1 LEST Reta igs DANA Pek Paine ns AR) re yo 3 on [Nel DMRS BoM Mba Mahe ex <}, ААУ #44 $ NE er K mite na + sus г a a CRE a Rae ne SE 7 Son ae а Se + een D ee a 1 CEUTA 7 SECTE 4 E EOS soe DOS HA +5 я У» À CAN a м Le A RR ne a т LAN x DH À ` 4 RARES MIN be sal HA A = 3 0 je a 4:0 HINA: baa IE A N ны АРА PACE ya, DAMAS 1. LORS LS DIEU LUE) ER AAA SEVEN D EN Ms al му ST G у ‘ OTAN TAS ЯНА À В Sion ME ko ALU E ® PS ET ' m e win 119 Ce оо ам AE EAN) data mer due AUS I TA OO MATE И на а . ‘ Hen ie iene oe La у» > q маи КИ ROA CENT SR “4 CUBE nats uk DELL QU à Dao De 46 dt On Ae g ATEN CHELLES AA О, A SR A EME Re DA Meath OTN MEM Re № A УЕ ве mentee party Dg NASA VAL, “y a ыы ай un water CUT у | PERRET ETAT Sr ÉCRIN 4 À NA AL Bb ARE OUT AT ERA aces ne Man Ernten een DEL tee ` ‘ Lund A, ` MAA May rio mA EN nea pl a aan y : aaa ‚ UA ’ N » A +, AONE nn nn! Len LAN Ne pus Ad | Papi à A - Pda AUN Al led, I + A EU Hg da As À A DDO мун ыы } A EEN nose луна Y . 4 ОНИ и: ale мм HARAS on EA WIR RIN AS ado Ea oe aa Xe emba cio pl o ON a RES Ey TEN: 7 ve x АНИ REDON DE CT serie m re Ed > ` he Ate Wea ДИНЫ ` ‘ A , Me deal RATES x AU { / ` ‘ К узы . Y ves a vn amas 1. Monte LT ‚Arad, ; ть PARTEIEN = rie му Fr | RIRE DER Mix x . ? : : E pde Seats “a Peras da LA Y Kr qye : NE RUN Dé Bol Le e Be VAN a pe a ! wera \ st CAEN Mate EEE ln cae LE APA A ails ba LM roe hak 4 À an . REN . 1? : .. 0 pl eee = к, RAA DE SIERT at 1 i, 4 ZR ois ola de ; à ; р À ais pia vf т en Lire a 11.2 ; N у DEM Yale A o Een x q A Mes . ик у sat e ere . > ; Ad ee A 4 ) НН у dy { ‘ 4 Ber Ira nt ; Ie DV EES. 1 НИНЫ Я ‘ ie e 7 ee Я per AU } очи АЙ fer Ny a A i 4 ARES ий ere В Burton ЦИ» AAA y PY > or DCS ' $ A PGA oe ayy. , ‘ > MES . ‘ «ту L pe миры } a Lh Ha à phe men À Bab ala we Faw ® Ноа NTM HE site Te SPINE YN » PATENT: O y OS gaie yá se ins rat ALAS и j aan. rahe LHI, Be pie EU PENH je MORE IO a IN ER En KR HARVARD UNIVERSITY la Library of the Museum of Comparative Zoology E, PMI A o А Er Г vs ih # № 1 7 NE u in i) 5 ña ie + | ti ot (7 . A | nt | A ih a | di | лия Lg VOL: 26 1985 MALACOLOGIA International Journal of Malacology Revista Internacional de Malacologia Journal International de Malacologie Международный Журнал Малакологии Internationale Malakologische Zeitschrift Publication date Vol. 25, No. 2—29 August, 1984 MALACOLOGIA, VOL. 26 CONTENTS . A. ABOUL-MAGD & S. A. SABRY Scanning electron microscopy of the body surfaces of Biomphalaria EA - AR RA NI RNA CIÓN AN 201 . M. BROWN, D. R. DEVRIES 4 B. K. LEATHERS Causes of life history variation in the freshwater snail Lymnaea elodes. . 191 . COOK Functional aspects of trail following by the carnivorous snail Euglandina ¡OSCAR RR Бе ее Ns RR ar renee oe ne PR en 173 . COOK The organisation of feeding in the carnivorous snail Euglandina rosea... 183 . У. DIMOCK, Jr. Quantitative aspects of locomotion by the mud snail /lyanassa obsoleta . 165 . C. EMBERTON Seasonal changes in the reproductive gross anatomy of the land snail Triodopsis tridentata tridentata (Pulmonata: Polygyridae) ............... 225 . FUJIOKA Population ecological aspects of the eulimid gastropod Vitreobalcis ICNNODIEUTICOIA RE RER NE erat Scere Pagal de oI 153 . HERSHEER Systematic revision of the Hydrobiidae (Gastropoda: Rissoacea) of the Guatro/Ciénegas Basin, Coahuila; Мехжсо. .-:.... elo mies eee 31 . E. JELNES & J.-P. POINTIER Taxonomie expérimentale de Biomphalaria (Gastropoda: Planorbidae)—1!. Mobilités enzymatiques considérées comme éléments de diagnostic pour les Biomphalaria Antillais. Etude de sept systemes enzymatiques....... 187 . С. КЕМК 8 В. В. WILSON А new mussel (Bivalvia, Mytilidae) from hydrothermal vents п the Galapa- GO SHITE ZO N ee a eet ese ts ee te 253 . R. KRAEMER & C. M. SWANSON Functional morphology of “eyespots” of mantle flaps of Lampsilis (Bivalvia: Unionacea): evidence for their role as effectors, and basis for hypothesis regarding pigment distribution in bivalve mantle tissue ................. 241 . М. LIVSHITS Ecology of the terrestrial snail Brephulopsis bidens (Pulmonata: Enidae): mortality burrowing andimigratory activi. een ana ee. 213 . 5. PREZANT Derivations of arenophilic mantle glands in the Anomalodesmata ....... 273 . 5. RICHARDS А new pigmentation mutant т Biomphalaria glabrata .................. 145 RS SEARY The pelagic genus Pterotrachea (Gastropoda: Heteropoda) from Hawaiian Waters mal TAXOMOMIG Кемер 125 . SOLEM & F. M. CLIMO Structure and habitat correlation of sympatric New Zealand land snail SPECIES о A Nt Lg ета а 1 WHY NOT SUBSCRIBE TO MALACOLOGIA? ORDER FORM Your name and address Send U.S. $17.00 for a personal subscription (one volume) or U.S. $27.00 for an institutional subscription. Make checks payable to "MALACOLOGIA. Address: Malacologia, Academy of Natural Sciences Nineteenth and the Parkway, Philadelphia PA: 1903; Ш.А. NTM L du 17 VOL. 26, NO. 1-2 1985 MALACOLOGIA MUS. COMP. ZOOL LIBRARY ‚JUL 1 5 1985 HARVARD UNIVERSITY в: : IE Е national Journal of Malacology | Revista Internacional de Molatalodia Journal International de Malacologie | Международный KypHan Малакологии Internationale Malakologische Zeitschrift X MALACOLOGIA GEORGE M. DAVIS Editors-in-Chief ROBERT ROBERTSON Editorial and Subscription Offices Department of Malacology The Academy of Natural Sciences of Philadelphia Nineteenth Street and the Parkway Philadelphia, Pennsylvania 19103, U.S.A. Associate Editors JOHN B. BURCH University of Michigan, Ann Arbor ANNE GISMANN Maadi, A. R. Egypt Editorial Assistants MARY DUNN KAREN BLAIR JOHN M. HARVEY MALACOLOGIA is published by the INSTITUTE OF MALACOLOGY, the Sponsor Members of which (also serving as editors) are: KENNETH J. BOSS Museum of Comparative Zoology Cambridge, Massachusetts JOHN B. BURCH MELBOURNE R. CARRIKER University of Delaware, Lewes GEORGE M. DAVIS Secretary and Treasurer PETER JUNG, Participating Member Naturhistorisches Museum, Basel, Switzerland OLIVER E. PAGET, Participating Member Naturhistorisches Museum, Wien, Austria ROBERT ROBERTSON CLYDE F. E. ROPER Smithsonian Institution Washington, D.C. W. D. RUSSELL-HUNTER, President-Elect Syracuse University, New York NORMAN F. SOHL United States Geological Survey Washington, D.C. SHI-KUEI WU, President University of Colorado Museum, Boulder J FRANCIS ALLEN, Emerita Environmental Protection Agency Washington, D.C. ELMER G. BERRY, Emeritus Germantown, Marland Copyright © 1985 by the Institute of Malacology 1985 EDITORIAL BOARD J. A. ALLEN Marine Biological Station Millport, United Kingdom E. E. BINDER Museum d'Histoire Naturelle Geneve, Switzerland A. J. CAIN University of Liverpool United Kingdom P. CALOW University of Glasgow United Kingdom A. H. CLARKE, Jr. Mattapoisett, Mass., U.S.A. B. C. CLARKE University of Nottingham United Kingdom C. J. DUNCAN University of Liverpool United Kingdom Z. A. FILATOVA Institute of Oceanology Moscow, U.S.S.R. Е. FISCHER-PIETTE Museum National d'Histoire Naturelle Paris, France М. ЕВЕТТЕВ University of Reading United Kingdom E. GITTENBERGER Rijksmuseum van Natuurlijke Historie Leiden, Netherlands A. N. GOLIKOV Zoological Institute Leningrad, U.S.S.R. S. J. GOULD Harvard University Cambridge, Mass., U.S.A. A. V. GROSSU Universitatea Bucuresti Romania T. HABE Tokai University Shimizu, Japan A. D. HARRISON University of Waterloo Ontario, Canada K. HATAI Tohoku University Sendai, Japan B. HUBENDICK Naturhistoriska Museet Goteborg, Sweden S. HUNT University of Lancaster United Kingdom A. M. KEEN Santa Rosa California, U.S.A. В. М. 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. Y. KONDO Bernice P. Bishop Museum Honolulu, Hawaii, U.S.A. J. LEVER Amsterdam, Netherlands A. LUCAS Faculté des Sciences Brest, France C. MEIER-BROOK Tropenmedizinisches Institut Tubingen, Germany (Federal Republic) Н. К. MIENIS Hebrew University of Jerusalem Israel J. E. MORTON The University Auckland, New Zealand 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. Е. PONDER Australian Museum Sydney A. W. B. POWELL Auckland Institute & Museum New Zealand R. D. PURCHON Chelsea College of Science & Technology London, United Kingdom O. RAVERA Euratom Ispra, Italy N. W. RUNHAM University College of North Wales Bangor, United Kingdom S. @ SEGERSTRALE Institute of Marine Research Helsinki, Finland G. A. SOLEM Field Museum of Natural History Chicago, U.S.A. F. STARMUHLNER Zoologisches Institut der Universitat Wien, Austria Y. |. STAROBOGATOV Zoological Institute Leningrad, U.S.S.R W. STREIFF Universite de Caen France J. STUARDO Universidad de Chile Valparaiso Т. Е. THOMPSON University of Bristol United Kingdom FESTOREOLEMIO Societa ltaliana di Malacologia Milano R. D. TURNER Harvard University Cambridge, Mass., U.S.A W. $. $. VAN BENTHEM JUTTING Domburg, Netherlands J. A. VAN EEDEN Potchefstroom University South Africa J.-J. VAN MOL Universite Libre de Bruxelles Belgium N. H. VERDONK Rijksuniversiteit Utrecht, Netherlands B. R. WILSON National Museum of Victoria Melbourne, Australia C. M. YONGE Edinburgh, United Kingdom Н._ ZEISSEER Leipzig, Germany (Democratic Republic) A. ZILCH Natur-museum und Forschungs-Institut Senckenberg Frankfurt-am-Main, Germany (Federal Republic) MALACOLOGIA, 1985, 26(1—2): 1-30 STRUCTURE AND HABITAT CORRELATIONS OF SYMPATRIC NEW ZEALAND LAND SNAIL SPECIES Alan Solem' & Frank М. Climo? ABSTRACT The land snail fauna of the Manukau Peninsula near Auckland, North Island, New Zealand, is known to consist of 89 species, only four of which are European imports. At least 71 of the native species could co-exist essentially microsympatrically in a single remnant bush patch. Sixty species have been found in one bush patch. On a typical adult of each species, measurements were made of shell height, diameter, whorl count, 'ımbilical width or condition, major radial ribs on the body whorl, spire height, and body whorl descension. Ratios were calculated for height/diameter, diameter/umbilical width, and radial ribs/mm on the body whorl. Status of whorl contour, periostracal sculpture extensions, and shell color were noted. The volume of space occupied by the shell was calculated. Taxonomic distributions of each measurement or character state were plotted, and an attempt was made to determine correlation between each major variation and shelter site preference of species. Punctids were found to be smaller than charopids, and more elevated, with fewer whorls, generally narrower umbilicus, more often without prominent radial sculpture, more frequently with angulated or carinated periphery, and more frequently monochrome in coloration. Shelter preference site correlations were few. Development of a peripheral keel is associated with sheltering in open ground space under deep, wet litter; periostracal fringes are most frequent in inhabitants of friable, broken-down litter; variegated color is represented better in arboreal taxa; and light or dark brown monochrome coloration in friable, broken-down litter. None of the above correlations are based on monophyletic lineages. Number and spacing of major radial ribs were not size-associated and did not correlate with habitat preference site. Compared with Northern Hemisphere faunas, the Manukau Peninsula species are small and when shell height is plotted against shell diameter, there is one scatter, rather than the dual scatter reported for most other land snail faunas. The distribution of shell volume among the species is not unimodal. Species with the same shelter site preference differ by at least 40% in volume. It is suggested that there is strong selective pressure for this pattern of shell volume difference, but in the absence of com- prehensive ecological data on all of the species, determination of the reasons for this are at present impossible. Key words: land snails; sympatric; structure and habitat; shell sculpture; shell shape. INTRODUCTION Land snail diversity on the Manukau Penin- sula, southwest of Auckland, New Zealand, was surveyed by Solem, Climo & Roscoe (1981). They concluded that more than 70 of the 85 resident species of native land snails and slugs could be expected to coexist in 2 hectare patches of generalized and un- disturbed lowland bush such as Jones Bush near Waiuku, from which 60 species have been collected. A diagram of typical bush facies has been published in Solem, Climo & Roscoe (1981: fig. 2) and descriptions of the special habitats are given in that report. From 45% to 75% of the total species present in a particular bush patch are found in each small (20 x 30 cm) bag of selected litter collected from an area of 0.2-0.3 m”. All such samples were taken under a single tree or clump of saplings belonging to one plant species. Many species were collected in all basic litter types, from the drier upper fringes to the very wet streamside piles of fern fronds. Very few taxa seem to be restricted to a specialized space niche, although most species have characteristic preferences as to moisture and space conditions (i.e., where they are most ‘Field Museum of Natural History, Roosevelt Rd. at Lake Shore Drive, Chicago, Illinois 60605-2496, U.S.A. “National Museum of New Zealand, Private Bag, Wellington, New Zealand. 2 SOLEM 8 CLIMO easily collected alive) (Solem, Climo 4 Ros- coe, 1981: appendix 1). The 70 species are essentially microsympatric. This level of sympatric land snail diversity so greatly exceeds that found in other areas of the world, which normally is five to twelve species (Solem, in preparation), that a pre- liminary review of correlations among habitat preferences, taxonomic groups, and aspects of shell size, shape, proportions, volume, sculpture, and color seemed desirable. Limitations of knowledge concerning the biol- ogy of individual species are a severe hand- icap. We do not know the life history of any of the species discussed. We have no data on seasonal abundance fluctuations, and no clear indication as to differentiation into shel- ter, foraging, or mating niches for any of the species. Our own sampling program included a mixture of shelter and foraging niches, since rains ended a short drought of about two weeks at the start of our field work and contin- ued at intervals through the program. Despite these limitations, we have de- veloped several testable hypotheses to ex- plain the nature of this high diversity (Solem, Climo & Roscoe, 1981). First, we have demonstrated that this fauna is 80.5% com- posed of species whose ranges extend con- siderably north and south of the study area, 19.5% of species whose ranges are at or near a north, south or west limit. Second, the fauna is not composed of a few species blooms. Only four of the 85 resident Manukau Penin- sula land snail species are interpretable as being products of a cladistic bifurcation. The other 81 species have separate “nearest rela- tives“ and often not even demonstrable “grandparental” ties. Thus, this faunal diver- sity is not packed through recent union of two or more allopatric faunas, nor is it the result of species swarms or blooms resulting from a few colonizations of virgin territory. Third, we suggest that this is a mature community of land snails, accumulated gradually over time, resulting primarily from highly favorable and stable conditions of moisture and space retention in the litter. This snail community ranges from somewhere north of Auckland south through the Central North Island cave region. Fourth, the number of sympatric land snail species declines sharply in other parts of New Zealand. We have presented hypoth- eses based on patterns of moisture interrup- tion, changes in litter space quality, recent- ness of area colonization, degree of exposure to desiccating winds, changes in rock sub- strate, topographic variation, and alterations in litter acidity to account for these reductions in diversity. In this paper we attempt to relate physical characters of the snail shells with observed habitat preferences. We regard shell size as a rough predictor of snail body size, since the snail can withdraw its body completely into the shell for all but a few of the New Zealand native taxa. We have chosen to exclude from this analysis the very large slug-like carnivore Schizoglossa worthyae Powell, 1949, which has an ear-shaped shell remnant into which the animal cannot possibly withdraw, and the medium to large, native, shell-less slugs of the family Athoracophoridae. We also are omitting three species that we did not collect personally on the Peninsula (Paryphanta bus- Бу!) [Gray, 1840], a human introduction; Egestula egesta [Gray, 1850], a northern spe- cies recorded once from the Manukau Penin- sula tip; and “Phrixgnathus” n. sp. 55, re- corded in beach wash), and two taxa whose absence from our collections surprised us (“Phrixgnathus” n. sp. 61 and Otoconcha di- midiata [Pfeiffer, 1853]). We thus are basing this analysis on 82 species of shelled land snails, four introduced from Europe (Table 1), instead of the 89 species reviewed by Solem, Climo & Roscoe (1981: appendix 1). The number of described New Zealand land snail species, 315, is much less than the 670 now represented in the collections of the National Museum of New Zealand (Climo, unpublished). Thus, many species are un- described and must be indicated by either a number or “n. sp. aff.” Each such undescribed taxon is referenced to a National Museum of New Zealand catalogued lot (Solem, Climo & Roscoe, 1981: appendix 1). Similarly, generic units in the Charopidae and Punctidae are in a State of flux. Current units are “form genera” without phyletic coherence. Citation of a spe- cies as “Charopa” is to give both an idea of general shell morphotype, and also to indi- cate that the species probably is not con- generic with the generotype. A monograph of the New Zealand Punc- tidae is partly completed (Climo, in prepara- tion) and data as to subfamily and generic units have been tabulated for use in testing structural associations. This is in too pre- liminary a form for publication and is subject to revision in final manuscript. Nevertheless, it does enable us to state that certain features of the punctid variation patterns are not monophyletic. Knowledge of the New Zea- SYMPATRIC NEW ZEALAND LAND SNAILS 3 land Charopidae is much less advanced, and we are not prepared to make phylogenetic predictions in relation to this complex. To our knowledge, this is the first such attempt to analyze sympatric land snail spe- cies associations in this manner. We recog- nize the defects inherent because life history data and precise ecological knowledge are lacking. We hope that this preliminary report will stimulate others to test our predictions through instigation of such studies and to apply the techniques we use here toward study of land snail faunas in other areas of the world. MATERIAL STUDIED All measured and observed specimens for this study are deposited in the National Museum of New Zealand, Wellington. Most of the specimens used are those reviewed by Solem, Climo & Roscoe (1981). There are a few exceptions. Certain species collected on the Manukau in low numbers were repre- sented only by dead, worn adults or subadult examples. Other species were represented only by specimens with the shell surface so covered by debris that observation and measurement of critical features was not possible. Jones Bush, near Waiuku, had the most diverse fauna of all the patches that we sampled, and is the primary source of materi- al used in this study. Where adequate speci- mens from Jones Bush itself were available, they were used. The many studies of Cumber (1960, 1961, 1962, 1964, 1967a-d) amply demonstrated that there is considerable local geographic variation in both size and sculp- ture counts within New Zealand land snail species. Therefore, when we lacked ade- quate Jones Bush or Manukau Peninsula material, we had to select material for measuring that fairly represented the Man- ukau morphotype. Recent studies on the Endodontidae (Solem, 1976), Pacific Basin Charopidae (Solem, 1983), and Western and central Aus- tralian Camaenidae (Solem, 1979, 1981a, b) have demonstrated that under desert, savan- nah, and tropical rain forest conditions, size distributions within large samples of shells are unimodal. The mean will shift from population to population as local variation determines, and allochronic variation is a real phenom- enon, but the principle of unimodal distribu- tion within a population is well established. Most species of New Zealand bush snails are collected in low numbers, so that sample size generally was small. For a few species (Table 11) we have measured series and provide variation data. For most species we have simply selected a typical adult example to represent the Manukau morphotype of that species. The authors consulted together on the selection of “an average adult” to represent the species. Most New Zealand land snail species show nondeterminate growth. There is, however, an easily recognizable short zone of gerontic shell growth and increased body whorl descension that represents “adult shell increment.” For the Charopidae, Punctidae, Achatinellidae, and Rhytididae, this is the common growth pattern. In the Liareidae, Hydrocenidae, Valloniidae, Cionel- lidae and Helicidae, termination of shell growth is marked by the formation of a re- flected lip. For the introduced Oxychilus, a definite narrowing of the shell aperture is an equivalent indicator of shell maturity. Thus, selection of an adult shell that is average in size and form is easily accomplished. All representative adult specimens were selected and measured before any analysis was attempted. We are confident that our representative specimens lie within 1.5 stan- dard deviations on each side of the population mean. Since we are concerned with size dif- ferences greater than 40% for shell volume and generally even larger differences for par- ticular measurements, we consider that our data base is adequate for the intended study. We lacked sufficient series of adults to make comprehensive measurements for each spe- cies. Appendix 1 reviews data on variation in some species in comparison with the average adult selected. BASIC MEASUREMENTS For the pulmonates, which are hermaphro- ditic, basic measurements were made on an adult shell; for the dioecious prosobranchs, measurements were taken on both a male, which is consistently smaller, and a female shell. Shells over 5mm in maximum dimen- sion were measured with a vernier caliper, shells under 5 mm with an ocular micrometer. Accuracy of measurement is as summarized by Solem (1976: 15). Fig. 1 indicates how the basic measurements were taken: SOLEM 8 CLIMO FIG. 1. Basic measurements of shells. Whorls were recorded to the nearest "sth; Shell diameter is A to B distance; Spire diameter is C to D distance; Shell height is B to G distance; Spire elevation is B to E distance; Body whorl height is E to G distance; Body whorl descension is E to F distance; Umbilical width is H to | distance. Two standard ratios were calculated: H/D ratio, a measure of proportionate shape obtained by dividing Shell height by Shell diameter; and D/U ratio, an indication of proportionate Umbilical width obtained by dividing Shell diameter by Umbilical width. Where the radial sculpture was clear enough and large enough to be counted at 60x magnification, the number of major ribs on the body whorl was counted, and then an index of rib spacing, Ribs/mm оп Body whorl, calculated by the formula: radial ribs on body whorl - = Ribs/mm т x shell diameter IE to provide data on rib frequency and spacing. The minor error inherent in this index has been discussed by Solem (1976: 42-43). The above measurements are standard in systematic malacology and are useful in systematic discrimination of taxa. They are of less utility in trying to indicate the relative size of species. Nobody has devised an acceptable volume measure of a crawling snail. The foot varies in length from roughly the diameter of the shell aperture in such taxa as Cytora, Laoma, and “Phrixgnathus” to the very long, slender foot with prominent mucous pore and “horn” seen in such genera as Allodiscus or Otoconcha. The body size of snails varies dramatically with water content. Temporary water volume loss of 30% with concomitant body shrinkage is not unusual. This fact produces an unac- ceptable margin of error in measuring an- imals. Material in preservative varies so much in degree of contraction that no comparable measurements can be made. Methyl alcohol produces three to four times the shrinkage occurring when ethanol is used. We have not attempted to make live or preserved speci- men body measurements. We suggest that a more accurate predictor of normal body size is to prepare estimates of shell volume. The reasons for this are simple. When the snail is extended and crawling, the SYMPATRIC NEW ZEALAND LAND SNAILS 5 pallial cavity inside the shell body whorl con- tains sufficient space for the head and foot to be withdrawn completely. If the animal is part- ly desiccated, it, plus a reserve store of water, can retract for a significant part of a whorl. If the snail is fully hydrated, it may have to eject water in order to complete withdrawal. Thus, the volume of space inside the shell may be taken as an average approximation of actual body volume. It is not a perfect estimate. Species with greatly increased whorl count, such as Laoma leimonias (Fig. 3d) with 7% whorls, may have the soft parts absent from the first two or three shell whorls, and in many species there is a fraction to a whole whorl of empty space above the apex of the soft parts. As the snail has added later whorls, the apical soft parts have been withdrawn from the orig- inal apical whorl(s). There is thus a small amount of empty space in the upper spire of the shell. Short of working out individual growth curves from analysis of cross-section- al views prepared of each species and correcting for this empty space, any internal volume estimate will be subject to major error. If we were dealing with actual biomass, then calculation of such internal shell space volumes would have been appropriate. Our concern is with occupied habitat space—how much space in the litter will be needed by the typical adult snail shell. We thus have pre- pared estimates of total shell volume for this study, rather than biomass or internal shell volume. While these total shell volume calculations probably are roughly proportional to snail biomass, differences in shell thickness, ex- ternal contours and growth pattern would cause variation. Total shell volume is in- tended to be an indication of the physical space needed by the snail in its habitat and should not be interpreted otherwise. The method of preparing shell volume es- timates is given in Appendix 1, together with evidence as to individual species size, shape and volume variability. TAXON-LINKED VARIATION IN BASIC PARAMETERS Most readers will be unfamiliar with the appearance of these species. Figs. 2—4 con- tain side view line drawings of some char- opids and all punctids mentioned in this re- port. Table 1 gives a reference to illustrations of all native New Zealand species, referring to Figs. 2—4, the standard monographs of Suter (1915) and Powell (1976, 1979), or a few technical reports when no other illustrations have been published. Inclusion also of top and basal views would have been useful, but this was not practical as no stock of such figures was available to us. Few of the figures are of Manukau Peninsula specimens. Some may reflect local geographic or clinal differ- ences from the discussed Manukau mor- photype. Nevertheless, the figures do give the general aspect of each species, although they are not intended to represent the precise de- tails of spire elevation, rib spacing, body whorl contour, or color patterns. Tables 2-6 outline taxonomically linked patterns of shell variation in the Manukau Peninsula fauna. Two families predominate, the Charopidae with 40 species and the Punctidae with 27 species. The 15 “other” taxa include four European imports (species 79-82 in Table 1), one Hydrocenidae (Om- phalorissa purchasi), six Liareidae (species 2—7 in Table 1), one Achatinellidae (Lamel- lidea novoseelandica), and three Rhytididae (two Delos and one Rhytida). Of these, the introduced Cionella, Omphalorissa, Lamel- lidea, both Liarea, Cytora pallida and C. tor- quilla are elongated to turritelliform in shape with H/D ratios of 1.358-2.075; the introduced Helix aspersa, and native taxa Cytora cytora and С. hedleyi have nearly equal heights and diameters; Rhytida greenwoodi has the height two-thirds of the diameter; the in- troduced Vallonia and Oxychilus, plus both Delos have more nearly planiform shells, with the height about half the shell diameter. None of these species show especially unusual shell dimensions, allowing for increased whorl count in lanceolate-shaped taxa (Liarea and elongated Cytora). Most of these species are strictly terrestrial, except for Lamellidea which is arboreal, and Omphalorissa which is both arboreal and terrestrial. There is no correlation between habitat preference and shell shape in this grouping. Total size, as indicated by the Ad- justed Shell Volume (ASV), has the arboreal Lamellidea (1.30 mm?), ambivalent Om- phalorissa (0.95 mm’), and terrestrial Cytora torquilla (1.01 mm?) at the lower size range. The carnivorous Rhytida greenwoodi (3,523 mm?) and herbivorous Helix aspersa (6,148 mm?) have ten to eighteen times the ASV of the next largest species (the charopid Allodiscus dimorphus, 340 mm?). Vallonia and two Cytora (cytora and hedleyi) are in the 6 SOLEM & CLIMO ZA HATE М | 7 FIG. 2. Shells of some Manukau Peninsula charopids: a, Allodiscus dimorphus; b, “Allodiscus” urquharti; с, Allodiscus п. sp. aff. granum; а, Allodiscus planulatus; e, Flammulina perdita; f, Charopa coma; g, “Charopa” п. sp. aff. pseudanguicula; h, “Charopa” pseudanguicula; i, “Charopa” costulata; |, “Charopa” ochra; К, “Charopa” pilsbryi, |, “Charopa” fuscosa; m, “Charopa” chrysaugeia; п, Huonodon pseudoleiodon; о, Huonodon hectori; p, Therasiella n. sp. aff. neozelanica; q, “Mocella” n. sp. aff. maculata; r, Fectola mira. Scale line equals 1 mm. Drawings by F. M. Climo from unpublished manuscripts. SYMPATRIC NEW ZEALAND LAND SNAILS 7. k | FIG. 3. Shells of Мапикаи Peninsula punctids: a, “Laoma” тапае; b, “Phrixgnathus” poecilosticta; с, “Phrixgnathus” moellendorffi; d, Laoma leimonias; e, Laoma п. sp. aff. marina; f, Laoma marina; д, “Phrixgnathus” erigone; h, “Phrixgnathus” levis; i, “Phrixgnathus” conella; j, “Phrixgnathus” ariel; k, “Phrix- gnathus” n. sp. 59; |, “Phrixgnathus” glabriusculus; m, “Phrixgnathus” pirongiaensis. Scale lines as marked. Drawings by F. M. Climo from monographic review of New Zealand punctids that is in preparation. 8 SOLEM & CLIMO all A = = \ \ 7 ках TS NE y \ \ = A \ \ АА \ « Y N FIG. 4. Shells of Manukau Peninsula punctids: a, Paralaoma caputspinulae; b, “Paralaoma” п. sp. 40; с, “Paralaoma’ п. sp. aff. 33; а, “Paralaoma” francesci; e, “Paralaoma” serratocostata; f, “Paralaoma” п. sp. 1; g, “Paralaoma” п. sp. 33; h, “Phrixgnathus” elaiodes; i, Obanella rimutaka; |, Pasmaditta jungermanniae; К, “Paralaoma" п. sp. 8; |, “Paralaoma’ п. sp. 38; m, “Paralaoma” п. sp. 29; п, “Paralaoma” lateumbilicata. Scale line equals 1 mm. Drawings by F. M. Climo from monographic review of New Zealand punctids that is in preparation. SYMPATRIC NEW ZEALAND LAND SNAILS TABLE 1. List of taxa discussed. Family Hydrocenidae 1. Omphalorissa purchasi (Pfeiffer, 1862) —(Powell, 1979: 85, fig. 13-1) Family Liareidae Liarea hochstetteri carinella (Pfeiffer, 1861) —(Powell, 1976: pl. 35, fig. 7) . L. egea egea (Gray, 1850) —(Powell, 1976; pl. 35, fig. 4) . Суюга cytora (Gray, 1850) —(Powell, 1979: 85, fig. 12-3) . С. hedleyi (Suter, 1894) —(Powell, 1979: 85, fig. 12-5) . С. pallida (Hutton, 1883) —(Powell, 1979: 85, fig. 12-2) . С. torquilla (Suter, 1894) —(Powell, 1979: 85, fig. 12-4) Family Achatinellidae 8. Lamellidea novoseelandica (Pfeiffer, 1853) —(Powell, 1979: 302, fig., 74-1) Family Rhytididae 9. Delos coresia (Gray, 1850) —(Powell, 1979: 348, figs. 81-2-4) 10. D. jeffreysiana (Pfeiffer, 1853) —(Powell, 1979: 348, figs. 81-1) 11. Rhytida greenwoodi (Gray, 1850) —(Powell, 1979: pl. 64, figs. 1-2) Family Charopidae 12. Cavellia buccinella (Reeve, 1852) —(Powell, 1979: pl. 57, fig. 8) 13. С. roseveari (Suter, 1896) —(Suter, 1915: pl. 10, figs. 5a, b) 14. Mocella eta (Pfeiffer, 1853)—(Climo, 1981: fig. 1A-C) 15. “М.” п. sp. aff. maculata (Suter, 1891) —Fig. 2q 16. “M.” n. sp. aff. manawatawhia (Powell, 1935)—(aff. Powell, 1979: pl. 57, fig. 10) 17. “Charopa” pseudanguicula lredale, 1913—Fig. 2h 18. “С.” chrysaugeia (Webster, 1904)—Fig. 2m 19. “С.” п. sp. aff. pseudanguicula Iredale, 1913—Fig. 2g 20. “С.” fuscosa (Suter, 1894)—Fig. 21 21. “C.” pilsbryi (Suter, 1894)—Fig. 2k 22. C. coma (Gray, 1843)—Fig. 2f 23. “С” costulata (Hutton, 1883)—Fig. 21 24. “C.” ochra (Webster, 1904)—Fig. 2) 25. Fectola mira (Webster, 1908)—Fig. 2r 26. F. unidentata Climo, 1978—(Climo, 1978: fig. 4) 27. F. infecta (Reeve, 1852)—(Climo, 1978: fig. 3) 28. Huonodon pseudoleiodon (Suter, 1890)—Fig. 2n 29. H. hectori (Suter, 1890)—Fig. 20 30. Allodiscus dimorphus (Pfeiffer, 1853)—Fig. 2a 31. A. tessellatus Powell, 1941—(Powell, 1979: pl. 58, figs. 1-2) 32. “A.” urquharti Suter, 1894—Fig. 2b 33. A. п. sp. aff. granum (Pfeiffer, 1857)—Fig. 2c 34. A. planulatus (Hutton, 1883)—Fig. 2d 35. Geminoropa cookiana (Dell, 1952)—(Climo, 1981: fig. 2A-C) 36. Serpho kivi (Gray, 1843)—(Powell, 1979: pl. 65, fig. 9) 37. Flammulina perdita (Hutton, 1883)—Fig. 2e 38. F. chiron (Gray, 1850)—(Suter, 1915: pl. 26, figs. 13a-b) 39. “F.” feredayi (Suter, 1915: pl. 9, figs. 10a-b) 40. “Thalassohelix’ ziczac (Gould, 1848)—(Powell, 1979: pl. 65, fig. 10) 41. Therasia decidua (Pfeiffer, 1857)—(Suter, 1915: pl. 50, fig. 8) 42. Suteria ide (Gray, 1850)—(Powell, 1979: pl. 65, figs. 1-2) 43. Phenacohelix giveni Cumber, 1961—(Powell, 1979: pl. 65, fig. 11) 44. P. pilula (Reeve, 1852)—(Suter, 1915: pl. 26, figs. 7a-b) 45. “Р.” п. sp.—(Suter, 1915: aff. pl. 26, figs. 5a-b—without color pattern) 46. P. ponsonbyi (Suter, 1897)---(Suter, 1915: pl. 26, figs. 8a-b) 47. Therasiella neozelanica Cumber, 1967—(Cumber, 1967d: fig. 1G-I) 48. T. serrata Cumber 1967—(Cumber, 1967d: fig. 1A-C) 49. Т. п. sp. aff. neozelanica Cumber, 1967d—fig. 2p 50. Т. celinde (Gray, 1850)—(Cumber, 1967d: fig. 3) 51. T. tamora (Hutton, 1883) —(Cumber, 1967d: fig. 2) NONADN SOLEM 8 CLIMO TABLE 1 (Continued) Family Punctidae 52. Obanella rimutaka Dell, 1952—Fig. 4i 53. “Laoma” mariae (Gray, 1843)—Fig. 3a 54. [. п. sp. aff. marina (Hutton, 1883) —Fig. 3e 55. L. marina (Hutton, 1883)—Fig. 3f 56. L. leimonias (Gray, 1850)—Fig. 3d 57. “Phrixgnathus” erigone (Gray, 1850)—Fig. 3g 58. “P.” ariel (Hutton, 1883)—Fig. 3j 59. “P.” elaiodes Webster, 1904—Fig. 4h 60. “P.” moellendorffi Suter, 1896—Fig. 3c 61. “P.” conella (Pfeiffer, 1862)—Fig. 3i 62. “Р.” poecilosticta (Pfeiffer, 1852)—Fig. 3b 63. “Р.” glabriusculus (Pfeiffer, 1853)—Fig. 81 64. “P.” pirongiaensis Suter, 1894—Fig. 3m 65. “P.” levis (Suter, 1913)—Fig. 3h 66. “P.” n. sp. 59 (= glabriusculus of authors)—Fig. 3k 67. Pasmaditta jungermanniae (Petterd, 1879)—Fig. 4j 68. “Рагааота” п. sp. 38—Fig. 41 .” п. sp. 29—Fig. 4m = Me sph 1—Figet Mn. sp. 8—Fig. 4k = о оо ‘P.” п. sp. 40—Fig. 4b 75. P. caputspinulae (Reeve, 1852)—Fig. 4a 76. “P.” francesci (Webster, 1904)—Fig. 4d 71. PR." n. sp. 33—Fig: 4g 18-MPAnASspralf 33—20} 46 INTRODUCED ТАХА 79. Vallonia spp. (Family Valloniidae) .” lateumbilicata (Suter, 1890)—Fig. 4n P 73. “P.” serratocostata (Webster, 1906) —Fig. 4e IR 80. Oxychilus cellarius (Müller, 1774) (Family Zonitidae) 81. Cionella lubrica (Müller, 1774) (Family Cionellidae) 82. Helix aspersa (Müller, 1774) (Family Helicidae) 3.2-5.6 тт? range, and the remaining seven species are in the 12.1-72.8 mm? classes. The punctids and charopids show more easily recognizable differences. Tables 2 and 3 show that the charopids are significantly larger in diameter (Table 2), but tend toward a lower H/D ratio (Table 3) because the shell height generally is in a reduced range in proportion to the shell diameter. The three most elevated charopid shells are those of Serpho kivi (H/D ratio 0.771), Allodiscus n. sp. aff. granum (Fig. 2c, H/D ratio 0.788), and “Phenacohelix’ n. sp. (H/D ratio 0.852). The first species is normally found foraging on large leaves or on tree trunks, the latter two are terrestrial. Those punctids with H/D ratio above 0.850, include the characteristic in- habitant of new leaves on top of the litter (Laoma leimonias, Fig. 3d, H/D ratio 1.438), ground surface under moist litter (“Phrix- gnathus” poecilosticta, Fig. 3b, H/D ratio 0.892), in drier friable litter (*Paralaoma” fran- cesci, Fig. 4d, H/D ratio 1.035), and the characteristic tree branch and sapling species (“Phrixgnathus” erigone, Fig. 3g, H/D ratio 1.026). The way in which these increased H/D ratios is achieved differs considerably. The four punctids have increased whorl counts and strong spire protrusion; Serpho kivi shows moderate whorl increment and strong spire protrusion; “Phenacohelix” n. sp. and Allodiscus п. sp. aff. дгапит show a combina- tion of spire protrusion and lateral compres- sion of the body whorl (Fig. 2c). The latter two species lack any trace of a keel or peripheral angulation; the others have sharply angled or keeled peripheries. Whorl count distribution (Table 4) confirms the difference, in that mean whorl count for the punctids is one interval less than for the charopids. Those few punctids with enlarged whorl counts, Laoma leimonias (Fig. 3d), “Phrixgnathus” erigone (Fig. 3g), and “P.” poecilosticta (Fig. 3b), also have the in- SYMPATRIC NEW ZEALAND LAND SNAILS 11 TABLE 2. Height and Diameter distributions. Height Interval Other Charopid Punctid Diameter Total Other Charopid Punctid Total 0.50— 1.00 1:50 151 2100 202150 2.51— 3.00 3.01— 3.50 3.51— 4.00 4.01— 4.50 4.51— 5.00 3:.01—9:50 5.51— 6.00 6.01— 6.50 Goi 7200 7.01— 8.00 8.01— 9.00 9.01-10.0 10: UA) = 14.2 1 = = 21-24 1 * NIN | — D © © O1 © © O1 ¡e = * | = | | + À © — © ls] | -=-+-nonwon- — N N = ма | Inn | — * | DEN D IN nm * | | *Indicates an introduced species. TABLE 3. Height/Diameter Ratio distributions. Interval Other 0.400—0.450 ile 0.451—0.500 1 0.501—0.550 1 0.551-0.600 qe 4 Charopid Punctid Total — = 0.601—0.650 0.651—0.700 0.701-0.750 0.751—0.800 = 0.801—0.850 = 0.851—0.900 = 0.901—0.950 = 0.951—1.000 1 1.001—1.050 1 1.051—1.100 1 1.101—1.500 2 = 4 4 | = | vw 1.501—1.800 1.801-2.100 IE een | aso=o=lulrodoron *Indicates an introduced species. creased H/D ratio discussed above. Reduc- tion in whorl count is more frequent in the Charopidae. The “Flammulina”-type taxa (Fig. 2e) are early stages in a slug lineage, and hence show initial stages in shell whorl reduc- tion. The reason why several of the “Charopa” group show only 3% or 3% whorls, while the median for the Charopidae is 4% whorls, is TABLE 4. Whorl Count distributions. Whorl counts grouped Other Charopid Punctid Total 27% — 1 — 1 314 24 3 1 6 3% 1 3 3 7 4 3 3 4 10 4% 2 8 6 16 4% 1 5 4 10 5% 1 11 2 14 5% 1 6 4 11 57 1 — — 1 6% 1 — 2 3 6% 1 — — 1 7 = = = = 7% 1 — 1 2 *Indicates an introduced species. unknown to us. Since the three Flammulina show different habitat preferences (tree trunks, undersides of logs, near ground sur- face in deep wet litter) and the “Charopa” with greatly reduced whorl counts are equally var- ied (chrysaugeia, Fig. 2m, probably arboreal; fuscosa, Fig. 21, under compressed broad leaf litter; pilsbryi, Fig. 2k, under loosened bark of fallen logs; costulata, Fig. 2i, in well- decomposed, powdery litter near logs), this 12 SOLEM & CLIMO type of change cannot be linked simply with habitat preference. Umbilical proportion distributions are charted (Table 5) only for the Charopidae and Punctidae. The other taxa have closed umbil- ici except for the three Rhytididae, Vallonia and Oxychilus. Those with open umbilici are in the 2.70-5.04 range. The dominant families show clear differences. Charopids tend toward widely open or narrow to closed umbil- ici, while the punctids show basically narrow to closed, with a few moderately open. The more openly umbilicated punctids are not tax- onomically clustered, and show different habi- tat preferences. Adjusted Shell Volume (ASV) distribution (Table 6) again shows the Charopidae as TABLE 5. Distribution of umbilical proportions. Interval Punctids Total 2.00-2.50 2.51-3.00 3.01-3.50 3.51—4.00 4.01—4.50 4.51-5.00 5.01-6.00 6.01-7.00 7.01-8.00 8.01-10.0 10.1-15.0 15.1-20.0 Lateral crack Closed Charopids |enr-o-om VOVOAD0DN==w0=| bh. SJ © ®@ + O1 D — O1 U1 ND) O — Oo Saw | TABLE 6. Adjusted Shell Volume (ASV) distribu- tion. Interval in mm? Other Charopid Punctid Total 0.25-0.50 = — 2 2 0.51-1.00 1 — 5 6 1.01-1.50 2 1 2 5 1.51-2.00 — — 2 2 2.01-3.00 — 5 5 10 3.01-5.00 2* 5 2 9 5.01-7.00 1 4 2 7 7.01-9.00 = 6 1 7 9.01-15.0 2. 