gland; cg, ciliated groove; cm, columellar muscle; cpc, cerebral-pedal connective; а, dialyneury; exh, exhalant siphon; Icg, left cerebral ganglion; llg, left lateral groove; Ipl, left pleural ganglion; m, mouth; mc, mantl cavity opening or inhalant siphon; mp, mantle papillae; op, operculum; Ovp, Ovipositor; pn, pallial nerve; rcg, right cerebral ganglion; rpg, right pedal ganglion; rpl, right pleural ganglion; s, sole; sbc, subesophageal connective; sbg, subesophageal ganglion; sec, supraesophageal connective; sn, snout; spg, supraesophageal ganglion; st, statocyst; t, tentacle; vg, ventral groove; z, zygoneury. 2F, ct). Hypobranchial gland white, narrow, about one-half length of rectum, and secret- ing great sheets of mucus over ctenidium and rectum. Rectum very large, wide, with in- terior epithelium forming deep transverse ridges and complex folds, producing large surface area. Anus located well behind front of ctenidium. Pallial gonoducts open, com- prising two major laminae, and parallel with rectum. Alimentary System: Snout tip with thick, fleshy, ovate pad surrounding vertical slit-like mouth. Pair of hook-like extensions (Fig. 2C) at top edge of oral tube. Pair of small jaws in sides of oral tube. Radula (Fig. 4A—D) very small, fragile, short, about 5 mm long, com- prising 80 rows of teeth, and about 3.3 per- cent of shell length. Rachidian tooth (Fig. 4B,C) oval, taller than broad with convex an- terior end and long, dagger-like basal plate; 296 HOUBRICK FIG. 4. Telescopium telescopium, scanning electron micrographs of radula. A, mid-section of radular ribbon with marginals spread open; Баг = 200 рт. В, detail of lateral and rachidian teeth; bar = 80 um. С, half row of radula showing elongate basal plate of rachidian tooth; bar = 90 um. D, detail of marginal teeth showing flange on outer marginal tooth; bar = 40 pm. E, half row; Баг = 100 рт. cutting edge with spoon-shaped central cusp flanked on both sides by three or four small denticles. Lateral tooth (Fig. 4B) longer than broad, having long, wide, rectangular basal plate, pointed posteriorly, and with slightly flaring outer side; cutting edge with large spatulate central cusp, two inner denticles and three outer denticles. Marginal teeth (Fig. 4C,D) with long narrow shafts and hook-like tips having long, pointed central cusps. Inner marginal tooth with four inner denticles and three outer denticles, outer marginal tooth lacking outer denticles, but with wide flange (Fig. 4D) along outer side of tooth shaft. Buc- cal mass very small, at snout tip. Small, paired, tubular salivary glands (Fig. 5A, sgd) beginning well anterior to nerve ring, each looping and folding along sides of buccal mass, and opening into anterior dorsal oral cavity. Anterior esophagus (Fig. 5A, eso) a simple tube having dorsal food groove, twist- ing at nerve ring, and becoming slightly wider behind it. Mid-esophagus (Fig. 5B) long, mod- erately wide, with large dorsal, ventrally lo- cated food groove (Fig. 5B, df), but not de- veloped into esophageal gland. Posterior esophagus very long, becoming narrower posteriorly, entering right lateral side of pos- REVIEW OF TEREBRALIA AND TELESCOPIUM 297, FIG. 5. Telescopium telescopium, anatomical features. A, representation of dorsal lateral view of buccal mass, dorsal epithelium partly opened to expose radula. B, cross-section through mid-esophagus showing dorsal food groove and cartilage-like tissue lining both sides of esophageal cavity. C, camera lucida drawing of central part of pallial oviduct. D, schematic drawing of pallial oviduct. ant, anterior part of pallial oviduct; ag, albumen gland; bm, buccal mass; с, cartilaginous tissue; cg, capsule gland; df, dorsal food channel; eso, esophagus; Il, lateral lamina; ml, medial lamina; osb, opening to spermatophore bursa; osr, opening to seminal receptacle; ov, coelomic oviduct; ovg, oviductal groove; prd, parallel ridges; r, radula; sb, spermato- phore bursa; sg1, outer sperm gutter; sg2, inner sperm gutter; sgd, salivary gland; sr, seminal receptacle; vf, ventral food channel. terior part of stomach. Stomach very large, elongate, about one and one-half whorls and comprised of three chambers: posterior sort- ing area with large ridge-like central pad and broad sorting area comprised of many small, transverse, epithelial folds; deep, transverse channel with centrally located cuticular gastric shield having denticulate margin; anterior por- tion of stomach long, tubular, consisting of very long style sac and opening to intestine. Style sac very long, adjacent to intestine, but separate from it, reaching anteriorly as far as mantle cavity to level of pericardium. Crystal- line style very long, club-shaped at gastric shield. Reproductive System: Female pallial oviduct (Fig. 5C,D) long, comprising two complex, major laminae; lateral lamina (Fig. 5C,D, Il), fused on one side to mantle wall, and medial free lamina (Fig. 5D, ml). Medial lamina edge trifurcate, comprising three parallel minor laminae and two ciliated sperm gutters (Fig. 5C,D sg1, sg2). Inner sperm gutter (Fig. 298 HOUBRICK 5C,D, sg2) extending for two-thirds of lamina length, and entering duct leading to posterior seminal receptacle (Fig. 5D, sr); outer gutter (Fig. 5C,D, sg1) extending for more than two- thirds of length of medial lamina, entering into spermatophore bursa (Fig. 5D, sb) in poste- rior third of lamina. Lateral lamina (Fig. 5C,D, ll) bifurcate at free edge, with one lamina ad- jacent to oviductal groove and other lamina comprising broad, twisted parallel ridges (Fig. 5D, prd) with which trifurcate edges of medial lamina interdigitate. Glandular parts of bases of both laminae formed of paired row of nu- merous thick, transverse ridges, bordering deep oviductal groove (Fig. 5C,D, ovg). Albu- men gland (Fig. 5D, ag) in posterior portion of oviduct. Capsule gland (Fig. 5D, cg) in thick- ened base of anterior third of oviductal groove. Male pallial oviduct unknown. Eus- permatozoa classified in structural group 2 (Healy, 1983:65,73). Paraspermatozoa with head region and tail tuft; assigned to struc- tural group 1 (Healy, 1986:187). Nervous System: Central nervous system (Fig. 3C) epiathroid. Cerebral ganglia (сд, rcg) with short commissure and with pleural ganglia (rpl, Ipl) closely joined. Short connec- tives (cpc) between cerebral and pedal gan- glia; pedal-pleural connectives thin, adjacent to pedal-cerebral connectives. Pedal ganglia (rpg) each with three nerves and with stato- cyst (st). Zygoneury (z) between subesoph- ageal (sbg) and right pleural ganglia (rpl). Very long connective (sec) between right pleural and supraesophageal ganglia (spg). Dialyneury (d) between supraesophageal nerve and left pallial nerve (pn). Labial nerves very long, matching snout length. Long vis- ceral loop extending back to visceral ganglion at posterior end of mantle floor. Remarks It is remarkable that no contemporary stud- ies have been made on the anatomy and life history of this large, common prosobranch. Aside from a few notes by Prashad (1925) on the mantle cavity and ctenidium, the only pub- lished anatomical account of substance is an early paper by Berkeley & Hoffman (1834) de- scribing the gross anatomy, but with several major errors. The nervous system was well described and figured by Bouvier (1887:145— 146, pl. 8, fig. 32). Shell: The shell of Telescopium telescopium is one of the most distinctive of all living prosobranchs, and is not likely to be confused with that of any other species except Campa- nile symbolicum lredale, 1917, with which it is convergent. The adult shell is normally 90— 100 mm long, but Brandt (1974:196) has re- corded a very large specimen reaching 130 mm in length. Telescopium telescopium is one of the few gastropods with an aperture tangential (perpendicular) to the shell axis. The anterior canal is nearly centrally located and weakly reflected dorsally (Fig. 1D,G,H). Sculptural elements occur on younger individ- uals (Fig. 1H-J), but are usually eroded in older shells (Fig. 1A-C). The columellar plait is strong at the aperture but becomes weaker internally (Fig. 1L). Telescopium mauritsi Butot, 1954, de- scribed from the Sunda Straits, Indonesia, as a new Recent species with smoother sculp- ture, was later shown by Brandt (1974:196) to be merely a gerontic, fully adult form of Tele- scopium telescopium, with its spiral sculpture partly eroded, and covered with something that Brandt (1974:196) called “a secondary layer of unknown substance.” Budiman (1988:240) also demonstrated that the two taxa are indeed synonymous and joined by a complete series of intermediate forms. He fur- ther showed that the secondary coating of the smooth form is not part of the shell, but a deposit of an organo-metallic complex (Fe, Mn), which is a common feature of epifaunal mollusks in this habitat. Anatomy: The ventral and lateral grooves (Fig. 3A,B, vg, llg) of the foot are unusual and not seen in other potamidids. The slit-like opening to the anterior mucous gland is un- usual in extending from the edge of the propo- dium along both sides of the sole, for half its distance (Figs. 2D, 3B, amg). Both sexes have a ciliated groove (Fig. 3A, cg) leading from the exhalant siphon down the right side of the foot to the edge of the sole, whereas in other potamidids and cerithioideans, the groove is seen only in females. Budiman (1988:244) suggested that males use this groove to transport and introduce sperm into the female aperture. However, if spermato- phores are present, as they are in all other known cerithioideans, this function is difficult to visualize. The groove in males might merely remove mucus and debris from the mantle cavity, transporting it in a mucous string to the foot margin. The ovipositor (Fig. 3A, ovp) is near the edge of the foot and, in section, is seen to comprise a glandular pad (Fig. 2A, ovp) that does not mark the entrance REVIEW OF TEREBRALIA AND TELESCOPIUM 299 to an internal chamber, as in Terebralia spe- cies. The pallial light-sensitive organ appears to have a lens, but is poorly organized. It is un- doubtedly homologous with similar organs in Terebralia species and with the well-devel- oped pallial eyes of Cerithidea species (Hou- brick, 1984:10-11). The osphradium (Fig. 2F,H, os), which is quite short in relation to the large ctenidium (Fig. 2F,G, ct), is reduced to an anterior, ex- tremely narrow ridge and is unique among potamidids in that it lies in an elongate pit (Fig. 2G,H) next to the efferent ctenidial sinus. To the right of the osphradium, the epithelium is folded into thick ridges, while to the left, lying between the osphradium and the effer- ent sinus, is a wide, ridge-like structure (Fig. 2F, r) of unknown function. Prashad (1925: 141) noted “a large elongate gland next to the osphradium which secreted a green sticky substance when the animal was disturbed”; this is probably the ridge-like structure | ob- served although | did not see any secretions. The ctenidium, poorly depicted by Berkeley & Hoffman (1834:pl. 20, figs. 3,5), has been ac- curately described and figured by Prashad (1925:140, fig. 3a). The change in the ctenid- ium from anterior, triangular leaflets (Fig. 2F) to posterior elongate, parallel ridges or corru- gations on the posterior pallial roof (Fig. 2G) is unusual, but a similar ctenidial modification also occurs in some Cerithidea species (Houbrick, 1984). As Prashad (1925:140- 141) noted, the afferent ctenidial vein is ab- sent in Telescopium, and hemolymph from the abdominal sinus is taken up directly by the ctenidial ridges from numerous sinuses into which the abdominal sinus divides after en- tering the mantle. Similar blood sinuses and vessels are known to occur in the mantle roof of Cerithidea species such as Cerithidea ob- tusa (Lamarck, 1822) in which the ctenidia are very much reduced (Houbrick, 1984:11). | have interpreted the long, narrow, white area of the mantle roof between the ctenidium and rectum, which secretes great sheets of mucus that coat the ctenidium, as a hypo- branchial gland. Prashad (1925:141) on the other hand, wrote that the hypobranchial gland had entirely disappeared, its place be- ing occupied by prolongations of the ctenidial filaments, but the ctenidium does not occupy the entire area of this part of the mantle roof. The supple, extensible snout (Fig. 3B, sn) is almost proboscis-like in function, and per- haps the longest among neotaenioglossans (mesogastropods) of the suborder Discopoda Fischer, 1884; it was well illustrated by Berke- ley & Hoffman (1834:pl. 20, fig. 1). The radula has never been well illustrated, but Das et al. (1988, pl. 3) recently depicted the cutting edges of the rachidian, lateral and some mar- ginal teeth in an SEM photomicrograph. The rachidian tooth is unique among those of pot- amidids in being ovoid-triangular, and in hav- ing an elongate, very narrow basal plate (Fig. 4B,C). The contrast between the very large snout and the very small buccal mass and radula is notable. Indeed, the small radula, jaws and buccal mass also contrast with the large size of the animal, and reflect the fine particulate food eaten by it. My observations on the arrangement of the salivary glands differ from those of Berkeley & Hoffman (1834:433, pl. 21, fig. 14), who de- scribed each duct as running back from the anterior buccal mass, tightly coiling and then joining the other duct just below the apex of the radula. | noted a pair of salivary glands, each originating in front of the nerve ring, coil- ing anteriorly, and exiting at the anterior end of the buccal mass (Fig. 5A, sgd), but did not find a connection between the two glands as depicted by Berkeley & Hoffman. The large, complex stomach of Tele- scopium is typical of members of the Pota- midinae, and is highly adapted to deal with the fine particulate matter ingested. The gas- tric shield, well illustrated by Berkeley & Hoff- man (1834, pl. 20, fig. 6), is unusual in having a denticulate margin. The exceptionally long crystalline style (about 50 mm long in adults), is held completely separate from the intestine in the heavily ciliated style sac (Alexander & Rae, 1974:56). The dilated, club-like posterior end of the style was noted by Seshaiya (1932: 171). The style follows and parallels the in- testine as far forward as the level of the mid- region of the mantle cavity. Alexander & Rae (1974) made a detailed study of the structure and formation of the crystalline style of Tele- scopium, and noted that the style persists even when the animal is starved for consid- erable periods. They showed that the style does not vary in size during the tidal cycle, unlike those in many style-bearing pelecy- pods in which the style might break down and reform, and that the style is present even in the smallest specimens of Telescopium tele- scopium. Enzymatic activity of the style of Telescopium telescopium has been studied by Alexander et al. (1979). The rectum is characterized by being very 300 HOUBRICK wide and by having its interior walls folded into complex ridges and pockets, thus in- creasing its surface area and perhaps func- tioning as a compactor of the ovoid fecal pel- lets. The large, complex, pallial oviduct (Fig. 5C,D) is basically an open, slit tube, compris- ing two laminae, the lateral (right) one at- tached to the mantle roof (Il), and the medial (left) one (ml) free—a typical cerithioidean plan. However, it departs from this layout in its complexity, having the two major laminae subdivided to form five minor parallel laminae, and in having two sperm gutters (sg1, sg2) in the major lamina. The tripartite condition of the pallial oviduct mentioned by Berkeley & Hoffman (1834:435), who throughout their text, refer to the oviduct as “the matrix,” is probably a misinterpretation, for there are re- ally three well-defined minor laminae on the medial lamina and two minor laminae on the lateral lamina, the fifth one being broad. Berkeley 8 Hoffman (1834) noted that the ovi- duct consisted of “. . . three strong folds which fit over a thick longitudinal wrinkled rib [the fifth, broad lamina] so closely that it appears like a simple sac and requires a minute in- spection to ascertain the real structure.” This description is fairly accurate, but incomplete inasmuch as it mentions neither the bursa or seminal receptacle nor the parallel ridges (prd) of the fifth, broad lamina next to the lateral lamina. | found it very difficult to interpret the arrangement of the pallial oviduct in preserved specimens of Telescopium telescopium, but dissections of several live specimens have clarified the complex details of the ducts and chambers within the laminae comprising the pallial oviduct; nevertheless, much remains unknown regarding the function of these parts. A diagram of the pallial oviduct is shown in Figure 5D. The medial lamina has undergone extensive modification in that it has become longitudinally trilobed, having two sperm gut- ters. The oviductal groove (ovg) lies between the inner lobe of the medial lamina and the attached lateral lamina, and it is here that the eggs presumably are fertilized, surrounded by albumen in the albumen gland (ag), encapsu- lated in the capsule gland (cg), and moved anteriorly to the ciliated groove on the foot and thence to the ovipositor. The edge of the lateral lamina (Il) is bifurcate, and one ridge forms small parallel longitudinal folds (prd) with which the inner lobe of the medial lamina interdigitates. Thus, although the pallial ovid- uct is anatomically open, it is functionally closed except anteriorly, owing to the close juxtaposition of the laminae. The pair of trans- verse glandular ridges lying along the length of the base of the oviductal groove undoubted- ly comprises the albumen (ag) and capsule glands. This area was thought by Berkeley & Hoffman (1834:435) to function as “many little bags for the reception of eggs”; although it is a transport area for eggs, the area should not be construed as a pallial brood pouch. These paired transverse ridges on each side of the oviductal groove also occur in members of the Turritellidae (Fretter & Graham, 1962:366; Carrick, 1980). The nervous system (Fig. 3C), well de- scribed by Bouvier (1887), is similar to those of Terebralia species, but differs from them in the greater length of the connective (sec) be- tween the supraesophageal and right pleural ganglia (see Bouvier, 1887, pl. 8, fig. 32). Both zygoneury and dialyneury are well es- tablished in Telescopium. Ecology: Several good papers exist on as- pects of the ecology of Telescopium tele- scopium. These include the studies of Lasiak & Dye (1986) in Queensland, Australia, the work of Wells (1986) in Western Australia, and studies made at Indonesian sites in Java, the Moluccas, and Irian Jaya (Western New Guinea) by Budiman (1988). Other, less com- prehensive accounts include some ecological notes of Butot (1954) on Prinsen Island in the Straits of Sunda, Indonesia, the observations of Benthem Jutting (1956) in Java, and those of Brandt (1974) in Thailand. The information from these studies and my own is summa- rized below. Telescopium telescopium lives intertidally on soft, nearly liquid, muddy substrates asso- ciated with mangrove forests in which it is frequently found in shady places in the more exposed parts of the mangrove and around the runoff gulleys common in these habitats. A good depiction of an individual plowing through muddy substrate in its natural habitat is given by Coleman (1981:36, fig. 81). The animals are shy to any movement and quickly retreat into their shells whenever approached. This shyness was also recorded by Berkeley & Hoffman (1834:431), who reported great difficulty in making drawings of the living an- imal. | observed Telescopium telescopium liv- ing in a mangrove habitat at Magnetic Island, Queensland, Australia, in which it coexists with large populations of Terebralia palustris. In this population, many individuals ranging REVIEW OF TEREBRALIA AND TELESCOPIUM 301 from adults to young snails (less than 5 mm long) occurred together, although the small- est appeared to prefer the higher ground around the bases of mangrove roots. Lasiak & Dye (1986:174) and Budiman (1988:240) observed the same phenomenon at Northern Australian and Indonesian sites, respectively, the latter author pointing out that both adults and juveniles are deposit feeders, having the same radular morphology. This is unlike the radular dimorphism seen between adults and juveniles of Terebralia palustris, and suggests that the segregation of adults and juveniles is not due to trophic factors. Budiman (1988) and Shokita et al. (1984) noted that Tele- scopium telescopium had a restricted distri- bution mostly at the middle and landward edge of the mangrove intertidal zone. According to Lasiak & Dye (1986:174) and Budiman (1988:242), Telescopium telesco- pium is active only when exposed by low tide, withdrawing into its shell whenever covered with water. Lasiak & Dye (1986:178) demon- strated that movement in Telescopium tele- scopium was related to the tidal regime and was not due to endogenous factors. The an- imal becomes active shortly before the ebbing tide uncovers the substrate, when it emerges from its partly burrowed position and begins creeping on the exposed mud. All activity ceases during hours of high tide, when the animals burrow into the muddy substrate, tak- ing a semi-vertical position just below the sur- face (Budiman, 1988:244). Benthem Jutting (1956:440) noted that the tip of the spire pro- jected when the shell was partly buried in the mud. During one tidal cycie Telescopium tele- scopium might be under water for as long as three or four hours (Alexander et al., 1979: 88). Telescopium telescopium is a deposit feeder, using its long, extensible snout to engulf fine mud and detritus from the surface of mudflats during low tide. Lasiak & Dye (1986:174) maintained that Telescopium tele- scopium can feed only at low tide; this obser- vation has been confirmed by Alexander et al. (1979:88), who have shown that there are re- duced enzyme levels in the crystalline style at high tide, in contrast to increased levels during low tide, indicating a fasting mode whenever snails are covered by water. Nevertheless, Bu- diman (1988:244) opined that Telescopium telescopium can maintain its feeding activity from its burrow by using its long, extensible snout, even when covered by high tide. This possibility led Budiman (1988:244) to suggest that exposed activity periods were more vital for sexual activity than for feeding. Telescopium telescopium is able to tolerate a wide range of salinities, from 15 ppt to full- strength sea water (Alexander & Rae, 1974: 56), and can withstand a considerable degree of desiccation. During dry, inactive periods, individuals frequently cluster together in shady refuge microhabitats beneath the man- groves, for extreme temperatures can result in high mortality. Although high temperatures might cause mortality, this species can en- dure astonishing periods of desiccation: Ben- son (1834) recorded that animals collected in India endured a trip to England lasting more than six months, with only occasional immer- sions in sea water, and lived. Predation has not been observed, and it is doubtful that large, thick-shelled adults can be attacked successfully and eaten by any pred- ators with the exceptions of the large, powerful mud crab, Scylla serrata (Forskal), and man, who is known to use Telescopium for food in Southeast Asia, Indonesia, Borneo and the Philippines (Tryon, 1882:259; Benthem Jut- ting, 1956:441; Brandt, 1974:196). Juveniles are more likely to fall victim to the predatory activities of mangrove crabs, birds and mam- mals. On the southeastern coast of India, this species is seldom eaten by man, but is ex- ploited in the lime industry. Kasinathan & Shanmugam (1988) have documented the overexploitation of Telescopium in this region: in a period of six months, nearly 22 large bags (500-900 kg/day) of Telescopium, each weighing 70-80 kg, were collected every week from the Pithavaram mangroves and Vellar Estuary, and sent to the lime industry. Reproductive Biology: Budiman (1988:244) noted that sexual pairing occurred at all times during periods of activity at low tide, when many paired individuals were seen. Accord- ing to his observations, the male holds the female’s shell with his foot, manouvering her so as to position her shell aperture against his shell aperture. When this position is attained, the male puts his head and foot into the fe- male’s shell aperture. Budiman (1988) sug- gested that sperm might be transferred by means of the groove on the right side of the male’s foot, but in light of the fact that other potamidids produce spermatophores (Houbrick, 1984:7; this study), sperm transfer probably is effected by production and trans- fer of spermatophores either along the male groove or by pallial currents, as has been ob- 302 HOUBRICK served in other cerithioideans (Houbrick, 1973). Spawn masses of Telescopium tele- scopium were not observed in the field nor were any preserved spawn examined. A good description of the spawn and spawning be- havior of Telescopium telescopium was pub- lished by Ramamoorthi & Natarajan (1973), who studied this species at Porto Novo, India. They recorded that in its natural habitat on the mudflats of the Vellar estuary, this animal spawned at the low water mark from April to July, but always just below the water mark. During this period snails were frequently seen associated in pairs. Animals kept in finger bowls and in experimental ponds on the mud flats attached their egg masses