6 3 11 15.1-40.0 2 3 3 8 40.1-75.0 3 4 — i 75.0-150.0 — 3 = S 150.1-350.0 — 3 = 3 3,000—6,000 23 —= — 2 *Indicates an introduced species. being, in general, larger than the Punctidae, but the dispersion for each family is clearly not unimodal. The probable reasons for this are discussed below. The only really small charopid is “Allodiscus” urquharti (Fig. 2b, 1.30 mm?), an inhabitant of broken-down lit- ter. The node of large punctids includes one tree trunk species (“Phrixgnathus” ariel, Fig. 3j, 12.3 mm?) and five ground surface under wet litter taxa that vary from 9.1-31.1 mms. Only one of these, “Phrixgnathus” poecilostic- ta, has a markedly increased whorl count (6/4) to explain partly the large size. How snails get big is not a simple matter. Increase in whorl count is an obvious means, but where this occurs, the animal often with- draws from the earlier whorls. These can be sealed off by a calcareous plug in such taxa as Urocoptidae, Rumina, some Clausiliidae, and New World Pomatiasidae, with the early whorls breaking off subsequently. Where the shell shape is lanceolate, increased whorl number is a very practical option, producing a very high H/D ratio and permitting total whorl counts in excess of 20 to be reached. For essentially planulate taxa or those with near globular shape, now useless early whorl de- collation is not a viable alternative. Massive withdrawal from early whorls has been demonstrated for the planulate helicarionid Coxia m. macgregori (Cox, 1870) and the endodontid Libera f. fratercula (Pease, 1867) (see Solem, 1976: 95, fig. 55). We predict this will be found also in such planulate taxa as the Brazilian Polygyratia polygyrata (Born, 1778) and the Mexican urocoptid Hendersoni- ella palmeri (Dall, 1905) (see Zilch, 1959— 1960: 529, fig. 1857; 604, fig. 2119). The only Manukau Peninsula species that has т- creased size in this manner is the com- paratively small Laoma leimonias (Fig. 3d), height 2.57 mm with 7% whorls. Live speci- mens show that the first two to three whorls are without soft parts, although whether this space contains an air bubble or a water reser- voir is unknown at present. Species with mod- erate whorl increments, such as “Phrix- gnathus” erigone (Fig. 3g), “P.” poecilosticta (Fig. 3b), “Paralaoma” francesci (Fig. 4d), Cavellia roseveari, “Charopa” pseudanguicu- la (Fig. 2h), C. coma (Fig. 2f), Allodiscus dimorphus (Fig. 2a), Serpho kivi, and “Phena- cohelix” п. зр., also can show gross in- crements with thickenings of the ribbing, enlagement of both the nuclear and post- nuclear whorls. More frequently, size in- crease is a result of the latter process. How- SYMPATRIC NEW ZEALAND LAND SNAILS 13 ever, none of the Manukau Peninsula taxa, except Laoma leimonias, show a clear and notable pattern of whorl increment change. While there are some trend differences be- tween the Charopidae and Punctidae in re- spect to individual size and shape paramet- ers, the above brief review of departures from the norm shows no simple and direct correla- tions between such variations and space or moisture preferences by the species involved. Some variations in shell contours are partly correlated with habitat specialization. De- velopment of a keel is restricted to the punc- tids, and Liarea hochstetteri carinella. In that species, Laoma leimonias (Fig. 3d), “Para- laoma” serratocostata (Fig. 4e), and “P.” fran- cesci (Fig. 4d), the keel is low on the whorl profile, a result of spire elevation and high H/D ratio. In “Laoma” mariae (Fig. 3a), L. n. sp. aff. marina (Fig. 3e), “Phrixgnathus” poecilosticta (Fig. 3b), and “P.” levis (Fig. 3h), the keel is medial on an obtusely angled periphery. Of the Laoma and “Phrixgnathus” taxa listed above, only “P.” ariel is arboreal, with the others mostly ground space associ- ated. The remaining “Phrixgnathus” lack keels, with the arboreal “P.” erigone (Fig. 3g, sharply angled periphery) and “P.” elaiodes (Fig. 4h, weakly angled periphery), sus- pended litter in bracts (“Р.” п. sp. 59, Fig. ЗК, weakly angled), ground surface in drier litter (“P.” moellendorffi, Fig. 3c, weakly angled) and spaces in wet litter above ground level (“P.” glabriusculus, Fig. 3l, rounded periph- ery), Suggesting that possibly possession of a keeled periphery is an advantage in occupy- ing the ground surface under moist to wet litter, although the association is far from complete. Other taxa with sharply angled pe- ripheries include the arboreal Serpho kivi, and the basically ground level Therasiella, Cytora pallida, Therasia decidua, Obanella rimutaka (Fig. 41), and “Paralaoma” п. sp. 38 (Fig. 41), which show different habitat preferences, but are terrestrial rather than arboreal. A few spe- cies show a weak mid-whorl angulation of the periphery—Rhytida greenwoodi, “Mocella” п. sp. aff. manawatawhia, Paralaoma caput- spinulae (Fig. 4a), “P.” n. sp. 33 (Fig. 4g), and “Р.” п. sp. aff. 33 (Fig. 4c), but the remainder show evenly rounded or slightly laterally flat- tened peripheries as in the larger Allodiscus. The 15 species of punctids with angulated to carinated peripheries are not closely re- lated. Climo (in preparation) refers them to 12 genera in three subfamilies. Five species in five genera in three subfamilies occupy ground surfaces under wet litter and have protruded thread-like keels. They are “Laoma” mariae (Fig. 3a), “Phrixgnathus” poecilosticta (Fig. 3b), “Paralaoma” serrato- costata (Fig. 4e), Laoma n. sp. aff. marina (Fig. 3e), and “Phrixgnathus” conella (Fig. 3i). The other punctids with thread-like keels are Laoma marina (Fig. 3f), which lives “on leaves and twigs in wet areas”; “Phrixgnathus” piron- giaensis (Fig. 3m), found in “wet deep litter”: Laoma leimonias (Fig. 3d), found “in wet, undecomposed, broad-leaf litter”; “Phrix- gnathus” ariel (Fig. 3j), found on “trunks and branches of larger trees”; and “Phrixgnathus” levis (Fig. 3h), found “under moist broadleaf litter” (Solem, Climo & Roscoe, 1981: appen- dix 1). The presence of a keel is associated with the ground surface habitat, but it is not restricted to this niche. Where there is relatively close relationship among Manukau Peninsula punctids, true Laoma for example, shelter site preferences are diverse. Laoma leimonias (Fig. 3d) lives at the top of the litter among newly fallen leaves; “Phrixgnathus” erigone (Fig. 3g) pre- fers encrusting materials on saplings and small tree trunks; L. marina (Fig. 3f) is on leaves and twigs in wet litter above ground level; and L. n. sp. aff. marina (Fig. 3e) is on the ground surface under deep litter. Thus, the similarity of keeled structure in ground space taxa is not caused by mono- phyly, and the only set of congeneric punctids differs radically in its shelter site preferences and shell forms. POST-APICAL SCULPTURE The post-apical sculpture found on the Manukau Peninsula species ranges from smooth, glossy surfaces from which even in- cremental growth irregularities are absent, to quite prominent periostracal extensions in a series of mostly unrelated taxa. This extreme is discussed first. Five species have periostracal “hairs” or “setae” extending upward from the side or tops of radial ribs. These are easily abraded and will be partly eroded or missing in many individuals. They are microscopic in size and regularly spaced on the main ribs of “Para- laoma” serratocostata (Fig. 4e). In “P.” fran- cesci (Fig. 4d) they are much longer and grouped into two spaced clusters of three or four each at intervals along the radial ribs above the periphery, but are single and at 14 TABLE 7. Radial periostracal sculpture. SOLEM 8 CLIMO Largest Periostracal Habitat Species dimension ASV structure preference 3 Liarea egea 6.95 mm 28.6 ridges wet piles 4 Cytora cytora 2.63 547: 0.2 mm hairs friable, well-drained 5 C. hedleyi 25 32 ridges rimu litter 6 C. pallida 48 18.5 ridges unknown 7 C. torquilla 23 1.2 ridges friable, well-drained 30 Allodiscus 9.2 340.2 ridges very rotten logs dimorphus 40 “Thalassohelix” 125 134.4 0.5 mm hairs on ground under logs ZICZAC 42 Suteria ide Vol 118.5 0.7 mm hairs on ground under logs 47 Therasiella 2.37 3.7 0.15 mm triangular friable, well-drained neozelanica peripheral 48 T. serrata 2.93 6.4 0.3 mm triangular friable, well-drained peripheral 49 T. п. sp. aff. 2.93 all 0.3 mm triangular friable, well-drained neozelanica peripheral 50 T. celinde 3.0 8.0 very short peripheral underside of logs 51 T. tamora 3.43 lie 0.2 mm fan-shaped friable, well-drained peripheral 73 “Paralaoma” 1825 0.6 microscopic hairs fine grain decomposition serratocostata 76 “P.” francesci 1.45 0.9 clumped hairs deep, friable podocarp regular intervals on the shell base. In Cytora cytora the setae are 0.2mm long, spaced singly along low radial ridges at intervals slightly less than their height, and all are slanted backward from the aperture at about a 45° angle. “Thalassohelix” ziczac has 0.5 mm long setae and Suteria ide has up to 0.7mm long setae that arise at regular intervals from the top of narrow, low ridges and point directly outward. The form of the setae compares quite well with those found in the Hawaiian Cookeconcha decussatulus (Pease, 1866) (see Solem, 1976: 36, figs. 26b—c). Simple periostracal radial ridge extensions, that are very subject to wear, are found in juveniles of Liarea egea, all ages of Cytora hedleyi, C. torquilla, C. pallida, Allodiscus di- morphus (Fig. 2a), Therasiella celinde and T. tamora. п the Therasiella, these ridges are extended slightly or into long curved pro- jections from the periphery (see Cumber, 1967d: 64, figs. 2А-С; 66, figs. ЗА-С). Other Therasiella show an intensification of this phenomenon. T. neozelanica has 0.15 mm long triangular projections from the periphery (Cumber, 1967d: 62, figs. 1G-I), they reach 0.3 mm long in 7. serrata (ibid., 62, figs. 1A— C), and form a nearly continuous fringe of about 0.3 mm long triangular projections in 7. n. sp. aff. neozelanica (Fig. 2p). The taxonomic spread of hairy shells is broad, and the appearance of “hairs” clearly is of independent origin. The functional and habitat associations of the periostracal ex- tensions are not clear. Table 7 gives the largest dimension, ASV, type and size of peri- ostracal extension, and habitat preference of each species. In the log habitat, the two spe- cies with longest hairs, “Thalassohelix’ ziczac and Suteria ide, occur on the ground under logs or in deep litter by logs. Their setae could be interpreted as an aid in minimizing litter accumulation on shell surfaces. The very large Allodiscus dimorphus (Fig. 2a) is log associated, but we cannot say if it prefers the underside of the log, as does Therasiella celinde, or the ground spaces under the log. It has very low periostracal extensions. The first three species are quite large, the latter of relatively small size. Intermediate-sized spe- cies that are log associated, such as Allodis- cus tessellatus in litter or under logs, Charopa coma (Fig. 2f) in decay spaces on the log itself, and A. planulatus (Fig. 2d) found on the ground under logs, show no such periostracal SYMPATRIC NEW ZEALAND LAND SNAILS 15 features. The latter three species range in the 11.8 to 63.4 тт? ASV level, thus being in- termediate in size. T. celinde, the only log dweller in its genus, has the least prominent periostracal extensions. According to Cumber (1967d: 61), species of Therasiella “... are rendered inconspicu- ous by the habit of gathering trash on all surfaces except those immediately adjacent to the aperture. It would appear that the membraneous plaits [= peripheral periost- racal projections] have adhesive properties when moist, and this soon gathers a mis- cellaneous covering of bits and pieces... .” We cannot comment on Therasiella tamora, which was a rare species on the Manukau (Solem, Climo & Roscoe, 1981), but the other three species quite commonly are collected sympatrically and Cumber (1967d) lists a number of same day-same locality records for tamora, serrata and neozelanica. The size of the peripheral protrusions is proportionate in the four species, allowing that 7. neozelanica is much smaller and that the shorter blades of T. tamora extend well above and below the actual periphery, thus offering greater surface area than the longer triangular projections of T. serrata and Т. п. sp. aff. neozelanica (Fig. 2p). All four Therasiella live in friable, broken- down litter, in which they are joined by the similar-sized Cytora cytora with 0.2 mm slanted hairs, the very small ridged C. tor- quilla, and the microscopically haired and very small “Paralaoma” serratocostata (Fig. 4e) and “P.” francesci (Fig. Ad). Our knowledge of their ecology does not permit defining exact niches for any of these species, but it must be emphasized that their preference sites are shared with species of similar size that do not have such special features (Solem, Climo & Roscoe, 1981: fig. 2). The occurrence of such periostracal de- velopments effectively spans the volume range of snail species found on the Manukau. For only two very large species, “Thalas- sohelix” ziczac and Suteria ide that live on the ground surface under logs, can we suggest an obvious function of the periostracal pro- trusions. For the other species, we record the existence of various types of projections in the same preferred environment, but have no knowledge of their function. Normal “endodontoid” sculpture consists of major radial ribs between which lie a complex microsculpture. In both the Pacific Island En- dodontidae (Solem, 1976) and Charopidae (Solem, 1983), there is a close correlation between shell size and spacing of the major ribs. Larger species have larger ribs that are spaced more widely; small species have smaller and more crowded ribs. For the En- dodontidae, the spacing pattern for 133 spe- cies was Summarized (Solem, 1976: 44, table XV) and then graphed for a monophyletic lineage (ibid.: 48, fig. 35). There are few de- partures from this normal pattern. For the 40 species of charopids recorded from the Manukau Peninsula, the five Ther- asiella have only periostracal extensions; Serpho kivi and Therasia decidua have only weak growth irregularities; Flammulina per- dita (Fig. 2e) has a glossy, smooth shell sur- face; both Flammulina chiron and “F.” feredayi have greatly reduced surface sculp- ture although reticulated remnants can be detected; and both “Charopa” pilsbryi (Fig. 2k) and “С.” costulata (Fig. 21) have the sculp- ture detectable as composed of major and minor elements, but too fine to count at 60 x magnification. Serpho kivi (ASV is 487.9 mm?) and Flammulina perdita (ASV is 72.5 mm?) are arboreal taxa; F. chiron (ASV is 116.3 mm?) is found on the underside of logs; “Е.” feredayi (ASV is 10.5 mm?) is in deep litter near the ground; Therasia decidua (ASV is 237.6 mm?) is a ground dweller around monocots in more open areas; “C.” costulata (ASV is 13.4 mm?) lives in well- decomposed powdery litter near logs; and “С.” pilsbryi (ASV is 4.9 mm?) is under the loosened bark of fallen logs. Sculpture reduc- tion in arboreal and very large species is typical in the Charopidae (Solem, 1983), but reduction occurring in the two “Charopa” and “F.” feredayi is unexpected and we can assign no reason for this. The 28 species of Charopidae in which the sculpture could be counted are graphed in Fig. 5. ASV has been used as a better in- dicator of size than any single dimension. With the exception of four taxa in the Allodis- cus group, species 30-33, which do show a good linear correlation of rib spacing and size, the amazing fact is that there is no pattern to the scatter. The four Phenacohelix (species 43-46) maintain essentially the same rib spacing despite a sixfold size differential (ASV 10.2-60.0 mm°), and the three Fectola (species 25-27) show only minor change. Similar rid spacing occurs at widely different volumes, and quite divergent rib spacing occurs at the same volume. Considering those taxa at the 2-3 тт? 16 SOLEM & CLIMO 1000.0 100.0 ALLODISCUS LINEAGE 10.0 ADJUSTED SHELL VOLUME 0.1 3 6 9 12.15 18 921 24 27 SOUMIS 5 SCR scam RIBS/MM. FIG. 5. Relationship between rib spacing and adusted shell volume in Manukau Peninsula Charopidae. Numbers are those used in Table 1. SYMPATRIC NEW ZEALAND LAND SNAILS 7 ASV level, “Charopa” n. sp. aff. pseudan- guicula (19, Fig. 2g) and “C.” chrysaugeia (20, Fig. 2m) are both tree trunk taxa and show the sculptural extremes; “С.” fuscosa (20, Fig. 21) prefers compressed broad leaf litter; Huono- don hectori (29, Fig. 20) is in suspended litter of tree axils and bracts; and Geminoropa cookiana (35) prefers the loam and grit next to the ground under logs. The slightly larger Huonodon pseudoleiodon (28, Fig. 2n) pre- fers older, but still not decayed fallen leaves, near the top of the litter. On the Pacific Islands, the Charopidae and Endodontidae live in situations with very few sympatric species (five to ten would be nor- mal) and their sculpture spacing is closely size linked. The Manukau Peninsula char- opids show little linkage between shell sculp- ture spacing and size. We can detect no habitat preference correlation with sculpture spacing. Size, prominence and shape of the radial sculpture require only a few comments. In the large species with periostracal hairs, Suteria ide and “Thalassohelix’ ziczac, the main ribs are very low and widely spaced, and they also are small in Allodiscus dimorphus (Fig. 2a). Charopa coma (Fig. 2f) is best described as “gross” in that the ribs are very thick and prominent. Its shell is significantly enlarged in diameter over probable relatives, but is not significantly altered in whorl count. Fectola mira (Fig. 2r, 25) and, to a lesser extent, Fectola infecta (see Climo, 1978: 193, fig. 3) have enlarged and strongly sinuated sculp- TABLE 8. Punctid radial sculpture. ture combined with a tendency toward lateral whorl flattening. An interesting parallelism to this was seen in “Charopa” n. sp. aff. pseudanguicula (Fig. 2g, 19), which has sculpture shape and proportion, plus lateral whorl flattening, that closely mimics the appearance of F. mira. On several occasions Frank Climo and David J. Roscoe have col- lected samples of F. mira in the field only to find out when sorting under a microscope that a small proportion of the specimens were the “Charopa”. Our Manukau samples showed that this association is less than perfect, since 9 out of 10 live “С.” п. sp. aff. pseudanguicula were picked off tree trunks, while F. mira was found in slimy interfaces of nikau boles. The situation in the Punctidae is more com- plex. Of the 27 species of Punctidae recorded from the Manukau Peninsula, only nine show countable radial sculpture. Two of these show periostracal extensions (see above). Only “Phrixgnathus” poecilosticta (Fig. 3b) shows raised, rounded calcareous radial ribs equiv- alent to those of the Charopidae. In appear- ance we rank it as a punctid equivalent of Charopa coma in appearing “gross,” both in size and features. It is a ground surface dwell- er in good litter. Table 8 presents the data on the species of ribbed punctids, together with habitat preference information. As in the Charopidae, we can suggest no clear indication of a correlation among sculp- ture spacing, size, and habitat preference. It is quite possible that a correlation does exist, but our knowledge of local ecology is in- Sculpture Habitat type preference calcareous, 10/mm sinuated, 9/mm, perio- periostracal, 33/mm, also spiral corrugations thin periostracal, 15/mm deciduous periostracal on low bases, 40/mm, spiral corruga- same as 71, higher, 12/mm ASV in Species mm? 62 “Phrixgnathus” 17.0 poecilosticta 64 “P.” pirongiaensis 1.8 stracal 69 “Paralaoma’ п. sp. 29 0.5 70 “P.” lateumbilicata 1.0 Tl PRES el 0.4 tions ПРЕ п. р. 8 0.5 ИА НР: 5р) 40 4.3 periostracal, 13/mm, strong spiral corrugations ground surface, wet litter wet deep litter slimy litter friable dry litter deep fine grain litter deep fine grain litter medium moist litter 18 TABLE 9. Punctids with reduced sculpture. SOLEM & CLIMO ASV in Habitat Species mm? preference 53 “Laoma” mariae AN slimy surfaces 54 L. n. sp. aff. marina 10.0 drier than species 53, but similar 55 L. marina 6.5 leaves and twigs in wet areas 57 “Phrixgnathus” erigone 2.5 tree trunks 58 “P.” ariel 12.3 more open tree trunks 59 “P.” elaiodes 2.8 trees with scaly bark 60 “P.” moellendorffi 6.8 ground surface under drier litter 61 “P.” conella 9.1 ground surface under moist litter 67 Pasmaditta jungermanniae 2.4 uncertain 68 “Paralaoma” п. sp. 38 ES slimy interfaces adequate to recognize and define any such association. Four punctids, Obanella rimutaka (Fig. 4i), Paralaoma caputspinulae (Fig. 4a), “Р.” п. sp. 33 (Fig. 4c), “Р.” п. sp. aff. 33 (Fig. 4g), have detectable traces of sculpture, but it is re- duced to the point that primary and secondary riblets cannot be clearly distinguished. The degree of reduction is equivalent to that found in Flammulina chiron. Four punctids, Laoma leimonias (Fig. 3d, on older leaves near top of litter), “Phrixgnathus” levis (Fig. 3h) and “Р.” glabriusculus (Fig. 31) (both moist litter gener- alists), and “P.” n. sp. 59 (Fig. 3k, suspended litter in bracts), have smooth, shiny shells. The remaining ten punctids have irregular growth striae or faint remnants of radial sculp- ture. They include a variety of habitats and sizes (Table 9), and we are not able to make firm functional associations. Of the non-punctids remaining, Liarea hochstetteri carinella has very weak sculp- tural features, the arboreal Lamellidea novoseelandica and the deep litter to arboreal Omphalorissa purchasi have smooth, shiny shells. Both species of Delos have shiny shells relieved by the incised lines typical of many smaller rhytidids (Solem, 1959: pl. 13, fig. 5). Rhytida greenwoodi and the shell rem- nant of Schizoglossa worthyae have malle- ated surfaces, rather than discrete radial or spiral sculpture. In conclusion, we are unable to associate patterns of primary shell sculpture with either shell size or habitat preferences of the spe- cies. Study of the microsculpture requires SEM analysis and is beyond the scope of this project. SHELL COLOR AND HABITAT A detailed breakdown of color variation is presented in Table 10. There is a slight taxonomic difference, with the punctids being 51.9% monochrome and 48.1% variegated, while the charopids are 40.0% monochrome and 60.0% variegated. Variegated color is achieved in a variety of ways. Most commonly there is “tiger-striping” of alternating red and yellow irregular radial flammulations, often at least partially zigzagged (Figs. 2f, 1, К, п, O, г, 3b-i). Variation of both actual and relative width of the color bands within samples of a species normally is very large and we could not discriminate classes of color band width to which species could be assigned. Nor could we categorize color brightness on a species basis. A few species are unusual in their color variegation. Suteria ide has narrow red ra- dials on yellow; Serpho kivi irregular red blotches on yellow; Cavellia buccinella, Allo- discus tessellatus and A. n. sp. aff. granum (Fig. 2c) white dots on red ground color; and Liarea hochstetteri carinella, Cytora pallida, and “Paralaoma” lateumbilicata become secondarily “flammulated” as wear and radially-oriented peel spots on the brown peri- ostracum become yellowish in tone. The last three taxa would be effectively monochrome as juveniles, variegated as adults. Particular color patterns or tones are not micro-habitat restricted. The unusual white dots on red ground color, for example, is found on the shells of dwellers in dry litter (Cavellia buccinella), under logs in wet litter (Allodiscus tessellatus), and in the upper sec- tions of mamaku piles (A. n. sp. aff. granum). There are slight proportionate differences in color, noticeable in respect to both arboreal and well-decomposed litter habitats. Of the 36 taxa with regular or irregular variegated color, six (16.7%) are arboreal and two (5.6%) in- habit arboreal litter (Huonodon hectori in tree forks or axils and “Phrixgnathus” n. sp. 59 in slime or wet debris of flax, nikau, or Frey- SYMPATRIC NEW ZEALAND LAND SNAILS 19 TABLE 10. Taxonomie distribution of shell color. Family group Color Other Rhytidids Charopids Punctids Total species White 1 — 1 — 2 Yellow-brown 3 = 6 11 20 Light brown 1 = 7 2 10 Dark brown 3 = 2 1 6 White dots on red = = 3 — 3 Red and yellow flammulations 2 1 19 12 34 Brown, wear spots yellow 2 = — 1 3 Irregular red on yellow = = 1 — 1 Greenish yellow — 2 — = 2 Narrow red radials on yellow = == 1 = 1 Totals 12 3 40 27 82 cinetia banksii), whereas of the 30 with yellow brown or light brown color, only three (10%) are arboreal (“Phrixgnathus” elaiodes, Flam- mulina perdita, Lamellidea novoseelandica). The latter two are found on exposed surfaces, while the first shelters under scaling bark. Two monochrome species (6.7%), “Charopa” chrysaugeia and Omphalorissa purchasi, are sometimes found on arboreal surfaces. Thus, only 16.7% of the monochrome taxa, but 22.3% of the variegated taxa are arboreal. Where the leaves, fronds and twigs have been fragmented into deep friable litter, a larger difference in color distribution exists. Of the 36 flammulated taxa, only “Charopa” cos- tulata is in wet powdery litter near logs. Four of the ten light brown colored species, two of the six brown taxa, but only one (“Paralaoma” n. sp. 8) of the 20 yellow brown taxa prefer this habitat. All major color variations are found in the wet litter habitats, but we could detect no significant proportionate differences. The predominance of the monochrome taxa in decomposed litter, a basically mono- chrome habitat, is not surprising, but the scat- ter of variegated taxa through exposed and cryptic ground surface habitats was un- expected. To what extent this will correlate with foraging into more exposed sites and sheltering in hidden, dark sites can only be determined after extensive studies of move- ments of species. Certainly the efficacy of variegated color in aiding concealment in many groups of animals is well documented, but we have no knowledge concerning preda- tors on the New Zealand bush snails, and the significance of this color pattern is unknown. Compared with other land snail faunas known to the authors, the high proportion of var- iegated color patterns is matched only by the New Caledonian fauna, which is predomi- nantly composed of charopids and rhytidids, compared with the charopid-punctid domi- nance in New Zealand. SHELL VOLUME AND HABITAT We are not aware that shell volume has been used previously in trying to establish niche preferences in land snails. When members of a fauna range in shell shape from planiform to globular to lanceolate, neither maximum shell diameter nor shell height will describe adequately actual shell size. We present two main analytic portraits of volume relationships. Fig. 6 graphs Adjusted Shell Volume (ASV) against shape, using H/D ratio as the shape indicator. Fig. 7 attempts to correlate shape and habitat preferences for the species actually collected in Jones Bush, the site with the largest number of species. The few truly planulated endemic snails, three Fectola (25-27, Fig. 2r) and “Charopa” n. sp. aff. pseudanguicula (19, Fig. 2g, which may have a mimetic relationship with F. mira, Fig. 2r, see Solem, Climo 8 Roscoe, 1981: 465), have a H/D ratio of less than 0.450, which also is the situation in the introduced Vallonia (79). Nine species are between 0.450 and 0.500; 15 between 0.500 and 20 SOLEM 8 CLIMO ° 82) ell 1000.0 °30 °36 04] °40 1000! “® 60 °38 10 37 31943 .22 2 = a +53 .3 2 03 5 °31 °65 . > .62 E 585 .21 13 044 a 3 . . я Ш 100 ge 51. +58 °6 81 т = 23e 34 54° 45 2 49 ‘+39 oq [a] 24 70 963 = wu 26° 16,159 60.49 17 E Aa} 55 o4 uN = 5 a < 8 .] sl AA a ee ee) .5 6 7 8 9 Lore ae 18 20 H/D RATIO FIG. 6. Relationship between shell shape and adjusted shell volume. Numbers are those used in Table 1. Circled numbers are introduced species. SYMPATRIC NEW ZEALAND LAND SNAILS 21 Hel) (U) UNDERSIDE OF LOGS (G) GROUND CAVITIES UNDER LOG | O CHAROPIDAE (GG) GROUND GRIT A PUNCTIDAE (LB) LOOSE BARK ON FALLEN LOGS O OTHER FAMILIES 1000.0 360 (G) 0400 100.0 0420 | 380° (С) | (U) | | | We 19.0 1310 = | | = | | , | | 2A 53.0 | O | > | | = | | | | (U) = 440 | ae | 210° | » e330 | | | 45,0 | | | 2 ee. [a] Е = A 500» | (de | 4940 | 60,4 0 | 17*0 26*О = 150 [480 | 55% | | | Y 14.0 | (LB) | | > 12.0 Thea | | oe AR te co hore 0210 |66^ | = 41.0 | | 280*--.-: on .280 | | a 40 | | Olano д.59 < | | AS nd goss: OU 57 A 20.0 (GG) °68A = 80 | | | | 1MeA | | | MA | | | | | | | | | | | | E |) | | | | <> сел №) Ze AGO REDES > LAN ER OA > < LOA > EN Е SE IT MEDIUM LITTER ARBOREAL FIG. 7. Adjusted shell volume distribution within shelter site niches for species found in Jones Bush. Numbers are those used in Table 1. 22 SOLEM 8 CLIMO 0.600; 24 between 0.600 and 0.700; 14 be- tween 0.700 and 0.800; two, “Phenacohelix” п. sp. (45) and “Phrixgnathus” poecilosticta (62, Fig. 3b) between 0.800 and 0.900; and 13 species above 0.900. The latter group includes the introduced Cionella lubrica (81) and Helix aspersa (82), two Liarea (2, 3), four Cytora (4-7), Lamellidea novoseelandica (8), Omphalorissa purchasi (1), and three punc- tids (Laoma leimonias [56], “Phrixgnathus” erigone [57], and “Paralaoma” francesci [76]). For the prosobranchs, both male (smaller) and female (larger) specimens have been plotted. There is a tendency for the very small, under 1 mm?, and very large, over 120 mm?, species to be globular, but no obvious pattern for those of intermediate size. There are no simple habitat-shape correla- tions in this fauna. Of the eight truly elongated species, H/D ratio over 1.200, two (Om- phalorissa purchasi sometimes, Lamellidea novoseelandica always) are arboreal, the other six (Cionella lubrica, Laoma leimonias, two Liarea, two Cytora) are strictly terrestrial. Of the five nearly globular species, Helix aspersa, “Paralaoma” francesci, and two Cytora are terrestrial, and only “Phrixgnathus” erigone is arboreal. The overall taxonomic distribution of ASV is shown in Table 6, with the native and in- troduced faunal elements separated. The in- tervals are not intended to be equal, but rather to suggest where natural breaks seem to be occurring and thus to emphasize natural rather than arbitrary clusters. The taxonomic distribution is significant, with the ASV of the Punctidae clearly less on the average than for the Charopidae. The only punctid species to exceed 30 тт? is “Laoma” mariae (Fig. 3a), an inhabitant of the wettest niche, slimy interfaces near the ground surface in very wet piles of litter. Both “Phrixgnathus” poecilosticta (Fig. 3b) and “P.” levis (Fig. 3h) are in the 17-19 тт? size range; Laoma n. sp. aff. marina (Fig. 3e), “Phrixgnathus” ariel (Fig. 3j), and “P.” conella (Fig. 3i) in the 9-12.3 тт? range; and the other 21 species are less than 7.5 mm’. The only very small charopid is “Allodiscus” ur- quharti (Fig. 2b), 1.4 mm?, while 10 species (25%) exceed 31 mm°. We do not comment on the probable reasons for this situation, since no published phylogenetic interpreta- tion of the charopid-punctid lineage exists, and this is not the place to try to fill that gap. When volume relationships (ASV) are or- ganized on a habitat basis for the 56 shelled species of native snails we actually collected in Jones Bush (Fig. 7), there is clear sugges- tion that species are rarely occupying the same space at the same size. Wherever there is less than a 40% volume difference, except for the tramp and dry fringe species, it is possible to show clear differentiation of pre- ferred habitat. Of those taxa with a preference for decomposed, friable litter, Therasiella ser- rata (48) prefers wetter heaps of litter than T. n. sp. aff. neozelanica (49), while Cytora tor- quilla (7) prefers well-drained slopes and “Allodiscus” urquharti (32) is in very damp litter. The ground or cavity taxa appear clumped, but significant preferences do exist. Kneeling while collecting Laoma п. sp. aff. marina (54) results in soaked pants, but hand- picking “Phrixgnathus” poecilosticta (62) and “P.” conella (61) produces only wetted pants. The latter two species are very different in volume (17.0 and 9.1 mm‘, respectively). L. marina (55) is on twigs and leaves above ground level in wet areas, while “Phrix- gnathus” moellendorffi (60) prefers the ground surface on drier slopes. Cavellia roseveari (13) is on soil surface in medium moisture conditions, but seems often to be absent when these larger punctids are present. “Phenacohelix” n. sp. (45) is on medium to wet slopes with nikau and flax, often in cavities. Thus, all seven species, although clumped in size and general type of habitat preferred, are recognized by skilled collectors as being most frequently found in slightly to moderately different moisture re- gimes, or if occurring in the same conditions, show clear size differences. Species inhabiting the upper part of the litter, freshly fallen to compacted but still un- decomposed leaves, sort out vertically (“Charopa” fuscosa, 20, on older leaves and Laoma leimonias, 56, on newly fallen leaves), or occur in a variety of wet litter types (Huono- don pseudoleiodon, 28), including under stones and under sides of logs, as well as the older leaf litter. Log or sticky pile taxa sort out nicely by shelter, except for the two largest species. Both “Thalassohelix” ziczac (40) and Suteria ide (42) are in cavities next to the ground under large logs, share long periostracal hairs, and reduced primary sculpture. The former is a quite rapidly moving species, while the latter is very slow moving (teste Winston Ponder), so that there may be highly signifi- cant foraging differences with only resting space clearly shared. SYMPATRIC NEW ZEALAND LAND SNAILS 23 The dry fringe and tramp species Cavellia buccinella (12), Mocella eta (14), “M.” n. sp. aff. maculata (Fig. 2q, 15), and Therasiella neozelanica (47) form a group, differing from nearest neighbor only by about 20% in volume. There are habitat differences under conditions of limited abundance (Solem, Cli- mo & Roscoe, 1981: 477—478), but they are not separable where common. The only other cluster, of three tree trunk taxa, may be an artifact caused by lack of knowledge. “Charopa” chrysaugela (Fig. 2m, 18), is a rare species that in Jones Bush was taken live on a tree trunk (one specimen) and on nikau boles (three specimens). “С.” п. sp. aff. pseudanguicula (Fig. 2g, 19) in Jones Bush had nine specimens from tree trunks and one from a nikau bole. “Phrixgnathus” elaiodes (Fig. 4h, 59) is generally found on tree trunks with scaling bark. Allowing for such imperfections of knowl- edge as have just been cited, it seems that ASV is a good indicator of niche specializa- tion, with taxa of similar volume showing dif- ferent specializations, and generally only taxa with more than 40% volume difference oc- cupying the same litter space by preference. There is no obvious differentiation of size among shelter habitats. Considerable study of exact preferences and spot abundance of these species will have to be made before this hypothesis can be confirmed or rejected, but it does suggest numerous opportunities for field study in the North Island of New Zealand. A number of other Manukau Peninsula species that might also be found in Jones Bush were identified by Solem, Climo & Ros- coe (1981). For 12 of these it was possible to specify habitat preferences. None of these conflict in volume with those currently re- corded from Jones Bush, although in the in- terest of simplicity and understandability of Fig. 7, species 5, 23, 24, 30, 31, 34, 39, 41, 46, 51, 64 and 76 have not been plotted. For species 67 and 77, habitat preference cannot be specified at present, and “Phrixgnathus” n. sp. 55, although recorded from drift at Wat- tle Bay, Manukau Peninsula by Norman Douglas, has not been included in our dis- cussions here. We thus suggest that a 71 native species land snail fauna for Jones Bush would be harmonious in that a “same site” volume dif- ferential of at least 40% would exist among species. Why this volume differential exists remains to be determined. An important in- dication that selective forces are involved comes from the spaced distribution of ASV (Table 6), which contrasts significantly with other parameters. The spaced nature of the dispersions for the Charopidae and Punctidae strongly suggest selection for differences. We do not think that competition for food is a meaningful possibility. The massive quanti- ties of litter available, small size of snails, presence of few carnivorous Snails (two Delos and one Ahytida), lack of dominant species in terms of individual numbers, and wide move- ments through microhabitats, all suggest general, rather than specialized, feeding be- havior. We suggest that investigation of ASV rela- tionships among sympatric taxa in other areas of the world may yield highly significant results in regard to coexistence of land snail species. SHELL SHAPE AND HABITAT The shape distribution of land snails in a number of faunas has been reviewed by Cain (1977, 1978a—b, 1981a-b). He demonstrated a typical bimodal pattern of scatter when max- imum height is plotted against maximum di- ameter. The high-spired, generally multi- whorled shells have the height significantly greater than the diameter and occupy an up- per scatter, while the discoidal to low-spired shells, generally with comparatively few whorls, occupy a low scatter. Except for the fauna of the Philippines and New Guinea (Cain, 1978a), which show т part an in- termediate scatter indicating a near globose shape, Cain's observations are of clear-cut shape dichotomy. His suggestion that the “tall” taxa are arboreal, or at least forage on vertical surfaces, has received some support in regard to the British fauna by the studies of Cameron (1978) and Cain & Cowie (1978). An equivalent diagram for the Manukau Peninsula taxa is presented in Fig. 8. Direct comparisons with Cain's diagrams are com- plicated by the necessity to use a double logarithmic plot for the Manukau taxa. Cain used arithmetic scales as he, in general, was dealing with relatively large taxa. The Helici- dae, for example, are in the 5-50 mm size range (Cain, 1981b; 151, 156, fig. 10). Even when faunistic diagrams include smaller Orthurethra and aulacopod taxa, for example Cain (1981b: 158-164, figs. 11-17), most of the species are ....yer than 10 mm in diam- eter. 24 [100.0 10.0 HEIGHT SOLEM 8 CLIMO 46 37 43 82e 3 15e “29 HR E: 32 73. 22680 18 25 то 72. *69 35*203.19 78* +71 1770 0.1 1.0 | 222] DIAMETER 10.0 100.0 FIG. 8. Relationship between shell height and shell diameter in Manukau Peninsula land snails. Numbers are those used in Table 1. The Manukau taxa, in contrast, are quite small. Only two native species, Rhytida greenwoodi (11) and Serpho kivi (36) plus the European Helix aspersa (82) exceed 10 mm in diameter, and two native species, Allodis- cus dimorphus (Fig. 2a, 30) and Therasia decidua (41) are 8-10 mm in diameter. Ofthe 82 species, 68 (82.9%) are less than 5 mm in diameter and 74 (90.2%) are less than 5 mm in height. The Manukau fauna is thus com- posed of much smaller snail species than in the faunas analyzed by Cain. Not only is there an order of magnitude size differential, but the Manukau taxa do not break up into separate scatters. Most species lie just below or mod- erately below the mid-line, significantly above the “low scatter” position of Cain's diagrams. The few taxa above the mid-line are phyleti- cally diverse, comprising the hydrocenid Om- phalorissa purchasi (1); both Liarea (2, 3) and two of the four Cytora (6, 7) in the Liareidae; the only achatinellid species, Lamellidea novoseelandica (8); a punctid, Loama leimon- газ (56); and the introduced Cionella lubrica (81). Of these eight “tall taxa,” only Lamel- lidea is primarily arboreal. Omphalorissa is occasionally arboreal and the remaining six species are litter dwellers. Since only about SYMPATRIC NEW ZEALAND LAND SNAILS 25 Уп of the Manukau taxa is arboreal, this habitat distribution of “tall taxa” is normal. The greater central tendency of the Man- ukau scatter and the absence of a clear di- chotomy is highly distinctive. We have very little understanding of the physical forces ex- erting selective pressures in the litter. It is quite probable that small sized species are more affected by adhesive tendencies of moist granules or decayed fragments of plants, can move through litter regardless of shell shape with relative ease, and are more sensitive to micro-differences to moisture than the generally larger taxa of the Northern Hemisphere faunas charted by Cain. The reasons for the difference in the Man- ukau fauna remain to be investigated, but the existence of this altered pattern is obvious. DISCUSSION Surprisingly few correlations between par- ticular structures and habitat preference were identified in this analysis. Presence of a sharp keel was associated with taxa occupying the ground surface under thick, moist to wet litter, although peripheral angulation of the shell was found in taxa occupying a variety of habitats. There was no phyletic basis to the keel-habitat association. Periostracal fringes or hairs occurred mostly in species sheltering in friable, broken-down litter, but many spe- cies in this habitat did not have periostracal extensions. There was a slight overrepresen- tation of shells with variegated color pattern among arboreal taxa, and a stronger associa- tion of light brown or dark brown monochrome taxa with friable, broken-down litter. We must stress that a majority of the spe- cies has been found in several microhabitats (Solem, Climo & Roscoe, 1981: 463, table 5) and our own collecting demonstrated that taxa sheltering in slime-filled interfaces of nikau boles or mamaku fronds at least some- times forage on tree trunks. Other taxa may be more sedentary, i.e., Therasiella, Liarea, Cytora, but until information is accumulated on species movements, more precise correla- tion of structure and habitat will not be possi- ble. Many structural features may be advanta- geous during foraging, but neutral to mildly disadvantageous in shelier or mating areas. Thus, we consider our results to be pre- liminary and of value primarily in pointing out where additional work is needed. We have chosen not to discuss in detail the possible ecological significance of the color and shape correlations as deduced from equivalent observations on insects and vertebrates, since predator, diel cycle movement, and feeding data are not available for the snails. More significance can be placed on the departures from the expected patterns that we demonstrate. In both the Endodontidae (Solem, 1976) and Charopidae (Solem, 1983) from the Pacific Islands, radial rib spacing and prominence correlates closely with shell size and habitat. Smaller species have more and more crowded ribs, larger species have fewer and more widely-spaced ribs. Large species and semiarboreal taxa have a tendency toward reduction or loss of shell sculpture. Discovering that rib prominence and spacing in the New Zealand Charopidae and Punc- tidae (Tables 7-9, Fig. 5) is not size or habitat correlated was surprising. Similarly, the absence of a bifurcated scatter to the plot of shell height and shell diameter (Fig. 8) opens up many intriguing possibilities for investigation both in New Zea- land and for parallel studies elsewhere. Is the smaller size of the New Zealand species a major factor? Do the New Zealand species have a greater vertical foraging zone than members of northern faunas? Where do the South African and Australian faunas plot out in relation to the two observed modes? What are the mechanics of movement through litter in terms of shell size and shape? To what extent is the bimodality of plots for species from other areas a function of shell geometry and whorl number factors? Finally, perhaps the most significant finding is the discovery of spaced distribution of shell volume (Table 6) in the Manukau fauna and the existence of clear volume differential among taxa in the same shelter preference site (Fig. 7). While it is tempting to invoke Hutchinsonian ideas of species size rela- tionships (Hutchinson & MacArthur, 1959) to explain the volume differential, the lack of data on feeding, movement, life history, and species interactions makes such an attempt premature. In addition, the recent review of Simberloff & Boecklen (1981) questioning the reality of a size ratio among sympatric taxa suggests caution in making sweeping con- clusions from our preliminary data. We choose to point out the phenomenon and suggest that investigation of volume rela- tionships in other faunas should yield impor- tant information and ideas. After completion of this manuscript, the re- 26 SOLEM 8 CLIMO port of Waldén (1981) on high diversity com- munities of land snails in talus and boulder slopes in Sweden was received. His con- clusions regarding accumulation of species in habitats parallels many of the conclusions we presented in Solem, Climo, & Roscoe (1981), but his study does not address questions we attempt to answer in this report. ACKNOWLEDGMENTS The help of Norman Douglas, Waiuku, and David Roscoe, Nelson, New Zealand, during the fieldwork, was fundamental to this project. Winston Ponder, Sydney, and Simon Tillier, Paris, gave many helpful suggestions during discussions of these studies. Figs. 1, 5-9 were very capably rendered by Gail Rogoz- nica-McKernin, Department of Exhibition, Field Museum of Natural History. We are indebted to Valerie G. Connor-Jackson, Divi- sion of Invertebrates, Field Museum of Nat- ural History, for her dying efforts in typing, retyping, editing for consistency, correction of grammar and spelling, and general all around nagging to make us finish this work. Financial support to Alan Solem from the American Philosophical Society (Grant #8848, Penrose Fund) and Field Museum of Natural History, and to both authors from the National Museum of New Zealand for the fieldwork is gratefully acknowledged. LITERATURE CITED 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., 1981a, Possible ecological significance of variation in shape of Cerion shells with age. Journal of Conchology, 30: 305-315. CAIN, A. J., 1981b, Variation in shell shape and size of helicid snails in relation to other pulmon- ates in faunas of the Palaearctic Region. Mala- cologia, 21: 149-176. CAIN, A. J. & COWIE, В. H., 1978, Activity of different species of land-snail on surfaces of different inclinations. Journal of Conchology, 29: 267-272. CAMERON, R. A. D., 1978, Differences in the sites of activity of coexisting species of land mollusc. Journal of Conchology, 29: 273-278. CLIMO, F. M., 1978, Classification of the New Zealand Arionacea (Mollusca: Pulmonata). A re- view of the New Zealand charopine snails with lamellate apertures. National Museum of New Zealand, Records, 1: 177-201. CLIMO, F. M., 1981, Classification of the New Zealand Arionacea (Mollusca: Pulmonata). VIII. Notes on some charopid species, with descrip- tion of new taxa (Charopidae). National Museum of New Zealand, Records, 2: 9-15. CUMBER, R. A., 1960, Riblet frequency as a taxonomic character in New Zealand terrestrial Mollusca. Transactions of the Royal Society of New Zealand, 88: 99-103. CUMBER, R. A., 1961, A revision of the genus Phenacohelix Suter 1892 (Mollusca: Flammulini- dae) with description of a new species, and studies on variation, distribution, and ecology. Transactions of the Royal Society of New Zea- land, Zoology, 1: 163-196. CUMBER, R. A., 1962, Paleogeographic history reflected in speciation trends of the New Zealand ribbed pulmonate Charopa coma (Gray). Char- opidae. Transactions of the Royal Society of New Zealand, Zoology, 1: 365-371. CUMBER, В. A., 1964, Regional variation in riblet frequency in the Ptychodon (Ptychodon) hectori- hunuaensis complex (Mollusca: Charopidae). Transactions of the Royal Society of New Zea- land, Zoology, 4: 161-166. CUMBER, В. A., 1967a, Variation in Laoma (Phrix- gnathus) mariae (Gray) (Gastropoda: Laomid- ae). Transactions of the Royal Society of New Zealand, Zoology, 9: 33-38. CUMBER, R. A., 1967b, Regional variation in Laoma (Phrixgnathus) sciadium (Pfeiffer) (Gas- tropoda: Laomidae). Transactions of the Royal Society of New Zealand, Zoology, 9: 181-186. CUMBER, В. A., 1967c, A new species of Laoma (Phrixgnathus) (Gastropoda: Laomidae) from the North Cape—Cape Reinga area. Transactions of the Royal Society of New Zealand, Zoology, 9: 187-188. CUMBER, R. A., 1967d, The genus Therasiella (Mollusca: Flammulinidae) in the North Island Mainland, with description of three new species. Transactions of the Royal Society of New Zea- land, Zoology, 10: 61-70. HUTCHINSON, G. & MACARTHUR, R. H., 1959, A theoretical ecological model of size distribution among species of animals. American Naturalist, 93: 117-125. POWELL, A. W. B., 1976, Shells of New Zealand. Ed. 5. Whitcoulls, Christchurch, New Zealand, 154 р., 45 pl: POWELL, А. W. В., 1979, New Zealand Mollusca. Marine, Land and Freshwater Shells. Collins, Auckland, xiv + 500 p., 82 pl. SIMBERLOFF, D. 8 BOECKLEN, W., 1981, Santa Rosalia reconsidered: size ratios and competi- tion. Evolution, 35: 1206-1228. SYMPATRIC NEW ZEALAND LAND SNAILS 27 SOLEM, A., 1959, On the family position of some Palau, New Guinea, and Queensland land snails. Archiv fur Molluskenkunde, 88: 151-158. SOLEM, A., 1976, Endodontoid land snails from Pacific Islands, Part I, Family Endodontidae. Field Museum of Natural History, Chicago, Illi- nois, U.S.A., xii + 508 p. SOLEM, A., 1979, Camaenid land snails from Western and central Australia (Mollusca: Pulmo- nata: Camaenidae). |. Taxa with trans-Australian distributions. Records of the Western Australian Museum, Supplement 10: 1-142. SOLEM, A., 1981a, Camaenid land snails from Western and central Australia (Mollusca: Pulmo- nata: Camaenidae). Il. Taxa from the Kimberley Amplirhagada lredale, 1933. Records of the Western Australian Museum, Supplement 11: 143-320. SOLEM, A., 1981b, Camaenid land snails from Western and central Australia (Mollusca: Pulmo- nata: Camaenidae). Ill. Taxa from the Ningbing Ranges and nearby areas. Records of the West- ern Australian Museum, Supplement 11: 321- 425. SOLEM, A., 1983, Endodontoid land snails from Pacific Islands. Part II. Families Punctidae and Charopidae, Zoogeography. Field Museum of Natural History, Chicago, Illinois, U.S.A., ix + 336 p. SOLEM, A., CLIMO, F. M. & ROSCOE, D. J., 1981, Sympatric species diversity of New Zealand land snails. New Zealand Journal of Zoology, 8: 453— 485. SUTER, H., 1915, Manual of the New Zealand Mollusca. Atlas of plates. Wellington: John Mackay, Government Printer. WALDEN, H. W., 1981, Communities and diversity of land molluscs in Scandinavian woodlands. |. High diversity communities in taluses and boul- der slopes in Sweden. Journal of Conchology, 30: 351-372. ZILCH, A., 1959-1960, Gastropoda, Teil 2, Eu- thyneura, (in) Handbuch der Paláozoologie (SCHINDEWOLF, O. H. ed.), 6: xii + 834 p. APPENDIX 1. Method of shell volume calculation and intrapopulational variability The protruding spire of the shell is consid- ered to be a cone, and its volume is computed using the Spire height (Fig. 1, B-E) and Spire diameter (C—D). If the spire is flat-sided as in Liarea and many of the larger punctids (Figs. За, €, +, g), this can be very accurate. When the whorls are rounded, as in many charopids (Figs. 2c, e, n), using the suture-to-suture distance results in the actual whorl profiles dipping in and out of the projected cone pro- file. These dips would tend to cancel each other, and we consider the variation in- troduced by this factor to be negligible. In only a few taxa, such as Laoma leimonias (Fig. 3d) with its “U”-shape, and Obanella rimutaka (Fig. 4i), which has concave sides to shell spire, is this calculation noticeably inaccurate. In most taxa, the spire forms a small propor- tion of the total shell volume, so even these errors are considered to be relatively minor. In a separate calculation, the body whorl is treated as a cylinder, using the measure- ments Body whorl height (Fig. 1, E-G) and Shell diameter (А-В), to calculate its volume. This includes two major sources of error: 1) the body whorl is one volution of a logarithmic spiral, and thus does not conform to a circle; and 2) there usually is clear descension of the body whorl from the previous suture (Fig. 1, E-F), so that the aperture is deflected signifi- cantly downward from the initial plane of body whorl coiling (see Figs. 2f, q, r, 3k, |, 4b). The curvature of the whorls produces corner volume that is not an actual part of the shell. There also is a slight increase in whorl cross- sectional areas from the beginning of the body whorl to the aperture. We consider that the changes in whorl cross-section and body whorl contour are minor and would be roughly equivalent for each species. Exceptions are the few species in which the whorls are later- ally flattened (Figs. 2r, 3d). Here there is less “corner volume” included in the volume calculation. Protrusion of the periphery into a keel (Figs. 2p, 3e, f, 4e, i) normally results in narrowing of the body whorl height, thus part- ly compensating for the increased “shell cor- ner volume.” For our purposes, these changes are considered to be minor. The error introduced by the coiling being a logarithmic spiral, rather than a circle, would be roughly the same in all species. The significant variable among species is the degree of Spire descension (Fig. 1, E-F). This ranges from none in the planulate Gemi- noropa cookiana (see Climo, 1981: fig. А-С), up to 44% of the Body whorl height in Serpho kivi, Cytora hedleyi, and C. torquilla (see Powell, 1979: figs. 12-3, 12-5, pl. 65, fig. 9). We have adjusted the body whorl volume by deducting the percentage descension of the spire from the raw calculated volume. The adjusted body whorl volume and the spire volume were then added to produce the Ad- justed Shell Volume (ASV). We utilize this figure as a complete volume estimate of the physical space occupied by the adult shell in the habitat. The included minor additional space resulting from peripheral contours and the difference between the spiral versus 28 SOLEM 8 CLIMO circular coiling can be viewed as necessary room for the foot to protrude partially and the snail to start crawling. The umbilical opening at the base of the shell is a small to significant volume of space. It is not enclosed by the shell in most New Zealand species. In rare cases, it is secon- darily narrowed and used as an egg brood chamber (such Pacific Island Endodontidae as Libera, Gambiodonta, Endodonta, Pseudolibera, and Taipidon semimarsupialis, see Solem, 1976, and the New Zealand “Fec- tola” marsupialis [Powell, 1941]), or it may be used as an egg deposition site by spiders or insects. The effect of umbilical size on Ad- justed Shell Volume is indicated in Fig. 9. Of the 78 native New Zealand species reported from the Manukau, 31 had a relatively open, “U-shaped umbilicus, and nine showed in central section a more nearly straight-sided “V”-shaped umbilicus. Seven species had the umbilicus completely closed by reflection of the columellar lip, three species had it present as a lateral crack, and 28 species had the umbilicus less than 0.3 mm wide with a D/U ratio of 9.7 to 100. The volume of their umbilici would be negligible (see Fig. 9). Calculation of the umbilical volume was done in two ways. For the taxa with “V”- shaped umbilici, Body whorl height less Body whorl descension plus Spire elevation pro- vides an indication of umbilical depth. The space approximates that of a cone, except that the very tip would be truncated by the apical whorls. We have not compensated for this, since the tip of the cone contains trivial volume. There would be greater error by | © > % REDUCTION IN SHELL VOLUME | —l .65* *57* 1 bringing the tip of the cone to the base of the apical whorls. We also recognize that the curved inner walls provide spaces that we are not measuring. Calculating the umbilicus as the volume of a cylinder provides comparative figures. Since the “V”-shaped umbilici are mostly narrow, their impact on total shell volume (Fig. 9) is quite small, and their shape produces a clear offset from the curve for the “U”-shaped umbilici. Calculating the volume of the “U”-shaped umbilicus as that of a cylinder presents addi- tional possibility for error. The upper end is clearly dome-shaped, not nicely truncated, and there will be some umbilical narrowing toward the apex. To compensate for this, we have reduced the “Body whorl height less Body whorl descension plus Spire elevation” measure by 20% to allow for these space losses. Utilizing these assumptions and ex- pressing Umbilical volume as a percentage of ASV, we plot this percentage against pro- portionate umbilical width (Fig. 9). When the umbilical opening is less than a fifth of the shell diameter, the volume of the umbilicus is less than a twentieth of the shell itself. Only when the umbilicus is a third or more of the shell diameter, does the umbilical area repre- sent a significant portion of the shell volume. We consider that ignoring this factor in our analysis is justified. Data on variability within populations is pre- sented for several species (Table 11) and compared with measurements of the repre- sentative adult used in the main analysis. The measured population samples are from the National Museum of New Zealand's mollusk * V-SHAPED UMBILICI == = == 2 3 4 5 6 7 D/U RATIO IL 2 — = 9 10 11 12 13 14 15 FIG. 9. Effect of umbilical size on adjusted shell volume. Numbers are those used in Table 1. 29 SYMPATRIC NEW ZEALAND LAND SNAILS м—з—жбФЖЩььььььу _—___ 5 84! (Meere) OSHO = Ste 00`01 (E’L1-70°2) gel + 898 198 (179-917) Зе == 95 Gye (002-521) LS LE = 2'251 vb'S (0S'5-89'€) 579`0 = 8€'p 051 (pL 2-07 1) 075`0 = ¿9'L ‚uw и! эшп|ол 119ys рэ1$пру 6 6$/`0 581 (8G 0] 9) (928'0-65/`0) (po'2-16'1) 19 9200 + 88/0 0$0`0 = 861 +26 61/`0 055 (AS 0} —56) (77/`0-099`0) (62'€-€6 2) AG S20'0 + 20/0 MAO! —AG 059'0 Ors ("AG 0} AP) (199'0-899'0) (SL S-+8'+) +824 L£O'O + SZ9'0 LSO 266 8AG 6/S'0 Gol (AS 01 -416) (L6S'0-6SS'0) (05'8-95`/ AG Z10'0 + 8/S'0 9860 = SOL Ar 0150 GG'e (“Av 0} -9€) (Lps'0-2Sv'0) (09'2-0€'Z) ++ 550`0 + 20S'0 v600 + Er? Yet S/6'1 021 (%G 0} -Yp) (962'©—898`1 (BE 1-92' 1) —G LOL'O + 4861 GGO'0 Feet SIOYM ones Q/H ww ul лээшер |jEUS ‘JO чоцемлар рлерие}з pue эбиел ‘ueayy E A Ыыыы Ее. = == === LEAL (99 1-sp' 1) 650`0 = 9S'L 082 (or 2-10°2) 0010 = 912 or’E (6/£-L0'€) 80£ 0 + 9€'€ 02$ (06'p=S0'p) 508`0 + zer oe 1 (se: L-S0 1) 160'0 + 221 Lem (€6'2-€€ 2) 0/L0 = +92 ww ul 34б!эч jays cl LL Ol 8 synpe рэлпзеэш © JoquunN VINSNIN3d ПУЯПМУИ WOH1 LINGV 0319313S (ешиедхе) 66759 W sısugeıbuond ‚snyJeußxuydg, 11Ndv 0319373S 06/15 W eunew ye ‘ds ‘и ешое7 INVASIONES 99EZS N ıAquosuod xıl3yo9euayd 11NAWV 03193738 8611$ N его: XI9YOSSE/EY! , 11NAV 95193155 06569 И 219 E/J990N пам‘ а=тэ==$ 76119 И ворив/эа5олои BapıjlaWwe7 Jaquinu бо|ееэ pue 5эюэа$ ‘зиоцепао4 pajoajas шцум ApiqeueA ‘LL 378VL 30 SOLEM 8 CLIMO collection and consist of dead adult shells. They are from Jones Bush, Manukau Penin- sula or near Waitomo Caves, except for “Phrixgnathus” pirongiaensis. Only two speci- mens of the latter species were taken from kauri litter on the Manukau (Solem, Climo & Roscoe, 1981: 470, 484). They were small and barely adult. Thus a very small adult was selected from an extralimital set with many specimens. Subsequent measuring of repre- sentative adults from that sample (Table 11) resulted in a much larger size. This example is included here to emphasize both the geo- graphic variability within New Zealand land snail species, and the slight intrapopulational variability shown by this inhabitant of deep wet litter. The choice of sets included in Table 11 was limited by: 1) availability of sufficient adult specimens; 2) an attempt to include a sample of the family taxa; and 3) to use materials representative of different habitat prefer- ences. Lamellidea novoseelandica is an ar- boreal achatinellid; Mocella eta is a charopid found in drier fringe area litter; “Thalas- sohelix” ziczac lives in deep wet litter that is well shaded and is one of the larger taxa; Phenacohelix ponsonbyi is a charopid from well-drained slopes in moderately wet, un- disturbed areas; Laoma n. sp. aff. marina typically lives on wet ground surfaces under broad leaf litter and is a large, keeled punctid; and “Phrixgnathus” pirongiaensis is a keeled, small punctid from very wet litter. Most measurements of the selected repre- sentative adults are well within one standard deviation of the population means. Since these specimens were selected prior to the population sample measuring and analysis, we are confident that the individuals used of all species fairly represent the size and shape of the Manukau morphotypes. The selected Lamellidea is a little smaller, and the Mocella a little larger than the means, but these differ- ences are much less than the basic 40% differences between species used in the anal- ysis. We consider that the data basis is adequate for this study. MALACOLOGIA, 1985, 26(1-2): 31-123 SYSTEMATIC REVISION OF THE HYDROBIIDAE (GASTROPODA: RISSOACEA) OF THE CUATRO CIENEGAS BASIN, COAHUILA, MEXICO Robert Hershler' Edwards Aquifer Research and Data Center, Southwest Texas State University, San Marcos, TX 78666, U.S.A. ABSTRACT This study gives detailed morphological descriptions, including aspects of shell and soft-part anatomy, for 12 species of nine genera of hydrobiid snails (Gastropoda: Rissoacea) from the isolated desert spring system of Cuatro Cienegas, Coahuila, Mexico. Snails were collected from 103 localities in the basin and summaries of the distribution and ecology for each species are given. One new genus and three new species are described. The six nominal species of Mexipyrgus are reduced to one variable species, Mexipyrgus churinceanus, as there are no suites of morphological features that can consistently define separate taxa when a large number of populations is studied. A multivariate morphological analysis of Mexipyrgus churinceanus, involving 20 morphological characters from 33 pop- ulations in the basin, shows that the trends of variation only partly follow the distribution of populations among the drainage systems of the basin. Contrary to previous thought, there are no subfamilies of hydrobiids endemic to the Cuatro Ciénegas basin; all taxa studied belong to either the Nymphophilinae or Littoridininae, widely distributed subfamilies. Five genera and at least nine species are endemic to the basin. Phenetic and phyletic analyses show that of the five endemic genera, three are more closely related to nonendemic genera found in the basin than to each other, suggesting a polyphyletic origin for the endemic snails. The endemic snails may also be of a more recent and local origin than once thought. Snail taxa from the Pliocene Pebas Formation of Peru, the shells of which are superficially similar to those of the Cuatro Ciénegas endemic taxa, are not Hydrobiidae and thus the conchological similarity is due to convergence. Four of the Cuatro Ciénegas hydrobiid genera are ovoviviparous. Anatomical studies show that the evolution of this reproductive mode in the Hydrobiidae has involved modifications of the female reproductive system to separate incoming sperm from outgoing embryos, increase the amount of space available for holding embryos, and allow for control of the release of young. Key words: Hydrobiidae; Cuatro Ciénegas; systematics; morphology; endemism; evolution; ovoviviparity. INTRODUCTION The small (30 by 40 km) desert valley of Cuatro Ciénegas, Coahuila, México (Fig. 1) harbors a remarkable endemic biota (Con- treras, 1978; Minckley, 1969). Most of the endemic taxa are associated with the exten- sive spring fed aquatic environments of this closed-drainage basin and include one genus and four species of crustaceans (Cole & Minckley, 1966, 1970, 1972; Holsinger & Minckley, 1971), eight-ten species of fishes (Minckley, 1977), and two species of turtles (Schmidt & Owens, 1944; Webb & Legler, 1960). In addition, three subfamilies, five genera, and 12 species of hydrobioid snails (those rissoacean snails that resemble Hy- drobia in shell, operculum, penis, or radula) have been considered endemic to the valley (Taylor, 1966). Apart from their high endemism, the hydro- bioid snails of Cuatro Ciénegas are of interest for the following reasons: 1) a number of taxa have large, sculptured or color-banded shells whereas most hydrobioids have small, smooth shells without color bands; 2) several taxa may have been involved in coevolution with snail-eating cichlid fishes in the valley (Vermeij & Covich, 1978); 3) the snails are deployed within a nearly unique variety of spring fed aquatic environments within the desert; 4) differentiated populations of snails are found among the various springs of the valley and offer the opportunity to study evolution in a natural laboratory (Taylor, 1966; Taylor & Minckley, 1966). The original description of the Cuatro Ciénegas hydrobioids (Taylor, 1966) stimu- lated this study and stands as an exemplary contribution for that time period. Credit should ‘Latest address: % Dr. Fred G. Thompson, Florida State Museum, Gainesville, FL 32611, U.S.A. 32 HERSHLER | В; /o G } SA | TEXAS TUS Ae | | | Del Rio | Acuna Rio Grande j Boquillas | E | | | @Eagle Pass | Piedras Negras | | | ! | COAHUILA, MEXICO | % | Muzquiz@ о e | Sabinas A > | © | | | | | ES O eS ee A AS E A AN << :5 O Cuatro Ciénegas” | | Monclova® | | | | | | N | | @San Pedro | Torreon | Parras | Saltillo e | | 2050 100 Кт FIG. 1. Map of Coahuila, México, showing the location of the Cuatro Ciénegas Вазт. The arrows indicate the direction of flow of the Rio Grande. CUATRO CIÉNEGAS HYDROBIIDS 33 be given to that author for recognizing the uniqueness of both the hydrobioid fauna and the environmental setting. The classification scheme within which the Cuatro Ciénegas hydrobioid snails were de- scribed was based on a character set re- stricted to shell, operculum, penis, and a few other aspects of external morphology (Taylor, 1966). It is now known, on the basis of overall soft part anatomy, that such characters have frequently converged in hydrobioid taxa that are not closely related (Davis, 1979). In a series of papers (Davis, 1968, 1979, 1980; Davis et al., 1976, 1982; Davis & Pons da Silva, 1984) it was shown that the Hydro- biidae (pre-1980) are polyphyletic and that study of all aspects of soft part morphology, particularly the entire female reproductive system, is necessary to recognize con- vergences and clarify the systematic rela- tionships of hydrobioid snails. As the original descriptions of the Cuatro Ciénegas hydro- bioids did not include aspects of internal an- atomy, these taxa's classification is suspect. The species descriptions were usually based on collections from single localities, and only 12 localities were sampled (Taylor, 1966). This paper: 1) presents detailed morpho- logical descriptions of the Cuatro Ciénegas hydrobioid taxa and assesses their systema- tic affinities, the results of which are frequent- ly at odds with previous classification (Taylor, 1966). One new genus and three new species are described. Morphological data from pop- ulations are analyzed to resolve species prob- lems; 2) summarizes the distribution and ecology of each species (103 localities sam- pled in the valley); 3) discusses the results of the above as they relate to the origin, evolu- tion, and endemism of the hydrobioid snails of Cuatro Ciénegas; 4) discusses the evolution of ovoviviparity' in hydrobioid snails. Environmental Setting Cuatro Ciénegas lies in the mideastern section of the Chihuahuan Desert (Miller, 1977; Fig. 1). The valley receives less than 200 mm of precipitation annually (Minckley, 1969). The mean annual temperature is 23°C (Morafka, 1977), with midday summer tem- peratures exceeding 40°C. The valley floor is 740 m above sea level, bounded on all sides by tall (to 3000 m) peaks of the Sierra Madre Orientale. The valley floor is relatively flat. The basin consists of two lobes, separated by the northern end of the Sierra de San Marcos (Fig. 2). There are far more springs in the eastern than in the western lobe. Springs are particularly concentrated around the Sier- ra de San Marcos. The springs vary: there are small seeps; small springs with spring pool areas of less than 10 т? that run for only tens of meters; and much larger springs with spring pool areas in excess of 900 m? and depths to 7m, whose outflows are large streams. While most of the springs are lim- nocrenes, with spring pools at the heads, there are also rheocrenes, where water rushes out of the ground as flowing streams. Other aquatic environments include playa lakes, receiving flow from large streams; spring fed pools that have no outflows; and extensive spring fed marshes. This great di- versity of desert spring fed aquatic environ- ments can only be matched in North America by that seen in the Death Valley-Ash Meadows area (Deacon & Minckley, 1974; Soltz & Naiman, 1978). Spring levels vary seasonally, as the water table rises in the winter and drops in the summer. The spring water is generally quite hard and high in sulphates (Minckley & Cole, 1968). Most of the springs are thermal (to 34°C), but cooler (14—25°C) springs are also found. The larger springs have fairly constant water temperatures throughout the year (Minckley, 1969), while spring runs and shal- low pools can be subject to considerable var- iation in water temperature. For example, North Spring has a spring pool area of about 230 m? and a maximum depth of 1.5 т. Its waters flow into a second pool and then run as a wide shallow stream for 77m before disappearing into a hole. The spring is ther- mal; 10 separate headspring temperature readings during 1981 gave a mean of 32.9°C (29.5-34.5°C). Maxi-mini thermometer read- ings for four days (beginning 6/18/81) had a variation of 31-34.5°C for the head and 20.6— 37.8°C at the hole where the water goes underground. Thus, during this time period, the headspring water temperature varied only 3.5°C while downstream it varied 17.2°C. The larger springs and their outflows have considerable microhabitat diversity, usually including several types of aquatic vegetation (sedges, Nymphaea, Chara, Utricularia); a soft sediment consisting of snail copropel and/or an algal-detritus mixture; a sand con- "By ovoviviparity | mean brooding young without direct tissue connection, following Van der Schalie (1936). 