to the sides of bowls and to cement slabs and stones, under water. Because Ramamoorthi & Natarajan (1973) did not mention its use during egg laying, the exact function of the ovipositor during this process remains unknown. The ciliated groove on the right side of the female’s foot begins just below the aperture of the pallial oviduct and terminates at the ovipositor, passing behind and under it (Fig. 3A, cg). The glandular composition of this large, padlike structure (Fig. 2A, ovp) suggests that it con- tributes to the secretion of the gelatinous coat Surrounding both the strings of egg capsules and perhaps also to the outer coat of the ge- latinous strings. Spawn is deposited in a con- tinuous, compact ribbon, folded closely upon itself such that the spawn mass is long and sheet-like upon completion (Ramamoorthi & Natarajan, 1973:159, figs. 1, 2). These au- thors noted that the egg mass is covered by a transparent, gelatinous matrix and that the eggs are light blue. Egg masses are very large, varying from 24-49 cm long and from 1.5-2 cm wide (n= 20), and each egg mass contains an average of 50,000 eggs. Individ- ual eggs are about 125 ¡um in diameter, and are surrounded by a fluid, presumably albu- men, and a thin, transparent capsule of 200 um (Ramamoorthi & Natarajan, 1973:158). At Porto Novo, free-swimming veliger larvae with smooth, transparent shells and a blue tint hatched from the egg masses 96 hours after spawning. To my knowledge, no other obser- vations have been made on spawning or lar- vae, and nothing is known of the duration of the free-swimming veliger stage prior to set- tlement, or of growth and longevity of adults. Because even the protoconchs of tiny juve- niles are very eroded, it is impossible to infer the length of planktonic life by observing the size and sculpture of protoconch-2, but the small diameter of the egg capsules and their great numbers in a spawn mass suggest a long planktotrophic larval stage. Fossil Record: Fossils of Telescopium tele- scopium occur in Upper Miocene through Ho- locene strata. This species has been found in many localities throughout the Indonesian ar- chipelago, in which it has been recorded from the Upper Miocene of Java (K. Martin, 1884: 145-146; 1889:220, pl. 33, fig. 509; Wissema, 1947:46) and E. Borneo (Wissema, 1947:48); from the Miocene of Java and Timor (Tesch, 1920:58—59, pl. 132, fig. 191); from the Qua- ternary of Nias (Wissema, 1947:47); the Pliocene of Nias (Tesch, 1920:59), Timor and Java (Wissema, 1947:46—47); from the Pleis- tocene of E. Java and Timor (Wissema, 1947); and from the Holocene of E. Sumatra and S.W. Celebes (for more detailed locality data, see Regteren Altena, 1941:13-14, and Wissema, 1947:46-47). In the Philippines, Telescopium telescopium has been recorded from the Miocene Vigo Formation (Popenoe & Kleinpell, 1978: pl. 2, fig. 13) and from the Upper Miocene of the Tartaro Formation of Luzon (Kanno et al., 1982:95, pl. 17, fig. 7). K. Martin (1889) described two fossil Tele- scopium species from the Pliocene of Java, Telescopium jogjacartensis and Telescopium titan, which are morphologically very similar to Recent Telescopium telescopium, and pos- sibly conspecific with it. Oostingh (1935:52) has recorded several localities for Tele- scopium titan on Java. Distribution: This species occurs in suitable habitats in the western Pacific from the Ryukyus south through Taiwan (Tapparone Canefri, 1883:58), the Philippines, New Guinea and throughout tropical Australia (Fig. 6). It is found throughout the Indonesian ar- chipelago and the coasts of Southeast Asia. In the Indian Ocean it occurs in the Mergui archipelago, the Andamans, and along the coasts of India and Ceylon west to Karachi. It has been cited from Reunion (Deshayes, 1863:95) and Madagascar (Rajagopal & Mookherjee, 1982:31), but the latter record is probably incorrect, because this region has been well collected and there are no other records. Although cited from both the Gulf of Oman (Bosch & Bosch, 1982:47), and the Persian Gulf (Smythe, 1979:65), these records were based on dead or subfossil specimens. A single museum record from the REVIEW OF TEREBRALIA AND TELESCOPIUM 303 RER DE М AU С 90° FIG. 6. Geographical distribution of Telescopium telescopium. Gulf of Aden (ANSP 191200) needs confirma- tion. Material Examined: ARABIA: Gulf of Aden, near Aden (ANSP 191200). PAKISTAN: Ma- jia Id, Karachi (MCZ 255730); Baba Id, Kara- chi (CNHM 69795); Patiani Creek, Karachi (USNM 633063); near Karachi (USNM 693979; AMNH 109797). INDIA: Calcutta (CNHM 5070; USNM 18585; MCZ 135602; AMNH 32128); Vengurla, N of Goa (USNM 443203); Colaba, Bombay (USNM 443317); Bandara, N of Bombay (USNM 443623); Bombay (ANSP 231872; AMNH 135320); Mulund, Bombay (AMNH 104797); Juhu, Var- sava (USNM 598139, 611579); Matapalem Creek, Godavary Estuary, Andhra, Pradesh (MCZ 260876); Cochin Harbor, Kerala (ANSP 303822); Canning, Bengal (USNM 638074). BANGLADESH: Burigoalni Sundarbans (MCZ 256420). CEYLON: Trincomallee (MCZ 101297). THAILAND: Pattani (MCZ 238425); Lem Ngob, Trad (USNM 384178, 776705); Narativat (USNM 776707); Lem Sing, Chant- aburi Prov. (USNM 419159; MCZ 238375); Goh Phi Phi (Pipidon) (USNM 661487); Chantabun (USNM 529520); Grabi (USNM 776706); Laam Yamu, E. Phuket Id (ANSP 285908, 286210). MALAYA: Singapore (ANSP 18003); Kuala Selangor, Selangor, Malaysia (ANSP 320715). VIET NAM: Nha Trang Harbor (AMNH 156581). INDONE- SIA: Tjiperwagaran, Bantam, Java (USNM 260571); Pasir Putih, Jailolo District, Halma- hera, Moluccas (USNM 837107); Papajato River, Celebes (USNM 244076). BORNEO: road to new port, Port Swettenham, Selangor, Malaysia (CNHM 140958); NE corner Palau Lumut, Port Swettenham, Selangor, Malaysia (USNM 661019); Po Bui Id, Sandakan, N. Borneo (USNM 244066; AMNH 150838); Sempora, N. Borneo (USNM 658194); Tan- jong Aru, Jesselton, N. Borneo (USNM 658594); W Marudu Bay, N. Borneo (USNM 632215; ANSP 255776); Jesselton, N. Bor- neo (ANSP 275078). NEW GUINEA: Daugo Id, Papua (CNHM 140290); Bosnek, Sorida, Papua (AMNH 99187); E Kaipoeri Village, Koeroedoi Id, Geelvink Bay, W Irian (ANSP 206509). AUSTRALIA: Bowen, Queensland (ANSP 234264); Cooktown, Queensland (ANSP 195583); Cairns, Queensland (USNM 603491, 795231; ANSP 204885); Cockle Bay, Magnetic Id, Queensland (USNM 828833, 828804, 842987); Broome, Northern Territory (ANSP 232006; MCZ 265997). PHILIPPINES: E of Nagsulu, Batangas Prov, Luzon (ANSP 229709); Batangas Bay, Luzon (USNM 846507); Camp Wallace, La Union, Luzon (USNM 233100); Bacoor Bay, Luzon (USNM 244060); Mariveles, Luzon (USNM 283258); Borogan Village, E side Samar (ANSP 223068); Mangarin, Mindoro (USNM 542677); Iloilo, Panay (USNM 385080, 304 HOUBRICK 383986); Banio Luloc, Panay (USNM 419549); mouth of Jardau River, Gimaras (USNM 243882); Tiglauigan Pt, Cadiz, Ne- gros (USNM 313027); Balolan, Negros (USNM 313233); Marongas, Jolo (USNM 233238); Port Dupon, Leyte (USNM 232964); Enrique Villanova, Siquijor (USNM 617818); Cebu, Cebu (USNM 244071); Palawan (USNM 303929); Nakoda Bay, Palawan (USNM 240238); Tilig, Lubang (USNM 243981); Panabuban Bay, Mindanao (USNM 244075); Dos Amigos Bay, Tawi Tawi Gp (USNM 214067); Port Tataan, Tataan, Tawi Tawi (USNM 244002, 244067). RYUKYUS: Miyako (CNHM 68787). TEREBRALIA SWAINSON, 1840 Terebralia Swainson, 1840:315 (Type spe- cies, by subsequent designation (Sacco, 1895:51), Strombus palustris Linne, 1758. Diagnosis: Shell moderately large, solid, brown, with numerous flat-sided whorls, and sculptured with spiral cords and axial ribs; ap- erture ovate, outer lip Sweeping in arc to join moderately-developed anterior canal, fre- quently forming complete peristome. Col- umella with paired internal plaits; internal palatal teeth opposite varices. Operculum corneous, round, multispiral with central nu- cleus. Mantle edge at inhalant siphon with pit- like light-sensory organ. Snout broad, muscu- lar, with large buccal mass. Anterior pedal mucous gland opening slit-like, extending halfway down sides of sole. Radula taenioglos- sate; rachidian tooth rectangular with large central cusp; outer marginal tooth with wide lateral lamella. Stomach elongate, with gas- tric shield, and long style sac. Right side of foot in females with large, complex ovipositor. Pallial oviducts with medial fusion. Nervous system epiathroid, zygoneurous. Spermato- phores crescentic, having transverse ridges. Eggs deposited in coiled gelatinous ribbons. Remarks: The genus Terebralia, sensu stricto, can be traced with relative certainty to the Early Miocene. Terebralia bidentata (De- france in Grateloup, 1832), from that epoch, differs little in shell morphology from Recent Terebralia species. Cossmann (1906:126) cited the genus from the Late Cretaceous (Maastrichtian), but the fossils from that pe- riod are at best equivocal, and it is unlikely that they belong to Terebralia as now under- stood. The genus was common in the Tethys Sea, from which it spread into the tropical At- lantic and Indo-Pacific Oceans. According to Woodring (1959:178), Terebralia species were widely distributed in Tertiary seas until about the end of the Miocene, but are today restricted to the tropical Indo-West-Pacific Oceans. There are several well-authenticated American Miocene species from the Carib- bean region, such as Terebralia dentilabris (Gabb, 1873) (see Hoerle, 1972:20-21, pl. 1, figs. 9-11). Although many fossil species have been correctly attributed to Terebralia, numerous others must be removed from this genus. For example, K. Martin (1916) de- scribed seven species from the Upper Mi- ocene of Java, none of which should be allo- cated to this genus, inasmuch as they are more like cerithiids or batillariids. A large number of fossils that appear to be true Te- rebralia species were described from the Late Miocene of Piemont, Italy, by Sacco (1895), and numerous typical Terebralia taxa have been described from Neogene formations of Southeast Asia. The numerous described Tertiary species have not been critically re- viewed. Of the three living species, Terebralia palustris and Terebralia sulcata have fossil records going back to the Miocene (see spe- cies accounts in this paper for details). In the older literature, Terebralia species are sometimes assigned to the genus Pota- mides Brongniart, 1810, or Pyrazus Montfort, 1810. Terebralia was placed in synonymy with Pyrazus by Adams & Adams (1858:291). Tryon also (1882:250) considered Terebralia to be a synonym of Pyrazus, which he re- garded as a subgenus of Potamides. The ge- nus Potamides was proposed to accommo- date some Paris Basin fossils, but the shell of the type species of Potamides, Potamides la- marckii (Brongniart, 1831), does not in the least resemble those of living Terebralia spe- cies, and although related, is probably not congeneric with them. It is, however, very probably a member of the family Potamid- idae. The type species of the genus Pyrazus, P. baudini Montfort, 1810 (now known as P. ebininus [Bruguiere, 1792]), is herein consid- ered a member of the subfamily Batillariinae, based on its anatomy (Houbrick, pers. obsr.), and not closely related to Terebralia. The anatomical information on Terebralia species presented herein is based primarily upon detailed dissection and study of Tere- bralia sulcata. The other two species, Tere- REVIEW OF TEREBRALIA AND TELESCOPIUM 305 bralia palustris (the type species) and Tere- bralia semistriata, are not Known in such great detail. The order of presentation below begins with brief morphological descriptions of Tere- bralia palustris and Terebralia semistriata; this is followed by a more detailed description of Terebralia sulcata. Terebriala palustris (Linné, 1767) (Figs. 7-9) Strombus palustris Linne, 1758:213 (Seba figure [1765, pl. 50, figs. 13, 14] here se- lected to represent lectotype; Type local- ity: here restricted to Singapore); Dodge, 1956:291. Trochus trisulcatus Forskal, 1775:126 (Lecto- type: Zoological Museum of the Univer- sity of California, Berkeley, 114 x 50 mm; Type locality: Lohaia, Red Sea). Yaron et al., 1986. Strombus agnatus Gmelin, 1791:3521 (Seba figure [1765, pl. 50, fig. 19] here selected to represent lectotype; Type locality: not given). Cerithium palustre Bruguiére, 1792:486 (Seba figure [1765, pl. 50, figs. 13, 14] here selected to represent lectotype; Type locality: East Indies); Kiener, 1841: 81-82, pl. 1. Cerithium carnaticum Perry, 1811: pl. 35, fig. 3 (Perry figure [pl. 35, fig. 3] here se- lected to represent lectotype). ? Cerithium crassum Lamarck, 1822:294 (nomen dubium). Cerithium sulcatum Bruguiere [sic]. Kiener, 1841: pl. 27, figs. 1, 2, 2a (in part: is mix- ture of three species; only fig. 2a is Te- rebralia palustris). Pyrazus palustris (Linné). H. & A. Adams, 1854:291, pl. 30, figs. 8, 8a, 8b; Tap- parone Canefri, 1874:41. Cerithium (Potamides) palustre (Linné). Sow- erby, 1855:883, pl. 185, fig. 261. Potamides (Pyrazus) palustris (Linne). Mar- tens, 1880:281; Tryon, 1887:160, pl. 32, figs. 41, 42. Potamides (Terebralia,) palustris Bruguiere. Fischer, 1884:681, fig. 447; K. Martin, 1907:217; Oostingh, 1925:46-50. Potamides (Tympanotonus) palustris (Linné). Tryon, 1887:160, pl. 32, figs. 41, 42. Cerithium palustre (Linné). Kobelt, 1898:35, pl. 8, figs. 1, 2; pl. 9, figs. 1, 2. Clava caledonica Jousseaume, 1884:191, pl. 4, fig. 12 (Syntypes: MNHNP; Type local- ity: New Caledonia). Potamides Caledonicus (Jousseaume). Tryon, 1887:160, pl. 23, figs. 43, 44. Cerithium (Pyrazus) palustre (Linné). Kobelt, 1890:35-37, pl. 8, figs. 1, 2; pl. 9, figs. 1, 2: Potamides (palustre var. ?) tryoni Kobelt, 1895:169, pl. 32, fig. 1 (no type; refers to C. caledonicum Tryon, 1887, pl. 32, fig. 43). Cerithium (Pyrazus) caledonicum (Jous- seaume). Kobelt, 1895:169, pl. 32, fig. 8. Potamides palustris (Linné). Martens, 1897: 176; Tomlin, 1932:317. Terebralia palustris (Linné). Cox, 1927:84, pl. 18, fig. 4; Regteren Altena, 1941:15-16; Rensch, 1934:338; Benthem Jutting, 1959:106; Brandt, 1974:194—195, pl. 14, 105. 57, 58. Synonymic Remarks: This species is the type species of Terebralia, by subsequent designation of Sacco (1895). A lengthy his- tory of the synonymy of Strombus palustris has been published by Dodge (1956). Al- though it is not in the Linnaean Collection in London and was not described in the “Museum Ulricae”, Dodge (1956:291-292) pointed out that the identification of this spe- cies by the immediate followers of Linné must have been based on the recognizable figures of the palustris of authors in its synonymy. The figure references of Linné (1758) all show Strombus palustris with sufficient accuracy to define it, and Seba’s (1765) pl. 50, figs. 13, 14 are here selected to represent the lectotype. The type locality, “In Indiae paludibus,” is here restricted to Singapore. Yaron et al. (1986:187-188, fig. 30) have found and fig- ured the type of Trochus trisulcatus Forskal, identifying it as a junior synonym of Terebralia palustris. Although the types of Strombus ag- natus Gmelin and Cerithium palustre Bru- guiere are missing, they both share the same figure references with Strombus palustris Linné, and are obviously conspecific with it. The figure of Cerithium carnaticum Perry is sufficiently characteristic to identify it unequiv- ocally with Terebralia palustris. Lamarck (1822) did not supply his description of Cer- ithium crassum with figure references, but noted that: “It is related to the cerithium [sic] palustre but differs from it by its aperture which is very narrow, the right margin being much inverted.” His description is unclear and because there are no figures or types, the name should be regarded as a nomen du- bium. Although C. crassum appears in virtu- 306 HOUBRICK FIG. 7. Terebralia palustris, shell and operculum. A-D, adult shell from Zanzibar, Zanzibar; 103.1 x 36.9 mm (USNM 604453). E-H, adult shell from Pulau Panjo, Sumatra, Indonesia; 121.9 x 49.7 mm (USNM 661891). |, juvenile shell from Magnetic Island, Queensland, Australia; shell whitened with ammonium chloride to enhance sculptural details (USNM 828805). J, shell fragment with aperture removed to show columellar and palatal folds; 58.7 mm long (USNM 808336). K-N, adult shell from Anse Vata, New Cale- donia; 61.9 x 23.3 mm (USNM 835665). O, P, free and attached sides of operculum; 15.7 mm diameter. ally every synonymy of Terebralia palustris (see Dodge, 1956:291), it is cited in the above synonymy with a query. Clava caledonica, de- scribed from New Caledonia, is merely a dwarfed form of Terebralia palustris (see below). Kobelt's (1895) description of Cerith- ium tryoni was based upon one of the figures of Clava caledonica presented by Tryon (1887). Description Shell: Shell (Fig. 7) elongate, thick, solid and turreted, comprising as many as 20 flat-sided whorls and reaching a length of 190 mm. Pro- toconch unknown. Early whorls (Fig. 71) sculptured with strong, colabral axial ribs; spi- ral incised lines appearing on ninth or tenth whorl, gradually increasing to three in num- REVIEW OF TEREBRALIA AND TELESCOPIUM 307 ber. Adult whorls sculptured with four equal- sized, flattened spiral cords, three deep spiral grooves, and overlain by broad axial ribs, pro- ducing weak, square nodules. Varices broad, prominent and randomly distributed. Suture deeply impressed. Body whorl large, wide, with large varix opposite outer lip of aperture. Shell base moderately constricted, sculptured with many small spiral cords and numerous, very small, weak axial striae. Aperture ovate, grooved within; columella concave with thick callus and weak plait at anal canal; outer lip sinuous and flaring at anal canal (Fig. 7B, F, L), sweeping in broad arc to anterior canal, although not joining it. Anterior siphonal canal (Fig. 7D,H,N) short, tubular and nearly closed at junction of outer lip. Columellar pillar with two internal folds (Fig. 7J). Palatal tooth op- posite each external varix. Shell brown to blu- ish-black, occasionally with wide lighter bands; aperture glossy brown, columella light tan. Operculum (Fig. 70,P) corneous, circu- lar, multispiral with central nucleus, transpar- ent and tattered at edge. External Anatomy: Head-foot dark brown; snout and tips of cephalic tentacles black. Head with muscular, broad, transversely- lined snout, and with broad cephalic tenta- cles, each bearing large eye at peduncular base. Foot large, having white sole and large anterior mucous gland extending around an- terior half of sole periphery. Dorsal surface of foot (mesopodium) deeply grooved, conform- ing to heavy columellar plaits of shell. Ciliated groove on right side of foot in females, emerg- ing from anterior pallial oviduct and leading to very large, bulbous, white ovipositor near base of foot; ciliated groove at posterior end of ovipositor opening into large, cylindrical, jelly-producing chamber in inner right part of foot, beneath ovipositor. Inner ovipositor chamber filled with plug of tissue arising from posterior wall of chamber; chamber extending posteriorly, terminating under operculum. Mantle green; mantle edge bifurcate with scalloped outer fringe and inner row of 15-20 papillae. Inhalant siphon marked by indenta- tion. Papillae at inhalant siphon white-tipped. Inhalant siphon thick, muscular, with black undersurface at edge; inner surface of inhal- ant siphon with black-pigmented area sur- rounded by yellowish ring and with white, deep, semicircular light-sensitive pit, bor- dered with transverse ridge. Light-sensitive pit with white, villous epithelium comprising many small indentations and tiny pits, under- lain by large vacuolated cells over layer of darkly pigmented cells, and forming pigment cup. Sensory area of inhalant siphon inner- vated by extension of left mantle nerve. Ex- halant siphon black, bulging beyond mantle edge. Mantle Cavity: Mantle cavity very deep, oc- cupying last two whorls. Osphradium com- prising extremely thin, straight ridge about 0.5 mm wide, with microscopic, transverse inden- tations along edge, extending two-thirds of ctenidial length. Osphradium separated from ctenidium by broad epithelial membrane. Ctenidium very broad, shallow, comprising boomerang-shaped filaments, each about 6.5 mm long. Leading (left) edge of each filament with thin, vitreous rod-like structure and with many small transverse muscle fibers; right side of filament gradually becoming long and shallow; tip of leading edge with small inden- tation. Hypobranchial gland thick, broad, white, transversely ridged, secreting copious mucus. Intestine large, broad. Pallial gono- ducts open, comprising two parallel laminae attached to mantle roof. Alimentary System: Large buccal mass; rad- ula dark brown anteriorly, having long alary processes. Pair of large jaws (1.7 mm long) with scale-like cutting surfaces, nearly fused where upper lateral edges meet in oral aper- ture. Radula (Fig. 8) dimorphic owing to ontoge- netic change between juveniles and adults. Adult radular ribbon (Fig. 8C,D) robust, broad, brown at anterior end, about one-fourth of shell length. Rachidian tooth (Fig. 8D) rectan- gular, having nearly smooth cutting edge with weak central, dull point; rachidian tooth much broader than high with very narrow basal plate; rachidian tooth asymmetrically orien- tated at 45-degree angle to other teeth. Lat- eral tooth (Fig. 8C,D) large, broadly massive, having angulate, concave anterior, and broad basal plate with long, medial, ventral exten- sions and thin lateral extension; cutting edge smooth, sinuous, with very large inward- pointing cusp. Marginal teeth (Fig. 8C) small, with narrow stick-like shafts and spatulate tips; inner marginal tooth with one weak outer denticle on each side of tip and with narrow inner flange; outer marginal tooth with broad, outer flange on tooth shaft and lacking outer denticle. Juvenile radula (Fig. 8A,B) with oval, rect- angular rachidian tooth (Fig. 8B) having large central cusp flanked by three smaller denti- 308 HOUBRICK FIG. 8. Terebralia palustris, scanning electron micrographs of juvenile (A, B) and adult (C, D) radulae showing ontogenetic change in morphology. A, juvenile radula with marginal teeth spread open; bar = 40 um. В, detail of juvenile radula; bar = 30 рт. С, adult radula showing small narrow marginal teeth spread open; bar = 0.4 mm. D, adult radula, details of massive lateral and narrow rachidian teeth; bar = 0.3 mm. cles on each side. Lateral tooth (Fig. 8B) with broad basal plate having long lateral exten- sion and serrated with one large cusp, one inner denticle and two or three outer denti- cles. Marginal teeth (Fig. 8B) large, elongate, having hook-like tips serrated with three inner denticles, long central cusp and one outer denticle; outer marginal tooth without outer denticles. Reproductive System: Female pallial oviduct as in Terebralia sulcata, but with very narrow medial fusion along free edges of laminae comprising oviduct. Nervous System: Cerebral ganglia ovoid with short, thick commissure. Pleural ganglia closely connected to cerebrals, having long connectives to parietal ganglia; pedal ganglia having short connectives to cerebrals and with three well-developed pedal nerves. Zy- goneury between subesophageal and right pleural ganglia, and dialyneury between left pallial nerve and left pleural nerve. Long vis- ceral loop; visceral ganglion beneath floor of posterior mantle cavity. Remarks Shell: Terebralia palustris is by far the largest prosobranch found in the mangroves, and, for a potamidid, it can attain a truly remarkable shell size: Benthem Jutting (1956) cited a shell 160 mm long from Java, but an even greater giant of 190 mm in length has been REVIEW OF TEREBRALIA AND TELESCOPIUM 309 recorded from Arnhem Land, Australia (Loch, 1987:4). Loch (1987:4) hypothesized that gi- gantic specimens were the result of parasitic castration, suggesting that in animals with de- stroyed gonads, the energy normally directed to reproduction is diverted to growth. Terebra- lia palustris has a shell three to four times the size of Terebralia sulcata, and about twice the size of large Terebralia semistriata. The only other mangrove snail that approaches it in size is Telescopium telescopium. Aside from its large size, there is nothing unusual about the shell except that in longi- tudinal section there are two columellar plaits, a strong central one and a weaker parietal one. Opposite these, on the inner shell wall, there are palatal folds wherever an external varix has been formed (Fig. 7J). An excellent depiction of these folds in cut shells of Tere- bralia palustris and Terebralia sulcata has been presented by Martens (1897, pl. 9, figs. 24, 27). According to Tryon (1882:250), this character was discovered by Brot, and does not occur in other potamidids. Juvenile shells are quickly eroded and partly dissolved in the very acidic environment of the mangroves; shells with extant protoconchs were not found during this study. Clava caledonica, described by Jous- seaume (1884) from New Caledonia, is a dwarfed form of Terebralia palustris. Exami- nation of eight syntypes (MNHNP) and of many dwarfed specimens from New Cale- donia (USNM 724043, 724536) did not reveal significant differences from larger, more typi- cal specimens. Jousseaume (1884:192) dis- tinguished this taxon by its elongate shell, its reddish color, and the “form of its aperture.” Populations of dwarfed and elongate, narrow individuals occur throughout the range of Te- rebralia palustris, precluding subspecific rec- ognition of the New Caledonian populations of dwarfs. Normal-sized individuals also occur in New Caledonia and the reddish coloration of New Caledonian specimens is probably due to the presence