34 HERSHLER SIERRA DE LA MENCHACA SIERRA DE LA mn N 800 m un CUATRO CIÉNEGAS J SIERRA DE SAN MARCOS SIERRA DE LA FRAGUA SIERRA SAN VICENTE SIERRA DE LA PURISIMA FIG. 2. Map of the Cuatro Ciénegas Basin, showing the portion of the basin that was intensively sampled (area enclosed by dashed lines). The numbers indicate the origins of the seven major drainages of the basin (from Minckley, 1969): 1, the Churince system; 2, the Becerra system; 3, the Rio Mesquites system; 4, Rio Puente Chiquito; 5, Tio Candido; 6, Santa Tecla Laguna; 7, Rio Salado de Nadadores. sisting of travertine pieces and shell debris; large travertine blocks; and the banks, which can be gently sloping or greatly undercut. The smaller springs and their outflows have fewer microhabitats, typically including a dark or- ganic mud, fine travertine sand, and occa- sional Chara mats. Five to seven major drainage systems occur in the valley (Fig. 2), with possible natural connections existing between them via underground rivers, or surficial waters dur- ing rainy periods (LaBounty, 1974; Minckley, 1969). A number of irrigation canals drain water from the large springs, lowering their levels and destroying peripheral aquatic habi- tats (Minckley, 1969). Irrigation canals from different drainage systems are often con- nected, offering opportunities for gene flow between previously isolated populations. The basin currently has no surficial connection to outside drainage, but water has probably drained from the basin to the nearby Rio Salado de Nadadores (Fig. 2, Locality 7) in the past (Miller & Minckley, 1963; Minckley, 1969). More information on the aquatic en- vironments of the basin can be found in Arnold (1972), Brown (1974), Deacon & Minckley (1974), and Minckley (1969). MATERIALS AND METHODS Localities The waters of only a portion of the valley drainage (dashed line in Fig. 2), encompass- ing parts of four of the basin drainages, were intensively sampled. One hundred collection localities from this area are shown in Fig. 3 and described in Appendix 1. Three other localities (101-103 in Appendix 1) from other areas were also sampled. The various locali- CUATRO CIÉNEGAS HYDROBIIDS 35 SIERRA DE SAN MARCOS N FIG. 3. Map of the portion of the basin drainage that was intensively sampled. The numbers (1-100) refer to collection localities. The dashed lines refer to irrigation canals. The arrows indicate waters that continue to flow toward the east. ties (and ANSP catalog numbers for the lots) Collection Methods for each species are given in Appendix 2. Snails were collected and studied in the valley Fine hand sieves were used to collect during September, 1978; April-August, 1979: snails from sediments ranging from flocculent April, 1980-June, 1981; and December, copropel to coarse travertine sand. In some 1981. cases, the sediment itself was collected and 36 HERSHLER examined under a dissecting microscope to determine whether very small snails, which would pass through the sieves, were present. Snails were picked off large pieces of travertine using tweezers. Samples of aquatic vegetation were collected and carefully washed in a bucket to remove clinging snails. The material was then sorted under a dissect- ing microscope. A number of snail taxa are suspected of living in groundwater outlets or subterranean waters of the valley (Taylor, 1966). To collect such snails, ordinary domestic mops were placed into small spring- heads, removed after a 24 hour period, and then washed in a bucket. The snails that colonized the mops were collected and sorted under a dissecting microscope (method sug- gested orally by Dr. W. L. Minckley; similar method described in Holsinger & Minckley, 1971). For each small springhead, one to ten mop samples, spread out over a period of months, were taken. This method is crude as it cannot distinguish between snails living at the groundwater outlet and those being washed out from underground. A superior method, sampling only snails from sub- terranean waters, would be to place a fine mesh net over the groundwater outlet, es- sentially filtering snails from the water stream; such a method has been successfully used to sample the fauna of artesian wells (Holsinger & Longley, 1980). Anatomy Dissection techniques are those of Davis (1979) and Davis & Carney (1973). Several other techniques were also employed. The penis was cut from the male, examined on both sides at 50x magnification, and then wet-mounted on a slide with cover slip for study using a compound microscope fitted with an ocular micrometer. The length of the pallial oviduct is the length from the anterior end to the posteriormost point, excluding any length bent back upon itself at the posterior end. The length of the duct from the seminal receptacle is the length from the seminal receptacle body to the junction with the ovi- duct or sperm duct. The nervous system was studied for species of all genera except Mex- istiobia n. gen. and Coahuilix and showed no variance except in size and concentration of ganglia. Therefore, the measurements of the nervous structures are not presented. For ovoviviparous species, shelled embryos were gathered by cracking the shell of an adult female, removing the brood pouch, and plac- ing it in a drop of CLOROX for several min- utes (Davis, 1969b). The embryonic shells were then counted and, in some cases, measured. In addition, the brood pouches of several snails were teased apart so the small, nonshelled embryos could be counted. Shell measurements for the various taxa are from mature adults (those having a complete aper- ture). The apical whorls of shells were meas- ured using the method of Davis (1967, pl. 3, fig. 6). Radulae and shells were studied and photographed using the SEM facility at the Academy of Natural Sciences of Philadelphia. Statistical techniques, for the most part, are restricted to t-tests and correlation coeffi- cients. The generic descriptions are necessarily brief, as anatomical data is generally avail- able for only one species per genus (due to monotypy or lack of studies of congeners). The descriptions will have to be altered as more data become available. Only character states unique to the various taxa, or of use in assessing their systematic status, are stressed. Other characters and their char- acter states that are standard for the Ris- soacea: Hydrobiidae (Davis, 1966, 1979; Hershler & Davis, 1980) are not mentioned and include, for example, the characteristic loop of the intestine above the style sac, the position of the salivary glands on top of the nerve ring, and the ovoid shape of the fecal pellets. Fifty-one characters and their character states that were used to distinguish genera and generic groups in the Hydrobiidae are listed in Appendix 3 with notations as to where they are figured. Common radular for- mulas for species studied are given in Table 1. Several characters may be unfamiliar to the reader and require explanation. The bolster and ventral channel of the pallial oviduct are defined and discussed in Davis et al. (1982), and Davis & Pons da Silva (1984). The caecal chamber (defined in Davis et al., 1982), while apparently present in all Hydrobiidae, is re- duced in some taxa so as not to project posterior to the stomach. Tentacle ciliation is discussed in Davis et al. (1982). The digestive gland of hydrobioid snails usually has finger- like tubercles projecting from the main body, but in small sized snails the tubercles may be mere swellings. In resolving species prob- lems, emphasis was placed on whether pur- ported species were sympatric or not, and whether consistent morphological differences between purported species could be found CUATRO CIÉNEGAS HYDROBIIDS 37 TABLE 1. Generalized cusp formulas for the four tooth types of the radula of all species studied. Inner Outer Species Central Lateral marginal marginal Nymphophilus minckleyi 4(5)—1—4(5) 2(3)-1-2(3) 12-17 15-20 3-3 Mexistiobia manantiali 4(5)-1-4(5 4(5)-1-3 19-24 22-26 1-1 Coahuilix hubbsi 3(4)-1-3(4) 5-1-3(4) 16-21 16-19 1-1 Paludiscala caramba 4(5)-1—4(5 4(5)-1-3 18-24 16-25 1-1 Cochliopina milleri 4(5)-1-4(5 3(4)-1-3 18-25 19-28 1-1 Mexithauma quadripaludium 4(5)-1-4(5 3(4)-1-3(4) 10-14 12-16 2(3)-2(3 Durangonella coahuilae 4(5)-1-4(5 4(5)-1—4(5) 19-27 20-27 1-1 Mexipyrgus churinceanus 4(5)-1-4(5) 21-36 24-38 4(5)-1—4(5 2(3)-2(3 when numerous populations were studied. The descriptions of taxa not named in this paper are modified from those of Taylor (1966) and Thompson (1979). For a multivariate analysis of Mexipyrgus churinceanus populations the computer pro- gram used was the June, 1974 version of the SUNY at Stony Brook numerical taxonomy program, NT-SYS (Rohlf et al., 1972). Characters were standardized in the usual manner (Sneath & Sokal, 1973). In the Q- mode analysis, a taxonomic distance matrix was generated, using the unweighted pair- group method with arithmetic averaging (UPGMA). The minimum spanning tree (MST) and “subsets” components of NT-SYS were used. For the R-mode analysis, character correlations were subjected to Prin- cipal Components Analysis (PCA), with the first three components used to yield a matrix of OTU projections in principal component space. These OTU locations in the three-di- mensional PCA space were used as the initial configuration for a nonmetric multidimen- sional scaling (MDS) placement of the Q- mode taxonomic distances between OTUs. The Prim Network was used. As the cluster analysis and phenograms generated are sub- ject to distortion, only the ordination and MDS are presented here. Components were ex- tracted until eigenvalues were less than 1.0. Subset solutions and the minimum spanning tree are superimposed on the ordination di- agrams. SYSTEMATIC FRAMEWORK The basic features of rissoacean snails are reviewed in Fretter & Graham (1962). Davis (1979) has listed the features that distinguish hydrobioid snails from other rissoaceans. The definitions below are modified from those of Davis (1979, 1980) and Davis et al. (1982). Family Hydrobiidae These include hydrobioids in which sperm enter the anterior end of the pallial oviduct and pass along an internal, ciliated ventral channel to the bursa copulatrix; or, in which sperm enter a separate spermathecal duct, presumably formed by separation of the ven- tral channel from the pallial oviduct (not to be confused with the convergent structure of the Pomatiopsidae), and pass through it to the bursa copulatrix. The spermathecal duct is never associated with either the kidney or the pericardium (contrast the Triculinae). The mode of reproduction is oviparity or ovovivi- parity. The penis may (Thompson, 1968, fig. 39) or may not (Thompson, 1968, fig. 371) have lobes. The penis may also be without specialized glands (Hershler & Davis, 1980, fig. 4D), or may have glandular ridges consist- ing of an elevated area in which rows of small glands discharge through a central slit (Thompson, 1968, fig. 42); apocrine glands (Andrews, 1977, p. 82, fig. A); glandular papil- lae (Hubendick, 1955, fig. 88); or mammiform 38 HERSHLER glands (Fig. 44). The latter two gland types are only borne on penial lobes, whereas glandular ridges and apocrine glands can be found on the penis as well (Thompson, 1968, figs. 44, 38, respectively). There is neither a pedal crease nor a suprapedal fold (contrast the Pomatiopsinae); the snails move by ciliary gliding. The mantle collar may or may not have a pallial tubercle, filament, or numerous papillations. The tentacles may or may not have hypertrophied ciliary tufts. The eyes are located in slight swellings at the base of the tentacles. The central tooth of the radula usually has pronounced lateral angles, giving the tooth a trapezoidal shape, and one or more pairs of basal cusps that usually origi- nate from the lateral angles (contrast the Pomatiopsidae). The stomach has a caecal chamber that usually protrudes posterior to the stomach chambers. An anterior digestive lobe may be present. The shell may or may not have wrinkled, pitted apical microsculp- ture. Subfamily Nymphophilinae These include Hydrobiidae in which the pallial oviduct has an internal ciliated ventral channel, and the penis is bilobed and bears one or more elevated glandular ridges. The tentacles do not have hypertrophied ciliary tufts. The apical whorl has wrinkled, pitted microsculpture. The only mode of reproduc- tion thus far reported for this subfamily is oviparity (Thompson, 1968, 1977, 1979). Subfamily Littoridininae These include Hydrobiidae in which there is a spermathecal duct separate (at least poste- riorly) from the pallial oviduct. The spermathe- cal duct may be short or long. The pallial oviduct often has three or four tissue types. The penis may be without specialized glands, or with large apocrine glands, glandular papil- lae, or mammiform glands. The tentacles often have hypertrophied ciliary tufts. The api- cal whorl may or may not have wrinkled, pitted microsculpture. The mode of reproduc- tion may be oviparity or ovoviviparity. The subfamilial placement of the Cuatro Ciénegas hydrobioids, based on anatomical study, is contrasted with that of Taylor (1966) in Table 2. DESCRIPTION OF TAXA Nymphophilinae Nymphophilus Taylor, 1966 Type-species: Nymphophilus minckleyi Taylor, 1966. Distribution: endemic to the Cuatro Ciene- gas Basin. Species included: N. minckleyi, N. acarina- tus п. Sp. Description Diagnostic features of Nymphophilus т- clude the large (length, 3.5-8.3 mm) trochoid TABLE 2. Subfamilial placement of the Cuatro Ciénegas hydrobiid genera (based on the results of this study) contrasted with that of Taylor (1966). Taylor (1966) Family Hydrobiidae Subfamily Cochliopinae Coahuilix* Cochliopina Subfamily Littoridininae Mexipyrgus* Durangonella Subfamily Nymphophilinae** Nymphophilus* Subfamily Mexithaumatinae** Mexithauma* Subfamily Paludiscalinae** Paludiscala* This study Family Hydrobiidae Subfamily Littoridininae Coahuilix* Paludiscala* Mexithauma* Cochliopina Durangonella Mexipyrgus* Subfamily Nymphophilinae Nymphophilus* Mexistiobia' Subfamily Unknown Orygoceras ?* "Genus endemic to the Cuatro Ciénegas basin. **Subfamily considered endemic to the Cuatro Ciénegas Basin by Taylor (1966). "New genus. “Systematic status uncertain as anatomy is not yet studied. CUATRO CIÉNEGAS HYDROBIIDS 39 FIG. 4. Shells of Nymphophilus minckleyi from Locality 76. The shell on the left is 8.0 mm long; the others are printed at the same enlargement. shell (Figs. 4, 9), multispiral operculum (Fig. 5B), elongate osphradium (30% of the cteni- dium length), and bush-like male gonad (not shown). The bursa (Bu) is positioned posterior to the pallial oviduct (Figs. 7A, В); the duct of the bursa to the common opening of the albumen gland and ventral channel is elongate; the seminal receptacle (Sr) and oviduct coils (Coi) are located anterior to the bursa (Fig. 7B); the bolster of the ventral channel is well- nn 1.0 mm developed (Bvc, Fig. 7D); the pallial oviduct opens laterally as acommon genital aperture (Cga, Fig. 7E); the penis (Fig. 8A) with mas- sive, folded penial lobe (Plo) bears one to four glandular ridges (Gir) on its ventral surface. Discussion Among nymphophilines, Nymphophilus is most similar to Marstonia, as both taxa have a penis with few glandular ridges (for Marsto- ATAR Tn Va 2 : E ÁS = ee \ A N RS | - и | SN Va | \ | à | | К о) ) ] \ ) SEN, 27 \ \ IE = / \ 5 a ae | / \ 7 a === AS 1 ge B 1.0 mm FIG. 5. Head and operculum of N. minckleyi. A. Head seen dorsally. B. Operculum, with dashed line indicating attachment area to operculigerous lobe. Ey—eye; Sn—snout; Tn—tentacle. 40 HERSHLER nia, see Thompson, 1977, figs. 5, 7, 11) and a large bursa positioned posterior to the pallial oviduct (for Marstonia, see Thompson, 1977, fig. 10). M MU a a ol I AC ALL FIG. 6. SEM photos of shell and radula of N. minckley!. A. Portion of body whorl showing the peripheral keel and wavy collabral microsculpture. B. Portion of radular ribbon. C. Several central teeth. Nymphophilus minckleyi Taylor, 1966 Holotype: UMMZ (University of Michigan Museum of Zoology) 220188. Type-locality: Locality 53. Habitat: Nymphophilus minckleyi is found in large springs and their outflows. Nymphophi- lus minckleyi was rarely taken from the small- er springs; mops yielded specimens from the heads of only three of 38 such springs. In the large springs, N. minckleyi was collected from aquatic vegetation (Nymphaea, Chara, Utric- ularia), travertine, and, to a lesser extent, from gentle, sloping banks. On a microhabitat scale, this species is occasionally sympatric with Mexithauma quadripaludium and Coch- liopina milleri. lt has been suggested (Arnold, 1972) that N. minckleyi, as well as Mex- ithauma and Mexipyrgus, is nocturnal, mov- ing about at night when the predaceous cich- lid fish are inactive. The egg capsules of this species were found on water lily (Nymphaea) leaves from many localities throughout the year. It was not uncommon to find over 100 capsules on a single 15 cm leaf. The capsules were rarely found on the shells of living snails. The egg capsule is hemispherical and is coated with detrital material along the sides, but the top of the capsule is clear of detritus and the yellow- colored embryo is visible inside. Usually, hy- drobioid egg capsules are completely coated with either sand or detritus. For 13 egg cap- sules from Locality 76, the egg capsule dia- meter is 0.52 + 0.40 mm. The height of the capsule is about 0.25mm. The embryonic shells inside have 1.00—1.25 whorls. Description Nymphophilus minckleyi is distinctive in having a large shell (Fig. 4), with 5.5-6.0 flattened to slightly rounded whorls. There is a strong spiral keel, at or just above the suture, that fades on the body whorl and is barely noticeable at the aperture. Shell The spire is moderately high, the base rounded, and the umbilicus narrow. The su- tures are shallow. The aperture is longer than it is wide, somewhat angled adapically, rounded abapically, and with a complete thickened peristome in adults. In adult shells, the aperture is adnate to or slightly separated from the penultimate whorl. The plane of the CUATRO CIÉNEGAS HYDROBIIDS 41 FIG. 7. Female reproductive anatomy of N. minckleyi. A. Snail uncoiled, exposing the ventral aspect without head and kidney tissue. Note the position of the bursa (Bu) posterior to the pallial oviduct (Apo + Ppo). B. Oriented as in A, but with a portion of the pallial oviduct cut away to reveal the bursa copulatrix complex, bolster (Bvc), and ventral channel (Vc). The left-hand dashed line indicates the posterior extent of the pallial oviduct. C. Oriented as in B, but with the bursa, its duct (Dbu) and the anterior portion of the duct of the seminal receptacle removed to reveal the oviduct coils (Coi). D. Cross-section of the pallial oviduct (looking anteriorly), cut where the right-hand dashed line indicates in B. Note the thickened bolster (Bvc) and well-developed ventral channel (Vc). E. Dorsal aspect of the capsule gland (Apo) showing the lateral opening of the common genital aperture (Cga). Apo—capsule gland; Ast—anterior stomach chamber; Bu—bursa; Bvc—bolster of ventral channel; Cae—caecum of stomach; Cga—common genital aperture; Cl—columellar muscle; Coi—coil of oviduct; Dbu—duct of the bursa; Dsr—duct of seminal receptacle; Emc—posterior end of mantle cavity; Go—gonad; In—intestine; Lpo—lumen of pallial oviduct; Ma—mantle edge; Ov—oviduct; Ppo—albumen gland; Pst—posterior stomach chamber; Sr—seminal receptacle; Sts— style sac; Vc—ventral channel. aperture is only slightly tilted away from the coiling axis. The shell is colorless and translu- cent. The pitted apical microsculpture is shown in Thompson (1979, figs. 4-7). Postembryonic whorls have coarse, wavy growth lines (Fig. 6A), giving the shell a satiny sheen. Shell measurements for three pop- ulations, all from large springs or streams, are given in Table 3. Shell lengths of males are significantly larger (р < .01) than those of females for all three populations. Shells were removed from egg capsules from Locality 76 and their apical whorls measured. For 16 shells, the width of the tip of the apical whorl 42 HERSHLER TABLE 3. Shell measurements (mm) of males and females from three populations of Nymphophilus minckleyi. Snails with the dominant maximum whorl number were used. N = 9, Mean + standard deviation. “p” refers to the significance level for the difference between shell lengths of males and females (t-test) for that population. Length of Length of Width of Whorls Length Width body whorl aperture aperture p Locality 76 3 6.0 7.88 10:50, 5/90/2042 059921049488 0.421 3532022752005 ? 5:5 6:89 101281 25:51 20:22 85:37, S083! 39220411 3213020 Locality 97 3 6.0 7.90 = 0730) 5.68`= 0.27 5.86 20:27, 4.05 = 0.583556 == 0.1510 2 6.0 749} = 0.3211 De = 0.20’ 5152.2 0126) 2.77 = 0:24 1 3 22 Ons Locality 53 3 55 8730 = 0:36) 6:52 =10:37 6:57 = 0.39, 4.81 = 0.27 | 3:93 Е 09005 2 55 72.19, ©. 18 6:33 0.277 6.08 = 0:14. 4.45 = О 23:20 averaged 0.135 + 0.017 тт; the width of the first whorl was 0.339 + 0.040 mm. The width of the first whorl for shells from the type- locality was 0.30 mm (Thompson, 1979). Nonreproductive Features Details of the anatomy are from the popula- tion from Locality 76 unless otherwise in- dicated. Measurements of organs and struc- tures are given in Table 4. The snout (Fig. 5A) is 1.77 mm long and relatively squat while the tentacles are thick and elongate (relative to the snout). The snout and tentacles have embedded in them yellow granules. The eyes are partially surrounded by clear, closely packed granules that extend back along the neck. A light dusting of melanin on the ros- trum and tentacles was occasionally seen. The foot is large (relative to that of other species), thickened and dusted with melanin on its dorsal surface and sides. Body pigmentation consists of a dusting of reddish melanin on both dorsal and ventral surfaces. The male gonad occasionally has very dark pigmentation on its ventral surface. The oper- culum (Fig. 5B) has 5.5-6.0 whorls, and the nucleus is positioned at 39% of the long axis Ы—мьмнтнЪУ Bay ee 1.0 mm FIG. 8. The penis of N. minckleyi. A. Dorsal aspect of the penis. Note the large penial lobe (Plo) with numerous folds in it. B, С. Ventral aspect of the penial lobe showing the glandular ridge(s) (Gir). Vd—Vas deferens. CUATRO CIÉNEGAS HYDROBIIDS 43 TABLE 4. Dimensions (mm) or counts of non- neural organs and structures of Nymphophilus minckleyi. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Body EMO'S0E 1083012402 Oi Gill filament number 46.8 + 2.17 Osphradium Е 10.89 10:13 Gonad 249 ESOO 4 Ors 0:59 Wile 09 == 10:09) 1.19) = 0:01 Prostate L 1E33==.0512 W 0.65 + 0.08 Penis E 4.10 + 0.46 W 10.15 Pallial oviduct [22961-0135 WEEZE OMS Bursa copulatrix 9008 W 0.44 + 0.08 Seminal receptacle L 0.51 + 0.06 (body) (N = 6) W 0.21 + 0.03 Seminal receptacle L 0.14 + 0.10 (duct) W 0.12 + 0.02 of the operculum. The operculigerous lobe has a dusting of melanin along its perimeter. The caecal chamber (Cae) extends posterior to the stomach chambers (Fig. 7A). Radula The radula is shown in Figs. 6B & C. There are three pairs of basal cusps on the central tooth, arising from the lateral angles (Fig. 6C). The central cusp of the central tooth is broad and large relative to the cusps on either side. The lateral tooth also has a massive central cusp (Fig. 6B). The marginals have relatively few cusps. Radular statistics and the various Cusp arrangements for the four tooth types are given in Tables 5 and 6, respectively. Female Reproductive Anatomy The ventral view of the uncoiled female is shown in Fig. 7A. The lobe-like gonad (Go) is short (27%) relative to body length. There are three to five gonad branches, each consisting of small lobes. The oviduct (Ov) passes beneath the pallial oviduct just at the end of the style sac. A short gonopericardial duct is present (Thompson, 1979, fig. 15). The pallial oviduct is 32% of the body length. The two sections of this organ, the anterior capsule gland (Apo) and the pos- terior albumen gland (Ppo), are easily distin- guishable even in unstained specimens. The posteriormost 20% of the pallial oviduct over- lies the style sac (Sts, Fig. 7A). The anterior pallial oviduct ends 1.14 mm from the mantle edge. The relationships between the bursa copulatrix complex and pallial oviduct are shown in Figs. 7B, C. The bursa (Bu) is sac-like and large; 34% the length of the pallial oviduct. The duct of the bursa (Dbu) is FIG. 9. Shells of Nymphophilus acarinatus from Locality 113. The shell on the left is the holotype (ANSP 355255) and is 4.25 mm long. On the right, printed to the same enlargement, is a paratype (ANSP 355256). 34 110% of the bursa length (Fig. 7B). The duct of the bursa, seminal receptacle (Sr), and oviduct coils (Coi) lie appressed to the dorsal surface of the albumen gland. The seminal receptacle and oviduct coils are largely anteri- or to the bursa and dorsal to the duct of the TABLE 5. Radular statistics from 12 individuals of Nymphophilus minckleyi. X = mean, S = standard deviation. Measurements are in mm. Radular feature x S Length 1.96 0.11 Width 0.274 0.019 Number of rows 60.7 3.47 Number of rows in formative stage 3.58 ‚lesil Width of central tooth (N = 28) 0.081 0.0054 HERSHLER bursa. The pear-shaped seminal receptacle is relatively large; 50% of the length of the bur- sa. The oviduct coils twice, with the first coil dorsal to the second one, before receiving the short duct ofthe seminal receptacle (Dsr, Fig. 7C) and joining the duct of the bursa at the opening to the pallial oviduct (Fig. 7B). The bursa copulatrix complex has a com- mon opening with the albumen gland and ventral channel (Vc) at the posterior end of the mantle cavity (Fig. 7B). The ventral chan- nel is considerably folded toward the ventral side of the pallial oviduct (Fig. 7D). The bol- ster of the ventral channel (Bvc) is rounded and thickened (Figs. 7B, D). Sperm masses were found in the ventral channel below the bolster. The walls of the ventral channel do not fuse anteriorly to form a tube separate from the capsule gland (compare Fig. 7C with Davis & Pons da Silva, 1984; fig. 6). The TABLE 6. The various cusp arrangements for the four tooth types in 12 radulae of Nymphophilus minckleyi, with the percentage of radulae showing that arrangement at least once. Central Lateral anterior cusps basal cusps % cusps 3-1-3 8 2-1-2 3-3 4-1-3 17 3-1-2 3-3 4-14 1174 3-1-3 3-2 4-1-4 8 3-1-4 3-3 AA 8 4-11 4-4 5-1-4 25 4-1-2 3—2 5-14 50 4-1-3 3-3 5-15 8 4-1-4 3-2 5-1-5 67 3-3 6-1-4 8 3-3 6-1-5 8 3-2 6-1-6 8 3-3 7-1-4 8 3-2 7-1-6 8 58 17 Inner Outer marginal marginal cusps % cusps % 9 18 15 50 11 25 16 50 12 42 Ue 50 13 42 18 58 14 58 19 25 15 58 20 25 16 58 21 117 1174 17 24 8 18 8 CUATRO CIÉNEGAS HYDROBIIDS 45 capsule gland does not open at its anterior tip, but opens as a common genital aperture (Cga), lateral to and 0.3 mm posterior to the tip (Fig. 7E). Male Reproductive Anatomy The male gonad is relatively long, 38% of the body length, and extends to the posterior end of the stomach. The gonad has seven branches, each with many small lobes, giving the organ a bush-like appearance. The pros- tate is quite small, 11% of the body length, and largely posterior to the end of the mantle cavity. The anterior vas deferens exits from the posterior portion of the prostate. The penis (Fig. 8) is relatively large and thickened, with an elongate penial filament. lt is neither ciliated nor does it have an evers- ible terminal papilla. The single penial lobe (Plo) is positioned on the inner curvature slightly more toward the base of the verge than toward the tip. Thompson (1979, figs. 11-14) illustrates a much stouter penial fila- ment than that shown here, possibly because he was studying preserved material. The penis has no pigment. The vas deferens (Vd) travels near the outer curvature of the penis and coils only during a portion of its length. The penis has numerous Gl, glands (see Davis, 1969a, for a discussion of gland types), particularly in the penial filament. The penis has no folds on its outer curvature, while the inner curvature has folds from the base to just beyond the penial lobe. The penial lobe is quite stout and does not taper appreciably towards its distal end. Numerous folds extend inwards from its sides. The lobe curves both ventrally and towards the tip of the penis. Viewed from the ventral aspect (Figs. 8B, C), the distal edge of the lobe appears as a narrow projection folded above the proximal portion of the lobe. The surface of this distal edge, which cannot TABLE 7. Percent of individuals (N = 25) with 1, 2, or 3 glandular ridge(s) on the distal edge of the penial lobe in three populations of Nymphophilus minckleyi. Number of ridges 1 2 3 Locality 76 84 16 0 Locality 97 52 28 20 Locality 53 76 12 12 Ыжз—ньъ — __ be seen in Fig. 8A, has one to three (see Table 7) glandular ridges (Gir) along its length (Figs. 8B, C). The third ridge (not shown) is often lateral to the other two. Taylor (1966, fig. 21) illustrates a fourth ridge near the base of the penial lobe (ventral surface); this was seen in only one of the 75 specimens studied from three poulations. Nymphophilus acarinatus Hershler, n. sp. Synonymy: Nymphophilus Hershler, n. sp. Hershler in press. Etymology: the species name comes from the acarinate shell. Holotype, ANSP 355255, Fig. 9A; para- types (11); ANSP 355256, Fig. 9B. Type-locality: Locality 98. Habitat: Nymphophilus acarinatus is known only from empty shells from the type-locality and several specimens collected live from Santa Tecla Laguna (Locality 101). Nym- phophilus acarinatus is allopatric to N. minck- leyi. Description While there are insufficient anatomical data for a detailed account comparable to that of N. minckleyi, this species is placed in Nym- phophilus because the organization of the bursa copulatrix complex and form of the penis are like those of N. minckleyi. The shell (Fig. 9) differs from that of N. minckleyi in that it is Somewhat smaller (length, 4.20 mm), has fewer whorls (to 4.8) that are quite rounded, and lacks a peripheral keel, even on early whorls. The growth lines are less pronounced than those of N. minck- leyi. Measurements of the type and paratypes are given in Table 8. Discussion While the differences between N. acarina- tus and N. minckleyi are few and restricted to shell features, there is no blurring of these differences in any of the populations studied. Specimens of N. minckleyi from small springs can be as small as N. acarinatus, but the whorls remain flattened and the peripheral keel is always present. The consistency of these differences suggests that the taxa are distinct species and not mere allopatric var- iants. 46 HERSHLER TABLE 8. Measurements (mm) of the shells of the holotype (ANSP 355255) and paratypes (ANSP 355256) of Nymphophilus acarinatus. All shells are from adults with 4.5-4.8 whorls. Shell Shell length width Holotype 4.25 3.33 Paratype 3.57 2.89 Paratype 4.13 3.26 Paratype 4.37 3.10 Mexistiobia Hershler, n. gen. Etymology: the name was formed by add- ing the prefix Mexi-, referring to distribution within México, to Stiobia Thompson & McCaleb, 1978, a very similar nymphophiline from the southeastern U.S.A. Type-species: Mexistiobia manantiali n. sp. Distribution: thus far known only from the Cuatro Ciénegas Basin and Durango, Mexico (U.S. National Museum of Natural History 351817, labeled “Valvata’). Species included: monotypic. The specific status of the Durango population is not known. Description Among nymphophilines, the unique fea- tures of Mexistiobia include the position of the small bursa (Bu) anterior to the seminal Length of Length of Width of body whorl aperture aperture 3:37 2.30 1.91 2.98 2.22 1:71 3.30 2.38 12911 3.45 2.50 1.99 receptacle (Sr, Fig. 14B), the very short duct of the bursa (Dbu), and the position of the male gonad overlying the posterior stomach chamber (not shown). The shell (Fig. 10) is minute (length, 1.20 mm) and broadly conical; the bolster of the ventral channel is weakly developed (Fig. 14C); the capsule gland opens at its anterior tip as a common genital aperture (Cga, Fig. 14D); the penis has an elongate penial lobe (with one fold in it) bearing a single glandular ridge along its ventral length (Glr, Fig. 13D). Discussion Mexistiobia manantiali bears a remarkable conchological resemblance to Stiobia Thomp- son & McCaleb, 1978, a monotypic genus endemic to a spring in Alabama, yet it differs in 11 morphological features (Table 9). Two of these features (4, 6) may have been in- correctly interpreted by Thompson & FIG. 10. SEM photos of paratype shells (ANSP A9887d) of Mexistiobia manantiali from Locality 51. The shell on the left is 1.15 mm long, that on the right is printed at the same enlargement. CUATRO CIÉNEGAS HYDROBIIDS 47 FIG. 11. SEM photos of the apical region of the shell of Mexistiobia manantiali, showing the wrin- kled, pitted microsculpture at several magnifica- tions. McCaleb as these would be highly unusual traits for hydrobioid snails. Separate generic status is suggested for these two taxa be- cause of the major differences in the position and organization of the bursa copulatrix com- plex, and in the form of the penis and number of glandular ridges. While the stunted appearance of the female reproductive anatomy of Mexistiobia is unique among nymphophilines, a penis with few glandular ridges is also seen in Nymphophilus and Marstonia (see above). Mexistiobia manantiali Hershler, n. sp. Synonymy: “Stiobia” Hershler, n. sp. Hersh- ler in press. Etymology: the species name is formed from the Spanish word manantial, meaning spring, and refers to the spring-fed habitats of this snail. Types: holotype, ANSP 355205; paratypes, A9887d, АЭ8881, 355204, Fig. 10. Because of their small size, the shells had to be photo- graphed using the SEM, which leaves gold coating on the specimens, so the holotype was not used. The paratypes look like the holotype. Type-locality: Locality 51, a small spring. This species was shown, but undescribed, by Taylor (1966, fig. 4). Habitat: Mexistiobia manantiali is restricted to the smaller springs of the valley. It is found in association with Durangonella coahuilae, Paludiscala caramba, and Coahuilix spp. in the headsprings, and is sympatric with D. coahuilae in the spring runs. In terms of mic- rohabitat, Mexistiobia manantiali is common in fine organic sediments and Chara mats, and prefers a finer sediment than does D. coahuilae. While found in mop samples from 20 of 38 small springheads, Mexistiobia manantiali probably does not live in subterra- nean waters as all specimens collected have eyespots and because the species is most adapted for life in open, downstream habitats (see below). Shell The colorless shell has rounded whorls and an open umbilicus. Adults have 3.00-3.3 whorls. The whorls sometimes have a very slight angulation at the shoulder. The latter portion of the body whorl frequently pulls away from the preceding whorl in adults. Postembryonic sculpture is restricted to strong growth lines (Figs. 10, 11A). The plane of the aperture is tilted about 10° toward the coiling axis. The apical whorl microsculpture is shown at several magnifications in Figs. 12A—C. Shell measurements from three pop- ulations are given in Table 10. The shell length for females is significantly larger than 48 HERSHLER FIG. 12. SEM photos of the radula of Mexistiobia manantiali. A. Part of the radula ribbon. B, C. Central teeth. D. Outer marginal teeth. that for males in two of the three populations (Table 10). Nonreproductive Features Observations and data on external features and anatomy are from the type population. Measurements of organs and structures are given in Table 11. The snout is elongate, while the tentacles are relatively short and thickened (Fig. 13A). A small concentration of granules partially surrounds the eyes. The snout and tentacles usually have a dark mela- nin dusting. Body pigmentation consists of a dark melanin dusting on the dorsal and ven- tral surfaces. Adults have only 10-13 gill fila- ments (Table 11). A prominent caecal cham- ber (Cae) extends posterior to the stomach CUATRO CIÉNEGAS HYDROBIIDS 49 TABLE 9. List of 11 morphological differences between Stiobia nana and Mexistiobia manantiali. The morphological information on Stiobia nana is from Thompson & McCaleb (1978). Stiobia nana Mexistiobia manantiali 1. Shell with two spiral keels 2. Wrinkled, pitted microsculpture on all shell whorls 3. Operculum with 2.5 whorls 4. Hypobranchial gland present 5. Lateral tooth of radula with massive, hoe-like central cusp 6. Bursa and duct of bursa imbedded in the pallial oviduct 7. Duct of bursa elongate 8. Seminal receptacle anterior to bursa 9. Penial filament 35% of penis length 0. Penial lobe massive and stout 1. Numerous glandular ridges over entire surface of penis Spiral keels absent Microsculpture restricted to apical whorl Operculum with 3.5 whorls Hypobranchial gland absent Central cusp small and dagger-like Bursa and duct dorsal to pallial oviduct Duct is short Seminal receptacle posterior to bursa Penial filament 0.45% of penis length Lobe is small and slender Single glandular ridge on ventral surface of penial lobe x-EKXESRQSRnOÓOQmÓÓQ—_————— —:): m ——_——о——_———_—___ TABLE 10. Shell measurements (mm) of males and females from three populations of Mexistiobia manantiali. Snails with the dominant maximum whorl number were used. N = 9, Mean + standard deviation. “p” refers to the significance level for the difference between shell lengths of males and females (t-test) for that population. Length of Length of Width of Whorls Length Width body whorl aperture aperture p Locality 51 д 3.0 1.191 ==10:08.. 0/99: 0/04 1.08 = 0.08 0:64 = 0.04. 053-00 >10 2 3:25 1e2ie== 0104) 1.04 0:07 1120.05 (0.63 = 0.037 0620103 Locality 65 3 3.0 0:98 = 0:04 0:95 +'0:04 0.84 = 0.02 0:54 +0:01 0.46 = 0.02 <.005 2 3.25 1.10220:05 1:03 - 0:05, 0:93 0104) 0:56 = 0.037 10:4710:02 Locality 68 3 3.0 ОО: 05 1:0 = 10:0 0/94/3004 20:59 0:03 10:53 = 0:02. 025 ? 3:25 ISO OS 1511200 0.95== 0/04 10/60/= 0103) 0:52 = 0103 (Fig. 14A). The operculum (Fig. 13B) has 3.5 Radula whorls, and the nucleus is positioned at 39% of the long axis of the operculum. Individuals with reduced body pigment were sometimes taken from the mops, while snails from downstream always have dark body pigment. The upstream pigment loss may be because the springhead is usually covered by riparian vegetation, and mimics a subterranean environment. A similar up- stream-downstream pigment change is re- ported for the amphipod Hyalella in Cuatro Ciénegas (Holsinger & Minckley, 1971). The usual dark pigmentation of the snails is prob- ably an adaptation to life in the open stream waters that are subject to great insolation. The radula is shown in Fig. 12. The central cusps of the central and lateral teeth are blade-like. The marginals have numerous cusps. The central tooth has a single pair of basal cusps originating from the lateral angles (Figs. 12B, C). Radular statistics and the var- ious cusp arrangements of the four tooth types are given in Tables 12 and 13, respec- tively. Female Reproductive Anatomy The organization of the female reproductive system is shown in Fig. 14. The female gonad 50 HERSHLER a ——— 0.5 mm С FIG. 13. Head, operculum and penis of Mexistiobia manantiali. A. Dorsal aspect of the head. В. Operculum. C. Dorsal aspect of the penis. Note the small, slender penial lobe (Plo). The dashed line indicates the region with melanin. D. Ventral aspect of the penial lobe showing the glandular ridge (Gir), with a close-up of the small glands. Ey—eye; Glr—glandular ridge; Plo—penial lobe; Sn—snout; Tn—tentacle; Vd—vas deferens. (Go), a single lobed mass, occupies only 17% of the body length. The oviduct (Ov) dis- appears beneath the pallial oviduct (Apo + Ppo) at the end of the style sac (Sts, Fig. 14A). The pallial oviduct is divided into two equal sections: the albumen gland (Ppo) and the capsule gland (Apo). The pallial oviduct constitutes 28% of the body length. The pos- teriormost 30% of the pallial oviduct overlies the style sac. The anterior end of the pallial oviduct is 0.3 mm from the mantle edge. The relationship between the bursa copu- latrix complex and the pallial oviduct is shown in Fig. 14B. The bursa copulatrix complex is CUATRO CIENEGAS HYDROBIIDS TABLE 11. Dimensions (mm) of counts of non- neural organs and structures of Mexistiobia man- antiali. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Body L 257+0.18 2.40 + 0.20 Gill filament number (N = 7) 11.3. = 0.95 Gonad (М = 7) Ё (0:45 = 0:06 0:94 =0.11 WE 0:30 = 10:02` 10!29)-£10/03 Prostate (М = 8) L 0.39 = 0.04 W 0.21 == 008 Penis [E ESOO W 0.34 + 0.04 Pallial oviduct 0.3 0:09 (М = 8) W 0.31 + 0.03 Bursa copulatrix ESOS == 0:02 (М = 6) W 0.07 + 0.02 Seminal receptacle L 0.10 + 0.02 (body) (N= 7) W 0.06 + 0.01 Seminal receptacle L 0.09 + 0.01 (duct) (М = 7) W 0.04 + 0.01 TABLE 12. Radular statistics from 12 individuals of Mexistiobia manantiali. X = mean, S = standard deviation. Measurements in mm. Radular feature x 5 Length 0.394 0.033 Width 0.069 0.005 Number of rows 56.5 6.5 Number of rows in formative stage 2.83 3.32 Width of central tooth (N = 8) 0.013 0.0004 51 dorsal to the pallial oviduct. The bursa (Bu) is 21% of the pallial oviduct length, and lies about 0.27 mm anterior to the end of the pallial oviduct. The seminal receptacle (Sr) and single oviduct coil are largely posterior to the bursa. In four of nine females dissected, the tip of the seminal receptacle protruded slightly posterior to the end of the pallial ovi- duct. The pouch-like seminal receptacle is similar in shape to and only slightly smaller than the bursa, but was easily distinguished by its pink sheen. The duct of the seminal receptacle (Dsr) is short. No gonopericardial duct was seen. The bursa copulatrix joins the common opening of the ventral channel and albumen gland just posterior to the end of the mantle cavity (Fig. 14B). The ventral channel is only slightly folded toward the ventral side of the pallial oviduct and the bolster is small (Figs. 14B, C). The walls of the ventral channel do not fuse anteriorly (Fig. 14D). Male Reproductive Anatomy The male gonad is lobed and is 38% of the body length. The seminal vesicle coils on the posterior stomach chamber. The prostate overlies the mantle cavity. The anterior vas deferens exits from the posterior portion of the prostate. The penis (Fig. 13C) has a slender, taper- ing penial filament, and the slender penial lobe is positioned on the inner curvature slightly closer to the tip than to the base of the penis. The penis has neither cilia, nor a ter- TABLE 13. The various cusp arrangements of the four tooth types of Mexistiobia manantiali, counted from 5 radulae using SEM, with the percentage of radulae showing that arrangement at least once. Central Lateral anterior cusps basal cusps % cusps 4-1-4 40 4-1-3 1-1 5-1-5 60 5-1-3 1-1 5-14 20 5-1-4 1-1 6-1-5 20 1-1 6-1-6 60 % 80 80 20 Inner Outer marginal marginal cusps % cusps % 19 20 22 40 20 40 23 40 21 60 24 60 22 60 25 80 23 40 26 60 24 20 52 HERSHLER FIG. 14. Female reproductive anatomy of Mexistiobia manantiali. A. Snail uncoiled, exposing the ventral aspect without head and kidney tissue. B. Oriented as in A, but with a portion of the pallial oviduct removed to reveal the bursa copulatrix complex, bolster (Bvc), and ventral channel (Vc). C. Cross-section of the pallial oviduct (looking anteriorly), cut just at the common opening of the bursa copulatrix complex and albumen gland. Note the small size of the bolster (Bvc) and reduced ventral channel (Vc). D. Dorsal view of the capsule gland (Apo), showing the opening of the common genital aperture (Cga) at the anterior end of the pallial oviduct. Apo—capsule gland; Ast—anterior stomach chamber; Bu—bursa; Bvc—bolster of ventral channel; Cae—caecum of stomach; Cga—common genital aperture; Cl—columellar muscle; Coi—coil of oviduct; Dbu—duct of the bursa; Dsr—duct of the seminal receptacle; Emc—posterior end of the mantle cavity; Es—esophagus; Go—gonad; In—intestine; Lpo—lumen of pallial oviduct; Ma—mantle edge; Ov— oviduct; Ppo—albumen gland; Pst—posterior stomach chamber; Sr—seminal receptacle; Sts—style sac; Vc—ventral channel. CUATRO CIÉNEGAS HYDROBIIDS 53 minal eversible papilla. The penial filament is darkly pigmented (pigmented area indicated by dashed lines in Fig. 13C) and occasionally the penial lobe is also pigmented. The vas deferens does not coil in the penis. The penis has Gl, and Gl, glands. While the outer cur- vature has no folds, the inner curvature has folds from the base to the penial lobe. The penial lobe is slender and tapers toward its distal end. Viewed from the ventral aspect (Fig. 13D), the single curved glandular ridge is seen. The ridge is elevated above the ventral surface of the penial lobe and consists of two rows of small glands that discharge through a central slit (see close-up, Fig. 13D). Littoridininae Coahuilix Taylor, 1966 Type-species: Coahuilix hubbsi Taylor, 1966 Distribution: endemic to the Cuatro Ciéne- gas Basin. Species included: Coahuilix hubbsi, Coahuilix landyei п. sp. Description Diagnostic features (unique among littoridi- nines) include a minute (width, 0.85- 1.40 mm) planispiral shell (Figs. 15A-G, I-K); intestine with a coil near its anterior end (Inc, Fig. 17C); basal cusps on the central tooth of the radula arising from the tooth face (Fig. 16C); coiling of the seminal vesicle on the posterior stomach chamber (Sv, Fig. 17C); and position of the prostate posterior to the mantle cavity (Fig. 17A). The apical whorl has pitted microsculpture (Fig. 16A); the animal is blind (without eyes) and unpigmented; the digestive gland tuber- cles are reduced to low swellings; the caecal chamber does not protrude posterior to the stomach; gonads of both sexes are a single non-lobed mass (Figs. 17B, C), with that of the female overlying the posterior stomach chamber; the pallial oviduct is divided into three tissue types (Fig. 17B); the oviduct coils, gonopericardial duct, and seminal receptacle are absent (Fig. 17B); the spermathecal duct is elongate and opens separately from the pallial oviduct (Sd, Fig. FIG. 15. SEM photos of Coahuilix hubbsi, Coahuilix landyei and Orygoceras (?) sp. Shells A-E are Coahuilix hubbsi from Locality 64; shells Е, С, 1, J, К are paratypes of Coahuilix landyei (ANSP 355211) from Locality 64; and shell His Orygoceras (?) sp. from Locality 67. Shell A is 0.871 mm wide, and all others are printed to the same enlargement except H, the tube of which is 2.26 mm long. 54 HERSHLER FIG. 16. SEM photos of shell and radula of Coahuilix hubbsi. A. Apical shell whorls, showing wrinkled, pitted microsculpture. B. Part of the radu- lar ribbon. C. Central teeth showing the origin of the basal cusps from the face of the tooth. 17B); the females are oviparous; the penis has a bulb-like lobe bearing a large apocrine gland (Agl, Fig. 17D). Discussion Within the Littoridininae, Coahuilix and Paludiscala form a subgroup as they share numerous features, many related to the small size of the snails and their unique habitat (see Tables 53-55, Figs. 49, 50). The female re- productive system of these snails is charac- terized by loss of the oviduct coils, gonoperi- cardial duct, and seminal receptacle. Coahuilix is distinguished from Paludiscala by the unique features listed above and by dif- ferences in the secondary sperm storage sacs (Coahuilix, absent; Paludiscala, present) and the condition of the openings of the spermathecal duct and pallial oviduct (Coahuilix, separate; Paludiscala, joined). Coahuilix hubbsi Taylor, 1966 Holotype: UMMZ 2220180 Type-locality: Pozo de la Becerra (Locality 10): only empty shells of this species have been found in this large spring. Habitat: Living Coahuilix hubbsi has been obtained only from mops placed into or just below small springheads. Downstream col- lecting efforts, with fine hand sieves, never yielded live specimens. Nor were they found when bottom material from the spring runs was collected and examined under the micro- scope. However, at Locality 64, mops were accidentally placed three meters down from the springhead, where the stream was still completely covered by riparian vegetation, and numerous living specimens were ob- tained. Coahuilix hubbsi was only moderately com- mon in the mop samples; while the species was found on mops from 15 of 38 small springheads, it never comprised more than 15% of the snails from the mops from any springhead. Only a few springs yielded more than 10 Coahuilix hubbsi per mop. The fact that Coahuilix hubbsi is blind and unpigmented, together with its apparent restriction to groundwater outlets, suggests that the species may also live in subterranean waters in the basin. Other snail taxa in Cuatro Ciénegas with a similar habitat are Coahuilix landyei, Paludiscala caramba, and Ory- goceras (?) sp. Description The shell (Figs. 15A—E) is less than 1.0 mm wide and has 2.3-2.5 whorls when adult. The last tenth of a whorl is slightly inflated (Fig. 15A). The aperture is inclined about 30° to the coiling axis. A small segment of the inner lip of the aperture is noticeably flared (Figs. 15B, C). Post-embryonic sculpture consists of CUATRO CIÉNEGAS HYDROBIIDS 55 RO PI | 0.1 mm FIG. 17. Aspects of the anatomy of Coahuilix hubbsi. A. Operculum. B. Ventral aspect of uncoiled female without head and kidney tissue. Note the simplified gonad (Go), lack of seminal receptacle, and differentia- tion of the capsule gland into two tissue types (Apo, and Apo). С. Ventral aspect of uncoiled male without head and kidney tissue. Note the simplified gonad (Go), coiling of the seminal vesicle (Sv) on the stomach, small prostate (Pr) and anterior intestine coil (Inc). D. Dorsal aspect of the penis showing the large bulb-like penial lobe (Plo) with a single apocrine gland (Agl). Agl—apocrine gland; Apo,—posterior capsule gland; Apo;—anterior capsule gland; Ast—anterior stomach chamber; Bu—bursa; Emc—posterior end of mantle cavity; Es—esophagus; Go—gonad; In—anterior intestinal coil; Ma—mantle edge; Ov—oviduct; Plo—penial lobe; Ppo—albumen gland; Pr—prostate; Pst—posterior stomach chamber; Sd—spermathecal duct; Sdu— sperm duct; Sts—style sac; Sv—seminal vesicle; Vd—vas deferens; Vd,—vas deferens from seminal vesicle to prostate; Vd.—vas deferens from prostate to penis. weak growth lines. The animal lacks gills, but (Localities 58, 38 and 67), shell widths for 20 has an osphradium. Shell Shell measurements for specimens from one population (Locality 64) are given in Table 14. The shell widths for females are greater than those for males (p < 0.05, Table 14). For each of three other populations adult specimens (Sexes mixed) were meas- ured with the following means and standard deviations: 0.822 + 0.055 mm, 0.829 = 0.056 mm, and 0.799 + 0.075 mm, respec- tively. The shell occasionally has a small spire (Fig. 15E). The last quarter whorl dips abapi- cally away from the preceding whorls. The peristome is complete and slightly thickened. 56 HERSHLER TABLE 14. Shell measurements (mm) of males and females of Coahuilix hubbsi (from Locality 64). N = 9. Mean + standard deviation. “p” refers to the significance level for the difference between shell widths of males and females (t-test). Length of Width of Whorls Shell length Shell width aperture aperture p ‹ 2.3-2.5 0.36 + 0.02 0.85 + 0.04 0.36 + 0.02 0.32 + 0.02 05, о 2:5 0:37 = 10103 0.89 + 0.05 0.36 + 0.02 033110102 TABLE 15. Dimensions (mm) of non-neural organs and structures of Coahuilix hubbsi. N = 5. Mean + standard deviation. L = length, W = width. Females Males Body L 1.25 = 0.08 1.24 + 0.08 Gonad L 0.36 + 0.04 0.29 + 0.03 W 0.16 + 0.01 0.16 + 0.004 Prostate L 0.19 + 0.01 W ONO ESOO Penis pá 0:35 10/01 W 0.10 + 0.01 Pallial oviduct Е 0.331 = 0:02 W 0.13 + 0.02 Bursa copulatrix EMO (040) W 0.07 + 0.02 TABLE 16. Radular statistics from 5 individuals of Coahuilix hubbsi. X = mean, S = standard devia- tion. Measurements in mm. Radular feature x S Length 0.214 0.011 Width 0.034 0.002 Number of rows 68.8 3.42 Number of rows in formative stage 5.48 0.55 No specimens were found with the extreme flaring of the aperture shown by Taylor (1966, figs. 9, 12). For nine specimens from one population (Locality 64), the width of the tip of the apical whorl averaged 0.081 + 0.006 mm; the width of the first whorl was 0.138 + 0.015 mm. No specimen seen had the strong growth lines shown by Taylor (1966, fig. 13). Nonreproductive Features Observations and data on external features and anatomy are from the population at Locality 64. Measurements of organs and structures are given in Table 15. The snout is squat and the tentacles are short and thick. The buccal mass, pink-red in color, is visible through the snout. There is a concentration of white granules and a slight pinkish color where the eyespot normally would be. The tentacles are without hypertrophied ciliary tufts. The operculum (Fig. 17A) has 3.3 whorls and the nucleus is positioned at 42% of the long axis of the operculum. A light pink color and scattered white granules are seen on the operculigerous lobe. Radula The radula is shown in Figs. 16B, C. The central tooth has well developed lateral an- gles, a small basal process, and a dagger-like central cusp. The basal cusp supports clearly arise from the face of the central tooth (Fig. 16C). Radular statistics and the various cusp arrangements for the four tooth types are given in Tables 16 and 17. Female Reproductive Anatomy The organization of the female reproductive system is shown in Fig. 17B. The gonad is 28% of the body length. The pallial oviduct extends to the anterior edge of the stomach and is relatively small, comprising 26% of the body length. The oviduct enters the posterior portion of the albumen gland (Fig. 17B). The capsule gland is composed of a large poste- rior white-colored section (Apo,, Fig. 17B) and a smaller anterior grey-colored section (Apo>). The sac-like bursa is positioned dorso-laterally to the pallial oviduct and has its posterior end even with that of the albumen gland. The bursa is 38% of the length of the pallial oviduct. A thin sperm duct (Sdu) issues from the anterior end of the bursa and joins the oviduct at the opening of the albumen gland (Fig. 17B). The posterior portion of the albumen gland, where the oviduct (Ov) and sperm duct (Sdu) jointly enter, has a pink sheen, indicating that sperm is inside. This region differs from the remaining albumen gland in that it is thin-walled and non- glandular: it may be a secondary sperm stor- CUATRO CIÉNEGAS HYDROBIIDS 5% TABLE 17. The various cusp arrangements of the four tooth types of Coahuilix hubbsi, counted from five radulae using SEM, with the percentage of radulae showing that arrangement at least once. Inner Outer Central Lateral marginal marginal anterior cusps basal cusps % cusps % cusps % cusps % + =3 100 5=1-3 100 16 40 16 20 —1 4-1-4 40 5-14 20 17 60 1174 60 1-1 6-1-4 20 18 80 18 80 19 20 ИИ 20 20 20 21 20 age area (as there is no seminal receptacle). The spermathecal duct (Sd) is usually tightly appressed to the pallial oviduct. Male Reproductive Anatomy The male gonad is 23% of the body length. The vas deferens branches off the anterior end of the gonad and the seminal vesicle consists of only a few coils (Sv, Fig. 17C). The anterior vas deferens exits from the anterior tip of the prostate. The penis has a short penial filament. The penial lobe (Plo, Fig. 17D) is located at 46% of the length of the penis from the base (on the outer curvature), and is slightly taller than it is wide, measuring 0.098 by 0.080 mm. Folds are seen on the outer curvature of the penis from the base to the penial lobe; the inner curvature has folds for 75% of the penis length from the base. The vas deferens does not coil in the penis. Infrequent con- centrations of Glo glands are seen in the penis. The penis is neither ciliated, nor does it have a terminal papilla. The single massive apocrine gland in the penial lobe occupies slightly more than one- half of the height of the lobe. The gland open- ing is clearly visible and almost circular in cross-section. Its detailed structure is the same as that of Heleobops (Thompson, 1968, figs. 38D, E). Coahuilix landyei Hershler, n. sp. Synonymy: Coahuilix, n. sp. Hershler, in press. Etymology: named after Mr. J. Jerry Land- ye, a student of the freshwater molluscs of the southwestern U.S.A., and México. Types: holotype, ANSP A9894n; paratypes (и). 355217195; 156 Gal К: Again, because of their small size the shells were photographed using SEM, and therefore the holotype was not used. The paratypes used look like the holotype. Type-locality: Locality 64. Habitat: Coahuilix landyei has been col- lected, with one exception, only from mops placed in small springheads. In one spring (Locality 63) several specimens were taken live from Chara mats five meters downstream from the groundwater outlet. Coahuilix land- yei was collected from mops from 13 of 38 small springheads. The species always com- prised less than 13% of the collection from any springhead, and never totalled more than four specimens per mop. Description While only a few specimens of Coahuilix landyei were dissected, all aspects of anat- omy seen were basically the same as those of Coahuilix hubbsi. The shell (Figs. 15F, С, I, J-L) differs from that of Coahuilix hubbsi in the following re- spects: 1) adults have one more whorl and are larger (width, to 1.31 mm) than Coahuilix hubbsi; 2) the last tenth of a whorl is much more inflated than that of Coahuilix hubbsi; 3) the growth lines of the body whorl are much more pronounced than those of Coahuilix hubbsi; 4) the last third of the body whorl 58 HERSHLER TABLE 18. Shell measurements (mm) of adult Coahuilix landyei (>3.25 whorls, sexes mixed) from two populations. Mean + standard deviation. The shells measured from Locality 64 are paratypes (ANSP 355211). Length of Width of Shell length Shell width aperture aperture Locality 64 N=9 0.50 + 0.04 1.28 + 0.02 0.52 + 0.03 0.39 + 0.04 Locality 67 N=7 0.49 + 0.03 1.31 + 0.06 0.49 + 0.03 0.41 + 0.03 ments). Shell measurements for two pop- 40 A ulations of Coahuilix landyei are given in Table 18. Discussion 30 JC. hubbsi The differences between these two species | are all associated with Coahuilix landyei hav- MC. landyei ing one more shell whorl than Coahuilix hubb- si. Separate specific status is suggested by 20 the two taxa being found together on mops at six localities (suggesting sympatry); in these cases the whorl count difference remained PES mm 0.653 0.832 101 119 127 1.54 B NUMBER OF INDIVIDUALS + 0.653 0.832 1.01 1.19 1.27 1.54 SHELL WIDTH FIG. 18. Shell width frequency distributions from Coahuilix hubbsi and Coahuilix landyei collected from mops from two localities. A. Locality 67. B. Locality 64. Almost all specimens taken were adults. Note the obvious size difference between the two sympatric species (analyzed in Table 12). overlaps the preceding whorl (Figs. 15F, J, K), while that of Coahuilix hubbsi merely touches the preceding whorl; 5) the aperture is much more inclined to the coiling axis than that of Coahuilix hubbsi; and 6) the inner lip of the aperture is much less flared than that of Coahuilix hubbsi. The animal has 10-12 gill filaments (Coahuilix hubbsi lacks gill fila- pronounced. The shell width frequency distri- butions (virtually all adult shells) for mop col- lections of the two species from two of these localities are shown in Fig. 18 and analyzed in Table 19. Note that the differences between the shell width means are highly significant for both localities. Immature Coahuilix landyei, with the same size and whorl number as adult Coahuilix hubbsi, are distinguishable from the latter as their apertures are neither thickened nor flared (Fig. 15G). Very small immature specimens (<2 whorls), were not found and probably could not be specifically identified. Other workers have noted two small, plani- spiral hydrobioid species in Cuatro Cienegas, and it is likely that the snail referred to but not described or figured as Hauffenia sp. (Holsin- ger & Minckley, 1971, p. 444) is Coahuilix landyel. Paludiscala Taylor, 1966 Type-species: Paludiscala caramba Taylor, 1966. Distribution: endemic to the Cuatro Ciéne- gas Basin. Species included: monotypic. Description Unique features include the two to three prominent swellings on the coilless oviduct (Fig. 22A) and the disc-like “pouch” that CUATRO CIÉNEGAS HYDROBIIDS 59 TABLE 19. Analysis of shell width frequency distributions shown in Fig. 18. “p” refers to the significance level for the difference between shell widths of the two sympatric species (t-test). Shell width (mean + N standard deviation) p Locality 67 C. hubbsi 20 0.799 + 0.075 .005 C. landyei 13 1.227 70.134 Locality 66 C. hubbsi 103 0.818 + 0.024 <.005 C. landyei 8 1.37 + 0.290 FIG. 19. SEM photos of shells of Paludiscala caramba. Shells A and В are from Locality 63, С and D are from Locality 38. Shell A is 2.42 mm long (B is printed to same enlargement), shell C is 1.58 mm long (D is printed to same enlargement). bulges from the ventral surface of the albu- men gland (Figs. 22A, C, D). The shell (Fig. 19) is small (length, 1.40- 2.60 mm) and turriform, with or without lamel- liform costae; the apical whorl has pitted mi- crosculpture (Fig. 20A); the animal is blind and unpigmented; the tentacles have Hydrobia-like hypertrophied ciliary tufts; the digestive gland tubercles are reduced to low swellings (Fig. 22B); the caecal chamber does not protrude posterior to the stomach (Fig. 22B); the pallial oviduct contains four distinct tissue sections (Fig. 22A); the seminal receptacle and gonopericardial duct are absent; the spermathecal duct (Sd) is elon- gate and has a common opening with that of the pallial oviduct (Fig. 22F); females are oviparous; the penis has a bulb-like lobe be- aring a large apocrine gland (Agl, Fig. 21C). Among littoridinines, Paludiscala is most similar to Coahuilix (see above; Tables 53- 55, Figs. 49, 50): 60 HERSHLER FIG. 20. SEM photos of the apical whorls and radula of Paludiscala caramba. A. Apical shell whorls, showing pitted microsculpture. B, D. Central teeth. C. Part of the radula ribbon. Paludiscala caramba Taylor, 1966 Holotype: UMMZ 220164. Type-locality: Locality 74. Living Paludisca- la caramba have not been found at this local- ity. Habitat: Paludiscala caramba was by far the most common species found in the small springheads; 32 of 38 of these springs yielded this species from mops. Of the 23 springs that yielded more than 100 snails from mop col- lections, Paludiscala caramba comprised greater than 10% of the collection for spring 18, and greater than 50% of the collection for 15. Perhaps more so than Coahuilix, Paludis- cala caramba can extend downstream when there is riparian vegetation covering the stream. At Locality 63, a small thermal (33-— 35°C) spring issues into a pool (see Brown, 1974, fig. 5), and then runs 170m before terminating in a marsh. Paludiscala caramba was very abundant on plant and rock surfaces for the upper 83 m, which had virtually com- plete vegetative cover. Below 83m the vegetative cover ended and no Paludiscala caramba were found, despite intensive col- lecting which yielded quantities of Duran- gonella coahuilae and Mexistiobia manantiali. Paludiscala caramba appears to have a sim- ilar pattern of distribution in other springs. Shell The shell of Paludiscala caramba has up to 7.5 rounded whorls. Shell measurements for two populations are given in Table 20. For CUATRO CIÉNEGAS HYDROBIIDS 61 0.5 mm 0.1 mm 0.25 mm B FIG. 21. Head, penis, and operculum of Paludiscala caramba. A. Dorsal view of the head. Note the Hydrobia-like hypertrophied ciliary tufts on the left tentacle and the lack of eyes. В. Operculum. С. Dorsal aspect of the penis showing the bulb-like penial lobe (Plo) with a single apocrine gland (Agl). The attachment area (Att) of the penial lobe to the penis is hidden by the lobe's curvature. Agl—apocrine gland; Att—attachment of penial lobe to penis; Plo—penial lobe; Sn—snout; Tn—tentacle; Vd—vas deferens. both populations, there was no significant sexual dimorphism in shell length (p > 0.1). For most populations sampled, the shells have fairly tall, thin costae that are curved in profile (Figs. 19A, B). The costae begin after 0.8 whorls and continue to the aperture. Cos- tae spacing is irregular: for 16 shells (sexes mixed) with 7.5 whorls from Locality 63, the penultimate whorl had 9.5 + 1.7 costae (range of 6-12), and the body whorl had 10.5 + 1.4 costae (range of 9-13). A few populations (Localities 38, 67) had smaller individuals (6.5 whorls, 1.4 mm shell length) with costae reduced or absent (Figs. 19C, D). The aperture is inclined only 10° to the coiling axis. The peristome is complete, slightly thick- ened, and adnate to or just free from the preceding whorl. The inner lip is slightly flared. Nonreproductive Features The anatomical description and data (Table 21) are from the population from Locality 63. The snout (Fig. 21A) is elongate, as are the tentacles. There are four or five ciliary tufts on the tentacles and the tufts are restricted to the outer edge of the left tentacle. There is a small concentration of white granules and a pink color in the areas where the eyespots normally would be. Crystalline granules are seen on the ventral body surface. The gills are reduced in number (Table 21). The oper- culum (Fig. 21B) has three whorls and the nucleus is positioned at 38% of the long axis of the operculum. The operculigerous lobe has a narrow band of crystalline granules and a small area of red-pink color. Radula The radula is shown in Figs. 20B-D. The central tooth has a single pair of basal cusps that originate from the lateral angles (Figs. 20B, D). Radular statistics and the various cusp arrangements for the four tooth types are given in Tables 22 and 23. Female Reproductive Anatomy The organization of the female reproductive system is shown in Figs. 22A, C, D, E, F. The 62 HERSHLER Sdu — A 0.5 mm FIG. 22. Aspects of the anatomy of Paludiscala caramba. A. Ventral aspect of the uncoiled female without the head and kidney tissue. A section of the oviduct (Ov) has been removed. Note the swellings of the oviduct, the albumen gland pouch (Agp), and capsule gland differentiated into three distinct regions (Apo,, Apo», and Apoz). В. Ventral aspect of the digestive gland (Dg), showing the tubercles reduced to mere swellings. C. Posterior pallial oviduct oriented as in A, but with the albumen gland pouch (Agp) folded ventrally to expose its basal connection to the main portion of the albumen gland. D. Cross-sectional view of the posterior end of the pallial oviduct showing how the albumen gland pouch (Agp) and oviduct (Ov) jointly open into the albumen gland proper. E. Base of the albumen gland pouch (Agp). Note that the oviduct (Ov) and sperm duct (Sdu) join together at the opening to the albumen gland, which is differentiated from the distal portion of the albumen gland pouch. F. Oriented as in A, but with most of the pallial oviduct cut away to reveal the bursa (Bu). Note that much of the bursa is anterior to the end of the mantle cavity (Emc) and that the spermathecal duct (Sd) joins the pallial oviduct at its anterior end. Agp—albumen gland pouch; Apo,—posterior capsule gland; Apo2—middle capsule gland; Apoz—anterior capsule gland; Ast—anterior stomach chamber; Bu—bursa; Cl—columellar muscle; Dg—digestive gland; Emc—posterior end of the mantle cavity; Es—esophagus; Go—gonad; In—intestine; Ma—mantle edge; Oo—oocyte; Oov—opening of the oviduct; Opo—opening of the pallial oviduct; Ov—oviduct; Ppo—posterior pallial oviduct; Pst—posterior stomach chamber; Sd—spermathecal duct; Sdu—sperm duct; Sts—style sac. gonad (Go) is a single non-lobed mass that comprises 15% of the body length. The pallial oviduct is 17% of the body length and extends slightly over the style sac. The albumen gland (Ppo) is of normal size. The capsule gland has three distinct tissue sections: a large pos- terior one (Apo,, Fig. 22A), a smaller grey- colored one (Apoz), and a somewhat larger white-colored one (Apo3). The sac-like bursa lies dorsolateral to the pallial oviduct, and does not extend posterior to it. The bursa is 58% of the pallial oviduct length, and its an- terior third lies anterior to the end of the mantle cavity (Figs. 22A, F). A thin sperm duct (Sdu) issues from the anterior end of the bursa and coils slightly on its ventral surface (Fig. 22F). The albumen gland “pouch” (Agp) appears as a disc appressed to the ventral surface of the albumen gland (Fig. 22A). When the pouch is pulled away from the pallial oviduct, its basal connection to the latter is readily seen (Fig. 22C). This is also seen in cross-section (Fig. 22D). The oviduct (Ov) and sperm duct (Sdu) jointly enter the CUATRO CIÉNEGAS HYDROBIIDS 63 TABLE 20. Shell measurements (mm) of males and females from two populations of Paludiscala caramba. Snails with the dominant maximum whorl number(s) were used. N = 9 unless stated otherwise. Mean + standard deviation. “p” refers to the significance level for the difference between shell lengths (t-test) for that population. Length of Length of Width of Whorls Length Width body whorl aperture aperture p Locality 63 2 (n= 10) 7.0 2.26 + 0.08 1.09+0.08 1.04+005 0.71 + 0.04 0.53 + 0.04 “5 РАДЕ 1009) 113 OA OA OOO 071 = 0103° 2053 0103 1 3 7.5 DA NEO NS IE 0:09) 1:0 10/05 0.710104 0153 10102 Locality 27 2 7.5 LOOSE 007 1080107 O74 01060.57 = 0103) =a 3 os 2.54 + 0.06 1.18+0.06 1.09 + 0.06 0.74 - 0.04 0.54 + 0.04 TABLE 21. Dimensions (mm) ог counts of поп- neural organs and structures of Paludiscala caram- ba. М = 5 unless stated otherwise. Mean + stan- dard deviation. L = length, W = width. Females Males TABLE 22. Radular statistics from 5 individuals of Paludiscala caramba. X = mean, S = standard Body Е 3.11 = 0.21 3.01 + 0.15 deviation. Measurements in mm. Gill filament number 13:6) = 0.89 Osphradium (N=6) L 0.15 + 0.02 Y Gonad o: A E E A A W 0.29+0.04 0.27 + 0.03 Length 0.343 0.016 Prostate IE 0.44 + 0.04 Width 0.050 0.004 W 0.19 + 0.03 Number of rows 72.0 2.0 Penis E 0.66 + 0.04 Number of rows in W 0.19 + 0.02 formative stage 5.2 1.6 Pallial oviduct EOFS 5210106 (N = 7) W 0.15 + 0.02 Bursa copulatrix ¡EROS 210103 (N = 8) W 0.14 + 0.02 TABLE 23. The various cusp arrangements of the four tooth types of Paludiscala caramba, counted from five radulae using SEM, with the percentage of radulae showing that arrangement at least once. Inner Outer Central Lateral marginal marginal anterior cusps basal cusps % cusps % cusps % cusps % 4 A 80 4-1-3 80 18 20 16 20 7 40 51153 20 19 20 17 20 : 1.3 80 4-1-4 20 20 20 18 20 ee : 20 21 40 22 40 ; 8 20 22 40 23 80 23 40 24 40 24 20 25 20 64 HERSHLER FIG. 23. SEM photos of Cochliopina milleri and Cochliopina riograndensis. Shells А-Н are Cochliopina milleri from Locality 38, shell | is Cochliopina riograndensis from Locality 101. Shell A is 3.26 mm wide; the others are printed to the same enlargement. albumen gland at the base of the pouch; this area is somewhat differentiated and appears to be a sphincter (Fig. 22E). There is no seminal receptacle of normal shape or posi- tion. The pouch, while non-glandular, is obviously of pallial oviduct origin, yet was seen to hold sperm and probably serves as a secondary seminal receptacle, as may the oviduct swellings. The spermathecal duct (Sd) may be tightly appressed to the col- umellar side of the pallial oviduct (Fig. 22A) or may be slightly separated from it (Fig. 22F). The male gonad is a single non-lobed mass. The prostate is relatively large and extends considerably anterior to the end of the mantle cavity. The anterior vas deferens exists from the anterior tip of the prostate. The penis is bluntly shaped and has neither cilia nor a terminal papilla. The penial lobe (Plo) is located at 67% of the penis length (on the outer curvature) from the base, and has a spherical shape with a diameter of 0.15 mm (Fig. 21C). The lobe overlies its attachment area to the penis, which is short and located on the outer curvature (Att, Fig. 21C). The apocrine gland (Ад!) extends for slightly more than one-half of the diameter of the penial lobe. The structure of the gland is precisely that of Coahuilix. Folds are seen on the outer curvature of the penis from the base to the penial lobe. The inner curvature has folds for virtually its entire length. The vas deferens (Vd) does not coil in the penis. Infrequent Glo glands are seen in the penis. CUATRO CIENEGAS HYDROBIIDS FIG. 24. SEM photos of shell and radula of Cochliopina milleri. A. Shell apex, showing wrinkled, pitted microsculpture. B. Portion of the body whorl showing spiral lines and collabral growth lines. C. Close-up of B. D. Central tooth of the radula. E. Outer marginal teeth. F. Part of radular ribbon. Cochliopina Morrison, 1946 Type-species: Cochliopina riograndensis (Pilsbry 4 Ferriss, 1906). Distribution: Rio Grande drainage, from Texas south to Panama. Species included: 20 species listed by Taylor (1966). Description Cochliopina has no distinctive unique fea- tures, but is recognizable by a combination of character states (see below). The shell (Fig. 23; Morrison, 1946, pl. 2, figs. 7-9, 11-13) is small (width, 5 mm) and planispiral to low-trochoid in form; the sculp- ture consists of spiral lines or cords, frequent- ly bearing periostracal bristles; the apical whorl has pitted microsculpture (Fig. 24A); the tentacles have Spurwinkia-like hyper- trophied ciliary tufts (Figs. 25A, B); females are ovoviviparous; the pallial oviduct has a slight posterior bend; the albumen gland is reduced in size (Ppo, Fig. 26A); the seminal receptacle opens into the oviduct (Figs. 26B, C); the oviduct and anterior end of the bursa are connected by a short sperm duct (Sdu, Figs. 26B, C); the non-muscular spermathe- cal duct opens just beyond the posterior end of the mantle cavity (Figs. 26A, C); the anteri- or end of the brood pouch is muscularized and coiled toward the columellar side (Fig. 26E); the penis is non-lobed, with an elongate penial filament, and lacks specialized glands (Fig. 25D; Morrison, 1946, pl. 3, figs. 9-15). Discussion Cochliopina and Mexithauma share numer- ous features (see Tables 53-55, Figs. 49, 50) relating to shell shape, sculpture, tentacle ciliation, reproductive mode, coiling of the pal- lial oviduct, reduction of the albumen gland size, connection between the seminal receptacle and oviduct, muscularization and coiling of the end of the brood pouch, and the form of the penis (and lack of specialized glands). Distinctive features of the female re- productive system shared by these taxa in- clude the slight posterior bend of the pallial oviduct, the opening of the seminal receptacle into the oviduct, and the well-developed muscularization and coiling of the anterior end of the brood pouch. The two taxa differ in the following features: 66 HERSHLER LE BE FIG. 25. Head, operculum, and penis of Cochliopina milleri. A. Dorsal aspect of the head. Note the Spurwinkia-like hypertrophied ciliary tufts (Ci) on the tentacles. B. Close-up of ciliation pattern on left tentacle. C. Operculum. D. Penis with penial folds (Pf), but no lobes. Ci—ciliary tufts on tentacle; Ey—eye; Sn—snout; Tn—tentacle; Vd—vas deferens. CUATRO CIÉNEGAS HYDROBIIDS 67 FIG. 26. Female reproductive anatomy of Cochliopina milleri. A. Ventral aspect of uncoiled snail without head and kidney tissue. Note the posterior bend of the pallial oviduct. B. Posterior region of the pallial oviduct, oriented as in A, but with a portion of the albumen gland (Ppo) and the bursa cut away to reveal the oviduct coils (Coi), seminal receptacle (Sr), sperm duct (Sdu), and opening of the oviduct into the albumen gland (Oov). C. Oriented as in B, but with the bursa (Bu) in place. D. The seminal receptacle (Sr) with pigment patches. E. Portion of the anterior end of the mantle cavity, showing the muscular coil (Mus) of the anterior end of the brood pouch (Bp). Ast—anterior stomach chamber; Bp—brood pouch; Bu—bursa; Cae—caecum of stomach; Coi—coil of oviduct; Dpe—gonopericardial duct; Dsr—duct of the seminal receptacle; Emb—embryonic shell; Emc—posterior end of the mantle cavity; Es—esophagus; Go—gonad; In—intestine; Ma—mantle edge; Mus—muscular section of the brood pouch; Oov—opening of the oviduct; Ov—oviduct; Ppo—albumen gland; Pst—posterior stomach chamber; Sd—spermathecal duct; Sdu—sperm duct; Sr—seminal receptacle; Sts—style sac. 68 HERSHLER condition of the mantle edge (Cochliopina, smooth; Mexithauma, papillate); apical micro- sculpture (Cochliopina, pitted; Mexithauma, absent); relative size of the female gonad (Cochliopina, large; Mexithauma, small); length of the spermathecal duct (Cochliopina, short; Mexithauma, long); insertion of the sperm duct from the oviduct (Cochliopina, to bursa; Mexithauma, to duct of bursa); and male gonad morphology (Cochliopina, simple lobes; Mexithauma, bush-like). Cochliopina milleri Taylor, 1966 Holotype: UMMZ 220182 Type-locality: Locality 53. Distribution: endemic to the Cuatro Ciéne- gas Basin. In the eastern lobe of the basin, it has not been found south of Locality 97 (Fig. 3). Habitat: Cochliopina milleri is most com- mon in the outflows of cool (<28°C) springs, and is usually sympatric with Nymphophilus minckleyi on aquatic vegetation, particularly Chara and Utricularia. Cochliopina milleri was rarely found in the large spring pools, and was never found in mop or sieve collections from the small springs. Description The generic placement of Cochliopina mil- leri is tentative as the detailed anatomy of the type-species, Cochliopina riograndensis, is not known. Cochliopina milleri does resemble the type-species in shell form and external anatomy. The shell (Figs. 23A-H) is thin, relatively small (length, 1.16-1.48 mm), and broadly conical to planispiral, with little whorl overlap; the narrow spiral cords are numerous, promi- nent, and fringed with light-colored periostra- cum; the aperture is nearly circular and adnate to or free from the penultimate whorl. Shell The shell has rounded whorls and deeply impressed sutures. The aperture is slightly angled adapically and inclined 30—45° to the coiling axis. For the population from Locality 38, coiling abnormalities were frequent and remarkable; varying from a slight loosening of the whorls (Fig. 23E), to a change in coiling direction near the end of the body whorl (Fig. 23H), to near-open coiling (Fig. 23C). Other populations show less coiling variation than that from Locality 38. There are generally 10—20 spiral cords, and numerous spiral lines on the last two whorls. Occasional specimens are almost smooth-shelled (Fig. 23F). The apical whorl microsculpture is somewhat coarser than that seen in other hydrobiid snails (Fig. 24A). Spiral sculpture begins after the first whorl. Close-ups of the spiral sculp- ture are shown in Figs. 24B, C. There are strong axial growth lines that become es- pecially prominent near the aperture (Fig. 24C). Shell measurements from the population from Locality 38 (excluding abnormally coiled specimens) are given in Table 24. Females have a greater shell width (p < .005) and are also relatively taller (t-test for difference be- TABLE 24. Shell measurements (mm) of males and females from one population of Cochliopina milleri and two populations of C. riograndensis. Snails with the dominant maximum whorl number were used. N = 9 unless otherwise indicated. Mean + standard deviation. “p” refers to the significance level for the difference between shell widths of males and females (t-test) for that population. Length of Length of Width of Shell length Whorls Length Width body whorl aperture aperture Shell width C. milleri (Locality 38) 3 3.0 |. 16 == 0:08 722922705157 ТРЕЕ 0.10’ 0965006 D SE 0107 ЕГО 2 3:9 1.780.107 3:08 = 041877 1.57 = 011 711920507 110 = 0106040158005 = <.005 С. riograndensis (Locality 101) д 4.0 2PSE ОИ ОЕ 0.09 1.48 + 0.09 1.22 + 0.07 0.82 + 0.06 ? 4.5 218 OS NA E AS e 0.15 1.59 == 0.19 1:36 == OSO OS ESIOIO6 С. riograndensis (Locality 102) 3 3:5 1545-10: 0252 2414708 1:09) = 0:09) 10/9352 0/08 О ИР ЕЕОТО5 ? 4.0 2.21 = 0.17 279205) 1.86 = 0.12 1.43 = 0108, 1.24 = 0108) NOMS EOS —4.0, —45 2.54 =015 (306 = 012 203 = 0108° 1.54 = 0.0387 1.33 270.0 7229183231008 CUATRO CIENEGAS HYDROBIIDS 69 TABLE 25. Dimensions (mm) or counts of non- neural organs and structures of Cochliopina milleri. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Body (N=6) [16:47 = 0:39 5.15 = 0:31 Gill filament number 269 == 2228 Osphradium (N=8) L 0.35 + 0.05 Gonad (N=6) 69) 0:09" 3280415 W 0.55+0.05 0.57 + 0.03 Prostate L 0.56 + 0.06 W 0.28 + 0.03 Penis LL. 1.24 + 0.05 W 0.36 + 0.04 Pallial oviduct [2.85 = 10:28 W 0.72 + 0.08 Вигза copulatrix L 0.27 = 0.04 W 0.10 = 0.01 Seminal receptacle L 0.10 + 0.01 (body) (М = 6) W 0.06 + 0.01 Seminal receptacle L 0.04 + 0.01 (duct) (N= 9) W 0.04 + 0.01 TABLE 26. Radular statistics from 12 individuals of Cochliopina milleri. X = mean, S = standard devia- tion. Measurements in mm. Radular feature x S Length 0.563 0.028 Width 0.106 0.006 Number of rows 45.6 1.96 Number of rows in formative stage 2.36 12 Width of central tooth (N = 21) 0.031 0.0014 tween means of shell length/width, p < 0.005) than males. This sexual dimorphism is seen by comparing Figs. 23A and B (adult female and male, respectively). Nonreproductive Features Details and data concerning anatomy are from the population from Locality 38. Mea- surements of organs and structures are given in Table 25. The snout is squat and the tenta- cles are elongate in comparison (Fig. 25A). The left tentacle has 10-12 hypertrophied ciliary tufts protruding from the outer edge and numerous ciliary tracts (Ci, Figs. 25A, B) Curving inward toward the center of the tenta- cle from both sides. These tracts are present for 67% of the tentacle length from the base. The tentacle also has a central ciliary tract along its length. The right tentacle alone has the central ciliary tract. The snout and tenta- cles are dusted with melanin to varying de- grees. A dark melanin patch at the base of each tentacle, across from the eyespot, is seen in most individuals. Occasional speci- mens had a dark melanin patch at the tenta- Cle tips. There is a cluster of dull white gran- ules around the eyes and smaller granules are found in the neck, snout, and tentacles. The sides of the head-foot are sometimes darkly pigmented, and in those specimens a non-pigmented strip was seen extending from the neck to the foot. Body pigmentation for the female consists of small melanin patches on the dorsal and ventral body surfaces that interdigitate with clusters of white granules, producing a mottled appearance. The male is similarly pigmented, but has solid dark mela- nin on the ventral surface of the gonad. The caecal chamber is prominent. The operculum (Fig. 25C) has 5.5 whorls and the nucleus is positioned at 42% of the long axis of the operculum. Pigmentation on the operculigerous lobe consists of two to three large melanin patches. Radula The radula is shown in Figs. 24D-F. The central tooth has one to three pairs of basal cusps that originate from the lateral angles. The central cusp of the central tooth is dagger-like. The marginal teeth have many cusps. Radular statistics and the various cusp arrangements for the four tooth types are given in Tables 26 and 27. Female Reproductive Anatomy The organization of the female reproductive system is shown in Fig. 26. The gonad (Go) has three lobate branches, and occupies 26% of the body length. The pallial oviduct occu- pies 44% of the body length and overlies most of the style sac. The anterior portion of the pallial oviduct is modified into a thin-walled, non-glandular brood pouch (Bp) for the stor- age of embryos. The pallial oviduct coils pos- teriorly, the distance from the posteriormost point of the pallial oviduct and the end of the coil being 0.67 mm. The small albumen gland (Ppo, Fig. 26A) constitutes this coiled portion. The oviduct (Ov) disappears beneath the anterior portion of the albumen gland (Fig. 26A). There is a short gonopericardial duct 70 HERSHLER TABLE 27. The various cusp arrangements for the four tooth types in 12 radulae of Cochliopina milleri, with the percentage of radulae showing that arrangement at least once. Inner Outer Central Lateral marginal marginal anterior cusps basal cusps % cusps % cusps % cusps % 4-13 8 4-1-3 83 17 8 19 40 2=2 ts 8 4-1-4 83 18 25 20 40 it 8 5-14 8 19 50 21 40 5153 42 SO 8 20 58 22 60 is lz SES 8 21 92 23 80 4-14 17 32123 42 22 33 24 60 3-3 14 174 23 58 25 60 2=2 = 8 24 8 26 20 3=2 1=5 17. 25 25 2 20 2—1 919 67 26 8 28 20 PE? 125 25 27 8 2=2 28 8 (Dpe). The bursa (Bu) is small, dorsal to, and Male Reproductive Anatomy almost entirely hidden by the albumen gland (Fig. 26A). The elongate seminal receptacle The male gonad consists of 10-12 lobed (Sr) is dorsal to and mostly hidden by the branches and fills the entire length of the bursa. There is usually a light melanin dusting digestive gland, covering the posterior stom- on the seminal receptacle (Fig. 26D). ach chamber. The gonad is 64% of the body The oviduct loops several times dorso- length. The prostate is relatively small, 11% of laterally to the bursa before receiving the the body length, but does overlap the mantle short duct of the seminal receptacle (Dsr, cavity. The anterior vas deferens exits from Figs. 26B, C). The minuscule sperm duct the anterior tip of the prostate. (Sdu) enters the oviduct just where the latter The penis (Fig. 25D) has numerous folds turns to open into the end of the albumen (Pf) on the inner curvature for one half of the gland (Fig. 26B). The spermathecal duct (Sd) penis length. While one of the folds was extends 0.14 mm beyond the posterior end of sometimes noticeably wider than the others the mantle cavity. (as in Fig. 26D), it did not project outward as a The well developed muscular loop (Mus) of penial lobe in any specimen. The outer curva- the anterior brood pouch is shown in Fig. 26E. ture is without folds. Scattered throughout the For 15 adult females, there was an average of penis are Gl, and Gl, gland types. The vas 11.6 + 2.97 shelled embryos in the brood deferens (Vd) coils and thickens from the pouch, as well as another eight to ten very base of the penis until it is even with the end small non-shelled embryos. For 108 shelled of the penial folds, after which it narrows and embryos, the range of shell width was quite stops coiling. The penis, exclusive of the long narrow, 0.317-0.475 mm, and the mean was penial filament, has a dark brown color that is 0.400 + 0.178mm. Embryonic shells have not an external melanin coating, but colored up to 1.3 whorls. tissue. CUATRO CIÉNEGAS HYDROBIIDS 71 Cochliopina riograndensis (Pilsbry & Ferriss, 1906) Holotype: ANSP 91324. Type-locality: debris of the Rio San Felipe near the Rio Grande, Val Verde County, Texas. Distribution: Rio Grande drainage of Texas and northeastern Mexico. While this species had been previously known from the Rio Salado de Nadadores, and an adjacent spring, both just east of the Cuatro Ciénegas Basin (Taylor, 1966), the author also col- lected it at Locality 101 in the southeastern lobe of the basin. Habitat: This species has been found in springs and spring outlets of various sizes (Fullington, 1978; Taylor, 1966). Locality 101 is a large spring pool (the Santa Tecla La- guna) and Cochliopina riograndensis was abundant on unidentified vegetation along its shallow edges. Description This species has been amply described by Fullington (1978) and Taylor (1966). Its shell is distinguished from that of Cochliopina mil- leri by its thicker, larger, and relatively taller appearance (see Table 24: note the ratios of shell length/width). Its whorls overlap greatly and it has a more conical shape than does the shell of Cochliopina milleri. The aperture is angled adapically and the inner lip is partly fused to the penultimate whorl. The spiral cords are few in number and lack promi- nence, but the periostracal bands are darker and wider (especially in the umbilical area) than those of Cochliopina milleri. Discussion The southeastern lobe of the basin has several nonendemic taxa of undisputed Rio Grande (= Rio Bravo) origin, including the cichlid fish, Cichlasoma cyanoguttatum (see Minckley, 1977). The distinctive Rio Grande aspect of the fauna from this portion of the basin contrasts with the more endemic aspect of the fauna from the remainder of the basin. Hubbs & Miller (1965) suggested that the southeastern lobe had a recent surficial con- nection to the Rio Salado de Nadadores (Fig. 2, Number 7, a Rio Grande tributary), thus explaining this pattern. The discovery of Cochliopina riograndensis in the southeast- ern lobe of the basin supports this hypothesis. The relationship between endemic Coch- liopina milleri and non-endemic Cochliopina riograndensis is unknown as the internal FIG. 27. Shells of Mexithauma quadripaludium from Locality 1. The shell on the left is 7.38 mm long, the other is printed at the same enlargement. 72 HERSHLER FIG. 28. SEM photos of shell and radula of Mexithauma quadripaludium. A. Apical whorls of shell. Note lack of microsculpture. B. Portion of penultimate whorl showing strong spiral cords and collabral microsculpture. C. Part of radular ribbon. D. Isolated central tooth. E. Outer marginal teeth. anatomy of the latter is not known. The dis- tinctively fragile and loosely-coiled shell of Cochliopina milleri may be associated with its restriction to fairly cichlid-free waters (i.e., low predation pressure). Mexithauma Taylor, 1966 Type-species: Mexithauma quadripaludium Taylor, 1966. Distribution: endemic to the Cuatro Ciéne- gas Basin. Species included: monotypic. Description Distinctive features of Mexithauma include the papillate mantle edge of both sexes (Pma, Figs. 30A, 31A) and the open channel (Oc) connecting the openings of the spermathecal duct and pallial oviduct (Figs. 30C, E). The shell (Fig. 27) is large (length, 7.0 mm), globose, without umbilicus, and with promi- nent spiral cords fringed with periostracum; the inner lip of the shell is thickened; the tentacles show Spurwinkia-like ciliation (Figs. 29A, B); females are ovoviviparous; the female gonad is very reduced in size (Go, Fig. 30A); a large pallial oviduct overlies the stom- ach and has a slight posterior bend (Fig. 30A); the seminal receptacle (Sr) is posi- tioned lateral to the bursa and opens into the oviduct (Fig. 30B); the oviduct connects with the duct of the bursa via a short sperm duct (Sdu, Fig. 30B); the anterior end of the brood pouch is muscularized and coiled (Figs. 30C, E); the male gonad is bush-like (Go, Fig. 31A); the penis is non-lobed and lacks spe- cialized glands (Fig. 31B). Mexithauma is most similar to Cochliopina (see above; Tables 53-55, Figs. 49, 50). Mexithauma quadripaludium Taylor, 1966 Holotype: UMMZ 220214. Type-locality: Locality 97. Habitat: Mexithauma quadripaludium has been found only in the larger springs and their outflows. It has been collected from all types of aquatic vegetation, sand (composed of CUATRO CIÉNEGAS HYDROBIIDS 73 1.0 mm en eSB ng DME ONE 0.2 mm FIG. 29. Head and operculum of Mexithauma quadripaludium. A. Dorsal aspect of head. Note the Spurwinkia-like hypertrophied ciliary tufts (Ci) and the central pigment streak (Pig) on the tentacles. B. Close-up of ciliation pattern on left tentacle. C. Operculum. The pigment pattern on the operculigerous lobe is also shown. Ci—ciliary tufts on tentacles; Ey—eye; Pig—pigment; Sn—snout; Tn—tentacle. 74 HERSHLER "0.2 mm FIG. 30. Female reproductive anatomy of Mexithauma quadripaludium. A. Ventral aspect of uncoiled snail without the head and kidney tissue. Note the very small gonad (Go) and large pallial oviduct (Bp + Ppo) with a posterior bend. B. Oriented as in A, but with most of the albumen gland (Ppo) cut away to expose the bursa copulatrix complex. C. Anterior portion of the mantle cavity showing the muscular bend (Mus) of the anterior end of the brood pouch. D. Oriented (and scale) as in C, but with a portion of the epithelium of the anterior end of the brood pouch cut away to expose the inner muscular layer (Iml). E. Oriented as in A, showing the openings of the spermathecal duct (Osd) and pallial oviduct (Opo) connected by an open channel (Oc). Ast—anterior stomach chamber; Bp—brood pouch; Bu—bursa; Cl—columellar muscle; Dbu—duct of the bursa; Dpe—gonopericardial duct; Dsr—duct of the seminal receptacle; Ef—efferent vessel; Emc—posterior end of the mantle cavity; Es—esophagus; Gf—gill filament; Go—gonad; Iml—inner muscular layer; In—intestine; Mus—muscular section of the brood pouch; Oc—open channel; Oov—opening of the oviduct; Opo—opening of the pallial oviduct; Osd—opening of the spermathecal duct; Ov—oviduct; Pe— pericardium; Pma—papillate mantle edge; Ppo—albumen gland; Pst—posterior stomach chamber; Sd— spermathecal duct; Sdu—sperm duct; Sr—seminal receptacle. on Nymphaea and Chara, while M. quad- ripaludium is most common in sand or on travertine blocks. However, when one of the species is of reduced abundance or absent, usually in a spring with low microhabitat di- travertine pieces and shell fragments), travertine blocks, and the gently sloping banks of spring pools. While M. quad- ripaludium and Nymphophilus minckleyi over- lap broadly in their microhabitat usage, within any given spring they are largely allopatric on a microhabitat scale. In springs with microha- bitat diversity, N. minckleyi is most common versity, the other species may “switch” to other microhabitats, including that usually occupied by the species that is rare or absent. CUATRO CIÉNEGAS HYDROBIIDS 75 Ast Pma FIG. 31. Male reproductive anatomy of Mexithauma quadripaludium. A. Ventral aspect of an uncoiled snail without head and kidney tissue. Note the bush-like gonad (Go). B. Dorsal aspect of the non-lobed penis. Ast—anterior stomach chamber; Cl—columellar muscle; Emc—posterior end of the mantle cavity; Es— esophagus; Go—gonad; In—intestine; Pf—penial fold; Pma—papillate mantle edge; Pr—prostate; Pst— posterior stomach chamber; Sts—style sac; Sv—seminal vesicle; Vd—vas deferens; Vd,—vas deferens from seminal vesicle to prostate; Vd,—vas deferens from prostate to penis; Ve—vas efferens. 76 HERSHLER TABLE 28. Shell measurements (mm) of males and females of Mexithauma quadripaludium from Locality 1. The shells measured are from the largest 10% of the population (>4.8 whorls). N = 9. Mean + standard deviation. “p” refers to the significance level for the difference between shell lengths of males and females (t-test). Length of Length of Width of Length Width body whorl aperture aperture p d 7.28 + 0.18 6.08 + 0.19 6.46 + 0.50 4.89 + 0.18 3:86 0/81 >.10 о 7.34 = 0.41 6. З0Е= 0.27 6.78 + 0.31 4.94 + 0.18 3.84 + 0.29 TABLE 29. Frequency distribution for number of spiral cords at the aperture of Mexithauma quadripaludium shells. Fifteen shells (sexes mixed) per whorl stage were used. The mean and standard deviation for the shell length of the shells used for each whorl stage are given. Whorls Shell length 7 8 9 10 4.0 3.30 + 0.32 1 3 2 4 4.5 SIREN = 5.07 7.44 + 0.34 ya — Shell Shell measurements for the population from Locality 1 are given in Table 28. The exact number of whorls in large adults cannot be determined as the apex is usually some- what eroded in such specimens. There is no significant sexual dimorphism in size of shell (Table 28). The spiral cords begin at 2.3 whorls and increase in number with shell size (Table 29). Adults have 15-22 cords on the body whorl that vary in height (Fig. 28B). Strong axial microsculpture, also fringed with periostracum, is present (Fig. 28B). The aper- ture is large, occupying more than one-half of the height of the body whorl, and is inclined only 10—20° to the coiling axis. The aperture is elliptical in shape, and is somewhat more angled above than below. While the inner lip is greatly thickened, the outer lip is thin. The apical whorl does not have microsculpture (Fig. 28А). Nonreproductive Features Anatomical descriptions and data (Table 30) are from the population from Locality 1. The snout (Fig. 29A) is relatively squat, while the tentacles are thickened and elongate. Each tentacle has a central dark pigment strip, extending from just beyond the eye to the tentacle tip (Pig, Fig. 29A). On the left tentacle, there are numerous hypertrophied ciliary tufts projecting from the outer edge, as Number of spiral cords 11 12 13 14 15 A OZ = 4 — —( = OS NS QS! DORE > TABLE 30. Dimensions (mm) or counts of non- neural organs and structures of Mexithauma quad- ripaludium. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Body Е 11.5 01485898 a= 032 Gill filament number 51.8. Es 179 Osphradium (N=7) L 0.76 + 0.08 Gonad Ё 0:88 = 007035410726 М/ 0.65 = 0.06 1.29 = 0.12 Prostate L 1.97 + 0.24 W 0.95 + 0.07 Penis pa 4.12 + 0.22 W 5 Ons) Pallial oviduct 76722085 W 1.91 + 0.34 Bursa copulatrix ESTO 2020103 W 0.19 + 0.01 Seminal receptacle L 0.16 + 0.01 (body) W 0.13 + 0.02 Seminal receptacle L 0.12 + 0.01 (duct) W 0.05 = 0.01 well as ciliary tracts curving inwards from both sides. The right tentacle lacks the ciliary tufts projecting from the outer side, and the ciliary tracts run along the length of the tentacle, rather than curving inward. The under-surface of each tentacle also has ciliary tracts running along its length (not figured). Small white granules are scattered in the neck, snout, and tentacles. The snout and neck have a light CUATRO CIÉNEGAS HYDROBIIDS dusting of melanin. The foot is large and thickened. There are distinctive pigment streaks just below the eyes and along the sides of the foot. The dorsal body surface has yellow and white pigment granules as well as melanin. The digestive gland is dark brown and has white granules scattered on its ven- tral surface. A prominent caecal chamber pro- trudes posterior to the stomach. The oper- culum (Fig. 29C) has 3.5 whorls and the nucleus is positioned at 41% of the long axis of the operculum. The characteristic pigment streaks on the operculigerous lobe are shown in Fig. 29C. Radula The radula is shown in Figs. 28С-Е. The central tooth usually has three pairs of basal 77 cusps arising from prominent lateral angles. The marginal teeth (Figs. 28C, E) have rela- tively few cusps. Radular statistics and the various cusp arrangements for the four tooth types are given in Tables 31 and 32. TABLE 31. Radular statistics from nine individuals of Mexithauma quadripaludium. X = mean, S = standard deviation. Measurements are in mm. Radular feature X 5 Length 1.42 0.058 Width 0.188 0.008 Number of rows 66.1 2.98 Number of rows in formative stage 4.44 1.01 Width of central tooth (N = 23) 0.047 0.0002 TABLE 32. The various cusp arrangements for the four tooth types in 11 radulae of Mexithauma quadripaludium, with the percentage of radulae showing that arrangement at least once. Inner Outer Central Lateral marginal marginal anterior cusps basal cusps % cusps % cusps % cusps % 3=1=3 9 2-1-3 9 10 27 11 9 22 3=1= 9 3-1-3 55 11 91 12 64 2=2 313 9 4-1-3 91 12 73 13 82 3—3 4-1-3 18 5-1-3 9 13 36 14 82 3-2 4-1-3 9 4-1-4 18 14 9 1S 45 3—3 4-14 36 16 9 16 18 2-2 4-1-4 45 ly 9 3-3 4-14 27 3—2 5—1-3 9 3=2 SES 18 22 ЕЕ) 18 3—3 5-14 27 3=2 5-14 18 3—3 5-14 9 2-2 GES 18 78 HERSHLER Female Reproductive Anatomy The female gonad (Go) occupies only 8% of the body length and is a mere terminal thickening of the oviduct with a few small lobes. The pallial oviduct is 58% of the body length. The posterior bend of the pallial ovi- duct extends for 2.5 mm, of which 1.2 mm is albumen gland (Ppo, Fig. 30A). The remain- der of the pallial oviduct is a thin-walled brood pouch (Bp). A gonopericardial duct (Dpe, Figs. 30A, B) is present. The sac-like bursa (Bu) is only 4% of the length of the pallial oviduct and is entirely dorsal to the albumen gland (Fig. 30B). The oviduct (Ov) coils once or twice before receiving the short duct from the seminal receptacle (Dsr, Fig. 30B) and then coils to enter the ventral surface of the end of the albumen gland. The spermathecal duct (Sd) is tightly appressed to the col- umellar side of the pallial oviduct (Fig. 30A). In Fig. 30D, the muscularized portion of the anterior brood pouch (Mus) is slit open and the epithelium has been cut away to reveal the inner muscular layer (Iml). For 15 adult females, the number of shelled embryos in the brood sac averaged 19.4 + 4.5 (range of 14-28), and there were an additional 22-35 small, non-shelled embryos packed in the posterior bend of the brood pouch. For 111 shelled embryos, the mean shell length was 0.48 + 0.37 mm. The range of shell lengths was four-fold, from 0.20 to 0.87 mm. The embryonic shells have up to 2.5 whorls. The embryos have red-brown pigment splotches and yellow-white granules on the dorsal body surface. Male Reproductive Anatomy The male gonad (Go, Fig. 31A) has five branches and occupies 29% of the body length, almost filling the digestive gland. The prostate overlaps the mantle cavity and the anterior vas deferens exits from the anterior tip of the prostate. The seminal vesicle coils (Sv) overlap slightly onto the stomach. The penis (Fig. 31B) has an elongate penial fila- ment. There are numerous folds on its inner curvature for slightly more than one half of its length. Where the folds end, the penis sud- denly narrows on the outer curvature, giving the penis a peculiar bulging appearance at this point. The penis has numerous Gl, and Glo glands. The vas deferens (Vd) only coils slightly in the penis. The penis has neither cilia nor a terminal eversible papilla. Durangonella Morrison, 1945 Type-species: Durangonella seemani (Frauenfeld, 1863). Distribution: Known from isolated drainage systems in arid north-central México. Species included: five species listed by Taylor (1966). Description Durangonella is distinguished from other littoridinine genera by a combination of character states (see below). The shell (Figs. 32, 33; Morrison, 1945, figs. 1-4) is smooth, slender, turriform, with five to eight slowly increasing, rounded whorls; the tentacles have Hydrobia-like cilia- tion (Fig. 35A); females are ovoviviparous; the female gonad is a very small, non-lobed swelling at the end of the oviduct (Go, Fig. 36A); the pallial oviduct is large and bends posteriorly with several loops in one plane (Fig. 36A); the albumen gland is reduced to a mere glandular smear on one of the loops (Ppo, Fig. 36A); the seminal receptacle (Sr) connects with the oviduct via a short sperm duct (Sdu, Fig. 36D); the spermathecal duct is elongate with an opening separate from that FIG. 32. SEM photos of shells of Durangonella coahuilae from Locality 6. The shell on the left is 3.60 mm long; the other one is printed at the same enlargement. CUATRO CIÉNEGAS HYDROBIIDS 79 of the pallial oviduct (Figs. 36A, F, G); the Discussion anterior end of the brood pouch is weakly coiled and muscularized (Figs. 36F, G); the Durangonella is most similar to Mexipyrgus penis has a blunt, ciliated tip, a terminal (see Tables 53-55, Figs. 49, 50), and these eversible papilla, and one (Fig. 35D) or two taxa share the following distinctive features: (Morrison, 1945, fig. 5) simple lobes that lack pallial oviduct with posterior coil in several specialized glands. loops; albumen gland reduced to a mere TX LA U FIG. 33. Camera lucida drawings of shells from six populations of Durangonella coahuilae. The shells are from the following localities: A-C, Locality 6; D-F, Locality 14; G-J, Locality 9; K-M, Locality 38; N-Q, Locality 13; R-U, Locality 65. Shell A is 2.42 mm long, and the others are at the same scale. 80 HERSHLER FIG. 34. SEM photos of shell and radula of Durangonella coahuilae. A. Apical whorls of the shell. Note the lack of microsculpture. B. Part of the radular ribbon. C, D. Isolated central teeth. glandular smear; seminal receptacle con- nected to the oviduct via a short sperm duct; anterior end of the brood pouch weakly mus- cularized and coiled; penis with a blunt, cili- ated tip and a terminal eversible papilla. Durangonella differs from Mexipyrgus in the following features: penis glands (Duran- gonella, absent; Mexipyrgus, mammiform); ciliary tufts on tentacles (Durangonella, Hydrobia-like; Mexipyrgus, absent); female gonad (Durangonella, small and non-lobed; Mexipyrgus, large and lobed); posterior coils of pallial oviduct (Durangonella, in one plane; Mexipyrgus, in several planes); duct of semi- nal receptacle (Durangonella, not coiled; Mexipyrgus, coiled); spermathecal duct (Durangonella, short; Mexipyrgus, long). CUATRO CIÉNEGAS HYDROBIIDS 81 FIG. 35. Head, operculum and penis of Durangonella coahuilae. A. Dorsal aspect of head showing the Hydrobia-like ciliation of the left tentacle and central ciliary bands (Ci) on both tentacles. B. Operculum. C. Tip of the penis showing the ciliated columnar epithelium. D. Dorsal aspect of the penis showing the blunt tip and small penial lobe (Plo). Note the reduced ciliation (compared to C) of this specimen. Ey—eye; Sn—snout; Tn—tentacle; Vd—vas deferens. Durangonella coahuilae Taylor, 1966 Holotype: UMMZ 220159. Type-locality: Locality 9. Distribution: endemic to the Cuatro Ciéne- gas Basin. Habitat: Durangonella coahuilae is found in the basin in a wide variety of aquatic environ- ments that include a playa lake, pools formed where the water table is at ground level, small spring-fed pits without outflows, marshes, and springs and streams of all sizes. Duran- gonella coahuilae was found on mops from 23 of 38 small springheads, but, as with Mexis- tiobia manantiali, it probably does not inhabit subterranean waters as it has eyespots and body pigment, and is very common down- stream. Durangonella coahuilae is most com- mon in soft organic sediments, but was also found in Chara mats and on marl pieces. More so than the other hydrobiids of the basin, D. coahuilae is found in waters whose temperatures fluctuate greatly on a diurnal and seasonal basis. For example, D. coa- huilae was collected from pools with sea- sonally fluctuating levels (Localities 7, 8) that had water temperature ranges of 9.5-35.6°C during 1980. Snails disappeared from the pools only when they went temporarily dry during an arid period. Description The shell (Fig. 33) is not readily distinguish- able from those of the other Durangonella spp. (Morrison, 1945, figs. 1—4): it varies in 82 HERSHLER 1.0 mm Dpe FIG. 36. Female reproductive anatomy of Durangonella coahuilae. A. Ventral aspect of uncoiled female without head and kidney tissue. Note the small gonad (Go), posterior coiling of the pallial oviduct (Bp + Ppo), and very small albumen gland (Ppo). B. Oriented as in A, but with the pallial oviduct cut away to expose the bursa (Bu). The spermathecal duct (Sd), which is elongate, has been cut. C. Oriented as in B, but with the bursa removed to expose the seminal receptacle (Sr), its duct (Dsr) and oviduct (Ov). D. Oriented as in C, but with the seminal receptale (Sr) rotated slightly to expose the short sperm duct (Sdu). E. Frequently-seen kink in the duct of the seminal receptacle. F, G. Variation in the position of the end of the spermathecal duct relative to the opening of the pallial oviduct (Opo). Ast—anterior stomach chamber; Bp—brood pouch; Bu—bursa; Cae—caecum of stomach; Cl—columellar muscle; Coi—coil of oviduct; Dbu—duct of the bursa; Dpe—gonopericardial duct; Dsr—duct of the seminal receptacle; Emc—posterior end of the mantle cavity; Es—esophagus; Go—gonad; In—intestine; Ma—mantle edge; Mus—muscular section of the brood pouch; Oov—opening of the oviduct; Opo—opening of the pallial oviduct; Osd—opening of the spermathecal duct; Osdu—opening of the sperm duct; Ov—oviduct; Ppo—albumen gland; Pst— posterior stomach chamber; Sd—spermathecal duct; Sdu—sperm duct; Sr—seminal receptacle; Sts—style sac. CUATRO CIÉNEGAS HYDROBIIDS 83 size, number and roundness of whorls, and depth of sutures to such an extent that in- dividuals can be found that correspond to the types of all of the nominal species. Durangonella coahuilae does differ from D. seemani (the type of the genus) in terms of number of penial lobes (D. coahuilae, 1; D. seemani, 2), and number of basal cusps on the central tooth of the radula (D. coahuilae, 1; D. seemani, 2). Shell The shell varies in length from 2.30- 5.10 mm and has 5.5-8.5 whorls. The aper- ture is crescent-shaped. The peristome is en- tire in adults and the end of the body whorl frequently pulls away from the penultimate whorl. The umbilicus varies from a slight chink to an open slit. Growth lines are prominent (Fig. 34A). Populations vary not only in size of shell and number of whorls, but also in rela- tive shell width, relative size of the body whorl, whorl roundness, suture depth and an- gle, and degree of sexual dimorphism (Fig. 33). The shell has a smooth apical whorl (Fig. 34A). Shell measurements from nine pop- ulations are given in Table 33. For all pop- ulations, females are clearly larger than males and have more whorls. Nonreproductive Features The measurements of organs and struc- tures for snails from four populations are TABLE 33. Shell measurements (mm) of males and females from nine populations of Durangonella coahuilae. Snails with the dominant maximum whorl number(s) were used. Mean + standard deviation. The localities are the following types of aquatic environments: 6, fluctuating pool semi-connected to stream; 9, playa lake; 13, 38, 8, large streams; 14, 51, 74, 65, 43, small streams. Length of Length of Width of Whorls N Length Width body whorl aperture aperture Locality 6 д 5.5 2256) =O alien 1610.07 1143 ESOO 0.32 ==10106 10:56 0103 ? 6.5 OM 3.54 = 017 | 1:45 = D CAN ESOO 1099-50105 0168 ==0104 Locality 9 3 6.0 97 25/5 == 20:07:50 == 0108 00/85 0/02 A0/6IE 10105 6.5 AAN 035 1:31 =10.05 1620097 0:91 0106. 0163== 0:06 2 6.5 99) 334001289) 144-210 1e O0 Ar | 1.01 =009 OOO Locality 13 3 5) 10 2.35 = 0.14 099+006 1.26 = 0.08 0.74+005 0:49 + 0:03 ? 7.0 1025375230416 313820107 1.65 = 0108 70:972=2.0:.05 0:66 ==10:05 7.5 94-15 0120143 E 002 OO 100 =0:06° 0169/10/03 Locality 14 3 5:5 я 2 = O11 1.07 = 0.08 1.37 + 0.06 0.80 = 0.04 0.58 + 0.04 6.0 й eya aos 1.09/=10:05: 14250108 0:84 = 0:06 0.56 = 0:04 о 7.0 Sy 3:85 == 0:17 1:51 810.06 1.851008 1108 0 05 10.74 == 0108 Locality 38 3 6.0 ¡04228508141 1.18 05 1.52 = 0108’ 091 = 0:04 00610102 g 8.0 8 2483-22 0:18 81561 1097 2:00 =0.10’ 114=0.07 079= 0104 8.5 10 5108==030158 07 199=010’ 1.11 = 0070 79ЕЕТ0:0А Locality 43 3 5.5 © 273=0 120 108 1.52 = 0:09’ 0:90 0105." 10761} = (0:04 Q 6.5 9 4.07 = 0.19 1.64 0х 2052 0110. 119=00” 0831 710104 Locality 51 3 7.0 SS == 0 1.26 = 0.06 1.58 = 0.08 089+0.04 0.63 + 0.02 2 75 SA EOI ESSE 0!05551:8930!09773108=310104 074/0104. Locality 74 3 6.5 TS OGC 0-18 2551514 = 0.073148 ==0:07 0/82 0/05 9058 == 0:04 о 7.0 93:96 0:27 1:47 = 0091: 86ЕЕ 008103 = 0:06 10:72 == 0704 Locality 65 3 6.5 ПРАВЕЕ ONO) a2 5 ЕТО 56 == 10107 0192 ==10 10/7 0/63 10:04 2 7.0 OMS O0 025 e ОЭ 1.060107 0.71005 84 HERSHLER TABLE 34. Dimensions (mm) or counts of non- neural organs and structures of Durangonella coahuilae from Locality 6. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Body | 5.04 = 039. 3:35 = 0412 Gill filament number 26.7 + 2.80 Osphradium MO == 100З Gonad ES 0'383== 0/08 si a3) == 0.09 W 0.08 + 0.02 0.38 + 0.004 Prostate (N = 6) Le 0.68 + 0.09 W 0.34 + 0.04 Penis L 0.55 + 0.05 W 0.21 + 0.03 Pallial oviduct 22020123 W 0.51 + 0.06 Bursa copulatrix L 0.33 + 0.04 (N = 7) W 0.22 + 0.01 Seminal receptacle L 0.11 + 0.01 (body) (N= 6) W 0.09 + 0.01 Seminal receptacle L 0.04 + 0.01 (duct) (N= 8) W 0.04 + 0.01 TABLE 35. Dimensions (mm) or counts of non- neural organs and structures of Durangonella coahuilae from Locality 9. N=5 unless stated otherwise. Mean + standard deviation. L = length, W = width. TABLE 36. Dimensions (mm) or counts of non- neural organs and structures of Durangonella coahuilae from Locality 13. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Body LE 5:50) 10727408 07256 Gill filament number 31.5 =1.76 Osphradium Е 0.21 = 0104 Gonad E 0:36 = 0:05 1.72 = 0718 W 0.10 + 0.02 0.44 + 0.03 Prostate Le 0.74 + 0.10 W 0,82=210:083 Penis E 0.42 + 0.04 W 0.16 + 0.02 Pallial oviduct р 2.39) = 2 W 0.47 + 0.05 Bursa copulatrix ESO SO ETOLOS (N = 9) М 9:17 = 003 Seminal receptacle L 0.13 + 0.01 (body) (N = 10) W 0.10 + 0.02 Seminal receptacle L 0.03 + 0.01 (duct) (N= 6) W 0.04 + 0.01 TABLE 37. Dimensions (mm) or counts of non- neural organs and structures of Durangonella coahuilae from Locality 14. N = 5 unless stated otherwise. Mean + standard deviation. L = length, W = width. Females Males Females Males Body L 5.09 = 0.43 4.46 = 0.30 Body L 6.44 = 0.63 4.76 + 0.34 Gill filamentnumber 26.4 + 1.82 Gill filament number 34.7 + 1.75 Osphradium L 0.20 + 0.03 Osphradium IS 0:21 = 0104 Gonad L 0.26+0.04 1.30 = 0.11 Gonad Е 0:41 = 002} OEOr23 W 0.09 + 0.01 0.43 + 0.03 W 0.09 = 0.02 0.47 + 0.04 Prostate (N = 6) L 0.72 + 0.06 Prostate (N = 7) L 1.00 + 0.09 W 0.34 + 0.04 W 0.40 + 0.04 Penis (N = 6) L 0.56 + 0.04 Penis L 0.46 + 0.04 W 0.21 + 0.03 W 0.21 + 0.04 Pallial oviduct 20021 Pallial oviduct EA SES 025 W1053 210/05 W 0.59 + 0.06 Bursa copulatrix 029/5002 Bursa copulatrix 0:34 = 0104 W 0.18 + 0.02 (N = 6) W 0.19 + 0.02 Seminal receptacle L 0.13 + 0.02 Seminal receptacle L 0.12 + 0.02 (body) (N= 8) W 0.10 + 0.02 (body) (N= 7) W 0.11 + 0.02 Seminal receptacle L 0.05 + 0.01 Seminal receptacle L 0.04 + 0.002 (duct) (N= 8) W 0.04 + 0.01 (duct) W 0.03 + 0.004 given in Tables 34-37. The anatomical draw- ings and radula photographs are from speci- mens from Locality 6. The snout (Fig. 35A) is elongate and the tentacles are relatively short. In addition to the seven to nine hyper- trophied ciliary tufts on the outer edge of the left tentacle, each tentacle has a central cili- ary tract (Ci, Fig. 35A). Small granules are seen around the eyespots and in the neck. The snout may or may not be dusted with melanin. The sides of the head-foot usually have a light melanin dusting, and an un- pigmented strip, extending from the eye to the base of the foot, can be seen. Body pigment can be red or black. The male gonad always has dark melanin on its ventral surface. Pop- CUATRO CIÉNEGAS HYDROBIIDS 85 ulations may have snails devoid of other body arrangements for the four tooth types (for pigment (Locality 6), or with a dark melanin specimens from Locality 6) are given in Table coating (Localities 9, 14), or a spotted pattern 39. The central tooth of the radula has one (Localities 13, 39) on the ventral body sur- (and occasionally a second) pair of basal face. The caecal chamber protrudes posterior cusps that arise from prominent lateral angles to the stomach (Cae, Fig. 36A). The pauci- (Figs. 34C, D). spiral operculum (Fig. 35B) has 3.3 whorls and the nucleus is positioned at 26% of the Female Reproductive Anatomy operculum length. The operculigerous lobe has several melanin streaks. The female gonad (Go, Fig. 36A) occupies only 6% of the body length. Oocytes were Radula frequently seen in the gonad throughout the year. The pallial oviduct occupies 34-44% of Radular statistics for specimens from the the body length, depending on the population. type-locality (Locality 9) and a second locality The length of the posterior bend of the pallial (Locality 6) are given in Table 38. The cusp oviduct is 0.6 mm, and the bend extends to TABLE 38. Radular statistics from individuals of Durangonella coahuilae from two populations. X = mean, S = standard deviation. Measurements in mm. Locality 9 (N = 13) Locality 6 (N = 5) Radular feature X 5 x 5 Length 0.444 0.035 0.380 0.023 Width 0.087 0.007 0.074 0.006 Number of rows 48.6 3:25 51.4 2.07 Number of rows in formative stage 3.0 1.3 4.0 1.2 Width of central tooth (N = 14) 0.020 0.001 TABLE 39. The various cusp arrangements for the four tooth types of Durangonella coahuilae, counted from 5 radulae using SEM, with the percentage of radulae showing that arrangement at least once. Inner Outer Central Lateral marginal marginal anterior cusps basal cusps % cusps % cusps % cusps % = A 20 3—1-3 20 19 40 20 20 — = 20 4-1-3 40 20 40 21 40 > A 60 4-1-4 40 22 40 22 60 eae 80 5-15 40 23 40 23 40 5=1=5 100 24 40 24 40 2-1 a8 20 25 40 25 40 1—5 20 26 20 26 20 2-1 6-1 20 27 20 27 20 86 HERSHLER within 0.13 mm of the end of the mantle cav- ity. The albumen gland (Ppo) is no more than a glandular smear on the posterior-most 0.28 mm of the pallial oviduct. The remainder of the pallial oviduct serves as a brood pouch (Bp). A gonopericardial duct (Dpe) is present (Figs. 36B-D). The sac-like bursa (Bu) is only 12-16% of the length of the pallial oviduct. The seminal receptacle (Sr), dorsal to the bursa, is circular in outline and 44% the length of the bursa. The oviduct (Ov) has a single coil dorso-lateral to the seminal receptacle (Figs. 36B, C) and then receives the short sperm duct (Sdu) from the duct of the seminal receptacle (Dsr). The length of the duct of the seminal receptacle to the opening of the seminal receptacle is short, 0.03-0.05 mm, and then the duct travels (dorsal to and hid- den by the bursa) for 0.41 mm until it joins the duct of the bursa (Dbu, Fig. 36C). This junc- ture occurs 0.20 mm anterior to the end of the mantle cavity. The duct of the seminal receptacle often has a kink in it just after the opening of the sperm duct (Fig. 36E). The opening of the spermathecal duct is 0.08— 0.20 mm posterior to that of the pallial oviduct (Figs. 36k, @). Data for number of shelled embryos brooded for females from six populations are given in Table 40. To these numbers can be added one to two non-shelled embryos that were dissolved in the Clorox. For 39 embryonic shells (Locality 6), shell length averaged 0.384 + 0.158 mm, with an eight- fold range in lengths from 0.079—0.693 mm. The largest embryonic shells have 2.5-2.8 whorls. Male Reproductive Anatomy The lobed male gonad has four to five branches and constitutes 29-43% of the body length. The prostate is 0.16-0.21% the body length, and overlaps the mantle cavity. The anterior vas deferens exits from the anterior tip of the prostate. The single penial lobe (Plo, Fig. 35D) is located at 61% the penis length from the base. Folds are present on the inner curva- ture from the base to just beyond the penial lobe. The blunt tip of the penis has tall col- umnar cells extending back 0.10mm to where the penial folds end. Ciliation of these cells is variable; for the population from Local- ity 6, one of the five penes studied had no cilia, and the other four had a small ciliated patch (Fig. 35D). Other populations, particu- larly those with large-sized males, usually had the entire columnar-celled area ciliated (Fig. 35C). The vas deferens (Vd) coils only slightly in the penis. The penis has both Gl, and Gl» gland types common. Some populations have males with a small pigmented patch near the penis tip. Discussion Among the Durangonella species, only D. coahuilae has received complete anatomical study. The penis and radula of D. seemani have been figured (see above), while the other four species are known only from the shell. Anatomical study of these allopatric species is necessary to resolve their syste- matic status. Durangonella coahuilae had been pre- viously known (Taylor, 1966) only from La- guna Grande (Locality 9), the playa lake that is the terminus of the stream from Laguna Churince, a large spring (Locality 1). It has been suggested that other populations in the basin may represent new Durangonella spe- cies (Holsinger & Minckley, 1971; Taylor, 1966). The author collected undoubted D. coa- TABLE 40. Data for number of shelled embryos brooded by females from six populations of Durangonella coahuilae. The mean shell length (for shells with maximum dominant whorl number) for adult females of each population is also given. Number of young/females 15 1.87 0.64 1-3 Mean shell length (mm) Locality 9 3.40 Locality 6 3.54 Locality 14 3.85 Locality 43 4.07 Locality 13 4.15 Locality 38 5.08 N x SD range 15 2.60 1.40 1-5 13 SA 1.69 1-6 14 8.14 1.41 6-10 18 5.64 1.67 2-8 15 5.60 1255 2-8 CUATRO CIÉNEGAS HYDROBIIDS 87 huilae not only from Laguna Grande, but also in pool areas along the stream feeding it (Localities 7, 8), groundwater-fed pools near the stream (Localities 4, 5, 6) and from a mop placed in a small seep near Laguna Churince (Locality 2). The shells of D. coahuilae from populations from the above localities of the Churince sys- tem (Figs. 33A-C, G-J), which is currently isolated from other waters of the basin, do differ from shells from other populations (Figs. 33D-F, K-U) in that the females have fewer whorls and smaller shells, and the shells are relatively wider with a relatively larger body whorl (Fig. 33; Table 33). The shell differ- ences may be partly allometric, as the Chur- ince shells are small, but in some cases Chur- ince shells that are smaller (in length) than those from other populations are also abso- lutely wider. Despite