of nickel in the mud. Brandt (1974:195) recorded the existence of two forms of Terebralia palustris throughout its range, which differed from each other in radu- lar morphology, the one having a cusped rachidian tooth, the other one having a smooth rachidian, and he doubted whether the two forms were conspecific. These two “forms” are merely the juveniles and adults of Tere- bralia palustris, and it is obvious that Brandt was unaware of the ontogenetic differences in radular morphology that exist in this species. Anatomy: An excellent illustration of a crawl- ing snail showing the head-foot and color of the living animal is in Kiener (1841, pl. 1). The gross anatomy of Terebralia palustris was briefly described by Starmühlner (1983:200— 203), and is basically the same as that of Te- rebralia sulcata, which is a much smaller snail. The large, muscular, darkly pigmented snout contains a massive buccal mass, dark- brown, robust radula, and a pair of large jaws, which allows adults effectively to consume fallen mangrove leaves. The pallial light-sensitive pit in the inhalant siphon was noted previously by Pflugfelder (1977:248-249), who referred to it as a sen- sory area. It probably functions as a pallial eye and is undoubtedly homologous with the similar organ in Telescopium telescopium and with the pallial eye of Cerithidea and Tympan- otonus species (Houbrick, 1984:10-11). The reproductive system of this species is very similar to that of Terebralia sulcata, but differs from that of Terebralia semistriata in having a more narrow medial closure of the laminae forming the open pallial oviduct. The ovipositor of Terebralia palustris is more pos- terior and closer to the operculum than in Te- rebralia sulcata (Fig. 15A, ovp). The nervous system has been well illus- trated by Starmühlner (1983: fig. 38), and both zygoneury and dialyneury are well es- tablished. The subesophageal ganglion com- prises two lobes, the configuration character- istic of cerithiid and potamidid taxa (Bouvier, 1887:131). The radula of Terebralia palustris differs considerably from those of its congeners. The small stick-like marginal teeth and massive lateral teeth are distinctive. One of the remarkable aspects of the on- togeny of Terebralia palustris is the trophic dimorphism between juveniles and adults and the correlative change in radular morphology between the two. Young snails are segre- gated spatially from adults and are deposit feeders, while adults eat only dead, decom- posing mangrove leaves. Juveniles have a radula (Figure 8A,B) similar to those de- scribed for some adult species of Cerithidea Swainson, 1840, subgenus Cerithideposilla Thiele, 1929 (see Houbrick, 1984). Sewell (1924:544) noted that adults of Cerithidea ob- tusa have radulae similar to those in the youngest stages of Terebralia palustris. The rachidian and lateral teeth have pointed cusps on their cutting edges (Fig. 8A,B), and the lateral teeth have long shafts and are 310 HOUBRICK spatulate and cusped at their tips. In adults, the buccal mass is very large, as are the jaws, and a large alary process anchors the radular ribbon to the odontophore. During radular transformation, the rachidian tooth becomes diminutive, compressed dorsoventrally, and asymmetrically placed at a 45-degree angle to the other teeth. The lateral tooth becomes dominant, massive and strongly pointed at its inner edge, and both the rachidian and lateral teeth lose their cusps. The marginal teeth re- main small, narrow and stick-like, and their curved tips lose some cusps. In short, the lat- eral teeth are the dominant instruments for dealing with mangrove leaves, while the rachidian and marginal teeth become small and vestigial (Fig. 8C,D). This transformation in radular morphology was first noted by Sewell (1924) and was studied further by An- nandale (1924), who gave a more detailed account of the process, and presented a fig- ure depicting the progression of changes in the radula from six snails of different ages (Annandale, 1924:550, fig. 11). Ontogenetic radular change appears to be limited to Tere- bralia palustris, for no such change has been observed in Terebralia sulcata, and Sewell (1924:544) did not detect radular differences in individuals of different ages of Telescopium or Cerithidea species. Reproductive Biology: Sewell (1924:542) noted that males appeared to be smaller than females of the same age in a population from the Nicobar Islands. Although spermato- phores were not found in the female gonoduct during the course of this study, it is assumed that they occur, and that they are similar to those described for Terebralia sulcata. The only account of spawning is that of Shokita et al. (1984:51-52), who studied this species in Thailand, where spawning occurs during the dry season, from December to January. They depicted spawning snails (p. 50, fig. 9) in ver- tical position, leaning against the mangrove prop roots. Egg capsules are deposited within gelatinous egg-strings that are attached to the roots of mangroves in the inner part of the mangroves. The use of the ovipositor in form- ing and depositing spawn masses was not discussed. Shokita et al. (1984) remarked that the same spawning habit has been ob- served in Okinawa, where the reproductive season lasts from May to October. Although the spawn masses of Terebralia palustris were not described in detail, photographs published by Shokita et al. (1984: fig. 9) show that they are nearly identical to the spawn of Terebralia sulcata, illustrated herein (Fig. 21D-F). A free-swimming larval stage in the development of Terebralia palustris was re- ported by Rao (1938), but details of the size and duration of the veliger stage and its set- tlement are unknown. Ecology: This large, conspicuous species oc- curs in great numbers in brackish water on coastal mudflats in mangrove regions, where it appears to prefer fine mud substrates (Wells, 1980:1-2). Ecological studies have been conducted in the Nicobar Islands (Sewell, 1924), the Andaman Islands (Rao, 1938), Java (Soemodihardjo & Kastoro, 1977), Okinawa (Nishihira, 1983), Thailand (Shokita et al., 1984), and in northwestern Australia (Wells, 1980, 1986). | observed a large population of Terebralia palustris in the mangroves of Magnetic Island, Queensland, Australia, where it is sympatric with Terebralia semistriata and Telescopium telescopium. The adult population was restricted to the higher, seaward portion of the mangrove for- est on a substrate of fine silty sand adjacent to a sandbar, and although very few individ- uals were found on open flats, they were abundant at tidal channels, where they ap- peared to have been washed out. Few young snails occurred in the open flats, but were common around mangrove roots. Segrega- tion of juveniles, which inhabit intertidal chan- nels and pools, from adults, which migrate into upper intertidal mangroves, was noted in New Caledonian populations by Plaziat (1984:122). In some areas, adult densities as high as 150/m* have been recorded (Plaziat, 1984:136). The change from juvenile to adult radula is undoubtedly a reflection of their change of diet. The apparent correlation between radu- lar morphology and diet is a trait | have noted in Rhinoclavis and Clavocerithium (Houbrick, 1986). This is in contrast to the situation in predatory thaidid gastropods (Thaidinae), in which Kool (1987) found that diet did not have a strong selective effect in the evolution of radular morphology. To date, the most com- prehensive work on trophic dimorphism and feeding in Terebralia is by Nishihira (1983). Trophic dimorphism has also been observed in populations from New Caledonia (Plaziat, 1977), and is probably characteristic of the species throughout its range. Terebralia palustris adults appear to be important in the in situ degradation of mangrove litter. Stom- REVIEW OF TEREBRALIA AND TELESCOPIUM 311 achs of dissected adults from Queensland, Australia, and Palawan, Philippines, were filled with small pieces of mangrove leaves. Plaziat (1977:37, fig. 32), Shokita et al. (1984: 51, fig. 10) and Nishihira (1983:52, 54-55, figs. 5, 6a, 6b) have depicted the grazed man- grove leaves eaten by adults. Nishihira (1983) showed that snails larger than 30 mm com- monly form feeding aggregations and graze directly on mangrove litter, including leaf, stip- ule, calyx, fruit and propagule. Large snails, including juveniles more than 30 mm long, can use both detritus and mangrove litter, while juveniles less than 30 mm in shell length eat detritus and do not graze on leaves. These observations correlate with the study of Rao (1938), who pointed out that juveniles preferred a muddy habitat to a rocky or sandy one. The oldest adults appear to occur on coarser-grained substrates than do subadults and juveniles (Shokita et al., 1984:42), which seem to prefer silt and mud. Budiman (1988: 240) pointed out that migratory behavior be- tween open mudflats and mangrove forests was related to change in diet, which, in turn, seems to be correlated with the morphologi- cal change of the radular teeth. The first worker to attempt to study growth in Terebralia palustris was Sewell (1924), who traced the growth of a population over a pe- riod of four years, in which shells attained an average length of 120 mm. Rao (1938) was unable to present a complete life history of a population he studied in the Andaman ls- lands, which consisted of mostly sexually im- mature individuals belonging to the growth stages of first- and second-year classes. Nev- ertheless, he showed that a free-swimming larval stage was present, for a large number of very young shells appeared in May, indicating a breeding season and release of larvae else- where in March to April. Shells were 5 mm long in the first year of growth, 18-20 mm long in the second year, and 40 mm in the third year. Growth under laboratory conditions was equally rapid. In Java, Soemodihardjo & Kastoro (1977) recorded that young individu- als gained an average additive length of 10 mm during a four-month period. Because no one has followed a population through to death, the lifespan of this species remains un- known. It appears that growth is determinate, because maturity is indicated by a thickening of the margins of the aperture, including the outer lip. Individuals more than 57 mm long had thickened lips in a Western Australian population examined by Wells (1980:2), but this feature probably varies throughout the geographic range, for populations of dwarfed adult individuals, less than 45 mm long, occur in New Caledonia (USNM 724043, 724536). Shokita et al. (1984) showed that two types of adult shells occurred in a population, a long type and a short type. This phenomenon does not indicate sexual dimorphism, but can be explained by physical factors: older shells are shorter owing to erosion of their upper whorls by the acidic environment of the mangrove forest floor. Terebralia palustris can tolerate consider- able environmental stresses. Rao (1938:203) showed that it can live without food for a con- siderable period (as long as four months), and Soemodihardjo & Kastoro (1977) found that it can live out of water and without food for as long as three months. As is the case with large adults of Tele- scopium telescopium, it is unlikely that many predators can successfully attack and eat adult individuals of Terebralia palustris, al- though the mud crab, Scylla serrata (Forskal, 1775), common in Indo-Pacific mangrove for- ests, is large and strong enough to crush the largest potamidids. Crushed Terebralia shells were observed around the mouths of burrows of these crabs in northern Australia (David Reid, pers. comm.). | did not see any empty shells with peeled apertures indicative of crab predation in Magnetic Island, Queensland, Australia. It is assumed that mortality due to predation occurs mainly in juveniles, but this has not been observed. Another predator of Terebralia palustris is man. Tryon (1882:250) remarked that“... in the Eastern Archipelago this species is assiduously collected by the natives, who roast them and suck the con- tents of the shell through an aperture made by breaking off the tip of the spire.” This species is also eaten by Australian aborigines in Arn- hem Land (Loch, 1987). Shells of Terebralia palustris often are en- crusted with oysters and/or barnacles in some areas, and the species frequently is corroded in the very acidic mangrove soils, which are rich in humic and fulvic acids. Scratches on the periostracum and borings by endolithic blue-green algae lead to in- creasing chemical dissolution of the shell, es- pecially the apex. Terebralia palustris reacts by abandoning damaged apical whorls and building successive menisciform septa. Pla- ziat (1984:132) noted that shell corrosion oc- curred in New Caledonia even at neutral pH values, indicating the primary role of boring 312 HOUBRICK FIG. 9. Geographic distribution of Terebralia palustris. algae in shell dissolution. Dissolution of the shell in older adults can result in the formation of dissolution bevels on the ventral shell sur- face, and an upper dissolution bevel can be formed by boring algae (see Plaziat, 1984: 135, fig. 22, for illustrations of this phenome- non). Post-mortem dissolution can reduce shells to their columellae. Fossil Record: Numerous fossils with wide geographic distributions occur in formations ranging in age from Early Miocene through Holocene. Sacco (1895:51, pl. 3, fig. 26) re- corded Terebralia palustris var. lineata Bor- son from the Late Miocene of Piemonte, Italy. His engraved figure of this fossil and the pho- tograph of the specimen later published by Mortara et al. (1985:199, pl. 36, fig. 7a,b) show that this fossil is indistinguishable from Recent specimens of Terebralia palustris. In East Africa, Terebralia palustris is listed from the Quaternary of Somalia (Regteren Altena, 1941:16; Wissema, 1947:48), Saudi Arabia (on raised beaches along the Red Sea, Regteren Altena, 1941:16; Smythe, 1982:46), Zanzibar (Cox, 1927:84, pl. 18, fig. 4), and Kenya (Regteren Altena, 1941:16). Crame (1986:191) found Late Pleistocene fossils in Kenya. Terebralia palustris also oc- curs in Pleistocene formations in Djibouti (Abrard, 1942:62, pl. 4, fig. 32; Wissema, 1947:48). This species is common as a fossil in Indonesia: K. Martin (1899:210-211, pl. 32, fig. 478) recorded it from the Miocene of Java, and Tesch (1920:57, pl. 131, figs. 183-184) recorded it from the Pliocene of Timor and also listed it from the Quaternary of the Celebes, the Pliocene of Nias and Java, and the Miocene of Java. Oostingh (1925:49—50) and Regteren Altena (1941:15-16) presented numerous fossil distributions, including the Pliocene of Timor and Java and the Pleis- tocene of New Guinea, Nias and the Celebes. This species has been recorded from the Phil- ippines by Tesch (1920:57), Oostingh (1925: 50), Regteren Altena (1941:16) and Wissema (1947:48). Fossil shells of Terebralia palustris are morphologically indistinguishable from Recent ones, indicating that this species has been in morphological stasis, with a fossil record going back to the Early Miocene. Distribution: This species has the widest range of any Terebralia species (Fig. 9). Inthe tropical Western Pacific, it occurs from the Ryukyus south through the Philippines, Bor- neo and New Guinea and throughout tropical Australia. It extends eastward to Palau and southeast to the New Hebrides (fide Oost- ingh, 1925) and to New Caledonia. Old REVIEW OF TEREBRALIA AND TELESCOPIUM 313 records citing the Gambier Islands in Polyne- sia (Oostingh, 1925:49) probably refer to Pseudovertagus clava (Gmelin, 1791), which resembles Terebralia palustris in size and su- perficially in sculpture. Terebralia palustris oc- curs in mangrove habitats throughout the In- donesian archipelago and in the estuaries of Southeast Asia, west to India and Ceylon. In the Indian Ocean, it is found in Nicobar and Andaman Islands, the Maldives, Mauritius, the Seychelles, the Amirantes and Madagas- car (Oostingh, 1925:49). Populations in east- ern Africa occur from South Africa northward to the Red Sea. Material Examined: RED SEA: Berenice, Egypt (ASNP 189208). ARABIA: Muscat, Oman (USNM 657262). EAST AFRICA: Dji- bouti, Djibouti (ANSP 194510); Mogadiscio, Somalia (USNM 673801); Malindi, Kenya (MCZ 106082); Mombassa, Kenya (USNM 707019); Kunduchi, Tanzania (USNM 703932; MCZ 271943); Bagamoyo, Tanzania (MCZ 261684); Kendawa Id, 4 m ESE of Dar es Salaam, Tanzania (MCZ 261683); Dar es Salaam, Tanzania (ANSP 156229; MCZ 109942); Mboa Maji, Tanzania (USNM 604600); Kijangiwani, S of Zanzibar, Zanzibar (USNM 604453; MCZ 280403); 1 mi N of Chukwani, W Zanzibar (MCZ 219010); Cape Inhambane, Mozambique (MCZ 234816); In- haca Id, near Lourenzo Marques, Mozam- bique (MCZ 201716); Bazaruto Id, Basaruto Bay, Mozambique (MCZ 234810); Durban Bay, South Africa (USNM 846524). INDIAN OCEAN ISLANDS: Ile Glorieuse (USNM 126217); Nosse Be, Madagascar (USNM 719443); Ambatozavavy, E Nossi Be, Mada- gascar (USNM 633349); Tulear (MCZ 261644); Grande Terre, Aldabra Atoll, Sey- chelles (USNM 836591); Seychelles (USNM 633323); Port Glaud, W Mahé, Seychelles (ANSP 311378); Curieuse Id, Seychelles (ANSP 298037); Mauritius (MCZ 1314); S half Kendikoln Id, Miladummadulu Atoll, Maldives (ANSP 305513); Port Blair, Andaman 14$ (USNM 609734); Ara Pt, Nilaveli, Ceylon (USNM 542240); Ceylon (USNM 90880, 18588; MCZ, unnumbered lot). THAILAND: Lem Sing (USNM 419160); Lem Ngob (USNM 384177); Welu River, near Ban Long Mai, Khlung, Chantaburi Prov. (MCZ 289125); Welu River, Trad Prov. (USNM 776701, 776703, 776702; MCZ 267508); Chantaburi River, Tachalaeb (USNM 776686); Klung Harbor, Chantaburi (USNM 776704). INDONESIA: Menscheneter ld, Java (ANSP 225485); Tjiperwagaran, Ban- tam, Java (USNM 260572); Kampong, Taber- fane, mouth Maikoor River, Aru, Moluccas (USNM 755620, 755615); Pruput, Java (USNM 260578); Pulau Siburu, N of Sipora, SW Sumatra (USNM 654762); Pulau Penju, SW Sumatra (USNM 661891); Mega, Men- tawii Ids, SW Sumatra (USNM 655078). BORNEO: W Marudu Bay, N Borneo (USNM 632189); Tanjong Aru, Jesselton, N Borneo (USNM 658481). NEW GUINEA: Milne Bay, Papua (USNM 543036). AUSTRALIA: Cockle Bay, Magnetic Id, Queensland (USNM 828832, 828805, 842986, 828801); East Arm, Darwin, Northern Territory (USNM 828803); Nightcliff Pt, Darwin, Northern Territory (USNM 602168); Mangrove Pt, Carnarvon, Western Australia (USNM 828811); Shark Pt, Barrow ld, Western Australia (USNM 691930); North West Pt, Bay of Rest, Western Australia (USNM 847078, 801600); Dampier Creek, Broome, Western Australia (USNM 828802). CAROLINE ISLANDS: Peleliu, Palau Ids (CMNH 25069). SOLOMON IS- LANDS: Pavuvu Id, Russell Gp (USNM 488321); Segi Pt, New Georgia (USNM 617791). NEW CALEDONIA: Tomo, Baie de St Vincent (USNM 725137); 3 mi N of Touho (USNM 631858); 3 mi E of Noumea (USNM 724043); St Marie, E side Noumea (USNM 724205); Pointe aux Long Cous, Noumea (USNM 724215); 2 km S of Conception (USNM 724536); near Port Laguerre (USNM 724778); E side Baie Boulare (USNM 664676); Dumbea River, Dumbea (USNM 724767, 724122, 724113); S of Ansi Vata (USNM 835665); San Gabriel (USNM 801591). PHILIPPINES: Pasacao, Luzon (USNM 240418); Sablayan, Mindoro (USNM 244073); E coast of Pollilo (USNM 311205); Catbalogau, Samar (USNM 243734); Tilig, Lubang Id (USNM 243681); Busuanga (USNM 244061); Jolo Id (USNM 243964); Port Busin, Burias (USNM 232927); Tara Id, Tapul Gp, Palawan (USNM 244069); Iwahig, 17 km W of Puerto Princesa, Palawan (FSM KA487); Mantagain Beach, Palawan (USNM 244068); Cape Melville, Balabac Id, Palawan (FSM KA438); 1 km SSE of Tapul, Polloc, Mindanao (USNM 244078); mouth of Mata- ling River, Malabang, Mindanao (USNM 243978); N of Mindanao River, Cobabato, Mindanao (USNM 244079); Zamboanga, Mindanao (USNM 243995); Port Tataan, Tataan, Tawi Tawi (USNM 243697). RYUKYUS: Okinawa (USNM 622035). 314 HOUBRICK Terebralia semistriata (Mörch, 1852) (Figs. 10-13) Strombus semistriatus Rodding, 1798:97 (Type not found, no figure reference, nomen nudum). Cerithium semistriatum Mórch, 1852:57. Cerithium semitrisculcatum Sowerby, 1855: 884, pl. 185, fig. 263 (Type: not located, Sowerby's fig. 263 selected to represent lectotype; Type locality: Port Essington, Queensland, Australia); Dautzenberg & Fischer, 1905:129. Pyrazus semitrisulcatum (Sowerby) Sowerby, 1865: pl. 1, fig. 4. Potamides (Terebralia) semitrisulcatus Morch. Tryon, 1887:160, pl. 32, fig. 45; Odhner, 1917:10. Pyrazus semisulcatus (Bolten). Dautzenberg & Fischer, 1905 (error for semitrisulcatus Sowerby, 1865). Synonymic Remarks: Strombus semistriatus Róding was introduced as distinct from Strombus mangos Roding (= Terebralia sul- cata [Born, 1778]), and is probably the entity now known as Terebralia semistriata (Mörch, 1852), a larger, smoother species than Tere- bralia sulcata. Roding (1798) cited no figure reference for Strombus semistriatus, but did refer to Gmelin’s Murex moluccanus, which is a synonym of Terebralia sulcata. Because there is no extant type and no figure refer- ence, Strombus semistriatus Roding must be regarded as anomen nudum. The figure cited for Cerithium semistriatum Mörch, 1852, clearly represents the larger, smoother spe- cies and Morch’s name is available, because Roding introduced the same name under Strombus. Sowerby (1855:899) listed Cerith- jum semistriatum Roding in his index, noting that it was “unidentified,” but did not include it in the monograph. Next in the index he listed Cerithium semitrisulcatum Sowerby, present- ing a full description and an excellent illustra- tion that is unequivocally conspecific with Cerithium semistriatum Morch. Thus, Sower- by's semitrisulcatum is a junior synonym of the larger, smoother species named semistri- atum by Mörch (1852). Dautzenberg & Fischer (1905:130), after examination of many specimens, remarked that although Kiener, Tryon and Sowerby maintained the separation of Terebralia semi- Striata (cited as semitrisulcata) from Terebra- lia sulcata, they were unable to find the line separating the two forms, and considered the former name to be a synonym. This study has shown that there is good conchological, rad- ular and anatomical evidence to accord spe- cific status to Terebralia semistriata. Description Shell: Shell large, solid, turreted , comprising 10 to 12 flat-sided to weakly-inflated whorls and reaching 57 mm in length (Fig. 10A- C,G,H). Protoconch unknown. Early whorls (Fig. 10H) sculptured with dominant axial ribs, but later developing spiral cords and becom- ing more cancellate. Adult shells (Fig. 10A-C) sculptured with broad, subsutural, flattened spiral cord and four or five smaller spiral cords diminishing in size abapically, and with weak axial ribs and incised lines, stronger in early whorls, becoming progressively weaker and narrower abapically, presenting overall smooth, tesselate appearance. Several va- rices randomly distributed. Suture distinct, slightly inset into each successive anterior whorl. Body whorl very large with strong varix opposite circular, flaring outer lip. Shell base (Fig. 10D) weakly constricted, sculptured with numerous spiral cords and weak, colabral, axial striae. Aperture large, circular-ovate, with concave columella having weak callus. Outer lip of aperture thick, smooth, circular, fusing with anterior canal to form complete peristome. Anterior canal centrally located, tubular, projecting through shell base. Anal canal weakly defined. Shell color dark- to light-brown. Aperture tan with shiny brown and whitish patches on columella and outer lip. Operculum (Fig. 10E,F) corneous, circu- lar, multispiral with central nucleus and ragged, transparent edge. Anatomy: Animal essentially same as other Terebralia species. Head-foot blackish, snout mostly black with some _ cream-colored blotches, upper foot gray, sole white. Tentac- ular peduncles thick and long; tentacle tips narrow, short. Inhalant siphon thick, fringed at edge; internally, with close-set papillae and with crescentic, lightly pigmented, slit-like, sunken sensory pit bordered with black pig- ment. Ridge-like osphradium very narrow, long. Ctenidium long, very wide, comprised of extremely shallow leaflets. Mantle tissue sup- porting ctenidium very thin. Rectum large, wide. Style sac extremely long, extending from stomach anteriory to pericardium. Kidney bright green. Female pallial oviduct with wide medial fusion. Male pallial gonoduct compris- ing two simple laminae. REVIEW OF TEREBRALIA AND TELESCOPIUM 315 FIG. 10. Terebralia semistriata, shell and operculum. A-D, adult shell from Darwin, Northern Territory, Australia; 56.7 x 25.8 mm (USNM 828831). E, F, operculum, showing free (E) and attached (F) sides; 11.1 mm diameter. G, early juvenile shell from Cockle Bay, Magnetic Island, Queensland, Australia, showing distinctive axial sculpture; 19.1 x 9.3 mm (USNM 828806). H, half-grown shell from Cape Bowling Green, Queensland, Australia, without varix on outer lip; 36.3 x 16.8 mm (USNM 622920). Alimentary System: Buccal mass relatively large. Pair of large jaws about 1 mm in length, comprised of scale-like plates attached to each other at anterior tips. Radula (Fig. 11A— D) moderately long, robust, brown anteriorly and with long alary processes. Radula with nine rows of teeth per mm. Rachidian tooth (Fig. 11C,D) rectangular-pentagonal with Straight anterior end and broad basal plate with weak, venirally located, central projec- tion; cutting edge with large, spade-shaped central cusp flanked on each side with one or two, rarely three, very small denticles. Lateral tooth (Fig. 11D) higher than wide, rectangular with basal plate