these differences, the Cuatro Ciénegas Durangonella is not being split into several species because: 1) there are no qualitative anatomical differences among the populations studied; 2) the Churince aquatic environments differ from those from which other populations were sampled; and 3) the above shell differences are not always pro- nounced and the author can not confidently separate out “species” when lots are mixed. Mexipyrgus Taylor, 1966 Type-species: Mexipyrgus carranzae Taylor, 1966. Distribution: endemic to the Cuatro Ciéne- gas Basin. Species included: reduced to monotypy (see below). Description Distinctive features of Mexipyrgus are as follows: 1) the massive pallial oviduct that extends onto the stomach and then bends into a series of loops that coil progressively dorsal to one another (Figs. 42, 43), partially enveloping the bursa and restricting the space for the kidney (Ki) and pericardium (Pe, Fig. 42A); and 2) the greatly coiled duct of the seminal receptacle (Dsr, Figs. 42B, D, E). The shell (Fig. 37) is large (number of whorls, 5.5-7.5; length, 3.03-8.45 mm), usu- ally thickened, and elongate-conic in shape; low spiral welts and noded ribs may or may not be prominent on the last two whorls (Figs. 37, 38); periostracal color bands may or may not be present and (when present) vary from one to thirty distinct bands, to a single wide solid band; the marginal tooth of the radula has numerous cusps (Figs. 39B, C); females are ovoviviparous; the albumen gland (Ppo) is reduced to a glandular smear on the dorsal- most loop of the pallial oviduct (Figs. 42, 43); the bursa is enlarged and elongate (Bu, Figs. 42А, 43); the seminal receptacle is dorsal to the bursa and connects with the oviduct via a short sperm duct (Sdu, Fig. 42C); the spermathecal duct (Sd) is short, muscular- ized, and separated from the bursa by a slight constriction (indicated by arrow in Fig. 42B); the anterior end of brood pouch is weakly muscularized and coiled (Mus, Fig. 41B); the penis (Fig. 44A) has a blunt, ciliated tip, ter- minal papilla, and lobes (outer curvature, one; inner curvature, one or two) bearing mammi- form glands (Mg, Fig. 44A). Mexipyrgus is most similar to Durangonella (see above; Tables 53-55, Figs. 49, 50). Mexipyrgus churinceanus Taylor, 1966 Holotype: UMMZ 220150. Type-locality: Locality 1. Synonymy: M. churinceanus Taylor, 1966 . escobedae Taylor, 1966 . lugoi Taylor, 1966 . carranzae Taylor, 1966 . mojarralis Taylor, 1966 . multilineatus Taylor, 1966 The name Mexipyrgus churinceanus, rather than M. carranzae (type of the genus in Taylor, 1966), is applied to this species be- cause the type population for the former has received the most morphological study. Habitat: Mexipyrgus churinceanus is found almost exclusively in the larger springs and streams of the basin. The species was never found in sieve collections from smaller streams, and only a single specimen was taken (Locality 65) from mops from 38 small springheads. Mexipyrgus churinceanus is restricted to soft sediments which appear (at 50x) to be composed of snail copropel or decaying plant matter. To a lesser extent, specimens were taken from a mixture of soft sediment and coarse travertine sand. Densi- ties of snails were determined using a box core sampler (22.5 cm square bottom) and ranged up to 49,000/m* (Locality 30). The only species found sympatric with M. churin- ceanus in its microhabitat was Durangonella coahuilae. == == 88 HERSHLER Wiss А ОЕ SAN MARCOS As N [a] po 5% I y | | Fe | | = NS 10 | = -@ ee ee 0 1 2 { EA e FIG. 37. Localities from which samples of Mexipyrgus churinceanus were taken for the multivariate analysis, with photos of shells from selected populations. The photos of shells from Localities 1, 5, 37, 50, 99, 76, 78, 79, 80, and 97 are printed at the same enlagement. The shell on the left for Locality 1 is 5.36 mm long. The photos of shells from Localities 10, 20, 30, 48, 73, 88, and 95 are printed at the same enlargement. The photo on the left for Locality 10 is 4.29 mm long. CUATRO CIÉNEGAS HYDROBIIDS 89 FIG. 38. Photos of shells of Mexipyrgus churinceanus. The top row has shells (periostracum removed) from Locality 1, showing sculptural variation. The shell figured on the left is 7.20 mm long. The others in this row are printed at the same enlargement. The bottom row has shells from Locality 90 (eastern lobe of the basin, Fig. 3) showing the thickened sutural periostracal band. The shell on the left is 3.76 mm long and the others in this row are printed to the same enlargement. Shell Shell measurements and other data for adults from 33 populations are given in Table 49. The shells are usually white-colored and Opaque, but in a few populations they are colorless and transparent. The whorls are flattened and the sutures are not very im- pressed. The apical whorl is smooth (Fig. 39A). Sculpture begins at or just before the beginning of the third whorl (Fig. 39A). Spiral cords dominate the third whorl while axial ribs predominate on the fourth whorl. Sculpture on the last two whorls is variable both within (Fig. 38, top row) and among (Fig. 37) populations. After the fourth whorl, noded ribs may be prominent, reduced, or absent; and the spiral sculpture usually consists of two low welts. Cancellate sculpture was rarely seen. In pop- ulations with adult shells having prominent axial sculpture, 12-20 ribs were seen on the penultimate whorl. In some cases, a small number of narrow spiral cords is seen below the suture on the body whorl. Sculpture is generally reduced on the body whorl relative to that of the penultimate whorl. Axial growth 90 HERSHLER FIG. 39. SEM photos of shell and radula of Mexipyr- gus churinceanus. A. Apical view of embryonic shell. Note smooth apex. B. Part of radular ribbon. C. Close-up showing central teeth. lines are usually prominent. Adult females generally have larger and relatively wider shells than males, as well as more prominent sculpture. The aperture is elongate and somewhat pyriform above. The outer lip usually has a pronounced sinuation. The umbilicus is a narrow slit. FIG. 40. Head and operculum of Mexipyrgus chur- inceanus. A. Dorsal aspect of the head. Note the concentration of glandular units (G) and darker color streaks. B. Operculum with the dashed lines showing the attachment area to the operculigerous lobe. Ey—eye; G—glandular unit; Sn—snout; Tn— tentacle. TABLE 41. Dimensions (mm) of non-neural organs and structures of Mexipyrgus churinceanus from Locality 1. N = 5. Mean + standard deviation. L = length, W = width. Females Males Body E 8.33 = 0:36 16:82 210.34 Osphradium L 0.30 + 0.02 Gonad Е 166 0235s 216323705111 W 0.59 + 0.03 0.75 + 0.08 Prostate ¡E 150230812 W 0.90 + 0.08 Penis L 26 2017 W 15005 Pallial oviduct [44.351 == 0:38 W 1.08 + 0.10 Bursa copulatrix Е 1.25 = 0.16 W 0.36 + 0.05 Seminal receptacle L 0.30 + 0.04 (body) W 0.09 + 0.01 Seminal receptacle L 0.13 + 0.01 (duct) W 0.06 + 0.002 CUATRO CIÉNEGAS HYDROBIIDS 91 Pst FIG. 41. General organization of the female reproductive anatomy of Mexipyrgus churinceanus. A. Ventral aspect of uncoiled snail without the head and kidney tissue. Note the posterior bend of the brood pouch (Bp). B. Portion of the anterior end of the mantle cavity showing the slight muscular twist (Mus) of the anterior end of the brood pouch. Ast—anterior stomach chamber; Bp—brood pouch; Bu—bursa; Cae—caecum of stomach; Cl—columellar muscle; Emc—posterior end of the mantle cavity; Es—esophagus; Go—gonad; In—intestine; Ma—mantle edge; Mus—muscular section of brood pouch; Ov—oviduct; Pst—posterior stomach chamber; Sd—spermathecal duct; Sts—style sac. 92 HERSHLER a PS \ Dpe Em— Y “ Osdu FIG. 42. Female reproductive anatomy of Mexipyrgus churinceanus. A. Oriented as in Fig. 41A, but with much of the brood pouch (Bp) cut away to expose the bursa (Bu) and continuation of the pallial oviduct coils dorsal to the part that has been cut away. Note the reduction of the size of the albumen gland (Ppo) and the infringement of the bursa (Bu) and pallial oviduct onto the space occupied by the pericardium (Pe) and kidney (Ki). B. Oriented as in A, but with portion of the brood pouch (Bp) and bursa (Bu) cut away to expose the oviduct (Ov), seminal receptacle (Sr), its duct (Dsr), and the opening of the duct of the seminal receptacle into the dorsal side of the bursa (indicated by small arrow). The large arrow indicates the constriction between the bursa (Bu) and spermathecal duct (Sd). Point C is included for comparison with Fig. 43. C. Oriented as in B, but with part of the oviduct cut to expose the slender sperm duct (Sdu). D, E. Oriented as in C to show variation in the coiling pattern of the duct of the seminal receptacle. Bp—brood pouch; Bu—bursa; Dpe—gonopericardial duct; Dsr—duct of the seminal receptacle; Ef—efferent branchial vessel; Emc— posterior end of the mantle cavity; Es—esophagus; In—intestine; Ki—kidney; Oki—opening of the kidney; Osd—opening of the spermathecal duct; Osdu—opening of the sperm duct; Ov—oviduct; Pe—pericardium; Ppo—albumen gland; Sd—spermathecal duct; Sdu—sperm duct; Sr—seminal receptacle. Nonreproductive Features Measurements of organs and structures from three populations are given in Tables 41—43. The description of external features and anatomy is largely based on study of the type population (from Locality 1). The snout (Fig. 40A) is elongate and the tentacles are thin and short by comparison. The tentacles are without hypertrophied ciliary tufts. Large, milk-white granules (G, Fig. 40A) are con- centrated in the rostrum and neck. A light dusting of melanin may or may not be seen on the rostrum and neck. The paucispiral oper- CUATRO CIÉNEGAS HYDROBIIDS 93 Bp A BES Ze >< = = о = N Е BZ у а} / | N | А \ | \ u = A Bu Ov р JE Sd > Emc 0d р С 8 С WE SS — CET SD Beh ur AQ E 1 © FIG. 43. The nature of the coils о the posterior pallial oviduct as revealed by progressive removal of portions of the coils. Orientation as in Fig. 42A. A. Pallial oviduct cut at point A. B. Part of the bursa (Bu) is cut away to expose the coils on its dorsal side. С. Section А-В’ of the pallial oviduct has been removed. D. Section C-C’ has been removed. The small arrows indicate the direction of movement of eggs and embryos through the pallial oviduct coils. Bp—brood pouch; Bu—bursa; Emc—posterior end of the mantle cavity; Ov—oviduct: Ppo—albumen gland; Sd—spermathecal duct; Osd—opening of the spermathecal duct. TABLE 42. Dimensions (mm) of non-neural organs and structures of Mexipyrgus churinceanus from Locality 30. N = 5. Mean + standard deviation. L = length, W = width. Females Males Body [5460514 5:01 = 0.21 Osphradium L 0.22 + 0.03 Gonad L 0.80+0.06 1.66 + 0.10 W 0.46+0.05 0.54 + 0.07 Prostate E 1.00103 W 0.53 + 0.06 Penis E 1.86 + 0.17 W 0.79 + 0.07 Pallial oviduct EN2S8E ОА W 0.70 + 0.13 Bursa copulatrix L 0.80 + 0.06 W 0.21 + 0.03 Seminal receptacle L 0.15 + 0.02 (body) W 0.07 + 0.02 Seminal receptacle L 0.10 + 0.01 (duct) W 0.06 + 0.01 TABLE 43. Dimensions (mm) of non-neural organs and structures of Mexipyrgus churinceanus from Locality 50. N = 5. Mean + standard deviation. L = length, W = width. Females Males Body ESOS 03390028 Osphradium Е 038 = 0.13 Gonad ES EOS 261076 W 0.73 = 0.08 0.76 + 0.09 Prostate L ИЕ. W 0.98 + 0.06 Penis E 3243023 W 1.52==10:23 Pallial oviduct 5431023 W 1.10 = 0.08 Bursa copulatrix ВР 010 W 0.38 = 0.05 Seminal receptacle | 0.36 + 0.07 (Боау) W 0.16 + 0.02 Seminal receptacle | 0.21 + 0.05 (duct) W 0.08 = 0.004 94 HERSHLER FIG. 44. The penis of Mexipyrgus churinceanus from Locality 1. A. Dorsal aspect of the penis. Note the folds in the penis, the blunt tip with ciliated columnar epithelium, and slender penial lobes (Plo) with mammiform glands (Mg). B. The tip of the penis showing the ciliated cells and pigment patch. Mg—mammiform gland; Plo—penial lobe; Vd—vas deferens. TABLE 44. Radular statistics for females from 6 populations of Mexipyrgus churinceanus. Measurements are in mm. Mean + standard deviation. “r” and “р” refer to the correlation coefficient (and significance level) between that particular radular feature and mean adult female shell length (from Table 33), for the 6 populations. Radular statistics Shell Width of central length N Length Width No. rows tooth (N) Locality 73 4.10 10 0:57 == 10102 0.09 + 0.001 49.4 + 1.49 0.030 + 0.0012 (32) Locality 95 5.29 9 0.65 + 0.03 0.11 + 0.006 53.3 + 2.96 0.033 + 0.0011 (33) Locality 76 6.72 9 0.78 + 0.03 0.13 + 0.005 5451572776 0.038 + 0.0017 (27) Locality 1 7.24 9 0.70 + 0.03 0.12 + 0.007 ESPESA 0.034 + 0.0013 (46) Locality 97 TES 9 0.71 + 0.02 0.12 + 0.005 54.7 + 1.80 0.034 + 0.0015 (29) Locality 50 8.25 9 0177-Е 002 0.14 + 0.005 54.1 = 1.96 0.039 + 0.0018 (22) r 0.860 0.833 0.758 p <.025 <.025 <.05 CUATRO CIENEGAS HYDROBIIDS 95 TABLE 45. The various cusp arrangements for the four tooth types in 18 radulae of Mexipyrgus churin- ceanus, with the percentage of radulae showing that arrangement at least once. Inner Outer Central Lateral marginal marginal anterior Cusps basal cusps % cusps % cusps % cusps % СЕНЕ: 6 4-1-4 58 21 11 24 11 2-2 4-14 17 5-14 58 22 6 25 11 3-2 14 28 5-1-5 58 23 33 26 22 3-3 5-14 33 6-14 25 24 11 27 44 3—2 5-14 44 8-14 8 25 33 28 33 3—3 55 54 22 6-15 42 26 56 29 33 = y 11 7-1-4 8 27 61 30 67 2 qe 6 6-1-6 8 28 56 31 56 5-15 28 29 39 32 33 22 5-1-5 33 30 39 33 39 3-2 9-1-5 44 31 17. 34 28 3-3 6-14 6 32 28 35 11 3-2 6-1-5 17 33 28 36 17 2-2 6=1=5 11 34 1174 ЗИ 17 3-3 1-6 6 35 11 38 11 2-1 6-1-6 6 36 11 2-2 6-1-6 11 3—2 1-6 28 3-3 7-1-5 6 —2 7-1-5 ih 3-3 7-1-6 6 7-1-6 li 3-3 7-1-7 6 3-3 7-1-7 6 3 8-1-5 6 3—3 6-14 6 96 HERSHLER TABLE 46. Common (=40% of radulae studied) central tooth cusp formulae for 6 populations of Mexipyrgus churinceanus. Formulae Locality 73 4-1-4/2-2, 5-1-4/2-2 Locality 95 5-1-4/2-2 Locality 76 5-1-5/2-2, 6-1-5/2-2 Locality 1 5-1-4/3-3, 5-1-5/3-3 Locality 97 5-1-5/3-2 Locality 50 5-1-4/3-3, 5-1-5/3-3 culum (Fig. 40B) has 2.0-2.5 whorls and the nucleus is positioned at 23% of the long axis of the operculum. The body pigmentation consists of thin bands of dark melanin on the dorsal surface, and yellow and white granules on the ventral body surface. The operculiger- ous lobe has several thin red-purple melanin streaks as well as a large central cluster of white granules. The caecal chamber extends posterior to the stomach (Cae, Fig. 41A). Radula The radula is shown in Figs. 39B, C. The central tooth has one to three pairs of basal cusps that arise from the lateral angles. Radular statistics for the type populations of the six nominal species are given in Table 44. Radulae were removed from large females of each population. As seen in Table 44, the length of the radula ribbon, number of rows of teeth, and width of the central tooth are all highly correlated with the average shell lengths of the females for the populations (p < 0.05). The cusp arrangement for the four tooth types for the population from Locality 1 are given in Table 45. Common formulae for the cusp arrangements for the central tooth for the same six populations as above are given in Table 46. Note that only the three populations with large-sized females com- monly have three pairs of basal cusps. Female Reproductive Anatomy The female gonad (Go) consists of four to six lobed branches (Fig. 41A) and is only 15-20% of the body length. The pallial ovi- duct is 44-57% of the body length, and the posterior coils extend to within 0.38 mm of the end of the mantle cavity. When part of the non-reflected portion of the pallial oviduct is dissected away, it is seen that the posterior bend continues to loop dorsally (Fig. 42A). Sections of these loops are progressively cut away in Figs. 43A—D to reveal their complex nature and the way that they partially envelop the bursa (Bu). The bursa is elongate, with a narrowed posterior section, and is 26-34% of the pallial oviduct length. The seminal receptacle (Sr) is appressed to the dorsal side of the bursa near its anterior end (Fig. 42B). The oviduct, after giving off a short gonopericardial duct (Dpe), disappears beneath the bursa and coils once before receiving the narrow sperm duct (Sdu) from the duct of the seminal receptacle (Dsr, Figs. 42B, C). The oviduct then loops back to the ventral side of the bursa to enter the posterior end of the pallial oviduct. The sperm duct is tightly appressed to the duct of the seminal receptacle. After the juncture with the sperm duct, the duct of the seminal recepta- cle coils variably several times before enter- ing the dorsal side of the anterior end of the bursa (Figs. 42B-E). The large bursa and massive coils of the pallial oviduct cover the kidney (Ki) and a small portion of the pericardium (Pe, Fig. 42A). The kidney and pericardium are rela- tively small and flattened compared to those of other hydrobiid snails. Data for number of embryonic shells obtained from adult females from 10 pop- ulations are given in Table 47. Non-shelled embryos were rarely seen in dissected speci- mens. The correlation between shell length and number of brooded young (data from Table 47) is 0.882 and highly significant (p < 0.005). For 100 embryonic shells from females from Locality 1, shell length averaged 0.386 + 0.190 mm, with a five-fold range from 0.119-0.634 mm. The embryos have red pigment on the dorsal body surface. Male Reproductive Anatomy The male gonad has five to six lobed branches, filling most of the digestive gland and comprising 33-39% of the body length. The prostate is 26-34% of the body length, overlaps the mantle cavity, and has the an- terior vas deferens exiting from its anterior tip. The penis is shown in Fig 44A. It has deep folds over most of its length. The single penial lobe on the outer curvature is located at 67% the penis length from the base. One or (more commonly) two lobes are on the inner curva- ture, again beginning at 67% of the penis length from the base (Fig. 45). In one popula- tion (from Locality 76) a few specimens had a CUATRO CIÉNEGAS HYDROBIIDS 97 A 1.0 mm FIG. 45. Variation in the penial lobation of Mexipyrgus churinceanus. A. The most common penis type with two lobes on the inner curvature and one lobe on the outer curvature. B. Presence of a second, “bud-like” penial lobe (with mammiform gland) on the outer curvature, seen only in specimens from Locality 76. С. Rare penis type with one lobe on both the inner and outer curvatures, seen only in specimens from Localities 73, 76, and 99. TABLE 47. Data for number of shelled embryos brooded by females from 10 populations of Mexipyrgus churinceanus. The mean shell length (for shells with the maximum dominant whorl number) for adult females for each population is also given. Shell length (mm) Locality 6 3.93 Locality 16 4.44 Locality 10 SO Locality 18 5.10 Localtiy 21 SAS Locality 11 5.90 Locality 5 5:99 Locality 22 6.04 Locality 1 7.42 Locality 50 8.25 Number of young/female N X SD range 16 3.56 1.67 0-7 15 3.60 3.00 0-12 14 9.21 3.07 6-17 17 8.88 3.95 4-17 15 8.87 3.14 6-18 13 723 2.81 3-14 14 9.14 1.03 8-11 16 4.75 1.29 3-8 15 19.1 3.08 14-24 15 22.1 5.44 14-35 second small bud-like lobe with mammiform gland on the outer curvature (Fig. 45B). The mammiform gland occupies about one- half the length of the penial lobe (Mg, Fig. 44A). While an apocrine gland is circular in shape and has both a very large central lu- men and large terminal opening (see Thomp- son, 1968, fig. 38E), the mammiform gland is more conical in shape and has a very narrow central lumen (surrounded by a muscular layer) and a small, pore-like terminal opening. The penis of Pyrgophorus coronatus also has glands that would be considered mammiform (see Fullington, 1978, fig. 16). The vas deferens (Vd) coils for most of the length of the penis. The tip of the penis has ciliated columnar epithelia extending back to the end of the folds on the inner curvature, 98 HERSHLER TABLE 48. Data matrix for the multi Character 1 5 10 11 18 20 21 22 37 38 43 46 47 1. Мах. no. whorls, 9 7.0 6.0 6.0 6.5 6.0 6.0 6.0 6.5 5.5 6.0 5.5 6.0 6.0 2. Мах. no. whorls, © 7:5 6.5 6.5 7.0 6.5 6.0 6.5 7.0 6.0 7.0 6.0 6.5 6.5 3. Shell length, ¢ 5.76 4.77 4.64 5:22 43514 4.25 4.73 548 (Ат STATS EAN RAZAS 4. Shell length, Y 7.24 15:99 5:10 5.90) 5:10’ 3:93 5.13, 6104 6160428 OOOO CODOS 5. Shell width, Y 3:35 3:04 2.72 3:05 272 2.24 276 316 (3:48 MAN OS SAC 21000219) 6. Length of body whorl 453 392 3147 3:83 3:41 |278 3:57 401 430 OOOO SOS SO 7. Length of aperture 2.18 (2.45 2:19. 2.33 217 1:74 2.28’ 2:36’ 2 NS CR 2 OOM Ao ZA 8. Width of aperture 1.79, 1.597 1:42 1:60’ 1:49) 1:24 1.51 1:74 OA TON ЗОО 9. No. of gill filaments 550 445 523 46.0 458 458 470 484 508 534 480 508 452 10. Body whorl with spiral cord 0 0 0 0 0 0 0 0 0 0 0 0 0 11. Shell with periostracal bands 2 0 0 0 1 0 0 0 1 1 2 0 2 12. Banded shells with thick sutural band 2 2 2 2 2 2 1 2 2 2 2 0 2 13. Penis with 1 lobe on inner curvature 0 0 0 0 0 0 0 0 0 0 0 0 0 14. Penis with 2 lobes on inner curvature 1 1 1 1 1 1 1 1 1 1 1 0 1 15. Rostrum pigmented 0 0 0 0 2 0 0 0 2 0 0 1 1 16. No. of periostracal bands 1 1 1 1 0 0 0 1 1 1 1 0 0 17. Freq. of sculpture score—1 02 0 14 02 03 33 .03 0 0 0 0 0 .02 18. Freq. of sculpture score—2 »122230 .33 .14 13 .48 .46 .02 10 0 .08 04 0 19. Freq. of sculpture score—3 54 50 49 72 83 19 51 62 .60 .66 .92 51 she. 20. Freq. of sculpture score—4 32 50 04 12 01 O 0 .36 .30 34 0 45 .26 and 0.3 mm back on the outer curvature (Fig. 44). A pigmented patch is sometimes seen near the tip of the penis (Fig. 44B.) The penis has Gl, and Gl, glands. Discussion While complete anatomical data are pro- vided for only three populations, specimens from numerous other populations (including those of the types for all nominal species) were dissected as well (see Table 48), yet no qualitative differences in soft-part anatomy were seen. The main difference between pop- ulations is in shell features, especially size, and anatomical features correlated with size, such as number of young brooded by fe- males, radular statistics, and number of gill filaments. The purported differences (shell and anatomy) between nominal species are blurred when numerous populations are stud- ied. For example, one of the diagnostic fea- tures of Mexipyrgus mojarralis (sensu Taylor, 1966), which is considered endemic to Local- ity 73), is a penis with a single lobe on the inner curvature (the usual number is two). Yet in the Mojarral East Laguna (Locality 76), which has a stream connection with Locality 73, individuals assignable to M. multilineatus (sensu Taylor, 1966) may also have this penis type (see Table 49). As there are no morphological criteria by which separate species can be recognized, the six nominal species are reduced to one, Mexipyrgus churinceanus. A detailed analysis of morphological variation of M. churince- anus, and its relation to the species problem, is presented below. Subfamily Unknown Orygoceras Brusina, 1882 Type-species: Orygoceras cornucopiae Brusina, 1882. Distribution: the single living species is re- stricted to two localities in the southwestern deserts of North and Central America (see below). Late Cenozoic fossils are known from eastern Europe (Brusina, 1882) and the northwestern United States (Dall, 1925; and others). Species included: the living species re- mains undescribed. Numerous fossil species have been described. Description The shell is variable in size (width, 2.0— 12.0 mm), but always uncoils after a whorl or so, producing a tube-like shape (Fig. 13H). Axial sculpture may (Brusina, 1882, pl. 11) or may not (Fig. 13H) be present. Orygoceras (?) sp. Distribution: restricted to Roaring Springs, Real County, Texas (Taylor, 1974); and a single spring in the Cuatro Ciénegas Basin (see below). Habitat: restricted to small springheads. Taylor (1974) found one living specimen close to the source of the spring after heavy rains. | collected three live specimens and two empty shells from mops placed at the small springhead at Locality 67. The rarity of live specimens at these localities and the fact that the snail is blind and unpigmented, suggest that its main habitat is subterranean. Description The shell (Fig. 13H) uncoils after 1.3 whorls and produces a tube shape that is very similar in all specimens seen. For two specimens from Cuatro Ciénegas, the shell lengths (par- allel to the coiling axis) are 1.70 and 1.58 mm, and the widths are 2.26 and 2.06 mm, respec- tively. Adult Roaring Springs specimens are about 2.0 mm wide (Taylor, 1974). The apical whorl has pitted microsculpture, while the la- ter whorls have strong growth lines. The oper- culum is paucispiral. The animal is blind and unpigmented. The buccal mass and operculigerous lobe have a red-pink color similar to that of Coahuilix and Paludiscala. The intestine has a loop near its anterior end (Taylor, 1974, fig. 1). While not dissected, the animal is hydrobioid in its ex- ternal appearance. Discussion There are three distinct groups of Ory- goceras species: 1) large (width, to 8 mm), CUATRO CIENEGAS HYDROBIIDS 99 variate analysis (measurements in mm). 48 49 30 50 78 #99 76 73 71 80 79 81 82 83 86 88 93 95 97 96 Poco cs 70 65 65 65 60 65 65 70 65 65 65 60 555 65 65 70 60 :3 70 65 15. u) 20 0 65 65 70 75 65 65 = 70" 65 155 47/0) O E10 m5) 4576" 417 7.03’ 6:23 5.37 6:03. 3:80 5:44 6:24 7.31 5.74 5.96 5.44 4.53 3.14 5.64 4.82 6.58 4.36 465 5.54 444 825 734 653 672 410 590 665 845 693 626 696 634 3.03 653 529 7.51 451 ВИ 273 412143, 4:28) 3191] 3:28 3139. 236 334 337 436 3.84 326 3.48 3:47 4174) 349 276 403 257 3.02 3.51 299 545 485 433 4.47 275 410 4.41 5.37 470 421 459 435 208 432 3.50 487 3.14 1.85 218 188 3.46 2.99 268 278 1.66 259 263 3.41 3.00 261 273 274 133 273 218 3.00 1.93 1.29 ЦЕ50 Esa. 2.20) 2:10 1.80 1.83 We 1.76 1.80 2.31 2.04 1.76 1.92 1.90 0.92 1.86 1.51 2.08 1:35 468 516 448 656 532 580 534 418 53.2 538 646 516 526 626 636 342 462 470 606 484 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 2 2 2 2 2 2 0 2 2 0 1 1 0 0 2 0 0 2 2 1 va 2 2 0 0 0 0 1 1 1 0 0 0 1 2 0 1 2 1 2 Го 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 1 0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 2 0 0 0 1 0 1 2 2 0 0 0 0 0 2 2 0 0 0 2 2 2 2 0 2 2 2 2 2 1 2 0 1 2 1 1 08 .04 0 RD TM ES A NO 03° 14 02" 020 0 DA 23 0 JO A O E ESE 31380002. ES A 56 06 08 0. 0 66 73 18 02 46 0 CO a a А eo) Я о р м 0 м 13 23 02 0 82 0 03 0 0 02 0 0 08 54 0 0 14 0 SG 87 TT sculptured species from Late Cenozoic Bal- kan lake beds; 2) large (to 12 mm), smooth- shelled species from Miocene-Pliocene Idaho lake beds; and 3) the very small (to 2 mm), smooth-shelled living species found in small springheads. The relationships among these three groups remain obscure: the living spe- cies is probably not closely related to the fossil taxa as it differs greatly in shell size and habitat. MORPHOLOGICAL DIFFERENTIATION AMONG POPULATIONS OF MEXIPYRGUS CHURINCEANUS Populations of Mexipyrgus churinceanus show considerable variation of shell features (Fig. 37). The six nominal species of Mexipyr- gus were described by Taylor (1966), all but one from single localities, based on charac- ters involving shell size, shape, sculpture, number and thickness of periostracal bands, and the number of penial lobes. | collected Mexipyrgus from over 40 localities in the basin, and noted patterns of variation in- consistent with the species concepts of Taylor (1966). To analyze these patterns of variation and to assess the similarities and differences among populations, a data base of 20 characters (16 from shell, one from body pig- ment, three from soft parts), including most of those employed in the diagnoses of the nominal species, from 33 populations (OTUs) was subjected to multivariate analysis. These 100 HERSHLER TABLE 49. Characters used in assessing similarities and differences between populations of Mexipyrgus churinceanus. Characters 3-9 represent means (of which 4-9 are for females only). In (0, 1) pairs, O represents absence of a character-state, 1 represents its presence. Characters 17—20 were also scored only from shells of females as those characters exhibit sexual dimorphism. 1. Maximum number of whorls, males; 2. Maximum number of whorls, females; 3. Shell length, males; 4. Shell length, females; 5. Shell width; 6. Length of body whorl; 7. Length of aperture; 8. Width of aperture; 9. Number of gills; 10. Prominent spiral cord on body whorl (Fig. 37, shells of Locality 73) (0, 1); 11. Shell with periostracal bands (0, 0-33% of shells banded; 1, 34-67%; 2, 68-100%); 12. Banded shells with thick sutural band (Fig. 38, bottom row) (0, 0-33% of banded shells with thick sutural band; 1, 34-67%; 2, 68-100%); 13. Penis with one lobe on the inner curvature (Fig. 45c) (0, 1); 14. Penis with two lobes on the inner curvature (Fig. 45a) (0, 1); 15. Rostrum pigmented (0, 0-33% of population; 1, 34-67%; 3, 68-100%); 16. Number of periostracal bands at the shell aperture (0, <8; 1, 8-14; 2, >14); 17-20. Frequency of shells with the following axial sculpture development at the end of the penultimate whorl; 17. Absent (see Fig. 37, Localities 76, 88); 18. Low ribs of low nodes without ribs (Fig. 37, Locality 48); 19. Moderately high, noded ribs (Fig. 37, Locality 1); 20. High, noded ribs (Fig. 37, Localities 30, 94). TABLE 50. List of Mexipyrgus churinceanus pop- ulations (used in the multivariate analysis) accord- ing to drainage systems. Drainage Localities (see Fig. 37) | 125 lla 10, Wd IIb 18,20; 21, 22, Зт, 38, 30; 43:46, 47, 48, 49 Ш 99 IVa 50.78, 71:73:76, 79. 8018182, 83, 86 IVb 88, 93, 95 V 96, 97 characters are listed and explained in Table 49. The entire data set is given in Table 48. The locations of the 33 populations, together with photographs of shells from many of them, are shown in Fig. 37. The various populations are listed (by local- ity number), drainage by drainage, in Table 50. Drainage 1, terminating in a shallow playa lake (Locality 9), is currently isolated from other waters of the basin. Drainage 2a con- sists of the large thermal limnocrene, the Pozo de la Becerra (Locality 10), and its outflow, the Rio Garabatal. Drainage 2b con- sists of the large number of springs in the area known as El Garabatal, to the north of the Pozo de la Becerra, and to the east of the Rio Garabatal. These springs all flow to the north or west and may have joined the Rio Garabatal in the recent past. While some of these springs flow into the waters of Drainage 4a (see below), the downstream portion of these spring outflows are fast flowing over a hard bottom and no Mexipyrgus was found in them during an intensive survey during 1981. Drainage 3 consists of the isolated Anteojo spring complex (Locality 99) which, prior to alterations, may have flowed south to join Drainage 4a. Drainage 4 consists of the large rheocrene, the Rio Mesquites (originating at Locality 50), and nearby springs that flow into it (together constituting Drainage 4a); and the large thermal limnocrene, Laguna Escobedae (Locality 95), and nearby springs that join the Rio Mesquites well downstream (together constituting Drainage 4b). Drainage 5 con- sists of the large thermal limnocrene, Laguna Tio Candido (Locality 97), and a nearby spring, that flow to the south of Drainage 4. Five principal components account for 79.93% of the variation. The first component accounts for 40.41%; the second 17.89% (accumulated 58.30%); the third 9.69% (accumulated 68.00%); the fourth 6.66% (accumulated 74.66%); the fifth 5.27%. Character loading for each component is given in Table 51. Characters are considered highly correlated if their load is greater than 0.60. Characters with values of greater than 0.50, but less than 0.60 were assigned to the principal component for which the characters had the highest value. Characters highly correlated within the first component are number of whorls for males (Character 1, Table 51), measurements from shells of females (Characters 3-8), gill num- ber (9), and number of periostracal bands on the shell (16). This component is one of size: the number of periostracal bands on the shell logically correlates with adult shell length. The second component has highly correlated characters of shell sculpture (10, 17, 19, 20), incidence of a thickened periostracal band (12), and number of penial lobes (13, 14). CUATRO CIÉNEGAS HYDROBIIDS 101 TABLE 51. Factor loading of characters for the five principal components that collectively account for 79.93% of the variation in the multivariate analysis. Principal components Character 1 2 3 4 5 1 0.653 — 0.080 0.061 — 0.004 0.401 2 — 0.406 0.180 0.251 —0.011 — 0.527 3 0.947 — 0.007 — 0.082 0.058 0.001 4 0.969 0.110 0.034 — 0.093 —0.106 5 0.951 0.173 0.076 —0.072 —0.132 6 0.971 0.124 0.031 — 0.077 — 0.158 7 0.958 0.151 0.031 — 0.085 — 0.156 8 0.945 0.181 0.060 — 0.081 —0.170 9 0.865 — 0.058 — 0.006 - 0.143 0.124 10 — 0.277 — 0.543 0.458 —0.152 — 0.006 11 0.147 0.061 0.546 - 0.395 0.512 12 — 0.477 0.597 0.095 - 0.430 0.114 13 — 0.006 — 0.764 0.421 — 0.128 — 0.227 14 0.121 0.591 — 0.593 — 0.011 0.238 15 0.234 0.248 0.177 0.646 — 0.022 16 0.765 — 0.137 — 0.071 0.004 — 0.079 17 0.388 — 0.826 — 0.060 —0.075 0.099 18 —0.158 —0.571 — 0.684 0.018 — 0.000 19 —0.301 0.690 — 01032 — 0.458 —0.261 20 0.016 0.520 0.544 0.511 0.159 These characters do not involve size. The third component includes highly correlated characters of incidence of periostracal band- ing (11), number of penial lobes (14) and shell sculpture (18). The sole characters highly correlated within the fourth and fifth com- ponents are pigmentation of the rostrum (15) and number of whorls for females (2), respec- tively. Ordination diagrams following non-metric three-dimensional scaling are given in Fig. 46 (1st vs. 2nd component), Fig. 47 (1st vs. 3rd component), and Fig. 48 (2nd vs. 3rd com- ponent). The various populations are referred to by locality numbers in the ordination di- agrams. The stress was low (0.0010). The matrix correlation between taxonomic dis- tance and distances in the three dimensional scaling was 0.988. In Fig. 46, the ordination of component 2 versus component 1, the smaller-shelled pop- ulations are in Quadrants | and IV, and the larger-shelled populations are in Quadrants II and Ill. Populations in Quadrants II] and IV are sculptured, have thickened sutural bands and are from four of the drainages, especially Drainages 1 and 2. Populations in Quadrant Il are with large, smooth shells, without a thick- ened sutural band, and sometimes with males having a single lobe on the inner curvature of the penis (from localities 76, 99); and are exclusively from Drainages Ill and Ма. In Quadrant | are populations with small-sized shells from Drainages II and IV. Of the nomi- nal species, the populations in Quadrant Il would be considered by Taylor as Mexipyrgus lugoi or M. multilineatus (type populations from localities 50 and 76, respectively). Quad- rants Ill and IV have the type population of М. churinceanus (1), M. escobedae (95) and M. carranzae (97), that differ largely in shell size and sculptural development. The type popula- tion of M. mojarralis (73) appears as an outlier in Quadrant | as its members have very small shells, with a prominent spiral cord on the body whorl, and males with a single lobe on the inner curvature of the penis. In Fig. 47, the ordination of component 3 versus component 1, the smaller-shelled pop- ulations are again in Quadrants | and IV, and the larger-shelled populations are in Quad- rants II and Ill. Populations with sculptured shells and a high incidence of periostracal banding are in Quadrants Ш and IV, while smoother-shelled populations with a lower in- cidence of periostracal banding are in Quad- rants | and Il. Quadrant Il has only pop- ulations from Drainage 4 while the other three 102 HERSHLER 0.968 0.700 0.432 0.164 O iOrS 2023721 -1.143 -0.886 -0.629-0.372-0.115 0.142 0.398 0.655 0.912 y FIG. 46. Ordination diagram of 1 x 2 principal components, drawn according to non-metric multidimensional scaling. The numbers refer to the 33 populations of Mexipyrgus churinceanus (Fig. 37). The minimum spanning tree and subset solutions have been superimposed. quadrants have populations from several drainages each. The population from Locality 88 appears as an outlier in Quadrant | be- cause its members have a very small, smooth, non-banded shell. In Fig. 48, the ordination of component 3 versus component 2, the influence of size is removed. In this case, the populations group to an even lesser degree by drainage. In Quardant |, smooth-shelled populations from Drainage 2 (10, 20, 21) group with those of Drainage 4. In quadrants Il and Ill, pop- ulations with sculptured shells with thickened sutural bands are found: note that these in- clude populations from Drainages Il, IV and V. The populations from Localities 99 and 73 are outliers in Quadrant IV as their shells are usually banded and the males have a single lobe on the inner curvature of the penis. The multivariate analysis indicates that, of the characters used, there is no consistent pattern of geographic variation among the various populations in relation to drainage. Of the eight subsets formed, three are formed among populations of Drainage 2 (10-21, 48-49, 18—47), three are formed among рор- ulations of Drainage 4 (50-79, 80-83, 78— 81), but two are formed among populations from widely separated drainages: 22-93 (Drainages 2b and 4b) and 30—96 (Drainages 2b and 5); and these two subsets illustrate the problem of recognizing different “species” of CUATRO CIÉNEGAS HYDROBIIDS 103 0.411 88 Il | Sl 20 0.244 83 Tete) \ Be 43 0.077 ee \ \ 95 \ A 18 TAE 3 78 O een ot, Ed | -0.089 SN 37 A | S IV 99 73 20.256 ay -0.422 = > = 1. 