having broad central pillar and pointed base; cutting edge with large, pointed, central cusp, one rounded inner den- ticle, and two small outer denticles. Marginal teeth (Fig. 11A,B) with long narrow shafts, flared, T-shaped tips. Inner marginal with long central cusp, two or three inner denticles, two outer denticles and narrow outer flange; outer marginal tooth same, but lacking outer denti- cles and with broad outer flange (Fig. 11B) on outer tooth shaft. Pair of tiny, tightly coiled, pinkish salivary glands anterior to nerve ring, not passing through it. Remarks Shell: The shell (Fig. 10A-D,G,H) of Tere- bralia semistriata closely resembles that of Terebralia sulcata, but is much larger and heavier, has weaker sutures, is weakly sculp- tured with flattened spiral cords and, except in juveniles, lacks axial ribs. The peristome is complete in adults. This species lacks the sculptural variability seen in the other two Te- rebralia species. Anatomy: The anatomy of this species (Fig. 12) was not studied in great detail, but the few dissections made show that it is nearly iden- tical to Terebralia sulcata, although twice the size of the latter. A distinguishing character is the wide medial fusion in the center of the open female pallial oviduct (me), which con- trasts sharply with the narrow medial fusion in Terebralia palustris and Terebralia sulcata 316 HOUBRICK FIG. 11. Radula of Terebralia semistriata. A, mid-radular ribbon with marginal teeth spread open; bar = 200 um. B, half row showing marginal teeth and flange on outer marginal tooth; bar = 80 um. C, detail of rachidian teeth; bar = 80 рт. D, detail of rachidian and basal plate of lateral teeth; bar = 40 um. (Fig. 18F, me). This wide medial fusion effec- tively closes the oviduct, forming a long me- dial oviductal passage (sections 2-3, op). The spermatophore bursa (sb) and seminal receptacle (sr) have a common opening. Reproductive Biology: Nothing is known about the reproductive biology, eggs, or lar- vae of this species. Ecology: This species was observed on the surface of open mudflats off the breakwater at Cairns, Queensland, and in the center of the mangroves at Magnetic Island, Queensland. In both places it was found on soft, sticky, cohesive mud. Fossil Record: Fossils have not been re- corded, and the geographic distribution is lim- ited. It is Known that the area of Australia in which it is common underwent significant ma- rine regressions in recent geological time, suggesting that Terebralia semistriata might have recently evolved. Distribution: Terebralia semistriata has the most limited distribution of any Terebralia species, and appears to be confined to trop- ical Australia and the southern coast of New Guinea (Fig. 13, stars). Although its develop- mental biology is unknown, its limited distri- bution suggests that this species might have direct development. Material Examined: AUSTRALIA: Bay of Rest, North West Cape, Western Australia (USNM 801606); Buccaneer Rock, Broome, Western Australia (USNM 631903); Broome, Western Australia (MCZ 265933); Dampier Creek, Broome, Western Australia (USNM 828829); Mangrove Point, Carnarvon, West- ern Australia (USNM 828814); Shark Bay, Western Australia (USNM 809759); Point Darwin, Darwin, Northern Territory (MCZ 100958); East Point, Darwin, Northern Terri- tory (MCZ 100958; AMNH 1013); Ludmilla Creek, 6 km N of Darwin, New Territory (USNM 828831); East Arm, 8 km ESE of Dar- win, New Territory (USNM 809767); Darwin Harbor, Darwin, New Territory (USNM 867709); Cooktown, Queensland (MCZ 265994); Cairns, Queensland (USNM 794875); Cape Bowling Green, Queensland (USNM 622920); Cockle Bay, Magnetic Id, REVIEW OF TEREBRALIA AND TELESCOPIUM 317 ovg FIG. 12. Diagrammatic figure of pallial oviduct of Terebralia semistriata; numbered arrows 1-5 indicate positions of transverse cuts through oviduct represented on left by numerals 1—5. ag, albumen gland; ant, anterior pallial oviduct; dsr, duct of seminal receptacle; Il, lateral lamina; me, medial fusion of pallial oviduct; ml, medial lamina; op, closed oviductal passage; osb, opening to spermatophore bursa; ov, oviduct; ovg, oviductal groove; sb, spermatophore bursa; sg, sperm groove; sr, seminal receptacle. 318 HOUBRICK FIG. 13. Geographic distribution of Terebralia semi- Striata (stars) and of Terebralia sulcata (shaded area). Queensland (USNM 828806); Townsville, Queensland (AMNH 171731); Port Curtis, Queensland (AMNH 14372); Mary River, Hervey Bay, Queensland (MCZ 104650); Long Beach, Kepple Bay, Queensland (MCZ 243490); Rockingham Bay, Queensland (MCZ 104649). NEW GUINEA: Maro River, Merauke, West Irian (MCZ 96915); Merauke, West Irian (MCZ 62358). Terebralia sulcata (Born, 1778) (Figs. 14-21) Murex sulcatus Born, 1778:324 (Holotype: Natural History Museum, Vienna, no. 5260; Type locality not given, here re- stricted to Ambon); 1780:320-321. Murex moluccanus Gmelin, 1791:3563 (Type not found: Lister's [1770] pl. 1021, fig. 85 selected to represent lectotype; Type lo- cality: Moluccas). Strombus mangos Róding, 1798:97 (Type not found: Lister's [1770] fig. 85 here se- lected to represent lectotype). Cerithium sulcatum Bruguiere [sic]. Kiener, 1841:89—90, pl. 27, figs. 1, 2 (in part, figs. 1, 2 only). Pyrazus sulcatus (Born). Tapparone Canefri, 1874:41. Cerithium (Pyrazus) semistriatus Mörch. Ko- belt, 1898:36, pl. 8, figs. 3, 4 (not semi- striata Mörch, 1852: is Terebralia sul- cata). Potamides (Terebralia) tenerrimus Schep- man, 1895:133-135, pl. 6 (Syntypes: [5]: ZMA 2.95.001, largest 10 x 6 mm; Type locality: Roti, Indonesia). Potamides (Terebralia) tenerrimus var. cos- tata Schepman, 1895:133 (Syntype: СМА 2.95.002, 19 x 6 mm; Type locality: Roti, Indonesia). Potamides sulcatus (Born). K. Martin, 1899: 211; 1911:21; Lischke, 1914:259; Ben- them Jutting, 1929:86. Potamides (Terebralia) sulcata (Born). Oos- tingh, 1925:50. Terebralia sulcata (Born). Oostingh, 1935:5; Regteren Altena, 1941:17; Benthem Jut- ting, 1956:442-443. Potamides (Terebralia) semitrisulcata (Bol- ten) Mórch. Odhner, 1917:10. Synonymic Remarks: Brauer (1878:170) has identified Born's original numbered specimen in the Vienna Museum, which is presumably the holotype of Terebralia sulcata. Born's (1780) references to the figures of Lister (1770:1021, fig. 85) and Martini & Chemnitz (1780, figs. 1484, 1485) clearly depict Tere- bralia sulcata of authors, although the Buo- nanni (1709) figure reference (fig. 68), which is listed first, depicts a short, strongly ribbed shell that only equivocally can be identified with it. Murex moluccanus of Gmelin (1791) and Strombus mangos of Röding (1798) are synonyms of Terebralia sulcata, because both authors referred to the same figure ref- erences as did Born (1778). Strombus semi- striatus was introduced by Róding (1798) as distinct from Strombus mangos Röding, 1798, but Gmelin’s (1791) moluccanus was also cited as a synonym of Strombus mangos. Be- cause no figure was cited for Strombus semi- Striatus, this taxon is herein considered a nomen nudum (see synonymic remarks un- der Terebralia semistriata). Potamides tener- rimus was described from a salt lake on Roti, Indonesia (Schepman, 1895). The syntypes and specimens from the type locality are small, black, slightly deformed and thin- shelled. The same is true for the varietal form named costata Schepman, 1895. These REVIEW OF TEREBRALIA AND TELESCOPIUM 319 FIG. 14. Terebralia sulcata, showing variable shell morphology. A-D, adult shell with strong axial sculpture from Okinawa; 46.9 x 19.5 mm (USNM 671076). E-H, adult shell with strong spiral sculpture from Tawi Tawi Islands, Philippines; 50.9 x 21.4 mm (USNM 233090). I, elongate, narrow adult shell from Hong Kong; 38.9 x 12.7 mm (USNM 858379). J, dwarf adult shell from Palawan, Philippines; 29 x 12.7 mm (USNM 808336). K, L, very inflated adult shell from Burias, Philippines; 45.9 x 21.5 mm (USNM 301727). M, juvenile shell from Masbate, Philippines; 21.2 x 12 mm (USNM 244022). specimens closely resemble small or imma- ture Terebralia sulcata, and are probably ecophenotypes or deformed individuals of this species. Specimens of Batillaria minima Mörch, 1852 and Cerithium lutosum Gmelin, 1791, from salt lakes in the Bahamas, have similar thin-walled, deformed shells thought to be caused by the atypical, hypersaline envi- ronment. The radula depicted by Schepman (1895, pl. 6) is merely an immature form ofthe radula of Terebralia sulcata. Description Shell: Shell (Fig. 14) moderately large, pen- dant-shaped, reaching 60 mm in length, and comprising about 12 weakly inflated to flat- sided whorls. Protoconch unknown. Early 320 HOUBRICK whorls (Fig. 14M) highly cancellate. Adult shells (Fig. 14A-L) with several wide, ran- domly placed varices, and sculptured with four or five flattened spiral cords and deeply incised spiral lines overlain by numerous axial ribs, forming overall sculpture of square nod- ules. Suture deeply incised. Body whorl wide, with expanded thickened outer lip. Aperture wide, ovate, slightly less than one-third the shell length, and with concave columella with broad columellar wash. Outer lip smooth to weakly crenulated, joining (fused to) base of columella just above short, centrally located, tubular siphonal canal (Fig. 14D,H). Body whorl sculptured with numerous beaded spi- ral cords. Shell dark brown, sometimes with lighter brown bands; varices whitish and beads sometimes light brown. Aperture shiny brown to cream. Operculum corneous, circu- lar, multispiral with central nucleus and ragged edge. External Anatomy: Animal (from Hong Kong) pigmented with yellow, dusky-brown blotches, flecked with bright yellow dots (Fig. 15A). Snout (sn) long, dark brown to black, with ir- idescent green, transverse stripes. Cephalic tentacles with broad peduncular bases and slender, long tips. Large black eye at anterior end of each peduncular base. Foot large, with deep groove on posterior propodium corre- sponding to columellar plait on shell. Sole fur- rowed with fine longitudinal folds. Opening to anterior pedal gland slit-like, extending poste- riorly along two-thirds of sole edge (Fig. 17B, amg). Females with deep ciliated groove (cg) leading from anterior pallial oviduct down right side of foot and around large, pad-like, bul- bous, cream-colored ovipositor (ovp) situated near medial foot edge. Small opening in ovi- positor leading into glandular chamber inside foot (cross-hatched area). Gonads located in upper visceral coils; ovaries bright green (eggs and spawn masses also green); testis orange- brown. Kidney brown, one whorl long, com- prising two lobes: large right lobe consisting of many fine lamellae; smaller left lobe with larger, coarser lamellae. Mantle skirt green, having bifurcate edge; outer (upper) edge scalloped; inner (lower) edge internally fringed with long, spade-shaped papillae having white tips (mp); ventral mantle edge smooth. Deep indentation at mantle edge adjacent to inhal- ant siphon (inh); exhalant siphon marked by minor indentation. Inhalant siphon thick, mus- cular, darkly pigmented along external edge, and with large, dark, inner papillae (Fig. 17C). в spg FIG. 15. Terebralia sulcata, anatomical features. A, right view of head-foot and mantle edge of female; crosshatched area represents internal ovipositor chamber. cg, ciliated groove; inh, inhalant siphon; mp, mantle papillae; op, operculum; ovp, ovipositor; pp, propodium; sn, snout. B, central nervous sys- tem; visceral ganglion and visceral loop not shown. cpc, cerebral pedal connective; d, dialyneury; Icg, left cerebral ganglion; In, labial nerves; Ipg, left pedal ganglion; Ipl, left pleural ganglion; pn, left pal- lial nerve; rcg, right cerebral ganglion; rpg, right pedal ganglion; rpl, right pleural ganglion; sbc, subesophageal connective; sbg, subesophageal ganglion; sec, supraesophageal connective; spg, supraesophageal ganglion; st, statocyst; tn, right tentacular nerve; z, zygoneury. Inner surface of inhalant siphon darkly pig- mented with large, semicircular, unpigmented sensory pit (Fig. 17C) innervated by a pallial nerve. Incross-section, sensory pit comprising thin layer of white tissue underlain by dark pigment (Fig. 17C). Sensory pit located sev- eral mm in front of osphradium and ctenidium. Mantle Cavity: Mantle cavity about two whorls long. Osphradium narrow, ridge-like with many small, tight folds along edge and sunken in trench-like pit. In cross-section, os- REVIEW OF TEREBRALIA AND TELESCOPIUM 321 phradium flanked on each side by tuft of cili- ated epithelium, innervated by large osphra- dial nerve located in center of osphradium base. Osphradium 3-4 mm behind mantle edge beginning about 1 mm behind anterior tip of ctenidium. Osphradial length about two- thirds ctenidial length, very narrow, separated from ctenidium by broad membrane. Ctenid- ium whitish-pink, long, extending length of mantle cavity, very wide and shallow, com- prising many thin, crescentic filaments; ctenidial filaments raised and triangular at left leading edge, becoming broad and shallow to right. Anterior ctenidial filaments more acutely triangular than broad, shallow, posterior ones. Tip of leading edge of filament directed to right, with small indentation. Ctenidial fila- ments each with numerous muscle bundles (strands), and having thin, vitreous rod-like el- ement supporting left leading edge. Hypo- branchial gland thick, wide, relatively thin anteriorly, comprised of transverse folds ar- ranged as shallow lamellae posteriorly. Hypo- branchial gland partly covering rectum, and secreting copious mucus. Rectum large, very wide, thin-walled, having internal epithelium with many transverse folds. Anus large, at an- terior end of rectum. Pallial gonoducts large, open, comprising two laminae with narrow, medial fusion in females. Alimentary System: Snout (Fig. 15A, sn) broad, bilobed at tip, thick-walled. Buccal hemocoel large, containing massive buccal mass attached to snout wall with numerous muscle bundles. Odontophore long, ovoid. Mouth located on ventral part of snout tip. Oral tube short, with pair of triangular chiti- nous jaws, each about as wide as width of anterior radula. Radular ribbon (Fig. 16) slightly more than one-fourth shell length, curving under buccal mass, and with pair of long narrow alary processes along anterior third of radular ribbon. Rachidian tooth (Fig. 16C) rectangular, forming semicircle, with narrow basal plate; cutting edge with long central cusp flanked on each side by two den- ticles. Lateral tooth higher than wide, rectan- gular with broad basal plate having broad central pillar and short lateral extension; cut- ting edge with large pointed major cusp and one to three outer denticles. Marginal teeth (Fig. 16A,B) with long narrow shafts and spat- ulate tips. Inner marginal tooth with broad central cusp, two or three inner denticles, two outer denticles and narrow outer flange. Outer marginal tooth with four cusps and broad outer flange. Salivary glands originat- ing immediately behind nerve ring, passing through it, comprising pair of tubes, looped along lateral sides of buccal mass, emptying into anterior part of oral cavity. Dorsal food groove in anterior esophagus large, twisting at nerve ring and becoming broad, shallow and ventral in midesophagus. Midesophagus (Fig. 18B, es) broad, dorso-ventrally flat- tened, and with interior epithelium consisting of numerous fine longitudinal folds and large dorsal fold. Stomach (Fig. 17A) large, occu- pying about one and one-half whorls of vis- ceral mass. Stomach with esophagus open- ing (eso) on right anterior side, adjacent to major typhlosole (t) and to large sorting area (sa). Large pad-like ridge arising from broad fold of stomach floor (rp) filling center and posterior of stomach, bordered along its pos- terior length by deep crescentic groove (cgr) and narrow ridge-like typhlosole (t) traversing its entire length, ending near gastric shield. Single opening to digestive gland (odg) lo- cated between typhlosole and ridge. Posterior part of stomach slightly constricted. Central- anterior part of stomach with large chitinous gastric shield (gs). Opening to intestine (int) located in anterior stomach adjacent to, but separate from, style sac opening (oss); style sac and intestine independent of each other. Style sac (ss) embedded in loose, spongy connective tissue anterior to stomach. Inter- nal epithelium of style sac heavily ciliated. Style sac and crystalline style very long, ex- tending from gastric shield, adjacent to kid- ney, as far anterior as pericardium, parallel to intestine in mid-region of the mantle cavity. Intestine long, looping back over anterior stomach before entering mantle cavity. Rec- tum (Fig. 18A-E, r) very wide, about one- fourth ctenidial width, having thick walls, and compressed into flattened tube; rectum hav- ing interior epithelium with thin ventral chan- nel bordered by one wall with numerous transverse, leaflet-like folds, and by another with wide thickened transverse ridges. Fecal pellets each compressed into long ovoid shape, stacked in large groups. Anus slightly detached from mantle wall, with broad open- ing. Reproductive System: Pallial oviduct (Fig. 18F) open along most of its length, but with narrow medial closure. Pallial oviduct consist- ing of two long, wide, thickened laminae: lat- eral lamina attached to mantle wall (Il), and free, medial lamina (ml). Laminae fused me- 322 HOUBRICK FIG. 16. Terebralia sulcata, scanning electron micrographs of radula from Hong Kong (USNM 858379). A, radular mid-section with marginal teeth spread open; Баг = 15 рт. В, mid-section of radular ribbon with marginal teeth partly opened showing basal plate of lateral tooth; bar = 20 рт. С, detail of half row showing rachidian teeth; bar = 10 pm. dially with thin sheet of mantle tissue binding together edges (me). Oviductal groove (ovg) at attached base of glandular portions of both laminae and bordered with many transverse ridges. Albumen gland (ag) white, opaque, beginning at posterior and extending anteri- orly to mid-pallial oviduct; darkly pigmented capsule gland (cg) anterior to albumen gland. Long sperm gutter (Figs. 18F, 19A, sg) lo- cated along edge of anterior medial lamina, opening into narrow bifurcate duct, leading into spermatophore bursa (Figs. 18F, 19D,E, sb) and seminal receptacle (Fig. 18F, sr). Sperm gutter highly ciliated, having many fine longitudinal folds. Spermatophore bursa (Fig. 18C-F, sb) large, elongate, lying in left, outer, posterior portion of non-glandular region of medial lamina. Spermatophore bursa interior epithelium bright yellow, ciliated, finely folded (Fig. 19D,E, sb), containing as many as three spermatophores. Seminal receptacle (Fig. 18C,D,F, sr) small, elongate, located in right inner wall of posterior medial lamina adjacent to spermatophore bursa. Seminal receptacle with large medial lobe having internal ciliated epithelium transversely folded. Oriented eupyrene sperm embedded in wall of seminal receptacle (Fig. 19E,F, sr). Non-glandular portion of lateral lamina (Fig. 18F, Il) consist- ing of thin sheet of tissue attached to floor of mantle cavity on right side of foot. Deep cili- ated groove (Fig. 15A, cg) emerging from an- REVIEW OF TEREBRALIA AND TELESCOPIUM 323 FIG. 17. Terebralia sulcata, anatomical features. A, diagrammatic representation of stomach; style sac folded back. cgr, crescentic groove; eso, opening of esophagus; gs, gastric shield; int, opening to intestine; odg, opening to digestive gland; oss, opening to style sac; rp, ridge-like pad; sa, sorting area; ss, style sac; t, typhlosole; 1, transverse cross-section through posterior stomach, shown below. B, sole of foot showing extent of opening to anterior mucous gland (amg). C, pallial light-sensitive organ at edge of inhalant siphon; arrow denotes plane of section through organ shown to left. terior pallial oviduct, running down right side of foot to white, bulbous ovipositor (Fig. 15A, ovp; 20A,B). Ovipositor with small opening (Fig. 20D,E) leading to narrow, interior, glan- dular chamber (Fig. 20F) and secreting milky, viscous fluid. Glandular tissue of ovipositor staining darkly in section. Ovipositor chamber a narrow slit, circular in cross-section (Fig. 20F,G), formed within thick glandular tissue, divided by longitudinal ridge emerging from anterior base of ovipositor on thin connecting ridge; epithelium of ovipositor chamber cili- ated. Ovipositor gland producing milky fluid, becoming viscous and swelling upon contact with water. Ovary green, producing bright green eggs about 200 um in diameter. Eggs surrounded by albumen and encapsulated, attaining di- ameter of 220 рт, arranged into long, twisted gelatinous strings in pallial oviduct, and emerging from it into ciliated groove (Fig. 15A, cg) on right side of foot, and moving down to large, glandular ovipositor near foot edge (Fig. 15A, ovp). Ovipositor secreting ge- latinous material surrounding egg capsules (Fig. 21F), forming elongate, thin, jelly tube, about 2.5 mm wide, encased within parch- ment-like Outer membrane, joined along its length by longitudinal suture (Fig. 21E). Elon- gate gelatinous tube containing egg capsules folded into loose coils forming twisted spawn mass about 40 mm long (Fig. 21D) and about 135 cm in length when unraveled. Spawn mass containing about 7,000 egg capsules, deposited on substrate, covered with adher- ing detrital particles and sand grains. Male pallial gonoduct open, comprised of two laminae; thin and pink anteriorly, thick 324 HOUBRICK FIG. 18. Terebralia sulcata, mantle cavity and pallial oviduct. A-E, transverse cross-sections, anterior to posterior, through mantle cavity showing relationship of oviduct to other organs. F, diagram of pallial oviduct; arrows show directions of ciliary currents, broken lines 1 & 2 and arrowheads show plane of transverse sections represented by sections 1 & 2. ant, anterior of pallial oviduct; ag, albumen gland; cg, capsule gland; cm, columellar muscle; ct, ctenidium; es, esophagus; hg, hypobranchial gland; k, kidney; ko, kidney opening; |, lateral lamina; me, medial fusion of oviduct; ml, medial lamina; os, osphradium; osb, opening to sper- matophore bursa; ov, coelomic oviduct; ovg, oviductal groove; pc, pericardium; r, rectum; sb, spermatophore bursa; Sg, sperm groove; sp, spermatophore; sr, seminal receptacle. and glandular posteriorly. Lateral lamina with wide, serpentine ridge along posterior-medial edge; ridge and thick glandular posterior gono- duct probably functioning as spermatophore- forming organ. Spermatophore (Fig. 21A-C) acellular, crescentic, fusiform, oval in cross-section, with external sculpture of many laminate ridges on one side (Fig. 21C). Spermatophore 0.5 mm wide, 5.4 mm long (n=3); opaque when filled with sperm, translucent when empty. Nervous System: Cerebral ganglia (Fig. 15B, Icg, rcg) ovoid, closely joined, and with short connectives to pleural ganglia (Ipl, rpl). Cere- bral ganglia each giving rise to three labial nerves and tentacular nerve (tn). Pedal gan- REVIEW OF TEREBRALIA AND TELESCOPIUM 325 FIG. 19. Terebralia sulcata, histological sections through pallial oviduct. A, mid-oviduct showing sperm gutter (sg), oviductal groove (og) and albumen gland (ag); bar = 0.5 mm. B, mid-oviduct showing opening to spermatophore bursa (osb), anterior spermatophore bursa (sb), duct to seminal receptacle (dsr), albumen gland (ag) and oviductal groove (og); bar = 0.5 mm. C, posterior seminal receptacle showing connective duct (cd) between spermatophore bursa (sb) and seminal receptacle (sr); bar = 0.5 mm. D, oriented sperm in seminal receptacle (sr) and connective duct (cd) to spermatophore bursa (sb); bar = 0.2 mm. E, section through bursa (sb) and seminal receptacle (sr) showing spermatophore (sp), in situ; bar = 0.5 mm. F, bursa containing disintegrating spermatophore and showing details of ridged epithelium lining bursa (wsb); bar = 0.5 mm. G, detail of epithelium separating seminal receptacle from bursa; note disintegrating spermatophore (dp), dispersed sperm (ds) in bursa and oriented sperm (os) in receptacle; bar = 0.2 mm. H, cross-section through disintegrating spermatophore (dp) showing keels (arrow heads); bar = 0.1 mm. 326 HOUBRICK RAT a A NS À OS FIG. 20. Terebralia sulcata, histological sections of head-foot. A—F, transverse sections through head-foot, anterior to posterior, showing external ovipositor gland (g) (A, B) and its interior chamber (F); note opening into gland (D) and formation of lumen (E, F); bars = 2 mm. G, section through critical point dried ovipositor gland showing lumen and plug-like longitudinal extension emerging from base of gland; bar = 1 mm. glia (Ipg, rpg) closely joined, each with basal pedal-pleural connectives thin, adjacent to statocyst (st); each ganglion giving rise to pedal-cerebral connectives. Subesophageal three nerves innervating foot. Pedal-cerebral ganglion with zygoneurous connection (z) to connectives (cpc) moderately short and thick; right pleural ganglion. Supraesophageal con- REVIEW OF TEREBRALIA AND TELESCOPIUM 327 FIG. 21. Terebralia sulcata. A, spermatophore (removed from bursa) in sea water; 5.4 mm long. B, C, critical point dried spermatophores, SEM; bars = 900 um and 300 um, respectively. D, coiled spawn mass; 30 mm long. E, F, detail of spawn mass showing outer opaque covering of jelly string and numerous egg capsules within; bar = 2 mm. nective (sec) long; dialyneurous connection (d) between supraesophageal nerve and left pallial nerve (pn). Subesophageal ganglion (sbg) with very short connective to left pleural ganglion. Long visceral loop; visceral gan- glion in floor of posterior mantle cavity. Remarks Shell: Of all the mangrove snails described herein, Terebralia sulcata has the thinnest shell, although it is by no means fragile. This shell is easily distinguished from that of Tere- bralia semistriata (Fig. 10), its morphologi- cally similar congener, by its smaller size, cancellate sculpture, and particularly by its prominent axial ribs. Shell size is very vari- able: some populations comprise only dwarfed individuals (Fig. 14J). Shell shape is also variable, especially between popula- tions: shells can be very squat and wide (Fig. 14K,L) or extremely tall and slender (Fig. 141). Shell sculpture is highly variable, especially in 328 HOUBRICK the number and prominence of axial ribs. In some phenotypes the spiral cords are flat and the shell appears nearly smooth, sculpture consisting of incised spiral and axial lines (Fig. 14A—C,!); other phenotypes have strongly developed spiral cords and axial ribs and appear very cancellate (Fig. 14J-L). Ju- venile snails have fine cancellate sculpture and deeply incised sutures (Fig. 14M). As in other Terebralia species, there are two plaits on the columellar pillar, extending up the en- tire shell, and opposite these, on the inner shell wall, there are teeth wherever an exter- nal varix has been formed. The shells of both Terebralia species are notable for the com- plete to nearly complete peristome, due to the fusion of the anterior outer lip to the anterior siphonal canal, and for the straight, short, tu- bular siphon, opening through the middle of the shell base (Fig. 14D,H). The complete peristome allows the animal to clamp down firmly on the substrate, and avoid desiccation and predators while maintaining communica- tion with the external environment through the tubular siphon. Anatomy: The darkly pigmented snout (Fig. 15A, sn) is one of the distinguishing features of this species. Although it can be consider- ably extended, the snout does not have the length or elasticity of the supple snout of Tele- scopium. The semicircular slit that forms a shallow pit on the underside of the inhalant siphon (Fig. 17C) is very similar to that of Terebralia palus- tris. Histological sections show that this si- phonal area is highly innervated, and although there is no evidence of a pallial eye as found in some other potamidids, this pit probably functions as a light-sensory organ. It is un- doubtedly homologous with the mantle eyes and light-sensory organs of other potamidids, for it is in the same place and innervated with the same nerves; it cannot compare in com- plexity with the well-developed pallial eye of Cerithidea and Tympanotonus species (Hou- brick, 1984, 1988). Prior to extending its head- foot from the shell, the animal extrudes the sensory pit of the mantle from the tube-like anterior canal of the shell (Fig. 14D,H), allow- ing it to detect shadows and movement. The osphradium resembles that of Tele- scopium in being sunken in a pit. A histolog- ical cross-section of this organ has been de- picted by Maeda (1986, pl. 1, fig. 2), who, in his survey of cerithioidean groups, desig- nated this kind of osphradium a Type-1 os- phradium (Maeda, 1986:35). The radula of Terebralia sulcata (Fig. 16A- C) most closely resembles that of Terebralia semistiata (Fig. 11A—D), but differs in having a narrower lateral lamella on the outer mar- ginals. Unlike the situation in Terebralia palustris, there is no ontogenetic change in radular morphology. The midesophagus is broad, and, as in Telescopium, is not devel- oped into an esophageal gland. A crop occurs in Cerithidea species (Houbrick, 1984), and a well-developed esophageal gland occurs in members of the family Cerithiidae (Houbrick, 1988). The raised pad on the ventral floor of the stomach (Fig. 17A, rp) is very large and unusual among cerithioideans. Ciliary move- ments in the stomach suggest that particles entering the stomach from the esophagus are moved in the crescentic groove (cgr) from the posterior stomach around the central pad to the opening to the digestive gland (odg), and onto the gastric shield (gs) and the rotating end of the crystalline style. It is here that amylase is thought to be secreted as part of the digestive process. The wide, thick-walled rectum, with its elaborately folded interior ep- ithelium, probably absorbs fluids from the fe- cal bolus, and molds and forms the fecal pel- lets. Reproductive Biology: Nothing has been re- corded in the literature about the reproductive biology, spawn or development of this spe- cies; thus a study of reproductively mature animals was undertaken in Hong Kong. The large cream-colored, pad-like ovipositor usu- ally bulges from the surface of the anterior right side of the foot in ripe females, and is much like the ovipositor of Terebralia palus- tris. The small opening at the posterior of the ovipositor (Fig. 20D,E) extends well into the foot, in which it forms a cylindrical chamber (Fig. 20F) that appears to produce a gelati- nous substance probably applied during final formation and deposition of the egg mass. The chamber has a longitudinal plug emerg- ing from its posterior wall and filling the lumen so that the cavity itself is narrow (Fig. 20F,G). In section, the extensive glandular part of the ovipositor consists of spongy-looking cells with small nuclei (Fig. 20C-F, g). The epithe- lium of the chamber consists of a layer of elongate cells having long, darkly staining nu- clei, and is ciliated, thus indicating an ecto- dermal origin. The ovipositor and chamber are similar in origin and placement, and prob- REVIEW OF TEREBRALIA AND TELESCOPIUM 329 ably homologous with similar structures de- scribed in other cerithioidean taxa (Houbrick, 1988:98, 101), but in Terebralia, the chamber does not function as a brood pouch. The ovi- positor and chamber of 7. sulcata are mor- phologically very similar to those described in Diastoma melanioides (Reeve, 1894), of which the chamber might or might not be a brood pouch. Spawn masses deposited by Terebralia sul- cata in Hong Kong during March occurred on the roots of mangroves, and were cryptic, be- ing covered by detrital particles (Yipp, pers. comm.). If the detrital particles are scraped from the jelly strings, the developing embryos appear bright green, matching the color of the ovaries of ripe female snails. Although the de- velopmental biology of Terebralia sulcata is unknown, its egg capsules are about equal in size to those of Telescopium telescopium, which is known to have free-swimming larvae; thus, it can be reasonably assumed that Te- rebralia sulcata also has free-swimming lar- vae. Terebralia sulcata is unusual in having an ornate, complexly sculptured spermatophore (Fig. 21A—C). The numerous parallel trans- verse ridges on the spermatophore appar- ently anchor it in the bursa (Fig. 19E,F). This kind of spermatophore sculpture may prove to be distinctive of the genus. Within a sper- matophore, a narrow elongate chamber holds both euspermatozoa and paraspermatozoa and other acellular elements of unknown composition. Longitudinal sections through a disintegrating spermatophore, in situ in the spermatophore bursa (Fig. 19E—H), show that the walls of the spermatophore are compart- mentalized and composed of a spongy, acel- lular, chitinous substance. The sections indi- cate that the sperm are liberated into the bursa as the walls of the spermatophore dis- integrate. The paraspermatozoa are probably resorbed in the bursa, for only euspermato- zoa are found in the seminal receptacle. The nervous system of Terebralia sulcata (Fig. 15B) differs little from that of Telesco- pium, although the supraesophageal connec- tive and left pallial nerve are shorter. Bouvier (1887:144—145, pl. 7, fig. 30) has described and illustrated in great detail the nervous sys- tem of Terebralia sulcata. Ecology: Terebralia sulcata is a hardy gener- alist, able to tolerate desiccation and a wide range of substrate types, and is able to ingest roughly equal portions of algae and vascular plants as well as large quantities of detritus and sand. In Hong Kong, Yipp (1980:705) identified four categories of plant materials, microalgae, filamentous algae, macroalgae and vascular plants, all of which underwent reduction on passage through the gut, with the possible exception of the filamentous al- gae. In contrast to Terebralia palustris, which is a much larger snail and which occurs on fine mud substrates, Terebralia sulcata prefers coarser substrates and attains its highest densities on them. | observed a population of this species in a stand of dwarf mangroves in Hong Kong, a habitat that has been thor- oughly described by Morton & Morton (1983: 222-223). This population, which also has been studied by Wells (1983) and Yipp (1980), occurs on intertidal sand and rocky habitats throughout the salt marsh and on the roots of the dwarf mangroves. Other Hong Kong populations of Terebralia sulcata occur in protected bays on similar substrates from which mangroves are absent (pers. obsr.; Wells, 1983:145). In contrast to the Hong Kong populations, Wells (1983:152) showed that this species is found only in mangroves in Western Australia and suggested that habitat segregation might differ in various regions. For instance, in the mangroves of the Bay of Rest, Western Australia, Wells (1980:2) found Terebralia sulcata was widely distributed throughout the seaward mangroves, Rhizo- phora stylosa and Avicennia marina, where it was common among the pneumatophores of the latter; however, in a mangrove forest in the Kimberly area, Western Australia, Tere- bralia sulcata was narrowly restricted to the floor of the Aegialitis zone (Wells & Slack- Smith, 1981). Wells (1986:88) remarked that of the mollusks living among Avicennia in the Bay of Rest, Western Australia, Terebralia sulcata dominated in terms of density and bio- mass, forming 50 percent of the total numbers and 85 percent of total biomass. Wells might not have discriminated Terebralia sulcata from Terebralia semistriata; consequently, his conclusions about habitat segregation might be erroneous and should be reconsidered. In Java, Benthem Jutting (1956:443) recorded this species living on mudflats, often attached to branches and roots of mangroves or on stones. It is likely that the microhabitat of this generalist species varies throughout its geo- graphic range. Little has been written about the predators of this species. Wells (1986:88) has sug- 330 HOUBRICK gested that in the Bay of Rest, Western Aus- tralia, Terebralia sulcata commonly lives among Avicennia pneumatophores for pro- tection from predatory rays, which are largely unable to feed among them. A small copepod lives in the mantle cavity of the Hong Kong populations (pers. obsr.). Fossil Record: Terebralia sulcata can be traced from the Late Miocene to the present and is well represented in the fossil records. Neogene records include Java, Sumatra, Nias, Timor, New Guinea and the Philippines (Regteren Altena, 1941:17). This species has been recorded as a Miocene fossil from Eni- wetok Atoll, Marshall Islands (Late Miocene; Ladd, 1972:27), from Java (Late Miocene; K. Martin, 1899:211; Wissema, 1947:48) and from the Philippines (Wissema, 1947:48). There are Pliocene records from Java (K. Mar- tin, 1899:211; Wissema, 1947:48), Sumatra (Wissema, 1947:48; Vlerk, 1931:25), Timor (K. Martin, 1899:211; Tesch, 1920:57, pl. 131, figs. 183,184; Vlerk, 1931:25; Wissema, 1947: 49), New Guinea (Wissema, 1947:49) and the Philippines (Wissema, 1947:49). Pleistocene records from Java were cited by Wissema (1947:49). There are also Holocene records from Nias and the Celebes (Wissema, 1947: 49) and from Taiwan (Regteren Altena, 1941: 17). During this study, fossils of Terebralia sul- cata from Niue were examined, indicating a previously wider geographic extension into the southern Pacific. Distribution: The easternmost extension of this species in the Pacific Ocean is in the west- ern Caroline Islands and in Guam (Roth, 1976: 8). Terebralia sulcata is common throughout the western Pacific from the Ryukyus south to Taiwan, China, Viet Nam, and throughout the Philippine archipelago (Fig. 13, shaded area). It also occurs in Borneo, New Guinea, and throughout tropical Australia. Terebralia sul- cata is common throughout the Indonesian Ar- chipelago, the Malayan peninsula, and in the estuaries and mangroves of Viet Nam, but ac- cording to Brandt (1974:195), has never been found alive in Thailand. Material Examined: MALAYSIA: Kranji, Sin- gapore (ANSP 239547; USNM 631935, 794078); Pulau Hanto, SW of Keppel Harbor, Singapore (USNM 660843). VIET NAM: Quiuhon (AMNH 86016). INDONESIA: Pulau Bai, Batu Gp, off Sumatra (USNM 654608); Pulau Siburu, N of Sipora, SW Sumatra (USNM 654700); Koeta Beach, Bali (USNM 617606); Limbe Id, Gulf of Tomini, Celebes (USNM 243938); Pasir Putih, Jailolo District, Halmahera, Moluccas (USNM 837034, 863460); Dorosago, Maba District, Halma- hera, Moluccas (USNM 837084); W side Mi- tak Id, Jamdena Strait, Tanimbar, Moluccas (USNM 747535); mouth of Maikoor River, Aru, Moluccas (USNM 755616). NEW GUINEA: Sowek, Soepoeri Ids, Schouten Ids, West Irian (ANSP 207871; USNM 835664); Ave Id, Geelvink Bay, West Irian (AMNH 128264). BORNEO: Tanjong Aru, Jesselton, N Borneo (USNM 658482); Port Essington, N Borneo (ANSP 225722); Po Bui Id, Sanda- kan, N Borneo (USNM 232866; AMNH 150829); Stanati, Kudat District, N Borneo (USNM 632194); N side Malawi Id, N Borneo (AMNH 106821); Taganak Id (USNM 243940). AUSTRALIA: Rowley Shoals, off Broome, Western Australia (USNM 847084); Bay of Rest, North West Cape, Western Aus- tralia (USNM 801606); creek, Darwin Harbor, New Territory (USNM 867710); Ludmilla Creek, 6 km N of Darwin, Northern Territory (USNM 828813); East Arm, 8 km of ESE of Darwin, Northern Territory (USNM 828812); Bickerton Id, Gulf of Carpentaria, New Terri- tory (USNM 602227); Thursday Id, Torres Strait, Queensland (USNM 613611, 603512; ANSP 242444). PHILIPPINES: Port Matalvi, Luzon (USNM 243616); Batangas,Luzon (USNM 233244); Port San Vicenti, Palaui Id,Luzon (USNM 232966); Kawit, Luzon (USNM 599751); Dumurug Id, Masbate (USNM 244022); Puerto Galero, Mindoro (USNM 777282); Alimango River, Burias (USNM 301727, 301726); Cebu, Cebu (USNM 419342); Siasi, Jolo (USNM 233234); Batag Id, Samar (USNM 472943); Puerto Princessa, Palawan (USNM 239799); Pala- wan (FSM 4168); Viejo Victorias, Negros (USNM 313250); Zamboanga, Mindanao (USNM 244035); Siminor Id, Tawi Tawi Gp (USNM 233090); Tataan, Simaluc Id, Tawi Tawi Gp (USNM 243699). CHINA: Tai Tam Harbor, Hong Kong (USNM 858379, 862676); Ting Kok, Tolo Harbor, New Territories, Hong Kong (USNM 858420). RYUKYUS: Okinawa (USNM 671076); Naha, Okinawa (USNM 632448); Orawan, Okinawa (USNM 593554); Saedake, Okinawa (AMNH 171733); Aha, Okinawa (AMNH 171732). CAROLINE IS- LANDS: Timil Harbor, Yap (USNM 485811); Yaptown, Yap (USNM 634251, 485850); Yap (USNM 634415); Garumisukan River, Kara- mando Bay, Babelthuap, Palau (USNM 620862; ANSP 200599); Babelthuap ld, REVIEW OF TEREBRALIA AND TELESCOPIUM 331 TABLE 1. Character comparison of Telescopium with Terebralia. Telescopium Shell shell aperture tangential peristome closed one columellar fold palatal teeth absent no varices on shell CUE CIN С > 5 atomy snout very long, supple snout tip with pad foot groove in both sexes ovipositor small 10. ovipositor lacking chamber 11. weak papillae on mantle edge 12. osphradium in deep trench 13. osphradium 1/4 gill length 14. osphradium next to ridge 15. radula & buccal mass small 16. oviduct open 17. oviductal groove complex хо чо Palau (USNM 631786); Koror, Palau (USNM 636196; AMNH 92748); Ngesias Village, Peleliu Id, Palau (USNM 656526); Arakitaoch Stream, Palau (USNM 656505); Ponape (AMNH 218504). DISCUSSION Comparisons of the morphologies, life his- tories and ecology of the species comprising the genera Telescopium and Terebralia and their relationships to other members of the Potamididae follow. The major characters dis- tinguishing the two genera are presented in Table 1. Juveniles of large mangrove-dwelling snail taxa, although superficially similar to one an- other, can be distinguished easily: Tele- scopium juveniles (Fig. 1H) are conical, sculptured with many spiral cords, and en- tirely lacking axial sculpture; Terebralia palus- tris juveniles (Fig. 71,J) are fusiform, have in- cised sutures, and are scuptured with wide, flat axial ribs; Terebralia sulcata juveniles (Fig. 14M) are fusiform, inflated, and have a deeply incised suture and a cancellate, beaded appearance; Terebralia semistriata juveniles have stocky, inflated shells with weak sutures, and in early stages have dom- inant axial sculpture (Fig. 10G), which later becomes more cancellate (Fig. 10H). Pallial siphonal eyes and light-sensory or- gans are common among members of both the Terebralia shell aperture normal peristome open two columellar folds palatal teeth present varices present snout short, robust snout tip without pad foot groove in females only ovipositor large ovipositor with chamber large papillae on mantle edge osphradium not in trench osphradium 2/3 gill length no ridge radula & buccal mass large oviduct with medial fusion oviductal groove simple Potamididae and other cerithioidean families. These pallial structures are thought to be ho- mologous because they all have the same lo- cation and innervation (Houbrick, 1984:10— 11). Most members of the family Potamididae appear to have pallial light-sensory organs lo- cated on the underside of the inhalant siphon. A deep sensory pit (presumably light-sensi- tive) in the inhalant siphon (Fig. 17C), occurs in Terebralia species. In Telescopium this structure is even more highly developed and contains a lens. A fully developed pallial eye with lens and cornea occurs in Cerithidea spe- cies (Houbrick, 1984:10-11) and in Tympan- otonus fuscatus (Linné, 1758) from West Af- rica (Johansson, 1956). In the family Batillariidae, pallial light-sensitive organs have been observed only in Pyrazus ebininus (Bru- guiere, 1792) (see Tenison-Woods, 1888: 175). Similar organs have also been found in species of Rhinoclavis Swainson (Houbrick, 1978) and of Gourmya (Houbrick, 1981b: 5-6), both of the family Cerithiidae Ferus-sac. The very long snout of Telescopium (Fig. 3B, sn) is comparable to the long extensible snouts seen in some Cerithidea species (Houbrick, pers. obsr.), but the radula is very small and weak. In contrast, the snout of Te- rebralia species (Fig. 15A, sn) is more robust, broader, somewhat shorter, and darkly pig- mented, and contains a long, robust radula. Both Telescopium and Terebralia species have propodial mucous glands with slit-like openings that are unusual in extending back 332 HOUBRICK to the middle of the foot (Figs. 3B, 17B, amg). This also occurs among members of the Planaxidae (Houbrick, 1987:445). In contrast to Terebralia species, the ovi- positor of Telescopium telescopium (Fig. 3A, ovp) is much smaller, located closer to the edge of the sole, and lacks an inner chamber. The posterior location of the ovipositor in Telescopium is unusual: in species of Cer- ithidea (see Houbrick, 1984) and in Terebra- lia, it is situated more medially on the foot (Fig. 15A, ovp). The very large ovipositor with interior chambers (Fig. 20) in female Terebra- lia species is an unusual structure, but not unique among cerithioideans, for a similar chamber in Diastoma has been found (Houbrick, 1981c:607—608). It is thought that secretions from these internal glandular chambers contribute to the coating of the spawn mass during deposition. The oviposi- tors with chambers in the foot of Terebralia species are homologous with, and a morpho- logical step toward, the cephalic brood pouches seen in planaxids, thiarids, siliquari- ids and fossarids (Houbrick, 1988:101). Cephalic brood pouches among cerithioide- ans appear to be modifications of the invagi- nated ectodermal ovipositor glands. Reduction of the ctenidium occurs among Potamididae, especially in the more highly am- phibious forms. The ctenidial filaments of Te- rebralia species are much reduced in height, and in Telescopium telescopium they have be- come aseries of fine ridges (Fig. 2F,G, ct). The extreme of this trend occurs among some members of the genus Cerithidea Swainson, 1840, in which ctenidia are so greatly reduced and degenerate that they are virtually absent in several species (Houbrick, 1984:11). Radular patterns of Potamididae members are all similar in that the rachidian tooth lacks basal cusps, the lateral extensions of the lat- eral tooth are of short to medium length, and the flanges of the inner and outer marginal teeth are of equal length. In contrast, batillari- ids have basal cusps on the rachidian, very long extensions on the lateral tooth, and the flange on the outer tooth is much shorter than that of the inner tooth. Perhaps the most striking internal anatom- ical feature of the large potamidids is the long style sac (Fig. 17A) and crystalline style, which protrude anteriorly as far as the poste- rior mantle cavity. Berkeley & Hoffman (1834: 436) were puzzled by the style in Tele- scopium, describing it as “. . . a cylindrical body consisting of a rather firm transparent jelly . . . ,” and suggested that it secreted something necessary for the eggs. Very long styles also occur among the Strombidae. The style of Telescopium is undoubtedly the long- est seen among potamidids, but in general form is similar to those of Terebralia species (this paper) and Cerithidea species (Hou- brick, 1984:7). All of the large potamidid taxa with very long crystalline styles also lack esophageal glands. Species of Cerithidea Swainson, 1840, also lack an esophageal gland, but have instead a dilated midesopha- gus, which presumably functions as a crop (Houbrick, 1984:5). Driscoll (1972:384) sug- gested a functional relationship between style length and the composition of ingested food, but this does not seem to be the case be- cause Telescopium ingests very fine particu- late matter, while Terebralia eats fallen man- grove leaves, and both taxa have long styles. The long style sac and the lack of an esoph- ageal gland are synapomorphies defining the family Potamididae (see Houbrick, 1988). The large raised pad dominating the floor of the stomach (Fig. 17A, rp) and the crescentic groove (cgr) bordering the pad are notable features of Telescopium and Terebralia spe- cies. Seshaiya (1932:174) pointed out that the pad (‘fleshy ridges”) in the stomach is the result of the typhlosole-like foldings of the ventral wall of the stomach. He noted that an examination of the stomachs of many style- bearing taxa does not lend support to the view that the crescentic groove on the stomach floor is a vestigial spiral caecum. Among cerithioideans, the pallial oviduct morphology of Telescopium is by far the most unusual and atypical. The complexity of the gutter system in the laminae (Fig. 5C,D) is equaled only by that of Modulus modulus (Houbrick, 1980:130, fig. 8). The numerous, thick transverse ridges on the glandular por- tions of the laminae bordering the oviductal groove (Fig. 5C,D, ovd) occur elsewhere only in turritellid pallial oviducts (Carrick, 1980: 245). The partial medial fusion (Figs. 12, 18F, me) of the pallial gonoducts of Terebralia spe- cies is unique among cerithioideans, but in other respects the gonoducts are typical of the superfamily. Complete fusion of the pallial oviducts occurs only in some thiarids such as Thiara Roding, 1798, Tarebia H. & A. Adams, 1854, and Melanoides Olivier, 1804. The spawn masses of members of both genera appear to be similar in morphology, and contain many small egg capsules typical of species having free-swimming larval stages. The wide distributions of species in both genera reflect good dispersal ability. REVIEW OF TEREBRALIA AND TELESCOPIUM 333 Nearly all cerithioideans are thought to pro- duce spermatophores, although few have been described in detail. The only cerithioid- ean spermatophore studied under SEM is that of Modulus modulus (Linne) (Houbrick, 1980: 127-129, figs. 6, 7), which is spindle-shaped, unsculptured and composed of a microscopic fibrous matrix. Spermatophores have been re- corded in numerous cerithioidean families (for summary, see Houbrick, 1988:111) and are usually simple, ovoid or crescentic structures having smooth surfaces and a few longitudinal keels. The spermatophore of Terebralia sul- cata (Fig. 21A-C) is by far the most elaborate observed among cerithioideans. The nervous systems (Figs. 3C, 15B) of all the large potamidids in this review are similar, differing only in minor details, such as the length of the supraesophageal connective and the length of labial nerves. Zygoneury and dialyneury are well established in both genera. The well-developed triple nerves em- anating from each pedal ganglion extend for- ward into the foot and innervate the ovipositor and glandular portions of the propodium. The major ganglia of the central nervous system are closely concentrated with the exception of the supra- and subesophageal and visceral ganglia, which are separated from the central nervous system by long connectives. The pot- amidid nervous system is thus unusual in be- ing simultaneously “close” and “loose.” All four of the potamidid taxa described in this study might occur in sympatry, but they usually differ in microhabitat. Terebralia palustris, the most widely distributed species, occurs with Terebralia sulcata and Tele- scopium telescopium in many parts of its range. Adult Telescopium telescopium and Terebralia palustris are frequently together on the mud, but have very different diets. Tere- bralia palustris prefers the shade of the man- grove canopy and is somewhat segregated from Telescopium telescopium, which ap- pears to prefer a more open habitat. In the Bay of Rest, Western Australia, Wells (1980: 2) found that Terebralia palustris was nar- rowly confined to the upper limit of the man- groves, whereas Terebralia sulcata was more widely distributed and occurred in the lower limit of the mangroves. In addition, the two congeners preferred different sediment grain sizes. Densities of Terebralia palustris were greatest in fine mud, while 7. sulcata had its greatest densities in coarser sediments. Wells (1980:4) pointed out that although Te- rebralia palustris occurs amongst mangroves of the genera Bruguiera, Ceriops and Avicen- nia, it avoids Rhizophora stylosa, and sug- gested that the Rhizophora stylosa sediments were probably too acidic for it. Although the two common Terebralia species might occur in sympatry, Wells (1980:1-2) pointed out that there are wide geographic gaps in the distributions of Terebralia sulcata and Tere- bralia palustris. As mentioned earlier, Wells might not have been aware of the existence of Terebralia semistriata, which is very common in the mangroves of this region, and might have lumped it with Terebralia sulcata. Tere- bralia semistriata occurs on soft mud; | have found it on mudflats of the Queensland coast devoid of mangroves. Terebralia is both the more species-rich and the older of the two genera, having a fos- sil record beginning in the Early Miocene. In terms of numbers of species, it appears to have reached its acme during the Middle to Late Tertiary, but now is represented by only three living species. In contrast, Telescopium, which can be traced back only as far as the Late Miocene, appears to have been a spe- cies-poor, highly apomorphic genus, repre- sented today by only one living species. In a previous phylogenetic analysis of fif- teen families of the superfamily Cerithioidea, | suggested that each of the subfamilies thought to comprise Potamididae, the Batil- lariinae and Potamidinae, be accorded full fa- milial status (Houbrick, 1988:114, 117). The morphological characters of Telescopium and Terebralia species described herein empha- size even more the great differences between these potamidid genera and the batillariids. It is therefore formally proposed that the Batil- laria group be excluded from the family Pota- mididae and be raised to familial rank as the Batillariidae. This action eliminates subfamil- ial categories from the Potamididae. The suggestion that the Cerithidea group be accorded familial status (as Cerithideidae; Houbrick, 1988:118) is herein revoked. Many shared anatomical characters such as those of the alimentary system, mantle cavity or- gans, and pallial eyes, unite Cerithidea with Telescopium and Terebralia in the Potamid- idae. Cerithidea differs from other potamidids in having two seminal receptacles, hardly rea- son for familial recognition. The exact phylogenetic relationship of Tele- scopium and Terebralia to the potamidid gen- era Cerithidea Swainson, 1840, Pirenella Gray, 1847, and Tympanotonus Schumacher, 1817, cannot be ascertained until anatomical study of the two latter genera is undertaken. The standing of Potamididae and Batillariidae 334 HOUBRICK in relation to other closely related families within Cerithioidea awaits a more formal phy- logenetic analysis of these groups in conjunc- tion with other unstudied cerithioidean fami- lies. ACKNOWLEDGMENTS Field studies for this research were sup- ported by the Smithsonian Secretary's Re- search Opportunity Fund. Field work in Aus- tralia was also supported by an Australian Museum Fellowship. Photography of speci- mens was done by Victor Krantz, Smithso- nian Photographic Services, and scanning electron micrography was provided by Su- sanne Braden of the Smithsonian Scanning Electron Microscope Laboratory. | thank the curators and collection managers of the fol- lowing institutions for providing me with spec- imens in their charge: Rüdiger Bieler (DMNH); Philippe Bouchet (MNHNP); George Davis (ANSP); William Emerson (AMNH); Sil- vard Kool (MCZ); lan Loche (AMS); John Tay- lor and Kathie Way (BMNH); Fred Wells (WAM). | am especially indebted to David Reid (BMNH) for logistic support and assis- tance at Townsville and Magnetic Island, Queensland, Australia. | also thank James Cook University for providing me with space at Townsville, Queensland, Australia. The manuscript was critically read by Kenneth Boss (MCZ), and the systematic portions by Rudiger Bieler (DMNH). LITERATURE CITED ABRARD, R., 1942, Mollusques pléistocènes de la côte française des Somalis recueillis par E. Au- bert de la Rue. Archives du Muséum National d'Histoire Naturelle, sixième série, 18:105 pp., 8 pls. ADAMS, H. & A. ADAMS, 1853-1858, The genera of Recent Mollusca; arranged according to their organization. London, J. von Vorst, 1(1-8): 1- 256 (1853); 1(9-15): 257—484 (1854); 2(19-24): 93-284 (1855). ALEXANDER, C. G. & J. C. RAE, 1974, The struc- ture and formation of the crystalline style of Tele- scopium telescopium (Linnaeus) (Gastropoda: Prosobranchia). The Veliger, 17(1):56-60. ALEXANDER, C. G., R. L. CUTLER & D. YELLOW- LESS, 1979, Studies on the composition and enzyme content of the crystalline style of Tele- scopium telescopium (L.) (Gastropoda). Compar- ative Biochemistry and Physiology, 64B:83-89. ALLAN, J., 1950, Australian shells. Melbourne, Georgian House, xix + 470 pp., 44 pls. ANNANDALE, N., 1924, Note on the radula of Pyrazus palustris. Records of the Indian Mu- seum, 26:549-551. ARCHIAC, A. D’ & M. J. HAIME, 1854, Description des animaux fossiles du Groupe Nummulitique de l'Inde precede d'un resume géologique et d'une monographie des nummulites. 2 vols. Paris, 373 pp., 36 pls. BASTEROT, B. DE, 1825, Description géologique du bassin tertiare du sud-ouest de la France. Premiere partie, comprenant les observations generales sur les mollusques fossiles (Descrip- tion des coquilles fossiles des environs de Bor- deaux). Mémoire de la Société d'Histoire Na- turelle de Paris, 2(1): 100 pp., 7 pls. BENSON, W. H., 1834, Note. Proceedings of the Zoological Society of London, Part 2, 1834:91— 92. BENTHEM JUTTING, W. $. $. VAN, 1929, A list of the land and freshwater Mollusca from Java. Treubia, 11:76-88. BENTHEM JUTTING, W. S. S. VAN, 1956, System- atic studies on the non-marine Mollusca of the Indo-Australian Archipelago. V. Critical revision of freshwater gastropods. Treubia, 23(2):259- 477. BENTHEM JUTTING, W. S. S. VAN, 1959, Cata- logue of the non-marine Mollusca of Sumatra and of its satellite islands. Beaufortia, 7(83):1-191, pis: 1, 2. BERKELEY, М. J. & С. H. HOFFMAN, 1834, A description of the anatomical structure of Cerith- ¡um telescopium, Brug. The Zoological Journal, 5:431-429, pls. 20, 21. BOETTGER, O., 1890, Ad. Strubell's Konchylien aus Java |. Bericht über die Senckenbergische Naturforschende Gesellschaft in Frankfort am Main, 1890:137-173, pls. 5, 6. BORN, |. VON, 1778, Testacea Musei Caesarei Vindobonensis. Vienna, J. P. Krauss, xxxvi + 442 + 17 pp., 18 pls. BORN, 1. VON, 1780, Testacea Musei Caesarei Vindobonensis, quae jussu Mariae Teresiae Au- gustae, disposuit et descripsit. Vienna, J. P. Krauss, xxxvi + 442 pp., 18 pls. BOSCH, D. & E. BOSCH, 1982, Seashells of Oman. New York, Longman, 193 pp. BOUVIER, E. L., 1887, Systeme nerveux, morpho- logie général et classification des gastéropodes prosobranches. Thesis préséntees a la Faculté des Sciences de Paris. Paris, G. Masson, 510 pp., 19 pls. BRANDT, R. A. M., 1974, The non-marine aquatic Mollusca of Thailand. Archiv fur Molluskenkunde, 105(1-4):1-423 pp., 30 pls. BRAUER, F., 1878, Bemerkungen über die im kai- serlich zoologischen Museum aufgefunden Orig- inal-Exemplare zu Ign. v. Born’s Testaceis Musei Caesarei Vindobonensis. Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften Mathematisch-Naturwissenschaftlichen Classe, 77(2):117-192. REVIEW OF TEREBRALIA AND TELESCOPIUM 335 BRONGNIART, M., 1810, Sur des terrains qui paraissent avoir été formes sous l'eau douce. Annales du Muséum d'Histoire Naturelle, Paris, 15:368. BRUGUIERE, J. G., 1789, 1792, Encyclopédie méthodique, histoire naturelle des vers. Paris, Panckoucke, 1 (1):1-344 (1789); (2):345-757 (1792). BUDIMAN, A., 1988, Some aspects on the ecology of mangrove whelk Telescopium telescopium (Linne, 1758) (Mollusca, Gastropoda: Potamid- idae). Treubia, 29(4):237-245. BUONANNI, P., 1709, Recreatio mentis et oculi, etc. Rome, Veresij, parts 1-3, text; pt. 4, pls. (not numbered). BUTOT, L. J. M., 1954, On Telescopium tele- scopium (Linne) and the description of a new species from P. Panaitan (Prinsen Island), Straits of Sunda. Basteria, 18(1-2):1-13, 2 pls. CARRICK, N., 1980, Aspects of the biology of Gazameda gunni (Reeve, 1849). Unpublished master’s thesis, University of Sydney. CERNCHORSKY, W. O., 1972, Marine shells of the Pacific. Sydney, Pacific Publications, 2:411 pp., 68 pls. CHAVAN, A., 1948, Nouveaux genres et sous- genres des mollusques. Compte rendu som- maire des Séances de la Société Géologique de France, no. 15-16:352-354. COLEMAN, N., 1981, What shell is that? Sydney, Lansdowne Press, 308 pp., illustrated. COSSMANN, M., 1906, Essais de paléocon- chologie compareé. Paris, F.R. de Rudival, 7:261 pp., 14 pls., illustrated. COX, L. R., 1927, Neogene and Quaternary Mol- lusca from the Zanzibar Protectorate. Zanzibar, Government of Zanzibar, 180 pp., 17 pls. CRAME, J. A., 1986, Late Pleistocene molluscan assemblages from the coral reefs of the Kenya coast. Coral Reefs, 4:183-—196. DAS, S., K. K. MISRA 4 K. C. GHOSE, 1988, Scan- ning electron microscope studies on radular teeth of three gastropod molluscs (Prosobranchia: Me- sogastropoda). Proceedings of the Zoological Society of Calcutta, 37:71-79, 3 pls. DAUTZENBERG, P. & H. FISCHER, 1905, Liste des mollusques récoltés par M. le Capitaine de Frégate Blaise au Tonkin, et description d’espe- ces nouvelles. Journal de Conchyliologie, 53(2): 85-234, pls. 3-5. DESHAYES, M. G. P., 1863, Catalogue des mol- lusques de l'Ile de la Reunion (Bourbon). Paris, Dentu, 144 pp., pls. 28-41. DODGE, H., 1956, A historical review of the mol- lusks of Linnaeus. Part 4. The genera Buccinum and Strombus of the class Gastropoda. Bulletin of the American Museum of Natural History, 111(3):157-310. DODGE, H., 1958, A historical review of the mol- lusks of Linnaeus. Part 6. The genus Trochus of the class Gastropoda. Bulletin of the American Museum of Natural History, 116(2):157-221. DRISCOLL, A. L., 1972, Structure and function of the alimentary tract of Batillaria zonalis and Cer- ithidea californica, style-bearing mesogastro- pods. The Veliger, 14(4):375-386. FISCHER, P., 1884, Manuel de conchyliologie et de paléontologie conchyliologique. Paris, Librairie F. Savy, 1369 pp., 23 pls. FORSKAL, P., 1775, Vermes 1: Mollusca. in: De- scriptiones animalium: avium, amphibiorum, pis- cium, insectorum, vermium quae in itinere orien- tali observavit Petrus Forskál. Copenhagen, Molleri, 19(1): xxxiv + 164 pp, 1 map. FRETTER, V. & A. GRAHAM, 1962, British proso- branch molluscs. London, Ray Society, 755 pp. GABB, W., 1873. Description of some new genera of Mollusca. Proceedings of the Academy of Nat- ural Sciences of Philadelphia. 24:270-274, pls. 9-11. GMELIN, J. F., 1791, Systema naturae, 13th edi- tion. Leipzig, J. B. Delamolliere, 1(6, Vermes): 3021-3909. GRATELOUP, M., 1832, Tableau (suite du) des co- quilles fossiles qu'on recontre dans les terrains calcaires tertiaires (faluns) des environs de Dax, département des Landes. Actes de la Société Linnéenne de Bordeaux, 5(29):261-282. GRAY, J. E., 1847, A list of the genera of Recent Mollusca, their synonyma and types. Proceed- ings of the Zoological Society of London, 15: 129-219. HEALY, J. M., 1983, Ultrastructure of euspermato- zoa of cerithiacean gastropods (Prosobranchia: Mesogastropoda). Journal of Morphology, 178: 57-75. HEALY, J. M., 1986, Ultrastructure of parasperma- tozoa of cerithiacean gastropods (Prosobran- chia: Mesogastropoda) Helgolander Meere- suntersuchungen, 40:177-199. HOERLE, S., 1972, Cerithiidae and Potamididae (Mollusca: Gastropoda) from the Chipola Forma- tion of northwestern Florida. Tulane Studies in Geology and Paleontology, 10(1):1-22, 2 pls. HOUBRICK, R. S., 1973, Studies on the reproduc- tive biology of the genus Cerithium (Gastropoda: Prosobranchia) in the Western Atlantic. Bulletin of Marine Science, 23:875—904. HOUBRICK, R. S., 1978, The family Cerithiidae in the Indo-Pacific. Part 1: the genera Rhinoclavis, Pseudovertagus and Clavocerithium. Mono- graphs of Marine Mollusca, No. 1: 130 pp., 98 pls. HOUBRICK, R. S., 1980, Observations on the anat- omy and life history of Modulus modulus (Proso- branchia: Modulidae). Malacologia, 20(1):117- 142. HOUBRICK, R. S., 1981a, Anatomy, biology and systematics of Campanile symbolicum with ref- erence to adaptive radiation of the Cerithiacea (Gastropoda: Prosobranchia). Malacologia, 21 (1-2):263-289. HOUBRICK, R. S., 1981b, Anatomy and systemat- ics of Gourmya gourmyi (Prosobranchia: Cerithi- idae), a Tethyan relict from the southwest Pacific. The Nautilus, 95(1):2-11. 336 HOUBRICK HOUBRICK, R. S., 1981c, Anatomy of Diastoma melanoides (Reeve, 1849) with remarks on the systematic position of the family Diastomatidae (Prosobranchia: Gastropoda). Proceedings of the Biological Society of Washington, 94(2):598— 621. HOUBRICK, R. S., 1984, Revision of higher taxa in genus Cerithidea (Mesogastropoda: Potamid- idae) based on comparative morphology and bi- ological data. American Malacological Bulletin, 2(1984):1-20. HOUBRICK, R. S., 1986, Discovery of a new living Cerithioclava species in the Caribbean (Mol- lusca: Prosobranchia: Cerithiidae). Proceedings of the Biological Society of Washington, 99(2): 257-260. HOUBRICK, R. S., 1987, Anatomy of Alaba and Litiopa (Prosobranchia: Litiopidae): systematic implications. The Nautilus, 101(1):9-18. HOUBRICK, В. S., 1988, Cerithioidean phylogeny. Malacological Review, Supplement, 4:88-128. HOUBRICK, R. S., 1989, Campanile revisited: im- plications for cerithioidean phylogeny. American Malacological Bulletin, 7(1):1-6. IREDALE, T., 1917, More molluscan name changes, generic and specific. Proceedings of the Malacological Society of London, 12:322- 330. JOHANSSON, J., 1956, On the anatomy of Tym- panotonus fuscatus (L.), including a survey ofthe open pallial oviducts of the Cerithiacea. Atlantide Report, 4:149-166. JOUSSEAUME, F., 1884, Description de mol- lusques nouveaux. Bulletin de la Société Zoolo- gique de France, 9:169-192, pl. 4. KANNO, S., 5. O'HARA & М. L. CAAGUSAN, 1982, Molluscan fauna from the Tartaro formation (Upper Miocene) of central Luzon, Philippines. Geology and Paleontology of Southeast Asia, 24: 51-128, Pls. 14-19. KASINATHAN, R. 8 A. SHANMUGAM, 1988, Over- exploitation of molluscan fauna in the Vellar es- tuary and Pitchavaram mangroves. Galaxea, 7: 303-306. KIENER, L. C., 1841-1842, Species général et ico- nographie des coquilles vivantes... genre cerite. Paris, Rousseau, 5, 104 pp., 32 pls. KOBELT, 2., 1888-1898, Die Gattung Cerithium. 297 pp., 47pls. in Е. Н. W. CHEMNITZ, Neues Systematisches Conchylien-Cabinet. . . . Nürn- berg, Bauer & Raspe, 1(26). KONINGSBERGER, J. C., 1915, Java zodlogisch en biologish. Dutch East Indies Department van Landbouw, Buitenzorg, Part 10: 663 pages. KOOL, S. P., 1987, Significance of radular charac- ters in reconstruction of thaidid phylogeny (Neo- gastropoda: Muricacea). The Nautilus, 101(3): 117-132. LADD, H., 1972, Cenozoic fossil mollusks from western Pacific islands, gastropods (Turritellidae through Strombidae). United States Geological Survey, Professional Paper 532: iv + 79 pp., 20 pls. LAMARCK, J. B. P. A., 1822, Histoire naturelle des animaux san vertèbres. . . . Paris, Verdiere, 7: 1-711. LASIAK, T. & A. H. DYE, 1986, Behavioural adap- tations of the mangrove whelk, Telescopium tele- scopium (L.), to life in a semi-terrestrial environ- ment. Journal of Molluscan Studies, 52: 174-179. LINNE, C., 1758, Systema naturae per regna tria naturae ... editio decima, reformata 1 (Regnum animale). Stockholm, Laurentius Salvus, 824 pp. LINNE, C., 1766-1767, Systema naturae per regna tria naturae . . . editio duodecima, reformata (Regnum animale). Stockholm, 1(2):533-1327. LISCHKE, M., 1914, Zur Molluskenfauna von Java und Celebes. Mitteilungen Naturhistorisches Mu- seum Hamburg, 31:203-284, pl. 1. LISTER, M., 1770, Historia sive synopsis metho- dicae conchyliorum et tabularum anatomicarum, Editio Altera. Oxford, Clarendoniano, 7 parts in one vol., 1081 pls. LOCH, 1., 1987, A tremendous Terebralia. Austra- lian Shell News, No. 58:4. MAEDA, T., 1986, Interrelationships of types of os- phradium, habitat and food habit of the Cerithio- idea and its near superfamilies (Mesogas- tropoda). Venus, 45:31-41. MARTENS, E. VON, 1880, Beiträge zur Meeres- fauna der Insel Mauritius und der Seychellen. Mollusken, pp. 181-352, pls. 19-22, in MOBIUS, Reise nach Mauritius. Berlin, O. Enslin. MARTENS, E. VON, 1897, Süss- und Brack- wasser-Mollusken des Indischen Archipels. Zo- ologische Ergebnisse einer Reise in Nederlan- dish Ost-Indien. Leiden, Brill, 4(1):1-355, 12 pls. MARTIN, K. 1884, Paläontologische Ergebnisse von Tiefbohrungen auf Java. Sammlungen des Geologisches Reichs-Museum in Leiden, 3(2-3): 43-184, pls. 1-9. MARTIN, K, 1889, Ein neues Telescopium und die Bezeihung dieser Gattung zu Nerinea. Paläonto- logische Ergebnisse von Tiefbonrungen auf Java. Sammlungen des Geologisches Reichs- Museums in Leiden, Ser. 1, 4(7):232-237, pl. 26. MARTIN, K., 1899, Die Fossilien von Java. 1, Gas- tropoda. Sammlungen des Geologisches Reichs- Museums in Leiden, neue Folge, 1(6-7): 133-221, pls. 31-45. MARTIN, K., 1907, Over Tertiaire en Kwartaire Vormingen van het Eiland Nias. Sammlungen des Geologisches Reichs-Museums in Leiden, 1. Beiträge zur Geologie Ost-Asiens und Aus- traliens, 8(3-4):204-252, pls. 14-18. MARTIN, K, 1911, Vorlaufiger Bericht über geolo- gische Forschungen auf Java. Sammlungen des Geologisches Reichs-Museum in Leiden, 1. Bei- träge zur Geologie Ost-Asiens und Australiens 9(1):108-200, pl. 9. MARTIN, K., 1914, Die Fauna des Obereocäns von Nanggulan, auf Java. A. Gastropoda. Sammlun- gen des Geologischen Reichs-Museums in Lei- den, neue Folge, 2(4):107-178, 6 pls. MARTIN, K., 1916, Die altmiocäne Fauna des REVIEW OF TEREBRALIA AND TELESCOPIUM 337 West-Progogebirges auf Java. Sammlungen des Geologischen Reichs-Museums in Leiden. neue Folge, 2(6):223-261, pls. 1-3. MARTINI, F. H. W. & J. H. CHEMNITZ, 1780, Neues Systematisches Conchylien-Cabinet . . . Nurenburg, Raspe, 4: 344 pp., illustrated. MONTFORT, D. DE, 1810, Conchyliologie syste- matique et classification methodique des co- quilles. Paris, F. Schoell, 2 vols.: 676 pp. MORCH, O. A. L., 1852, Catalogus conchyliorum quae reliquit D. Alphonso d'Aguirra 84 Gadea Comes de Yoldi. Part 1, Cephalophora. Hafniae, L. Kleine, 74 pp. MORTARA, E. F., L. MONTEFAMEGLIO, M. NO- VELLI, G. OPESSO, G. PAVIA & R. TAMPIERI, 1984, Cataloghi VII. Catalogo dei tipi e degli es- emplari figurati della collezione Bellardi e Sacco. Torino, Silvestrelli & Cappelletto, Part 2, 484 pp., 56 pls. MORTON, B. & J. MORTON, 1983, The sea shore ecology of Hong Kong. Hong Kong University Press, 350 pp., 28 pls. NISHIHIRA, M., 1983, Grazing of the mangrove lit- ters by Terebralia palustris (Gastropoda: Po- tamididae) in the Okinawan mangal: preliminary report. Galaxea, 2:45-58. ODHNER, N. H., 1917, Results of Dr. E. Mjobergs Swedish scientific expedition to Australia, 1910— 1913. Volume 17, Mollusca. Kungliga Svenska Vetenskapsakademiens Handlunger, 52(16):115 pp., 3 pls. OLIVIER, G. A., 1804, Voyage dans l'Empire Otho- man, l'Egypte et la Perse. Paris, H. Agasse, 3 vols. + atlas: xxxi + 595 pp. OOSTINGH, C. H., 1923, Recent shells from Java. Part 1—Gastropoda. Mededeelingen van de Landbouwhoogeschool te Wageningen (Neder- land), 26(3):1-174, figs. 1-15. OOSTINGH, C. H., 1925, Report on a collection of Recent shells from Obi and Halmahera (Moluc- cas). Mededeelingen van de Landbouwhooge- school te Wageningen (Nederland), 29(1):1-362. OOSTINGH, C. H., 1935, Die Mollusken des Pliozäns von Boemiajoe (Java). Wetenschappe- like Mededeelingen Dienst van den Mijnbouw in Nederlandsch-Indié, 26: viii + 247 pp., 17 pls. ORBIGNY, А. D. D’. 1834-47. Voyage dans l'Amé- rique meriodionale. Mollusques. Vol. 5 and atlas. xliii + 758 pp., 85 pls. PERRY, G., 1811, Conchology, or the natural his- tory of shells. London, W. Miller, unpaginated, 61 pls. PFLUGFELDER, O., 1977, Phaosomaugen. Mikro- kosmas, 66(8):246-249, 6 figs. PLAZIAT, J. C., 1977, Les cérithidés tropicaux et leur polymorphisme lié à l'écologie littorale des mangroves. Malacologica, 16(1):35—44. PLAZIAT, J. C., 1984, Mollusk distribution in the mangal. In: POR, D. & I. DOR, eds., Hydrobiol- ogy of the mangal: The ecosystem of the man- grove forests. Boston, W. Junk, x + 260 pp. POPENOE, W. P. & R. M. KLEINPELL, 1978, Age and stratigraphic significance for Lyellian corre- lation of the fauna of the Vigo Formation, Luzon, Philippines. Occasional Papers of the California Academy of Sciences, No. 129:1-73, 18 pls. PRASHAD, B., 1925, Respiration of gastropod mol- lusks. Proceedings of the Twelfth Indian Science Congress, July, 1925:126-143. QUOY, J. R. C. & J. P. GAIMARD, 1832-1835. Voyages de découverts de l’Astrolabe, . . . sous le commandement de M. J. Dumont d’Urville. Zoologie. vol. 2:1-320 (1832), 321-686 (1833); vol. 3:1-366 (1834), 367-954 (1835), 107 pls. RAJAGOPAL, А. S. 8 H. P. MOOKHERJEE, 1982, Contribution to the molluscan fauna of India. Part 2. Marine molluscs of the Coromandel coast, Palk Bay and Gulf of Mannar-Gastropoda: Meso- gastropoda (partim). Records of the Zoological Survey of India. Miscellaneous Publication, Oc- casional Paper No. 28: 1-53. RAMAMOORTHI К. & В. NATARAJAN, 1973, Spawning in Telescopium telescopium (Lin- naeus) (Potamididae-Gastropoda). Venus, 31(4): 158-159. RAO, H. S., 1938, Observations on the growth and habits of the gastropod mollusc, Pyrazus palus- tris (Linné) in the Andamans. Records of the In- dian Museum, 40(2):193-206. REEVE, L. A., 1894, Mesalia. Conchologia iconica. London, Reeve Bros., 5: 1 pg., 1 pl. REGTEREN ALTENA, C. O. VAN, 1941, The ma- rine Mollusca of the Kendeng Beds (East Java) Gastropoda, Part Il (families Planaxidae-Natici- dae inclusive) Leidsche Geologische Med- edeelingen, 12:1-86. RENSCH, B., 1934, Zur Kenntnis der Land- und Susswassermollusken Sumatras. Archiv für Mol- luskenkunde, 66(6):313-339, pl. 14. RODING, P. F., 1798, Museum Boltenianum: pars secunda continens conchylia. Hamburg, J. C. Trappii, 199 pp. ROTH, A., 1976, Preliminary checklist of the gas- tropods of Guam. University of Guam Marine Laboratory, Technical Report No. 27:99 pp. SACCO, F., 1895, Cerithiidae, Triforidae, Cerithiop- sidae, e Diastomidae. / Molluschi dei terreni Ter- ziarii del Piemonte e della Liguria. Turin, Stam- peri Reale, 17:83 pp., 3 pls. SCHEPMAN, M. M., 1895, A new Potamides (P. (Terebralia) tenerrimus) from Roti. Notes from the Leyden Museum, 16:133-135, pl. 6. SCHUMACHER, C. F., 1817, Essai d'une nouveau système des habitations des vers testacés. Copenhagen, Schultz, 287 pp., 22 pls. SEBA, A., 1765, Locupletissime rerum naturalium thesauri accurata descriptio . . . (Atlas). Copen- hagen, Wetstenium & Smith, 4 vols., 449 pls. SESHAIYA, M. A., 1932, The style sacs of some more gastropods. Records of the Indian Mu- seum, 34(2):171-175. SEWELL, S., 1924, Observations on growth in cer- tain mollusks and on changes correlated with growth in the radula of Pyrazus palustris. Records of the Indian Museum, 26:529-548. SHOKITA, S., S. LIMSAKUL & C. KARNJANAGE- 338 HOUBRICK SORN, 1983-1984, Distribution and abundance of the giant potamidid snail Terebralia palustris (Gastropoda) in the Thai mangal. Pp. 39-53, in Mangrove estuarine ecology in Thailand, Thai- Japanese cooperative research project on man- grove productivity and development, 1983-1984. SHUTO, T., 1978, Notes on Indonesian Tertiary and Quaternary gastropods mainly described by the late Professor К. Martin Il. Potamididae and Cerithiidae. Geology and Paleontology of South- east Asia, 19:113-160, pls. 15-18. SMYTHE, К. R., 1979, The marine mollusca of the United Arab Emirates, Arabian Gulf. Journal of Conchology, 30:57-80. SMYTHE, K. R., 1982, Seashells of the Arabian Gulf. London, Allen & Unwin, 123 pp., 20 pls. SOEMODIHARDJO, S. & W. KASTORO, 1977, Notes on the Terebralia palustris (Gastropoda) from the coral islands in the Jakarta Bay area. Marine Research in Indonesia, No. 18:131-148. SOWERBY, G. B., 1855, Monograph of the genus Cerithium, Adanson. Thesaurus conchyliorum. London, Sowerby, 2: 847-899, pls. 176-290. SOWERBY, G. B., 1865, Monograph of the genus Cerithium. Conchologia iconica. London, Reeve Bros., 15: unpaginated, 20 pls. STARMUHLNER, F. 1983, Results of the hydrobi- ological mission 1974 of the Zoological Institute of the University of Vienna. Part 8, Contributions to the knowledge of the freshwater gastropods of the Indian Ocean islands. Annalen des Naturhis- torischen Museum Wien, 84/B:127-249. SWAINSON, W., 1840, A treatise on malacology or shells and shell-fish. London, J. Taylor, viii + 419 pp. TAPPARONE CANEFRI, C., 1874, Zoologia del vi- aggio intorno al globo della Regia Magenta du- rante gli anni 1865-1868. Malacologia. Memorie della Reale Accademia delle Scienze di Torino, Serie 2, 28:1-161. TAPPARONE CANEFRI, C., 1883, Fauna malaco- logica della Nuova Guinea e delle isole adiacenti. Parte 1. Molluschi estramarini. Genova, Instituto Sordo-Mundi, 313 pp., 11 pls. TENISON-WOODS, J. E., 1888, On the anatomy and life history of Mollusca peculiar to Australia. Journal and Proceedings of the Royal Society of New South Wales for 1888, 22:106-187, pls. 3— 14. TESCH, P., 1920, Jungtertiäre und Quartäre Mol- lusken von Timor. Paläontologie von Timor, 8: 41-121, pls. 128-160. THIELE, J., 1929, Handbuch der Systematischen Weichtierkunde. Jena, Gustav Fischer, 1(1):376 рр. TOMLIN, J. В. LE B., 1932, Shells from а cave at Buang Bep, Surat, peninsular Siam. Journal of the Siam Society, Natural History Supplement, 8(4):315-317. TRYON, G. W., 1882, Structural and systematic conchology. Philadelphia, 1:453 pp., 140 pls. TRYON, G. W., 1887, Solariidae, lanthinidae, Tri- chotropidae, Scalariidae, Cerithiidae, Rissoidae, Littorinidae. Manual of Conchology, Philadelphia, 9:433 pp., 71 pls. VLERK, I. M. VAN DER, 1931, Pliocene of Sumatra and Timor. Onze Paleontologischen Kennis van Nederlandish Oost-Indi& in 1930—Cenozoic Am- phineura, Gastropoda, Lamellebranchiata, Sca- phopoda. Leidsche Geologische Mededeelingen, 5:206-296. WELLS, F. E., 1980, A comparative study of distri- butions of the mudwhelks Terebralia sulcata and T. palustris in a mangrove swamp in northwest- ern Australia. Malacological Review, 1980, 13: 1-5. WELLS, F. E., 1983, The Potamididae (Mollusca: Gastropoda) of Hong Kong, with an examination of habitat segregation in a small mangrove sys- tem. Pp. 139-154 in: MORTON, B. & D. DUD- GEON, eds., Proceedings of the Second Interna- tional Workshop on the Malacofauna of Hong Kong and Southern China. Hong Kong University Press. WELLS, F. E., 1986, Distribution of mollusks across a pneumatophore boundary in a small bay in northwestern Australia. Journal of Molluscan Studies (1986), 52:83-90. WELLS, F. E. & $. SLACK-SMITH, 1981. Zonation of molluscs in a mangrove swamp in the Kim- berly, Western Australia. Pp. 265-274, in Biolog- ical survey of Mitchell Plateau and Admiralty Gulf, Kimberly, Western Australia. Perth, West- ern Australian Museum. WENZ, W., 1938-1944, Gastropoda, Teil 1: Allge- meinen Teil und Prosobranchia. in: SCHINDE- WOLF, O. H., ed., Handbuch der Paläozoologie. Berlin, Gebrüder Bornträger, 6(4):721-960, Abb. 2084-2787 (1940). WISSEMA, G. G., 1947, Young Tertiary and Qua- ternary Gastropoda from the island of Nias (Malay Archipelago). Leiden, 212 pp., 6 pls. WOODRING, W. P., 1959, Geology and paleontol- ogy of the Canal Zone and adjoining parts of Panama. Description of Tertiary mollusks (gas- tropods: Vermetidae to Thaididae). United States Geological Survey, Professional Paper, 306- B:147-239, pls. 24-38. YARON, 1., Т. SCHIOTTE & G. WIUM-ANDER- SON, 1986, A review of molluscan taxa de- scribed by P. Forsskál and C. Niebuhr with cita- tion of original descriptions, discussion of type- material available and selection of some lectotypes. Steenstrupia, 12(10):157-203. YIPP, M. W., 1980, The distribution of ground- dwelling gastropods in a small mangrove stand in Hong Kong. Pp. 705-720, in: MORTON, B. & C. K. TSENG, eds., Proceedings of the First Inter- national Marine Biological Workshop: The marine flora and fauna of Hong Kong and southern China, Hong Kong, 1980. Hong Kong University Press. Revised Ms. accepted 23 October 1990 MALACOLOGIA, 1991, 33(1-2): 339-344 KARYOTYPIC EVOLUTION IN PLEUROCERID SNAILS |. PLEUROCERA, GONIOBASIS, AND JUGA Robert T. Dillon, Jr. Department of Biology, College of Charleston, Charleston, South Carolina 29424 U.S.A. ABSTRACT Rather little variation was noted among the karyotypes of three species of Pleurocera, seven species of Goniobasis, and one species of Juga. Most species have diploid numbers of 34, the remainder show 36. Pleurocera unciale seems to differ from Goniobasis simplex by but a single pericentric inversion, while the karyotypes of Juga hemphilli and Goniobasis livescens are indistinguishable. There is some evidence that the most ancestral karyotype belongs to one of these four species. Greater variation was noted among species within Goniobasis and Pleuro- cera, with numerous inversions and occasional translocations apparent. Most species of Go- niobasis seem to be equally related to several others in the genus, implying a uniform time of origin. Key words: snails, freshwater, Pleuroceridae, cytogenetics, chromosomes, karyotypes, evo- lution. INTRODUCTION In the first paper of this series, | pointed out that although freshwater snails of the family Pleuroceridae have figured prominently in many evolutionary and ecological studies, their systematics are poorly understood (Dil- lon, 1989). | reviewed data from several pre- vious studies (Patterson, 1969; Chambers, 1982; Dillon, 1982) suggesting that pleuro- cerid snails may be karyotypically variable and proposed a survey of karyotypic variation over all six genera. | then used flow cytometry to demonstrate that total genomic DNA is con- stant in pleurocerids, a result suggesting kary- otypic conservation, at least in regard to such phenomena as large-scale gene duplication. This would not preclude the occurrence of Robertsonian fusions or fissions, inversions, translocations, or any other such structural changes as might be useful in reconstructing the evolutionary history of a group of organ- isms. White (1973, 1978) has reviewed many very successful applications of cytogenetics to these purposes, not only in the polytene dipterans such as fruit flies, midges, and mos- quitos, but in other insects, plants, amphibi- ans, reptiles, and mammals. Among freshwa- ter snails, most attention has focused on the pulmonates (Patterson & Burch, 1978), espe- cially medically important planorbids (Gold- man et al., 1983, 1984). Recently, Nakamura & Ojima (1990) have added data on cellular DNA content to the already extensive infor- 339 mation on karyotypic evolution in Japanese Semisulcospira (Burch, 1968). Three of the six pleurocerid genera, Pleu- rocera, Goniobasis, and Juga, are character- ized by shell lengths much greater than shell width. Although few members of this family are characterized by preference for soft sub- strates, these three genera do not seem as restricted to rocky bottoms as do /o, Lithasia, and Leptoxis. Pleurocera, Goniobasis, and Juga would thus seem to constitute a natural subgroup of pleurocerids, and their karyotypic morphology is the subject of this investigation. METHODS The karyotypes of the following ten species are newly reported here: Goniobasis acuto- carinata (Lea) from southwestern Virginia, G. alabamensis (Lea) from Alabama, G. cate- naria dislocata (Reeve) from South Carolina, G. livescens (Menke) from Michigan, G. prox- ima (Say) from North Carolina, G. simplex (Say) from southwestern Virginia, Juga hemphilli (Henderson) from Oregon, Pleuro- cera acuta Rafinesque from Michigan, P. canaliculatum (Say) from Tennessee, and P. unciale (Reeve) from Tennessee. Photo- graphs and full locality data are given in Dillon (1989). In addition, | have included in this analysis the karyotype of Goniobasis floriden- sis (Reeve) as published by Chambers (1982), taken from an original photograph kindly provided by the author. 340 DILLON The primary challenge of this investigation was to obtain preparations with sufficient con- centrations of cells at mitotic metaphase. Strik- ing differences between individual snails were noticed in this regard, as well as short-term temporal and seasonal differences (Summer collections being best). Snails were generally held in their native water at environmental tem- perature and treated as soon as possible after collection. Preparations were generally made from a large number of males (20-40) and preliminarily screened to identify five from which karyotypic data could most easily be obtained. Snails were transferred to beakers of native water (usually about 10-20 individuals in 200 mi), with a pinpoint bubbler for aeration, and allowed to come to room temperature gradu- ally. Colchicine was added directly to the wa- ter at a concentration of 0.5 mg/ml (0.05%, w/v). Later in the study, it was discovered that the addition of a lectin, “pokeweed mitogen” (Sigma L-9379), effectively increased the mi- totic index at a concentration of 10 ywg/ml. Snails were removed from this solution at in- tervals from 12 to 24 hours, with about 18 hours usually providing best results. The cytogenetic techniques employed here are an original modification of standard clini- cal methods. Snails were washed, cracked with pliers, and females and parasitized indi- viduals discarded. The testes were dissected from healthy males and placed in about 1-2 ml of distilled water in a clear plastic culture tube. (Initial experiments showed that stan- dard hypotonic treatments, such as 0.075 M KCl, had little effect on these cells.) The tis- sue was disrupted by drawing the entire vol- ume in and out of a Pasteur pipette rapidly and repeatedly. The sample was then incu- bated at 37°C for 30 minutes. The sample was acidified with a drop of fixative and gently centrifuged (5 minutes at about 500 rpm on a large, swinging-bucket rotor). The supernatant was aspirated and the pellet resuspended in freshly prepared modi- fied Carnoy’s fixative (3 methanol:1 glacial acetic acid). To prevent clumping, the pellet was agitated with a pinpoint bubbler as the fixative was added dropwise. After 15 min- utes, the sample was recentrifuged and re- suspended in fresh fixative. It was found that cells fixed in this fashion could be held under refrigeration for at least two weeks. Clean slides were kept in distilled water at 3°C. Cells in fresh fixative were dropped onto these slides from a distance of about an arm's length, treated briefly with steam, and dried at 55°C. Then a syringe was used to layer fresh Wright's stain over each slide for 20-60 sec- onds. The stain was a dilution of 1 part Wright's stock (0.25% in methanol) to 4 parts phosphate buffer (0.025 M Na¿HPO,, 0.025 M KH,PO,, pH 6.86). Slides were washed in water, blown dry, dipped in xylene and mounted in Permount (Fisher). Early in this study, | experimented with a number of chromosome banding techniques, including the “АЗС” and “BSG” methods (Sumner et al., 1971; Sumner, 1972) and trypsin banding (Seabright, 1971). But the fairly lengthy colchicine treatment required to obtain a reasonable mitotic index had the un- desirable side-effect of yielding rather con- densed chromosomes. Thus banding was not reliably induced, and this approach was dis- continued. Idiograms were constructed in standard fashion, clipping and measuring photographs. The summed length of all chromosomes was held constant, given the findings of Dillon (1989). Chromosomes were assigned a qual- itative size category: “small” if representing less than 5% of the haploid genome, “medium” if 5-10%, and “large” if over 10%. Each chromosome was also assigned a cat- egory following Levan et al. (1964): metacen- trics showing an arm-length ratio of 1.0—1.7, submetacentrics 1.7-3.0, acrocentrics 3.0— 7.0, and telocentrics greater than 7.0. Chro- mosomes in typical preparations ranged only from about 2-6 um, rendering measurements of arm-length ratio approximate in many cases. For convenience | have considered any change in arm-length ratio as due to a peri- centric inversion, although certain transloca- tions could also account for a shift in centro- meric position. It should be emphasized that only a small proportion of all structural varia- tion is detectable by these techniques. Each karyotype was compared to that of all other species to identify similarities, and a hypoth- esis formed to account for observed variation assuming minimum rearrangement, as has become standard for studies of this sort (White, 1973, 1978). RESULTS AND DISCUSSION The example karyotypes from the three genera shown in Figure 1 are strikingly simi- lar. Goniobasis simplex and Pleurocera un- KARYOTYPIC EVOLUTION IN PLEUROCERIDS 341 Pleurocera unciale Sum we’ x ey Ving M КА ХХХ хх кххикавх и 3 > Se VK x И an s 84 ÑARX wx т. METZ № .^. “№ ABABA AA Bana Goniobasis simplex +.^ RU Rebates xa nuxttoux qe? % s Aa Un ax a a VU Y, Baba ла длал = Juga hemphilli e al „ХА ККХХ MX KKK aK 3” Х ху JD) Y Y sat ХУ ". BK mx xa E Fy, * ча + > + q a NADA AA ARARABAA FIG. 1. Exmaple karyotypes for three genera of pleurocerids. For each preparation, chromosomes are sorted by size in three categories: metacentric (M), submetacentric (S), and acrocentric (A). ciale both have diploid numbers of 34, and 36, differing from G. simplex only by an ap- seem to differ by as little as a single pericen- parent centric fusion/fission event involving a tric inversion in one of their medium-sized medium-sized metacentric and two small ac- chromosomes. Juga has a diploid number of rocentric chromosomes. 342 DILLON P. canaliculata P. acuta 2 21 у | DUDA magia P. unciale J. hemphilli G. livescens | G. floridensis o G. simplex 11,11 1£,11,11 G. catenaria dislocata ea era 1f, 11,11 4i G. proxima G. alabamensis El 1% Titre | ne FIG. 2. Idiograms for the 11 pleurocerid species treated in this study, together with their hypothesized relationships. Centromeric positions are approximate, especially in the smaller chromosomes. Ticks on the ordinate mark 5% and 10% of the haploid genome. Arrows show inversions (i) and stars show trans- locations (t) relative to the karyotype of Goniobasis simplex, the standard. Centric fusion/fission events are designated f. KARYOTYPIC EVOLUTION IN PLEUROCERIDS 343 Idiograms for all 11 species are displayed in Figure 2, linked together with karyotypically similar species. Karyotypes are characterized by chromosome size and centromeric posi- tion in Tables 1 and 2. Only Juga, G. lives- cens (karyotypically indistinguishable from Juga), G. catenaria dislocata, and G. floriden- sis (as previously reported) have 2N = 36. The remainder of the taxa show 2N = 34. The G. simplex karyotype was designated as archetypical or “standard,” because of its central position in Figure 2. (Coincidentally, this species also served as the standard in the isozyme study of Dillon & Davis, 1980.) Arrows locate apparent pericentric inversions and stars locate reciprocal translocations, rel- ative to G. simplex. Given previous reports of karyotypic varia- tion in the Pleuroceridae, this degree of con- servation was unexpected. Chambers (1982) reported that two populations of G. floridensis from the Florida panhandle differed from each other by at least one chromosomal rearrange- ment, and differed from the peninsular karyo- type shown in Figure 2 by a minimum of three rearrangements. Here it is reported that no greater difference is apparent between pleu- rocerid genera. But the degree of structural variation in karyotype is clearly underesti- mated by the techniques employed here. Chambers (1987) has argued that snails are generally not much more karyotypically con- servative than other animals, including the mammals. The apparently greater rates of chromosomal evolution that have been attrib- uted to mammals may be a function of their geological youth. Although the relationship among the Pleu- rocera species shown in Figure 2 is rather “treelike,” relationships among most of the Goniobasis species are unresolved. The karyotypes of G. acutocarinata, G. alabam- ensis, G. catenaria dislocata, and G. proxima each contain distinctive elements while re- maining as similar to G. simplex as to any other taxon. All five of these Goniobasis spe- cies could plausibly be related to at least two other taxa. This suggests both that these spe- cies of Goniobasis may have diverged at about the same time, and that some of the approximately 80 Goniobasis species unana- lyzed may have karyotypes intermediate be- tween these five. It should also be noted that implicit in Fig- ure 2 is the assumption that all chromosome rearrangements are equally diagnostic. But the pericentric inversions in particular are dif- TABLE 1. Goniobasis karyotypes, categorized by the criteria described in text. M—metacentric, SM—submetacentric, A—acrocentric, T—telocen- tric, Lg—large, Md—medium, Sm—small. Centromeric Position Ehrom. esse a er Species Size M SM Ares; G. acutocarinata Lg 1 1 Md 3 3 1 Sm 4 3 1 G. alabamensis Lg 1 1 Md 4 2 1 Sm 5 1 2 G. cat. dislocata Lg 1 2 Md 4 1 1 Sm 5 2 1 1 G. floridensis Lg 1 1 Md 4 1 2 Sm 4 2 2 1 G. livescens Lg 1 1 Md 2 1 3 Sm 5 1 4 G. proxima Lg 1 1 Md 3 1 3 Sm 3 3 2 G. simplex Lg 1 1 Md 4 1 2 Sm 4 1 3 TABLE 2. The karyotypes of Pleurocera and Juga, categorized by the criteria described in the text. M—metacentric, SM—submetacentric, A—acrocentric, Lg—large, Md—medium, Sm—small. Chrom Cent. Position Species Size M SM A P. acuta Lg 2 1 Md 1 2 2 Sm 3 3 3 P. canaliculata Lg 2 Md 2 2 3 Sm 3 2 3 P. unciale Lg 1 1 Md 3 2 2 Sm 4 1 3 J. hemphilli Lg 1 1 Md 2 1 3 Sm 5 1 4 ficult t0 score, and may not contain as much phylogenetic information as translocations or fusion/fission events. Thus one might reason- ably hypothesize, for example, that G. flori- 344 DILLON densis and G. catenaria dislocata share a common ancestry, by virtue of their shared chromosome number and size. If the currently accepted systematic rela- tionships among pleurocerid taxa accurately reflect their evolutionary history, one of the three karyotypes displayed in Figure 1 (P. un- ciale, G. simplex, or J. hemphilli/G. livescens) would seem to be ancestral. Assigning ances- tral status to any of the other eight karyotypes would imply polyphyletic genera. White (1978) has noted that the “overriding major- ity’ of cytogeneticists tend to assume that centric fusion is more common than fission, and hence that higher chromosome numbers are ancestral. But White offers evidence to the contrary from both the well-studied butter- flies and the Australian morabine grasshop- pers. Thus it would be premature to draw a “root” on Figure 2, or to speculate regarding the direction of karyotypic evolution implied, absent data on the other three genera in this family, Leptoxis, Lithasia, and lo. Such data will be forthcoming in the final paper of this series. ACKNOWLEDGMENTS | thank Dr. Gary Lamberti and Randy Wild- man for providing the Juga, Steve Ahlstedt for some Pleurocera, Dr. Steve Chambers for providing the G. floridensis karyotype, and John Wise and Robert T. Dillon, Sr., for help collecting many of the rest. Dr. Denise Smith provided technical assistance. Amy Wething- ton was a great help screening the slides, as were Kee Stewart and Josh Spruill. | am es- pecially grateful to Dr. Wayne Stanley of the Medical University of South Carolina, in whose laboratory | learned clinical cytoge- netic technique. LITERATURE CITED BURCH, J. B., 1968, Cytotaxonomy of Japanese Semisulcospira (Streptoneura: Pleuroceridae). Journal de Conchyliologie, 107: 1-51. CHAMBERS, S. M. 1982, Chromosomal evidence for parallel evolution of shell sculpture pattern in Goniobasis. Evolution, 36: 113-120. CHAMBERS, S. M., 1987, Rates of evolutionary change in chromosome numbers in snails and vertebrates. Evolution, 41: 166-175. DILLON, R. T., Jr., 1982, The correlates of diver- gence in isolated populations of the freshwater snail, Goniobasis proxima. Ph.D. dissertation, University of Pennsylvania, Philadelphia. 183 pp. DILLON, В. T., Jr., 1989, Karyotypic evolution in pleurocerid snails. |. Genomic DNA estimated by flow cytometry. Malacologia, 31: 197-203. DILLON, R. T., Jr. & G. M. DAVIS, 1980, The Go- niobasis of southern Virginia and northwestern North Carolina: Genetic and shell morphometric relationships. Malacologia, 20: 83-98. GOLDMAN, M. A., Р. Т. LOVERDE & С. L. CHRIS- MAN, 1983, Hybrid origin of polyploidy in fresh- water snails of the genus Bulinus (Mollusca: Planorbidae). Evolution, 37: 592-600. GOLDMAN, M. A., Р. T. LOVERDE, С. L. CHRIS- MAN & D. A. FRANKLIN 1984, Chromosomal ev- olution in planorbid snails of the genera Bulinus and Biomphalaria. Malacologia, 25: 427—446. LEVAN, A., K. FREDGA & A. A. SANDBERG, 1964, Nomenclature for centromere position in chromosomes. Hereditas, 52: 101-220. NAKAMURA, H. K. & Y. OJIMA, 1990, Cellular DNA contents of the freshwater snail genus Semisulcospira (Mesogastropoda: Pleurocer- idae) and some cytotaxonomical remarks. Amer- ican Malacological Bulletin, 7: 105-108. PATTERSON, C. M., 1969, Chromosomes of mol- luscs. Pp: 635-686, In Proceedings of the Sym- posium on Mollusca, Part II. Marine Biological Association of India. Bangalore, India. PATTERSON, C. M. & J. B. BURCH, 1978, Chro- mosomes of pulmonate molluscs. Pp. 171-217, In Pulmonates, Vol. 2A, Systematics, Evolution, and Ecology, V. FRETTER & J. PEAKE, eds. Academic Press, NY. SEABRIGHT, M., 1971, A rapid banding technique for human chromosomes. Lancet, 2: 971-972. SUMNER, A. T., 1972, A simple technique for dem- onstrating centromeric heterochromatin. Experi- mental Cell Research 75: 304-306. SUMNER, A. T., H. J. EVANS & R. A. BUCKLAND, 1971, New technique for distinguishing between human chromosomes. Nature New Biology 232: 31-32. WHITE, M. J. D., 1973, Animal cytology and evo- lution. 3rd ed. Cambridge University Press, Lon- don. 961 pp. WHITE, M. J. D., 1978, Modes of speciation. W. H. Freeman, San Francisco. 455 pp. Revised Ms. accepted 1 April 1991 MALACOLOGIA, 1991, 33(1-2): 345-351 INDEX Taxa in bold are new; page numbers in bold indicate pages on which new taxa are described; pages in italics indicate figured taxa. Acanthochitona 227, 228 acuta, Pleurocera 339, 342, 343 acutocarinata, Goniobasis 339, 342, 343 agnatus, Strombus 305 Akromesodon 163 alabamensis, Goniobasis 339, 342, 343 alexandrina, Biomphalaria 43-54 Allogona 75, 77, 122 Allogona profunda 76, 118, 120 altivagus, Mesodon 73, 75, 76, 93, 108, 110: 112, 114, 116, 120, 131,133, 147, 164, 166 altivagus, Mesodon (Akromesodon) 74, 92, 128, 129, 145 Amblyplax 221 Amicula 221, 223, 226, 227, 233 andrewsae, Mesodon 76, 80, 89, 108, 110, 112,114, 116, 120, 123, 124, 126, 130, 147, 164 andrewsae, Mesodon (Akromesodon) 74, 92, 128, 129, 145 Aphalogona 72, 74, 78, 162-163 Appalachina 78, 162, 170, 174, 175, 177 appressa, Patera 73, 75, 76, 108-110, 120, 122, 126, 141, 156, 166 appressa, Patera (Patera) 74, 97, 98-99, 127, 129, 145 appressus, Mesodon 114, 116 approximans, Inflectarius 75, 76, 105, 106, 108-110, 120, 125, 137, 159 approximans, Inflectarius (Inflectarius) 74, 87, 90, 127, 129, 145 approximans, Mesodon 114, 116 arbustorum, Arianta 56 archeri, Fumonelix 74-76, 81, 85, 107, 109-112, 120, 128, 129, 145, 161 arenaria, Mya 68 Argopecten irradians 189 irradians concentricus 189 irradians irradians 188 Arianta arbustorum 56 Arion ater 217, 265 Ashmunella 75, 118 danielsi 115, 120, 166 danielsi dispar 76 Ashmunellilnae 75 ater, Arion 217, 265 Australorbis glabratus 187 balthica, Macoma 68 Basiliochiton 221 Bathyteuthis 251 Batillaria minima 318 Batillariidae 333 Batillarinae 304, 333 baudini, Pyrazus 304 Benthoctopus 251 bidentata, Terebralia 304 bimaculatus, Octopus 251 bimaculoides, Octopus 242, 251 binneyana, Patera 73, 75, 76, 78, 80, 104, 108-112, 120, 128, 129, 142, 143, 146, 154 binneyana, Patera (Vesperpatera) 74, 100- 103, 101-102, 127, 129, 145 binneyanus, Mesodon 114, 116 Biomphalaria alexandrina 43-54 glabrata 43-54, 279 pfeifferi 46, 49, 53 straminea 43-54 Bradybaenidae 281 briareus, Octopus 242-250 bucculentus, Mesodon thyroidus 93 Bulininae 279 Bulinus globosus 52, 53 tropicus 52, 53, 273-280 burryi, Octopus 243-251 caledonica, Clava 305, 309 caledonicum, Cerithium (Pyrazus) 305 caledonicus, Potamides 305 californianus, Mytilus 37 californica, Nuttallina 229 Callochitonidae 231 Camaenidae 281 Campanile 291 symbolicum 291 Campanilopsis 291 canaliculata, Pleurocera 339, 342, 343 carnaticum, Cerithium 305 caverna, Lepidochitona 229 Cepaea nemoralis 55, 58 ceres, Cerithium 291 Cerithidea 299, 309, 310, 328, 331-333 obtusa 299 Cerithideposilla 309 Cerithiidae 289, 331 Cerithioidea 333, 334 Cerithium carnaticum 305 ceres 291 charpentieri 291 crassum 305 laeve 291 lutosum 318 palustre 305 semistratium 314 sulcatum 305, 318 telescopium 291 charpentieri, Cerithium 291 chilensis, Mytilus 68 chilhoweensis, Mesodon 75, 76, 104, 110, iit, 1414. 116; 120124126, e134 147: 162 chilhoweensis, Mesodon (Appalachina) 74, 95, 96, 128, 129, 145 Chitonidae 230-232 Chlamys delicatula 188 Chloraea 284 Choneplax 234 346 INDEX christyi, Fumonelix 73, 74, 76, 82, 86, 107, 109, 111, 120, 128, 129, 135, 145, 161 christyi, Mesodon 114 clarki, Mesodon 114, 116 clarki, Mesodon (Mesodon) 173 clarki, Patera 73, 75, 76, 80, 108-112, 120, 123, 124, 133, 147, 146, 157 clarki, Patera (Patera) 74, 97, 99, 127, 129, 132, 143, 145 clausus, Mesodon 73, 75, 76, 78, 110, 1127114116, 120, 131. 132, 1837147, 165 clausus, Mesodon (Mesodon) 74, 94, 97, 128, 129, 145 Clava caledonica 305, 309 clava, Pseudovertagus 313 Clavocerithium 310 clenchi, Mesodon 114 clenchi, Patera 73, 75, 76, 78, 104, 108- 112,120, 122, 123,. 144. 155 clenchi, Patera (Vesperpatera) 74, 102, 103, 127, 129, 145 clingmanica, Fumonelix wheatleyi 86, 144 clingmanicus, Fumonelix 73 Componochiton 234 concentricus, Argopecten irradians 189 corneus, Planorbis 188 coronae, Phreatodrobia 29 costata, Potamides (Terebralia) tenerrimus 318 Craspedochiton 221, 223, 227, 228, 234- 236 Craspedoplax 221 Crassostrea gigas 188 virginica 67, 68, 189 crassum, Cerithium 305 Crepidula 61 Cryptochiton 221 Cryptoconchus 221, 227, 228, 234 Cryptoplacidae 221-224, 227, 228, 234-236 Cryptoplax 222, 227, 228, 234, 235 Cyclophoridae 282, 284 danielsi, Ashmunella 115, 120, 166 defilippi, Octopus 243-251 delicatula, Chlamys 188 Dendrochiton 221 dentiens, Lepidochitona 224 Deroceras laeve 197 reticulatum 197, 199-220, 255-272 digueti, Octopus 242, 244 dislocata, Goniobasis catenaria 339, 342- 344 dispar, Ashmunella danielsi 76 downieanus, Inflectarius 75, 76, 104, 105, 110-112, 119, 120, 125, 139, 146, 157- 158, 76 downieanus, Inflectarius (Hubrichtius) 74, 85, 87, 88, 127, 129, 145 downieanus, Mesodon 172 Dreissena polymorpha 31-42, 36, 38-40, 63-70, 179-191 ebininus, Pyrazus 304, 331 edentatus, Inflectarius 104, 108-112, 120, 137, 142, 158 edentatus, Inflectarius (Inflectarius) 74, 87- 88, 90, 127, 129, 145 edentatus, Mesodon 114, 116 edulis, Mytilus 37, 41, 68, 188, 189, 265 edulis, Ostrea 68 elevatus, Mesodon 73, 75, 76, 78, 79, 81- 84, 108; 114, 120, 122. 123, 12451333 163 elevatus, Mesodon (Aphalogona) 74, 93-94, 128, 129, 132, 144, 145 Enoplochiton 231 Euaxoctopus 251 pillsburyae 243-246 eulaliae, Humboldtiana 284 Euprymna 251 fabula, Tellina 68 ferrissi, Inflectarius 73, 76, 81, 104, 106, 120, 133, 138, 140, 160 ferrissi, Inflectarius (Inflectarius) 74, 88, 91, 128, 129, 143, 145 ferrissi, Mesodon 111, 112, 114, 116 ferrissi, Mesodon (Mesodon) 171 filosus, Octopus 243, 246-251 floridana, Stenoplax 224, 225, 228, 229 floridensis, Goniobasis 339, 342, 343 Fumonelix 72, 74, 118, 123, 124, 130, 133, 140, 141, 143, 144, 146-148, 160 archeri 74-76, 81, 85, 107, 109-112, 120, 128, 129, 145, 161 christyi 73, 74, 76, 82, 86, 107, 109, 111, 120, 128, 129, 135, 145, 161 clingmanicus 73 jonesiana 73-76, 83, 86, 104, 107, 109- 112, 120, 128, 129, 136, 145,161 orestes 74-76, 85, 86, 107, 109-112, 121, 128, 129, 133, 136, 143,161 wetherbyi 72, 74, 76, 85, 87, 107, 109, 111, 112, 121, 128, 129, 1332733 144, 145, 160 wheatleyi 74-76, 80, 85, 86, 107, 109- 112, 121, 128, 129, 133, 136, 143; 145, 161 wheatleyi clingmanica 86, 144 fuscatus, Tympanotonus 331 fuscum, Telescopium 292 Galiteuthis glacialis 251 georgianus, Viviparus 66 gigas, Crassostrea 188 glabrata, Biomphalaria 43-54, 279 glabratus, Australorbis 187 glacialis, Galiteuthis 251 globosus, Bulinus 52, 53 Goniobasis 339-344 acutocarinata 339, 342, 343 alabamensis 339, 342, 343 catenaria dislocata 339, 342-344 floridensis 339, 342, 343 livescens 339, 342-344 proxima 339, 342, 343 simplex 339-344 Gourmya 331 Hachijomopalia 233 Helicina 282 Helicinidae 282 helicinum, Mesodon 77 Helicostylinae 281, 283, 284, 286 Helisoma trivolvis 188 INDEX Helix 284 pomatia 265 thyroidus 77 Helminthoglyptidae 284 hemphilli, Juga 339, 342-344 Horatiini 27 Hubrichtius 157 Humboldtiana 284, 286 eulaliae 284 Humboldtianidae 284 hummelincki, Octopus 248 Hyala vitrea 25 Hydrobiidae 1, 26 Hydrobiinae 27 Illex 251 indianorum, Inflectarius 128, 129 indianorum, Mesodon [MS.] 114 indianorum, Patera 75, 76, 104, 108-112, 120, 123, 128, 129, 154, 104, 73 indianorum, Patera (Vesperpatera) 74, 102, 103, 127, 129, 145 indicator, Telescopium 291 Inflectarius 78, 126, 133, 141-143, 146, 157, 170-172 approximans 75, 76, 105, 106, 108- 110, 120, 125, 137, 159 downieanus 75, 76, 104, 105, 110-112, 119, 120, 125, 139, 146, 157-158, 76 edentatus 104, 108-112, 120, 137, 142, 158 ferrissi 73, 76, 81, 104, 106, 120, 133, 138, 140, 160 indianorum 128, 129 inflectus 73, 75, 76, 87, 104-106, 108- 112, 120, 135, 140, 159 kalmianus 89, 104, 108-112, 119, 120, 124-126, 130, 131, 142, 146, 158 magazinensis 73, 76, 105, 106, 108- 102, 121, 125, 132, 133.137. 158 rugeli 75, 76, 106, 110-112, 121, 123, 125, 138, 140, 159 smithi 76, 87, 104, 107-112, 121, 132, 135, 159 subpalliatus 73, 76, 106, 110-112, 121, 123, 132, 133, 138, 140, 146, 160 verus 73, 108-112, 121, 128, 129, 132, 133, 146, 159 Inflectarius (Hubrichtius) 72, 74, 124, 125, 130, 131, 133, 148, 157 Inflectarius (Inflectarius) 119, 125, 130, 132, 133, 144, 148 inflectus, Inflectarius 73, 75, 76, 87, 104- 106, 108-112, 120, 135, 140, 159 inflectus, Inflectarius (Inflectarius) 74, 88, 90, 127, 129, 145 inflectus, Mesodon 114, 116 lo 339, 344 irradians, Argopecten 189 irradians, Argopecten irradians 188 Ischnochitonidae 221, 232 jogjacartense, Potamides 291 jogjacartensis, Telescopium 302 jonesiana, Fumonelix 73-76, 83, 86, 104, 107, 109-112, 120, 128, 129, 136, 145, 161 347 joubini, Octopus 242-250 Juga 339-344 hemphilli 339, 342-344 juvenisanguis, Moitessieria 12-13, 18, 25 kalmianus, Inflectarius 89, 104, 108-112, 119, 120, 124-126, 130, 131, 142, 146, 158 kalmianus, Inflectarius (Hubrichtius) 74, 87, 127, 129, 145 kalmianus, Mesodon 114 kalmianus, Mesodon (Mesodon) 172 Katharina 221, 234 tunicata 221 kiowaensis, Mesodon 114 kiowaensis, Patera 75, 76, 104, 108-112, 120, 144, 155 kiowaensis, Patera (Vesperpatera) 74, 102- 103, 103, 127, 129, 145 knoxvillina, Helix (Helicodonta) 77 laeve, Cerithium 291 laeve, Deroceras 197 laevior, Mesodon 114, 116 laevior, Patera 73, 75, 76, 104, 108-112, 121, 123, 142, 156, 75776 laevior, Patera (Patera) 74, 98, 99, 127, 129, 145 lamarckii, Potamides 304 lamellosa, Thais 188 lapillus, Thais 187, 188 Lartetia rayi 26 leatherwoodi, Mesodon 114 leatherwoodi, Patera 104, 108, 111, 112, 121. 1837 143, 146, 155 leatherwoodi, Patera (Vesperpatera) 74, 100, 103, 127, 129, 145 Lehmannia marginata 193-198 Lepidochitona 221 caverna 229 dentiens 224 Leptochitonidae 230, 231 Leptoxis 339, 344 lescherae, Moitessieria 12-13, 18, 25 leucodon, Mesodon 77 Limax maximus 198, 215, 216 pseudoflavus 213 lineolata, Moitessieria 25 Liolophura 231 Lithasia 339, 344 Littorina 61 livescens, Goniobasis 339, 342-344 locardi, Moitessieria 12-13 Lorica 234 Loricella 234 lutosum, Cerithium 318 Lymnaea 39 natalensis 52, 53 palustris 187 Macoma balthica 68 macropus, Octopus 243-250 Macrotritopus 251 maculatum, Mesodon 77 magazinensis, Inflectarius 73, 76, 105, 106, 10811221217 125, 1325 1335 137.158 magazinensis, Inflectarius (Inflectarius) 74, 90, 90, 127, 129, 145 magazinensis, Mesodon 115 348 INDEX magellanicus, Placopecten 68 mangos, Strombus 314, 318 margaritifera, Margaritifera 68 Margaritifera margaritifera 68 marginata, Lehmannia 193-198 martensii, Pinctada 68 massoti, Moitessieria 10-17, 21-25; 15-17, , mauritsi, Telescopium 292, 298 maximus, Limax 198, 215, 216 maya, Octopus 243, 245-250 Е Taonius 251 Melanoides 332 tuberculata 43-54 Mercenaria mercenaria 66-68, 188 mercenaria, Mercenaria 66-68, 188 Mesodon 77, 78, 123, 126, 133, 140, 141, 14316270 7117404175 177 altivagus 73, 75, 76, 93, 108, 110, 1127114, 116, 120, 71817188, 147. 164, 166 andrewsae 76, 80, 89, 108, 110, 112, 1124, 116, 1201123: 1245126; 730 147, 164 appressus 114, 116 approximans 114, 116 binneyanus 114, 116 chilhoweensis 75, 76, 104, 110, 111, 114, 116, 120, 124, 126, 134, 147, 162 christyi 114 clarki 114, 116 clausus 73, 75, 76, 78, 110, 112, 114, 1MOML2OMSI 928183147165 clenchi 114 downieanus 172 edentatus 114, 116 elevatus 73, 75, 76, 78, 79, 81-84, 108, 114, 120; 122, 123, 124,133, 163 ferrissi 111, 112, 114, 116 helicinum [М$.] 77 indianorum 114 inflectus 114, 116 kalmianus 114 kiowaensis 114 laevior 114, 116 leatherwoodi 114 leucodon 77 maculatum 77 magazinensis 115 mitchellianus 75, 76, 109, 110-112, 115. 119; 121, 122, 132: 147. 163 normalis 73, 75, 76, 79, 89, 104, 108, 110-112 115, 116, 121. 124. 126, 164 orestes 115 panselenus 115, 116 pennsylvanicus 115 perigraptus 115, 116 roemeri 115, 117 rugeli 115, 117 Sanus: 75716, 110 112, 115 4121/0197, 165 sargentianus 115, 117 sayanus 73, 76, 104, 110, 111, 115, 117, 121, 123, 124, 126, 183: 134: 147, 162 smithi 115, 117 subpalliatus 115, 117 thyroidus 72, 73, 75, 76, 110, 112, 115, 117. 121. 125, 131 133.165 thyroidus bucculentus 93 trossulus 73, 75, 76, 94, 110, 112, 121, 192, 133, 147, 165 verus 117 wetherbyi 115 Wheatleyi 115, 117 zaletus 73, 75, 76, 78-80, 81-84, 104, 108, 110, 115, 117, 119, 1219122: 133, 134, 163 Mesodon (Akromesodon) 72, 74, 119, 122, 125, 128, 129, 131, 133, 147, 1148: 163 Mesodon (Aphalogona) 72, 74, 130, 133, 147, 148 Mesodon (Appalachina) 72, 74, 123, 126, 130, 132, 133, 141, 142, 144, 148 Mesodon (Mesodon) 119, 125, 126, 128- 130, 132, 133, 144, 148 Mesodontidae 78 Mesodontini 71-178 minima, Batillaria 318 mitchellianus, Mesodon 75, 76, 109, 110- 112, 115, 119, 127, 122. 713214741163 mitchellianus, Mesodon (Aphalogona) 74, 94, 94, 128, 129, 143, 145 Modulus modulus 332, 333 modulus, Modulus 332, 333 Moitessieria 1-30 juvenisanguis 12-13, 18, 25 lescherae 12-13, 18, 25 lineolata 25 locardi 12-13 massoti 10-17, 21-25; 15-17, 19-24 ollieri 25 puteana 12-13, 25 rayi 12-13, 25 rollandiana 5, 12-13, 25 simoniana 2-9, 11; 3-5; 12-13 vitrea 12-13, 25 Moitessieriidae 221-225, 227, 230, 233, 236 Moitessieriinae 26 moluccanus, Murex 318 Mopalia 221, 226, 227, 233, 234 Mopaliidae1, 26 Murex moluccanus 318 sulcatus 318 Mya arenaria 68 ytilus 37, 39, 41, 187 californianus 37 chilensis 68 edulis 37, 41, 68, 188, 189, 265 nantahalae, Patera clarki 146 natalensis, Lymnaea 52, 53 nemoralis, Cepaea 55, 58 Neohelix 72 nitidosa, Nucula 68 normalis, Mesodon 73, 75, 76, 79, 89, 104, 108, 110-112, 115, 116, 121, 124, 126, 164 INDEX normalis, Mesodon (Akromesodon) 74, 92, 128, 129, 145 Notoplax 227, 228, 234 Nucula nitidosa 68 Nuttallina californica 229 Е 27 obtusa, Cerithidea 299 Octopodinae 241-253 Octopus 251 bimaculatus 251 bimaculoides 242, 251 briareus 242-250 burryi 243-251 defilippi 243-251 digueti 242, 244 filosus 243, 246-251 hummelincki 248 joubini 242-250 macropus 243-250 maya 243, 245-250 ornatus 246 rapanui 246 vulgaris 243-250 zonatus 243-246 ollieri, Moitessieria 25 Onithochiton 231 orestes, Fumonelix 74-76, 85, 86, 107, 109-112, 121, 128, 129, 133, 136, 143, 161 orestes, Mesodon 115 Orientalinidae 27 omatus, Octopus 246 Ostrea edulis 68 Paladilhia 1-30 pleurotoma 1, 18, 26, 28 palustre, Cerithium 305 palustre, Cerithium (Potamides) 305 palustre, Cerithium (Pyrazus) 305 palustris, Lymnaea 187 palustris, Potamides 305 palustris, Potamides (Pyrazus) 305 palustris, Potamides (Terebralia) 305 palustris, Potamides (Tympanotonus) 305 palustris, Pyrazus 305 palustris, Strombus 305 palustris, Terebralia 289, 290, 301, 304- 313, 306, 308, 328, 331, 333 panselena, Patera 76, 104, 108-111, 121, 140, 145, 146, 156 panselena, Patera (Patera) 74, 99, 99, 127, 129 panselenus, Mesodon 115, 116 Papuininae 281, 283, 286 Patera 78, 126, 133, 141-143, 153, 170, 17, 193, 1757 appressa 73, 75, 76, 108-110, 120, 122, 126, 147; 156, 166 binneyana 73, 75, 76, 78, 80, 104, 108-112, 120, 128, 129, 142, 143, 146, 154 clarki 73, 75, 76, 80, 108-112, 120, 123, 124, 133, 141, 146, 157 clarki nantahalae 146 clenchi 73, 75, 76, 78, 104, 108-112, 120.122; 123, 144; 155 indianorum 75, 76, 104, 108-112, 120, 349 123, 128, 129, 154, 104, 73 kiowaensis 75, 76, 104, 108-112, 120, 144, 155 laevior 73, 75, 76, 104, 108-112, 121, 123, 142, 156, 75, 76 leatherwoodi 104, 108, 111, 112, 121, 133, 143, 146, 155 panselena 76, 104, 108-111, 121, 140, 145, 146, 156 pennsylvanica 73, 75, 88, 104, 107, 109, 111, 112, 121, 133, 139, 145, 154 perigrapta 73, 75, 76, 104, 108-112, 121, 123, 140, 145, 146, 155-156 roemeri 73, 76, 104, 108, 111, 112, 121, 133, 144, 146, 155 sargentiana 73, 76, 104, 108-112, 121, 123, 141, 156 Patera (Patera) 72, 74, 122, 123, 130, 132, 133, 144, 148, 155 Patera (Ragsdaleorbis) 72, 74, 124, 126, 130, 131, 133, 148 Patera (Vesperpatera) 72, 74, 123, 126, 130, 132, 133, 144, 148, 154 Pelycidiidae 29 pennsylvanica, Patera 73, 75, 88, 104, 107, 109, 111, 112, 121, 133, 139, 145, 154 pennsylvanica, Patera (Ragsdaleorbis) 74, 100-101, 127, 129 pennsylvanica, Ptera 76 pennsylvanicus, Mesodon 115 perigrapta, Patera 73, 75, 76, 104, 108- 112, 121, 123, 140, 145, 146, 155-156 perigrapta, Patera (Patera) 74, 99, 99-100, 127, 129 perigraptus, Mesodon 115, 116 pfeifferi, Biomphalaria 46, 49, 53 Phreatodrobia 26, 27 coronae 29 pillsburyae, Euaxoctopus 243-246 Pinctada martensii 68 Pirenella 333 Placiphorella 221, 223, 226, 227, 231, 233, 234 Placopecten magellanicus 68 Placophoropsis 221 Planorbinae 279 Planorbis corneus 188 Plaxiphora 221, 223, 226, 227, 233, 234 Pleurocera 339-344 acuta 339, 342, 343 canaliculata 339, 342, 343 unciale 339, 340, 342, 343 Pleuroceridae 339-344 pleurotoma, Paladilhia 1, 18, 26, 28 Polygyra 75, 77, 105, 106-112 Polygyridae 71-178 Polygyrinae 152-165 polymorpha, Dreissena 31-42, 36, 38-40, 63-70, 179-191 Polyplacophora 221-240 pomatia, Helix 265 Potamides 291, 304 caledonicus 305 jogjacartense 291 lamarckii 304 350 INDEX palustre tryoni 305 palustris 305 sulcatus 318 tenerrimus 318 Potamididae 289-338 Potamidinae 333 Praticolella 75, 105-107, 109-112 profunda, Allogona 76, 118, 120 proxima, Goniobasis 339, 342, 343 pseudobeliscus, Telescopium 291 pseudoflavus, Limax 213 Pseudovertagus clava 313 Ptera pennsylvanica 76 puteana, Moitessieria 12-13, 25 Pyrazus 292, 304 baudini 304 ebininus 304, 331 palustris 305 sulcatus 318 Ragsdaleorbis 78, 153 rapanui, Octopus 246 rayi, Lartetia 26 rayi, Moitessieria 12-13, 25 reticulatum, Deroceras 197, 199-220, 255- 272 Rhinoclavis 310, 331 Rissoidae 26, 29 Rissooidea 26 roemeri, Mesodon 115, 117 roemeri, Patera 73, 76, 104, 108, 111, 112, 121, 133, 144, 146, 155 roemeri, Patera (Vesperpatera) 74, 100, 103, 127, 129, 145 rollandiana, Moitessieria 5, 12-13, 25 rugeli, Inflectarius 75, 76, 106, 110-112, 121, 123, 125, 138; 140, 159 rugeli, Inflectarius (Inflectarius) 74, 90, 91, 128, 129, 145 rugeli, Mesodon 115, 117 sanus, Mesodon 75, 76, 110, 112, 115, 121, 131, 165 sanus, Mesodon (Mesodon) 74, 93, 97, 128, 129, 145 sargentiana, Patera 73, 76, 104, 108-112, 121: 123: 141; 156 sargentiana, Patera (Patera) 74, 97, 100, 127, 129, 145 sargentianus, Mesodon 115, 117 sayanus, Mesodon 73, 76, 104, 110, 111, 115, 117, 121, 123, 124, 126,1133,.-134; 147, 162 sayanus, Mesodon (Appalachina) 74, 95, 96, 97, 128, 129, 145 Schizochitonidae 231, 234 semistratium, Cerithium 314 semistriata, Terebralia 289, 290, 310, 314- 318, 315-317, 328, 329, 331 semistriatus, Cerithium (Pyrazus) 318 semistriatus, Strombus 314, 318 semitrisulcata, Potamides (Terebralia) 318 simoniana, Moitessieria 2-9, 11; 3-5; 12-13 simoniana, Paludinella (Moitessieria) 2, 25 simplex, Goniobasis 339-344 smithi, Inflectarius 76, 87, 104, 107-112, 121, 132, 135, 159 smithi, Inflectarius (Inflectarius) 74, 90, 127, 129, 144, 145 smithi, Mesodon 115, 117 Spiralix 26 Stenochiton 225, 233 Stenoplax 224, 225, 228, 233 floridana 224, 225, 228, 229 Stenotrema 75, 105-112 straminea, Biomphalaria 43-54 Strombus agnatus 305 mangos 314, 318 palustris 305 semistriatus 314, 318 stultorum, Tivela 189 subpalliatus, Inflectarius 73, 76, 106, 110- 112, 121, 123, 132, 133, 738; 1407148: 160 subpalliatus, Inflectarius (Inflectarius) 74, 91, 97, 128, 129, 145 subpalliatus, Mesodon 115, 117 subpalliatus, Mesodon (Patera) 171 sulcata, Potamides (Terebralia) 318 sulcata, Terebralia 289, 290, 304, 315, 318-330, 319, 320, 322-327 sulcatum, Cerithium 305, 318 sulcatus, Murex 318 sulcatus, Potamides 318 sulcatus, Pyrazus 318 symbolicum, Campanile 291 aonius 251 megalops 251 Tarebia 332 Telescopioidea 291 Telescopium 289-338 fuscum 292 indicator 291 jogjacartensis 302 mauritsi 292, 298 pseudobeliscus 291 telescopium 289-292, 293-297, 298- 304, 309-311, 329, 332, 333 titan 302 telescopium, Cerithium 291 telescopium, Potamides (Telescopium) 292 telescopium, Telescopium 289-292, 293- 297, 298-304, 309-311, 329, 332, 333 telescopium, Trochus 291 Tellina fabula 68 tenuis 68 tenerrimus, Potamides (Terebralia) 318 tenerrimus, Potomides 318 tenuis, Tellina 68 Terebralia 289-338 bidentata 304 palustris 289, 290, 301, 304-313, 306, 308, 328, 331, 333 semistriata 289, 290, 310, 314-318, 315-317, 328, 329, 331 sulcata 289, 290, 304, 315, 318-330, 319, 320, 322-327 Thais lamellosa 188 lapillus 187, 188 Thiara 332 Thiaridae 43 thyroidus, Helix 77 thyroidus, Helix (Helicodonta) 77 INDEX thyroidus, Mesodon 72, 73, 75, 76, 110, MA 5 7» 2. 125, 131, 133, 165 thyroidus, Mesodon (Mesodon) 74, 93, 97- 98, 128, 129, 145 titan, Telescopium 302 Tivela stultorum 189 Tonicia 231 Triodopsini 75, 128, 142, 148 Triodopsis 72, 77 trisulcatus, Trochus 305 trivolvis, Helisoma 188 Trochus telescopium 291 trisulcatus 305 tropicus, Bulinus 52, 53, 273-280 trossulus, Mesodon 73, 75, 76, 94, 110, 2 121, 7932, 133, 147. 165 trossulus, Mesodon (Mesodon) 74, 94, 98, 128, 129, 145 Truncatelloidea 26 tryoni, Potamides palustre 305 tuberculata, Melanoides 43-54 tunicata, Katharina 221 Turritellidae 300 Tympanotonus 328, 333 fuscatus 331 unciale, Pleurocera 339, 340, 342, 343 Vanikoridae 29 verus, Inflectarius 73, 108-112, 121, 128, 129, 132, 133, 146, 159 verus, Inflectarius (Inflectarius) 74, 91-92, 92, 128, 129, 145 verus, Mesodon 117 Vesperpatera 154 virginica, Crassostrea 67, 68, 189 vitrea, Hyala 25 vitrea, Moitessieria 12-13, 25 Viviparus georgianus 66 vulgaris, Octopus 243-250 Webbhelix 72 wetherbyi, Fumonelix 72, 74, 76, 85, 87, 1107, 109, 111, 112, 121, 128, 129, 133, 135, 144, 145, 160 wetherbyi, Mesodon 115 wheatleyi, Fumonelix 74-76, 80, 85, 86, 107, 109-112, 121, 128, 129, 133, 136, 143, 145, 161 wheatleyi, Mesodon 115, 117 Xolotrema 72 zaletus, Mesodon 73, 75, 76, 78-80, 81-84, 104 108; 110, 115, 117, 119, 121, 122, 133.134. 163 zaletus, Mesodon (Aphalogona) 74, 94-95, 95, 128, 129, 145 zonatus, Octopus 243-246 351 WHY NOT SUBSCRIBE TO MALACOLOGIA? ORDER FORM Your name and address Send U.S. $21.00 for a personal subscription (one volume) or U.S. $35.00 for an institutional subscription. Make checks payable to “MALACOLOGIA.” Address: Malacologia, Academy of Natural Sciences 1900 Benjamin Franklin Parkway Philadelphia, PA 19103-1195, U.S.A. MALACOLOGIA, 1991, 33(1-2): 33(1-2) INSTRUCTIONS FOR AUTHORS 1. MALACOLOGIA publishes original re- search on the Mollusca that is of high quality and of broad international interest. Papers combining synthesis with innovation are par- ticularly desired. While publishing symposia from time to time, MALACOLOGIA encour- ages submission of single manuscripts on diverse topics. Papers of local geographical or systematic interest should be submitted elsewhere, as should papers whose primary thrust is physiology or biochemistry. Nearly all branches of malacology are represented on the pages of MALACOLOGIA. 2. Manuscripts submitted for publication are received with the tacit understanding that they have not been submitted or published elsewhere in whole or in part. 3. Manuscripts may be in English, French, German or Spanish. Papers in lan- guages other than English must include a translation of the Abstract in English. Authors desiring to have their abstracts published in other languages must provide the translations (complete with main titles). Include all foreign accents. Both American and British spellings are allowed. 4. Unless indicated otherwise below, con- tributors should follow the recommendations in the Council of Biology Editors (CBE) Style Manual (ed. 5, 1983) available for U.S. $24.00 from CBE, 9650 Rockville Pike, Bethesda, MD 20814, U.S.A. 5. Be brief. 6. Manuscripts must be typed on one side of good quality white paper, double-spaced throughout (including the references, tables and figure captions), and with ample margins. Tables and figure captions should be typed on separate pages and put at the end of the manuscript. Make the hierarchy of headings within the text simple and consistent. Avoid internal page references (which have to be added in page proof). 7. Choose a running title (a shortened version of the main title) of fewer than 50 letters and spaces. MALACOLOGIA 1991 8. Provide a concise and informative Ab- stract summarizing not only contents but re- sults. A separate summary generally is super- fluous. 9. Supply between five and eight key (topic) words to go at the end of the Abstract. 10. Use the metric system throughout. Mi- cron should be abbreviated рт. 11. Illustrations are printed either in one column or the full width of a page of the journal, so plan accordingly. The maximum size of a printed figure is 13.5 x 20.0cm (preferably not as tall as this so that the cap- tion does not have to be on the opposite page). 12. Drawings and lettering must be dark black on white, blue tracing, or blue-lined paper. Lines, stippling, letters and numbers should be thick enough to allow reduction by Y. or Ys. Letters and numbers should be at least 3mm high after reduction. Several drawings or photographs may be grouped together to fit a page. Photographs are to be high contrast. High contrast is especially im- portant for histological photographs. 13. All illustrations are to be numbered sequentially as figures (not grouped as plates or as lettered subseries), and are to be ar- ranged as closely as possible to the order in which they are first cited in the text. Each figure must be cited in the text. 14. Scale lines are required for all nondi- agrammatic figures, and should be conve- nient lengths (e.g., “200 um,” not “163 um”). Magnifications in captions are not accept- able. 15. All illustrations should be mounted, numbered, labeled or lettered, i.e. ready for the printer. 16. A caption should summarize what is shown in an illustration, and should not dupli- cate information given in the text. Each let- tered abbreviation labeling an individual fea- ture in a figure must either be explained in each caption (listed alphabetically), or be grouped in one alphabetic sequence after the Methods section. Use the latter method if many abbreviations are repeated on different figures. 17. Tables are to be used sparingly, and vertical lines not at all. 18. References cited in the text must be in the Literature Cited section and vice versa. Follow a recent issue of MALACOLOGIA for bibliographic style, noting especially that se- rials are cited unabbreviated. Supply pagina- tion for books. Supply information on plates, etc., only if they are not included in the pagination. 19. In systematic papers, synonymies should not give complete citations but should relate by author, date and page to the Litera- ture Cited section. 20. For systematic papers, all new type- specimens must be deposited in museums where they may be studied by other scien- tists. Likewise MALACOLOGIA requires that voucher specimens upon which a paper is based be deposited in a museum where they may eventually be reidentified. 21. Submit each manuscript in triplicate. The second and third copies can be reproduc- tions. 22. Authors who want illustrations returned should request this at the time of ordering re- prints. Otherwise, illustrations will be main- tained for six months only after publication. REPRINTS AND PAGE COSTS 23. When 100 or more reprints are or- dered, an author receives 25 additional cop- ies free. Reprints must be ordered at the time proof is returned to the Editorial Office. Later orders cannot be considered. For each au- thors’ change in page proof, the cost is U.S. $3.00 or more. 24. When an article is 10 or more printed pages long, MALACOLOGIA requests that an author pay part of the publication costs if grant or institutional support is available. SUBSCRIPTION COSTS 25. For Vol. 32, personal subscriptions are U.S. $21.00 and institutional subscriptions are U.S. $35.00. Back issues and single vol- umes: $35.00 for non-institutional purchaser; $40.00 for institutional purchaser. There is a one dollar handling charge for all purchases of single volumes. Address inquiries to the Subscription Office. VOL. 33, NO. 1-2 MALACOLOGIA CONTENTS MARCO BODON & FOLCO GIUSTI The Genus Moitessieria in the Island of Sardinia and in Italy. New Data on the Systematics of Moitessieria and Paladilhia (Prosobranchia: Hydrobiidae) (Studies on the Sardinian and Corsican Malacofauna, IX) .................... ULRICH BIELEFELD Histological Observation of Gonads and Digestive Gland in Starving Dreissena DONDE III) Ая td E ао TEE J. P. POINTIER, J. L. TOFFART 8 M. LEFEVRE Life Tables of Freshwater Snails of the Genus Biomphalaria (B. glabrata, B. alexandrina, B. straminea) and of One of its Competitors Melanoides tuberculata UnderLaboratory Conditions RAR oh Wes (oss a Pat yw adie sacs Ba ee $ Е. М. COOK Fluctuations and Immobility in Age-Structured Snail Populations .............. MARTIN SPRUNG Costs of Reproduction: A Study on Metabolic Requirements of the Gonads and Fecundity of the Bivalve Dreissena polymorpha ............................. KENNETH C. EMBERTON The Genitalic, Allozymic and Conchological Evolution of the Tribe Mesodontini (Pulmonata: Stylommatophora: Polygyridae) ................................ MARTIN SPRUNG & JOST BORCHERDING (4 Physiological and Morphometric Changes in Dreissena polymorpha (Mollusca; Bivalvia) During a Starvation Period... fie ok caera e dere ee MICHELE K. SURBEY & C. DAVID ROLLO Physiological and Behavioural Compensation for Food Quality and mai) in the Slug Lehmannia Marginala 2.2. Ns aa mado Sy ol onl LA RE C. DAVID ROLLO Endogenous and Exogenous Regulation of Activity in Deroceras reticulatum, A Weather-Sensitive Terrestrial Slug ..................................,... Pee G. THOMAS WATTERS Utilization of a Simple aaa by Polyplacophorans and Its Evolutionary Ae оз MA e PO Es an PN ENS ieee SI CEST ST IE JANET R. VOIGHT | Morphological Variation in Octopod Specimens: Reassessing the Assumption of | Preservation-Induced Deformation:: 2:52 LT Es a a SOS | RITA TRIEBSKORN The Impact of Molluscicides on Enzyme Activities in the Hepatopancreas of de. Berocetasretieckiatım MUI SEE sa ir toes e a odds NE ANN THERESE D. BRACKENBURY 4 C. C. APPLETON y Morphology of the Mature Spermatozoon of Bulinus tropicus (Krauss, 1848) (sastropoda >: Planombidae) ен... ол nme ae a were A у BARRY ROTH Tropical “Physiognomy” of a Land Snail Faunule From the Eocene of Southern L CAO Sa ee O a elas RARE A Re dy RICHARD S. HOUBRICK Systematic Review and Functional Morphology of the aan Snails Terebra- 4 lia and Telescopium (Potamididae; Prosobranchia) .......................... ROBERT T. DILLON, JR. 4 Karyotypic Evolution in Pleurocerid Snails Il. Pleurocera, Goniobasis, and — И A O SI A E ES ЗИ 2 3 2044 072 160 542 ww CREER w ER EEE Turn ne urn. m ng