143 -0.886 —-0.629-0.372-0.115 0.142 0.398 0.655 0912 : FIG. 47. Ordination diagram of 1 x 3 principal components. Mexipyrgus. The two populations of Drainage 5 are highly sculptured and are referable to M. carranzae (sensu Taylor, 1966). Yet one sees a remarkably similar-shelled population (30) from Drainage 2b, on the other side of the tall Sierra de San Marcos, with a low probability of previous connection between the two drainages (Fig. 37). In the other case, while populations from Drainages | and Il, in gener- al, are moderately sculptured, with thickened sutural bands, and are referable to M. churin- ceanus (sensu Taylor, 1966), populations with similar features are seen in Drainage 4 (71, 78, 81, 86, 93, and Locality 90, Fig. 38, bottom row), supposedly the drainage harbor- ing M. lugoi (sensu Taylor, 1966, smooth- shelled, without a thickened sutural band). While some of these Drainage 4 populations (particularly 71) are located close enough to the low, northern tip of the Sierra de San Marcos (Fig. 37) that they could have been founded by snails from western lobe waters (and hence their M. churinceanus-like fea- tures) given a previously different topography, the other populations (86, 93) are consider- ably to the south and east of the mountain tip and probably could not have been founded in such a fashion. One must conclude, therefore, that despite some geographic differentiation of M. churin- ceanus populations, separate species cannot be distinguished, as similar morphological features involving shell size, sculpture and banding pattern have apparently been in- dependently acquired in separated pop- ulations. The pattern seen in the ordination 104 HERSHLER 0.411 88 0.244 010177 0/50/89 =072/5/6 46 -0.422 OL Oe OOo 2a O От 2 99 IV 0.266 0.426 0.585 0.745 10.904 FIG. 48. Ordination diagram of 2 x 3 principal components. diagrams, with many populations clustered together and a few outliers with unusual fea- tures, is one of a single variable species. While the six nominal species must be re- duced to one, there may be three races or subspecies, corresponding to M. churin- ceanus, M. lugoi and M. carranzae (sensu Taylor, 1966), as populations assignable to each are concentrated in different portions of the basin drainage; Drainages | and Il, Ill and IV, and V, respectively. The shell features of Mexipyrgus churin- ceanus may be somewhat plastic phenotypi- cally and dependent on environmental fac- tors. While populations located close to one another may be quite similar (such as subset 48-49), in other cases populations from close-by springs (even those with an aquatic connection) are quite dissimilar on shell characters, especially when the springs differ in size, temperature, or substrate type. For example, the populations from Localities 80 and 79 are from springs (one small-sized and cold, the other large and warm) separated by only 100 m of stream, but differ greatly in shell size and frequency of a thickened sutural band (Table 49, Fig. 37). Populations from Localities 76 and 73 are also from close-by springs (one large and warm, the other small and very warm) with a short stream connec- tion, but differ enough to have been placed in different species (Fig. 37; Taylor, 1966). How- ever, when | collected and studied snails from a transect along the stream connection be- tween the springs, intergradation of the two forms was apparent (Hershler, in prepara- tion). In several cases | visited springs that had been altered by dredging. In these cases CUATRO CIÉNEGAS HYDROBIIDS 105 the living specimens differed greatly in size and sculpture from sub-fossil specimens found in the dredged sediment alongside the spring, suggesting that (as the dredging was recent) the habitat change quickly resulted in a shell change. Shell features may also depend on whether the population lives in a lentic (spring) versus lotic (stream) habitat: note that the spring- head and downstream populations in two in- stances (from Localities 1 and 5, 10 and 11) differ greatly in size, sculpture, and banding fequency. Banding frequency corrrelates with substrate color: populations from springs with light-colored sediments are generally un- banded while those from springs with darker sediments are usually banded. A banded shell may appear cryptic in dark sediment, making it more difficult to be seen by pre- daceous cichlid fish as they disturb and search the sediment when feeding. RELATIONSHIPS AMONG THE CUATRO CIENEGAS HYDROBIIDS The results of anatomical study of the Cua- tro Ciénegas hydrobiids show that all of the taxa belong to the Nymphophilinae or Littor- idininae (see Table 2) are widely distributed subfamilies. Thus while there are endemic genera, there are no subfamilies of hydrobiids endemic to the Cuatro Ciénegas Basin. While the relationships among the various hydrobiid taxa of Cuatro Ciénegas are discussed be- low, and shed some light onto the origin of the endemic taxa, the discussion is necessarily limited as the hydrobiids of the southwestern United States, Mexico, and Central America are almost entirely unknown in terms of soft part anatomy, with many taxa still unde- scribed. The two nymphophiline genera of the basin, Mexistiobia and endemic Nymphophi- lus, differ in at least 10 morphological features (Table 52). However, many of these dif- ferences may be simple correlates of the great size difference between snails of these taxa, and the two genera may, in fact, be closely related. A comparison of the six littoridinine genera of the basin, involving 36 characters, is given in Table 53. Of these characters, six (17%) are from the operculum or shell, seven (19%) are from nonreproductive aspects of an- atomy, eight (22%) are from the male repro- ductive anatomy, and fifteen (42%) are from the female reproductive anatomy. A matrix of percent difference between these taxa is given in Table 54 and was constructed by simply counting differences between taxa pairs and dividing by the total number of characters shared by the two taxa. In in- stances where for a given character two taxa both share a character state and have a dif- ferent character state (e.g., 1, 2 versus 1, 3), the difference is scored as 0.5. A phenogram, based on simple averaging of differences be- tween taxa, is given in Fig. 49. The phenogram (Fig. 49) indicates that there are three groups of littoridinines in the basin, each constituting a pair of genera. Two of the groups consist of very similar genera (=29% difference), the groups themselves linking at 46% difference. The taxa of the third group, Paludiscala and Coahuilix, link at 41% TABLE 52. List of 10 morphological differences between Mexistiobia and Nymphophilus. Mexistiobia . Operculum with 3.5 whorls . Osphradium short . Central tooth or radula with 1 pair of basal cusps . Male gonad overlaps stomach . Male gonad a single lobed mass . Penis with single glandular ridge . Penial lobe slender, with single fold . Bolster and ventral channel poorly developed . Bursa small (21% of pallial oviduct length), dorsal to pallial oviduct, with a short duct 10. Opening of common genital aperture at end of pallial oviduct CO D — OI BR Nymphophilus 5.5—6.0 whorls Osphradium elongate 3 pairs Male gonad posterior to stomach Male gonad bush-like 1-3 glandular ridges Penial lobe stout, with many folds Bolster and ventral channel well-developed Bursa large (32% of pallial oviduct length), posterior to pallial oviduct, with a long duct Opening of common genital aperture lateral to pal- lial oviduct 106 HERSHLER TABLE 53. Comparison of the six littoridinine genera of Cuatro Ciénegas involving 36 characters. Pal. = Paludiscala, Coah. = Coahuilix, Cochl. = Cochliopina, Mexith. = Mexithauma, Dur. = Durangonella, Mexip. = Mexipyrgus. Character Pal. Coah. Cochl. Mexith. Dur. Mexip. Shell 1. Shape: 3 0 0,1 1 3 2 a) planispiral (0) b) trochoid-globose (1) c) ovate-conic (2) d) turriform (3) 2. Sculpture: 0,2 2 1 1 2 0,1 a) ribs (0) b) spiral cords (1) c) absent (2) 3. Apical whorl microsculpture: 0 0 0 1 1 1 a) pitted (0) b) absent (1) 4. Shell with periostracal bands (0,1) 0 1 1 0 1 5. Shell aperture flared (0, 1) 0 1 0 0 0 0 External Features 6. Tentacle ciliation: 1 0 2 2 1 0 a) absent (0) b) Hydrobia-like (1) c) Spurwinkia-like (2) 7. Snail blind, unpigmented (0,1) 1 1 0 0 0 0 8. Mantle edge papillate (0, 1) 0 0 0 1 0 0 9. Position of operculum nucleus along long axis: 1 1 1 1 0 0 a) <0.30 (0) b) =0.30 (1) Digestive System 10. Digestive gland tubercles as low swellings (0, 1) 1 1 0 0 0 0 11. Intestine with anterior loop (0, 1) 0 1 0 0 0 0 12. Caecal chamber extends posterior to stomach (0, 1) 0 0 1 1 1 1 13. Origin of basal cusps of central tooth of radula: 1 0 1 1 1 1 a) from face of tooth (0) b) from lateral angles (1) Male Reproductive Anatomy 14. Male gonad morphology: 2 2 0 1 0 0 a) simple lobes (0) b) bush-like (1) c) non-lobed mass (2) 15. Seminal vesicle coils on stomach (0, 1) 0 1 0 0 0 0 16. Prostate posterior to end of mantle cavity (0, 1) 0 1 0 0 0 0 17. Penis with slender penial filament (0, 1) 0 1 1 1 0 0 18. Penial lobe(s): 1 1 0 0 2 2 a) absent (0) b) bulb-like (1) c) simple (2) 19. Penis ciliated (0, 1) 0 0 0 0 1 1 20. Penis with terminal eversible papilla (0, 1) 1 0 0 0 1 1 21. Penis with specialized gland(s) (0, 1) 1 1 0 0 1 1 CUATRO CIÉNEGAS HYDROBIIDS 107 TABLE 53 (Continued) Character Pal. Coah. Cochl. Mexith. Dur. Mexip. Female Reproductive Anatomy 22. Female gonad overlaps stomach (0, 1) 0 1 0 0 0 0 23. Female gonad: a) relatively large, lobed (0) b) relatively large, nonlobed (1) с) relatively small, a mere thickening of oviduct (2) 24. Reproductive mode: 0 0 1 1 1 1 a) oviparity (0) b) ovoviviparity (1) 25. Length of pallial oviduct/length of body: 0 0 1 1 1 1 a) <0.30 b) =0.30 26. Length of bursa/length of pallial oviduct: 2 1 0 0 0 1 a) <0.20 (0) b) =0.20 <0.40 (1) c) =0.40 (2) 27. Albumen gland: 0 0 1 1 2 2 a) normal size (0) b) reduced in size (1) c) very reduced in size (2) 28. Posterior pallial oviduct: 0 0 1 1 1 2 a) with bend (0) b) with simple bend (1) c) with complex bend in more than one ER = о № № о plane (2) 29. Normal seminal receptacle present (0, 1) 0 0 1 1 1 1 30. Secondary seminal receptacle present (0, 1) 1 0 0 0 0 0 31. Oviduct without coil (0, 1) 1 1 0 0 0 0 32. Spermathecal duct: 0 0 1 0 0 1 а) long (0) b) short (1) 33. Of ovoviviparous taxa, anterior end of pal- lial oviduct: = — 1 1 0 0 a) with slight muscular coil (0) b) with well-developed muscular coil (1) 34. Of taxa with a normal seminal receptacle, the seminal receptacle opens into: = — 0 0 1 1 а) the oviduct directly (0) b) the oviduct via a sperm duct (1) 35. Of taxa with a normal seminal receptacle, the length of the seminal receptacle/length of bursa: = = 1 2 1 0 a) <0.30 (0) b) =0.30 <0.50 (1) c) =0.50 (2) 36. Of taxa with a long spermathecal duct, the openings of the spermathecal duct and pallial oviduct are: 1 0 — 2 0 = a) separate (0) b) joined (1) с) separate, but with an open channel between them (2) 108 HERSHLER TABLE 54. A matrix of percent difference between pairs of the 6 littoridinine genera from Cuatro Ciénegas (based on data from Table 53). Paludiscala Coahuilix Cochliopina Mexithauma Durangonella Mexipyrgus Paludiscala = 41 69 73 59 67 Coahuilix = Ths 82 82 84 Cochliopina = 19 43 44 Mexithauma — 44 53 Durangonella = 29 Mexipyrgus = Mexithauma 8 SU .6 .5 .4 .3 -2 ail FIG. 49. Phenogram based on distance values derived from Tables 53 and 54. Hydrobia Spurwinkia Pal. Coah. Mexip. Mexit. FIG. 50. Cladogram based on character-states listed in Table 55. Pal. = Paludiscala, Coah. = Coahuilix, Mexip. = Mexipyrgus-Durangonella group, and Mexit. = Mexithauma-Cochliopina group. Paludiscala and Coahuilix share character- states 5-7 (hence the circle enclosing both taxa). difference. This group links with the other four genera at 75% difference. An hypothesis of the phyletic relationships among Hydrobia (Hydrobiinae), the six Cua- tro Ciénegas littoridinine genera, and Spur- winkia, the only other North American littori- dinine known from entire soft-part anatomy (see Davis et al., 1982), is shown in Fig. 50. The numbers indicate presumed derived character states, listed in Table 55, used to define clades. Several of these character states, relating to the female reproductive system, are illustrated in Fig. 51. “A” represents an hypothetical ancestral hydro- biid, with the female reproductive anatomy of Hydrobia. In Hydrobia (Fig. 51A), sperm pass along the ciliated ventral channel of the pallial ovi- duct, which connects with the lumen of the pallial oviduct via a narrow slit. The ventral channel bifurcates at the posterior end of the mantle cavity; one branch leads to the bursa and the other is the anterior end of the ovi- duct. This groundplan of the female reproduc- tive system is found in all Hydrobiinae, Nym- phophilinae and Lithoglyphinae (Davis et al., 1982). Davis et al. (1982) suggest that Spurwinkia (Fig. 51B) evolved from an ancestor with an Hydrobia-like female reproductive system by having the ventral channel close off and partly separate from the pallial oviduct (character state 1). As the eggs need to reach the albu- men gland, a connection to the albumen gland from the oviduct formed (character state 2); and the duct from that point to the duct of the bursa became an extension of the duct of the seminal receptacle (character state 3). The ventral channel is only partly separated from the pallial oviduct, suggesting that Spurwinkia is only a step removed from a snail with an Hydrobia-like female reproduc- tive system. Spurwinkia has the additional derived feature of holding egg capsule chains in the anterior end of the capsule gland (character state 4). While in Paludiscala the ventral channel is still only partly separated from the pallial ovi- duct, in Coahuilix the channel has separated entirely and constitutes a spermathecal duct (character state 8). Coahuilix and Paludiscala CUATRO CIÉNEGAS HYDROBIIDS 109 TABLE 55. Presumedly derived character states serving to define clades as shown in Fig. 50. © № — . Partial separation of the ventral channel from the pallial oviduct. . The oviduct enters the posterior portion of the albumen gland. . The duct of the seminal receptacle elongates, connecting to the duct of the bursa. A sperm duct connects the oviduct with the duct of the seminal receptacle. Loss of eyes and pigment. . Loss of oviduct coils. Loss of seminal receptacle. ODNAOA . Egg capsule chains are retained within the anterior end of the capsule gland. . Complete separation of the ventral channel from the pallial oviduct, forming a spermathecal duct. . Assumption of ovoviviparity: the anterior pallial oviduct is enlarged and modified into a thin-walled brood pouch, with the albumen gland reduced in size, and with the development of a muscular sphincter at the anterior end of the brood pouch. 10. Shift of ducting: the duct of the seminal receptacle shortens to open directly into the oviduct, and a short duct forms between the bursa (or duct of the bursa) and oviduct. Apo Ppo Bu D FIG. 51. Schematic drawings of the female reproductive morphologies of Hydrobia (A), Spurwinkia (B), Coahuilix (С), and Mexithauma (D). Lettering as on earlier figures. are considered closely related and special- ized; the evolution of these taxa has involved loss or reduction or morphological features, either associated with the very small size of the snails (character states 6, 7) or associ- ated with the unusual groundwater habitat shared by these taxa (character state 5). Hy- drobioid snails with a presumed groundwater habitat are known from many parts of the world and are frequently small-sized, blind, and without body pigment (Boeters, 1979; Climo, 1974; Ponder, 1966). The bursa copu- latrix complex of Paludiscala and Coahuilix (Fig. 51C) could have been derived from that of Spurwinkia by loss of the seminal recepta- cle and elongation of the sperm duct so as to 110 HERSHLER connect with the oviduct at the opening into the albumen gland. The remaining four littoridinines of Cuatro Cienegas all have the female reproductive anatomy modified as the result of the assumption of ovoviviparity (character state 9). In the Mexipyrgus-Durangonella group, the organization of the bursa copulatrix com- plex is basically as in Spurwinkia: a short sperm duct connects the duct of the seminal receptacle with the oviduct. In the Mexithauma-Cochliopina group duct- ing is different: the seminal receptacle opens directly into the oviduct and a short duct con- nects the bursa (Cochliopina) or duct of the bursa (Mexithauma, Fig. 51D) and the ovi- duct. These taxa share a puzzling mosaic of character states: while the opening of the seminal receptacle directly into the oviduct is also seen in Hydrobia (but not Spurwinkia; see above) and is arguably a primitive character state suggesting a close relation- ship among these taxa, the spermathecal duct is entirely separate from the pallial ovi- duct, a derived character state indicating that Mexithauma and Cochliopina evolved from a Spurwinkia-like intermediate. It seems prob- able that the opening of the seminal recepta- cle directly into the oviduct, within the littoridi- nine groundplan, is a derived condition and convergent with that of Hydrobia. Note that, among littoridinines, this condition has only been found in Mexithauma and Cochliopina, whereas in other littoridinines from various parts of the world for which the anatomy has been studied, the seminal receptacle con- nects with the oviduct via a sperm duct (Davis et al., 1982). The latter, more widespread condition is probably the primitive one. The condition seen in Mexithauma and Cochlio- pina could have been derived from that of Spurwinkia by a shortening of the duct of the seminal receptacle, so as to open into the oviduct, and the development of a duct from the oviduct to the bursa or duct of the bursa (character state 10). Character states from the female reproduc- tive system have been emphasized in the phyletic analysis because of the complexity of the system (relative to other organ systems in hydrobioid snails), with several organs and ducts functionally organized to receive and hold sperm, fertilize eggs, and provide pas- sage for the eggs or embryos through the pallial oviduct. Precise convergences should be unlikely in such a system. Yet the phyletic analysis indicates that, with the present data base, character states from the female repro- ductive system cannot always be confidently scored as primitive or derived; the hypothe- sized phylogeny is therefore tentative. An- atomical data are needed for more taxa to refine the phyletic analysis. If, for instance, littoridinine taxa are found that have the semi- nal receptacle opening into the oviduct as in Mexithauma and Cochliopina, but also with the ventral channel only partly separated from the pallial oviduct (as in Spurwinkia) then a rearrangement is suggested so that Mex- ithauma and Cochliopina are placed closer to “A” than Spurwinkia. Convergence is not entirely unknown among features of the female reproductive system of hydrobioid snails: the Pomatiopsi- dae and Littoridininae both have spermathe- cal ducts, but of separate ontogenetic origins. The duct of the Pomatiopsidae forms as a bud from the bursa (Davis et al., 1976) while that of the Littoridininae presumably forms as the ventral channel closes off and separates from the pallial oviduct (Davis et al., 1982). The organization of the bursa copulatrix complex of Mexithauma (Fig. 30B) is, in fact, virtually identical to that of Pomatiopsis (Davis, 1967, pl. 8). This must be because of convergence, as Mexithauma lacks the following diagnostic pomatiopsine features: eyes in pronounced swellings at the bases of the tentacles, pres- ence of a pedal crease and suprapedal fold, and basal cusps arising from the face of the central tooth of the radula (Davis, 1979). Character states involved with brooding young are unreliable, in themselves, for de- fining clades as the evolution of ovoviviparity involves simple, functionally correlated morphological changes (see below), and has occurred iteratively among many groups of gastropods (Fretter & Graham, 1962). Among hydrobioids, ovoviviparity has evolved at least twice: in the littoridinines and in Potamopyr- gus which have an Hydrobia-like female re- productive system (Fretter & Graham, 1962, fig. 186H). Other characters whose character states could be scored as primitive or derived with even less confidence, and hence were excluded from this analysis, include penial form and gland type, length of the spermathe- cal duct, and tentacle ciliation pattern. The data (Summarized in the phenogram and cladogram) suggest a polyphyletic origin for the endemic hydrobiids of Cuatro Ciéne- gas. Of the five endemic genera, each of three (Nymphophilus, Mexithauma, Mexipyr- gus) is more similar to a non-endemic genus CUATRO CIÉNEGAS HYDROBIIDS 111 found in the basin than to the other endemic taxa, suggesting that the endemic hydrobiid fauna may be comprised of at least four sepa- rate lineages. ORIGIN OF THE ENDEMIC SNAILS OF CUATRO CIENEGAS The idea that the endemic hydrobiids of Cuatro Ciénegas are an ancient fauna, with most taxa not closely related to other hydro- biids of the region, is prevalent in the literature (Minckley, 1969, 1977; Taylor, 1966). This hypothesis is based on the supposed endem- ism of subfamilies of snails that have diverged from a common ancestor, necessitating an origin dating back to the Tertiary period based on the usual slow rate of freshwater snail evolution (Taylor, 1966). Does information on the geological history of the region, coupled with the (above) results of systematic study of the snails, support this hypothesis? It is known from study of fossil plants from packrat middens that the Chihuahuan Desert is of recent origin, the change from wood- lands to desert having occurred during the past 12,000 years (Van Devender, 1976, 1977; Wells, 1977). There is faunal and structural evidence that a number of internal drainages of the Chihuahuan Desert once integrated with the Rio Grande system, and have since been isolated, perhaps due to decreased discharges associated with recent aridity (Morafka, 1977; Smith, 1981). The fauna of these now-isolated drainages is characterized by relictualism and local endemism (Miller, 1977; Milstead, 1960; Morafka, 1977). A good summary of the geological history of the Cuatro Ciénegas area and its effects on isolation of the basin drainage is given by Minckley (1969). While it is known that the Sierra Madre Orientale chain began to form in the early Tertiary, the age of the Cuatro Ciénegas Valley is unknown. It is known, from a study of pollen from cores taken from the valley, that aquatic environments have ex- isted in the valley for at least 40,000 years (Meyer, 1972, 1973). The basin waters have had past connections with the Rio Grande drainage via the Rio Salado de Nadadores, which heads just east of the valley. The fish fauna of Cuatro Ciénegas has numerous Rio Grande elements (Minckley, 1977). Of the snails, Cochliopina riograndensis, a species with a Rio Grande distribution, is found in the basin; and Nymphophilus, one of the endemic genera, has been found as a fossil from Pleistocene-Holocene deposits alongside the Rio Monclova (a Rio Grande tributary), 70 km east of Cuatro Ciénegas (J. Landye, personal communication, 1981). Waters from the southern Rio Nazas-Aguanaval system may have also connected with the Cuatro Ciéne- gas drainage in the past (Conant, 1977; Minckley, 1969). Two of the non-endemic genera of the basin, Durangonella and Mexis- tiobia, are known from the Rio Nazas- Aguanaval drainage, but not the Rio Grande. The above evidence, suggesting that the waters of the basin have had a recent con- nection to outside drainage, coupled with the discovery of a lower level of endemism than once thought, with no endemic subfamilies and three of five endemic genera closely re- sembling non-endemic taxa found in the basin, suggests that the endemic snails may be of a more recent and local origin than previously thought. The Rio Grande drainage of Texas and Mexico, and other waters of southwest Texas, do harbor littoridinine and nymphophiline taxa. Genera from this area assigned to the Littoridininae, on the basis of a penis with stalked, specialized glands (not glandular ridges), include Texadina (penis figured in Andrews, 1977: 82-83), Littoridinops (An- drews, 1977: 84), and Pyrgophorus (Fulling- ton, 1978, fig. 16). The distinctive penis type shared by Mexistiobia and Nymphophilus, with an elongate penial filament and small number of glandular ridges, is seen in Fonte- licella (penis discussed in Gregg & Taylor, 1965; figured in Russell, 1971), recently found in a Rio Grande tributary not far from Cuatro Ciénegas (Lytle, 1972). Anatomical study of the above taxa is needed to help determine the origin of the endemic hydro- biids of Cuatro Ciénegas. | predict that two of the endemic genera, Paludiscala and Coahuilix, will eventually be found living in waters outside of the basin (see below for discussion of possible Coahuilix from Texas). Mexico is undercol- lected for fresh-water snails and most work- ers have not employed the methods neces- sary to collect tiny snails from groundwater outlets. This prediction is supported by the fact that of the three blind, unpigmented crustacean genera originally described from (and considered endemic to) groundwater outlets in Cuatro Ciénegas, two, Mexiweck- elia and Mexistenasellus, were later dis- 112 HERSHLER covered in cave waters in more southerly parts of Mexico (Argano, 1974; Holsinger, 1973; Magniez, 1972). The discovery of Ory- goceras (?) sp., previously known from a single spring in southwest Texas, in Cuatro Ciénegas further attests to the potential for a widespread distribution of groundwater- dwelling taxa. Of the other three endemic genera, one, Nymphophilus, may be relict as it has been found fossilized outside the basin, and the other two could have evolved in the basin from the non-endemic taxa that they closely resemble. It is now known that the rate of evolution of fresh-water snails can be quite rapid (Davis, 1979, 1981; Stanley, 1979) and thus one need not invoke an ancient origin for these endemic genera. The amount of differentiation seen among the snail genera of the basin is slight (only 12 species are known); no genus has more than two species, and only one genus has sym- patric congeners. Such minimal differentiation does not support the idea of an ancient snail fauna present in the basin for tens of millions of years, although perhaps in even such a long time span one would not expect great differentiation in so small a basin with appar- ently plastic drainage patterns. Members of the endemic Cuatro Ciénegas snail fauna have been linked with those of two other faunas and these possibilities are now discussed. Numerous peculiar-shelled snail taxa have been described from the Pliocene Pebas and other formations from the Upper Amazon Valley in Peru (Boettger, 1878; Con- rad, 1871, 1874a, b; Gabb, 1869; de Greve, 1938; Pilsbry, 1944). Some of these taxa have not only been placed in the Hydrobiidae, but have also been considered closely related to some of the Cuatro Ciénegas endemic taxa (Kadolsky, 1980; Parodiz, 1969). These taxa include Tropidebora (Pilsbry, 1944), similar to Nymphophilus; and Eubora Kadolsky, 1980 (= Ebora Conrad, 1871), similar to Mexi- thauma. However, examination of types of the Pebas taxa shows that they cannot be hydro- biids as they have a siphonal notch (Fig. 52), or an otherwise peculiarly-angled aperture unknown in living hydrobiids. Similarities be- tween Pebas and Cuatro Ciénegas taxa must therefore be due to convergence. The Edwards Aquifer in southwest Texas harbors one of the world’s most diverse sub- terranean aquatic faunas (Longley, 1981). In- cluded in this fauna are a number of tiny, blind, unpigmented snails (Karnei, 1978) in- cluding species with lamelliform costae on the shell that were assigned to Paludiscala (Ful- lington, 1978, fig. 17). Alcohol specimens of these species, stored at Southwest Texas State University, were studied during Jan- uary, 1982. These snails differ from Paludis- cala in at least seven features, with the female reproductive anatomy still unstudied: 1) the shell is much smaller (length 1.1 mm) and has only 3.3-3.5 whorls; 2) strong spiral lines, not seen in Paludiscala, run between the costae (Fullington, 1978, fig. 17); 3) the aperture is greatly flared all around; 4) the operculum has a slight internal swelling or peg, not known for any other North American hydrobioid; 5) the intestine has an anterior loop; 6) the penis has neither lobes nor spe- cialized glands; 7) there is no ctenidium. These differences rule out there being Palu- discala. The shell of these snails is much more similar to that of Lanzaia Brusina, 1906 (figured in Bole, 1970, fig. 6), a European hydrobiine. While Paludiscala may not exist in the wat- ers of the Edwards Aquifer, it is possible that Coahuilix does. The type of Horatia micra (Pilsory & Ferriss, 1906), described from stream drift (probably washed out of a spring) of the Guadalupe River, New Braunfels, Texas, is remarkably similar to Coahuilix hubbsi, with similar slight apertural flaring. Horatia micra is also reported from the arte- sian well at San Marcos, Texas, and a sub- terranean stream in Manitou Cave, near Fort Payne, Alabama (Hubricht, 1940). Un- described Horatia are reported from Sala- mander Cave, Travis County, Texas (Reddell, 1965) and Roaring Springs, Real County, Texas (Taylor, 1974). EVOLUTION OF OVOVIVIPARITY IN HYDROBIOID SNAILS It has been suggested that the initial step in the evolution of ovoviviparity in hydrobioid snails involves a simple morphologic change: separation of the ventral channel from the pallial oviduct to form a spermathecal duct, thus keeping separate the functions of receiv- ing sperm and storing embryos (Davis et al., 1982). The need for such a prerequisite is suggested by there being numerous taxa worldwide which lack this separation, only one (Potamopyrgus) is known to brood young. Spurwinkia and Littoridinops are ap- parent intermediates in the evolution of CUATRO CIÉNEGAS HYDROBIIDS 113 FIG. 52. Holotype (NYSM 9193) of Eubora bella (Conrad, 1871). The shell is 7.58 mm long. Note the siphonal notch in the basal view (lower magnification). ovoviviparity, as they have spermathecal ducts (although that of the former is still con- nected anteriorly to the pallial oviduct) and, while they do not brood young, they hold egg capsules in the anterior end of the pallial oviduct before depositing them on the sub- strate (Davis et al., 1982). The littoridinines of Cuatro Ciénegas include taxa that have fur- ther modification of the female reproductive system associated with the assumption of ovoviviparity. Of the six littoridinine taxa of Cuatro Ciéne- gas, two, Coahuilix and Paludiscala, are egg layers, but do not hold egg capsules in the anterior pallial oviduct. The other four share features associated with the evolution of ovoviviparity from a Spurwinkia-like condition. All have an enlarged pallial oviduct (to 58% of the body length; for Spurwinkia it is 40%, Davis et al., 1982) that often overlaps part of the stomach, and that bends posteriorly to varying degrees (Spurwinkia has a slight pos- terior bend of the pallial oviduct). Mexipyrgus represents the pinnacle of this trend as the length of the brood pouch is vastly increased by a series of coils that are progressively dorsal to one another. Whereas egg laying hydrobioids have a thick walled, glandular pallial oviduct with a slit-like central lumen, the ovoviviparous littoridinines all show at least some reduction in the size of the albu- men gland, with the other (much larger) sec- tion of the pallial oviduct modified into a thin- walled, non-glandular brood pouch. These features collectively increase the amount of space available in the pallial oviduct for hold- ing embryos. In addition, all of these taxa have the anterior end of the brood pouch muscularized, giving it the ability to stretch as large sized embryos are released, and per- haps the ability to control timing of embryo release. Ovoviviparous snails may not need great egg production at any one time and this may explain why female gonads of three of 114 HERSHLER the four ovoviviparous taxa are unusually small, filling only a portion of the length of the digestive gland at all times of the year. The ovoviviparous Cuatro Ciénegas hy- drobiids appear to represent two different brooding strategies, each represented by two taxa of the same clade. Cochliopina and Mexithauma brood a relatively large number of similar-sized embryos, with only a two- and four-fold range in embryo shell lengths, re- spectively. Durangonella and Mexipyrgus brood relatively fewer young with a greater (five- and eight-fold, respectively) range of embryo shell lengths. The latter two genera, perhaps holding embryos for long time peri- ods (hence the great range in embryo shell lengths), may require a very long brood pouch, and they do, in fact, have a greater complexity of pallial oviduct coiling than the other two genera. It is not known why the anterior end of the brood pouches of Coch- liopina and Mexithauma is much more re- flected and muscularized than in Duran- gonella and Mexipyrgus: the former two taxa do not release relatively large young, but per- haps they release numerous young at the same time, necessitating greater stretching of the brood pouch. Thus the evolution of ovoviviparity in these littoridinines has involved modifications of the female reproductive anatomy to increase the amount of space available for holding em- bryos, and to allow and control the release of large-sized embryos. Some of these brooding features parallel those described for other fresh-water gastropods. The two brooding strategies outlined above were described for species of the cerithiacean Semisulcospira (Davis, 1969b). The great development of posterior pallial oviduct coiling in Mexipyrgus, while unique among hydrobioid snails, is par- alleled by that seen in several Viviparidae (Rohrbach, 1937). ACKNOWLEDGMENTS My co-advisors, Drs. Steven M. Stanley and George M. Davis, provided constant encouragement and financial support during tenure of this project. Dr. Davis suggested the project to me, shared his knowledge of hydro- bioid snails with me, and provided the space and facilities at the Academy of Natural Sci- ences of Philadelphia necessary for comple- tion of the project. Without his interest and support, the project would not have been completed. | thank the Mexican government for provid- ing the necessary permits for collecting fresh- water snails in their country. Fieldwork in Cuatro Cienegas was aided by Dr. Curtis Dunn, Mr. Daniel Bereza and numerous local townspeople, including Pepe Lugo, Hector Ruiz, Refugio Guajardo, Benustiano Guajar- do, Gerardo Sauceda and Modesto de la Garza. The people of Cuatro Ciénegas, par- ticularly the families of Antonio Palacios and Modesto de la Garza, made my stay there pleasant. Dr. W. L. Minckley and Mr. J. Jerry Landye lent me aerial photographs of Cuatro Ciéne- gas and spent time discussing the area and fauna with me. Dr. Irving Kornfield discussed aspects of the project with me and lent the box core sampler. Dr. Glenn Longley per- mitted me to study his specimens of hydro- bioid snails from the Edwards Aquifer. The following institutions lent me material that was of use in the study: Academy of Natural Sciences of Philadelphia (ANSP), Field Museum of Natural History (Chicago), New York State Museum (NYSM), and Smithsonian Institution (USNM). Financial support came from Sigma Xi, the National Capital Shell Club, the Theodore Roosevelt Memorial Fund, the Geological Society of America, and the National Institutes of Health (grant no. TMP-11373 to Dr. Davis). A Jessup Scholarship partly supported my stay at the Academy of Natural Sciences of Philadelphia. Research for this paper was done in partial fulfillment for a Doctor of Philosophy degree at Johns Hopkins University, Baltimore, Mary- land. | thank Drs. G. M. Davis, W. F. Ponder, and F. G. Thompson, and an anonymous reviewer for their comments on the manu- script. LITERATURE CITED ANDREWS, J., 1977, Shells and shores of Texas. University of Texas, Austin, ed. 2, 365 p. ARGANO, R., 1974, Mexistenasellus magniezi n. sp., a blind aquatic isopod from Veracruz, Méx- ico (Crustacea). Quad. Accademia Nazionale dei Lincei, Prob. Atti Sci. Cult., 171: 97-103. ARNOLD, A. E., 1972, Behavioral ecology of two pupfishes (Cyprinodontidae, genus Cyprinodon from Northern México. Ph.D. dissertation, Arizo- na State University, Tempe, 138 p. BOETERS, H. D., 1974, Horatia Bourguignat, Plagigeyeria Tomlin und Litthabitella Boeters. Archiv für Molluskenkunde, 104: 85-92. BOETERS, H. D., 1979 [“1977”"], Species concept of prosobranch freshwater molluscs in Western CUATRO CIÉNEGAS HYDROBIIDS 115 Europe, |. Proceedings of the Sixth European Malacological Congress. Malacologia, 18: 57- 60. BOETTGER, O., 1878, Die Tertiarfauna von Pebas am oberen Maranon. Jahrbuch der Kaiserlich- Koniglichen Geologischen Reichsanstalt, 28: 485-504. BOLE, J., 1970, Beitrag zur Kenntnis der Anatomie und Taxonomie der unterirdischen Hydrobiiden (Gastropoda, Prosobranchia). Slovenska Akademija Znanosti in Umetmosti Academia Sci- entiarum et Artium Slovenica. Razprave Dis- sertationes 13/2: 27 p. BROWN, W. S., 1974, Ecology of the aquatic box turtle Terrapene coahuilae (Chelonia, Emydidae) in northern Mexico. Bulletin of the Florida State Museum, Biological Sciences, 19: 1-67. BRUSINA, S., 1882, Orygoceras, eine neue Gastropoden-Gattung der Melanopsiden-Mergel Dalmatiens. Beitrage Palaontologie Osterreich- Ungarns und des Orients, 2: 33-46. CLIMO, F. M., 1974, Description and affinities of the subterranean molluscan fauna of New Zea- land. New Zealand Journal of Zoology, 1: 247- 284. COLE, G. A. & MINCKLEY, W. L., 1966, Speo- cirolana thermydronis, a new species of cirolanid isopod crustacean from central Coahuila, Méx- ico. Tulane Studies in Zoology and Botany, 13: 17-22. COLE, G. A. & MINCKLEY, W. L., 1970, Sphaero- lana, a new genus of cirolanid isopod from north- ern Mexico, with description of two new species. Southwestern Naturalist, 15: 71-81. COLE, С. A. & MINCKLEY, W. L., 1972, Stena- sellid isopod crustaceans in the Western Hemi- sphere—a new genus and species from México—with a review of other North American freshwater isopod genera. Proceedings of the Biological Society of Washington, 84: 313-326. CONANT, R., 1977 [*1974”], Semiaquatic reptiles and amphibians of the Chihuahuan Desert and their relationships to drainage patterns of the region. Transactions of the Symposium on the Biological Resources of the Chihuahuan Desert Region, United States and Mexico. National Park Service Transactions and Proceedings Series, 3: 455-491. CONRAD, T. A., 1871, Description of new fossil shells of the Upper Amazon. American Journal of Conchology, 6: 192-198. CONRAD, T. A., 1874a, Remarks on the Tertiary clay of the Upper Amazon, with descriptions of new shells. Proceedings of the Academy of Nat- ural Sciences of Philadelphia, 26: 25-32. CONRAD, T. A., 1874b, Description of two new fossil shells of the Upper Amazon. Proceedings of the Academy of Natural Sciences of Phil- adelphia, 26: 82-83. CONTRERAS, S. B., 1978, Biota endemica de Cuatro Ciénegas, Coahuila, México. Memoria del 1er. Congreso Nacional de Zoologia, 9-12 Octobre de 1977, Chapingo, México, р. 106- vs: DALL, W. H., 1925, Discovery of a Balkan fresh- water fauna in the Idaho Formation of Snake River Valley, Idaho. United States Geological Survey Professional Paper, 132: 109-115. DAVIS, С. M., 1966, Notes on Hydrobia totteni. Venus, Japanese Journal of Malacology, 25: 27- 42. DAVIS, G. M., 1967, The systematic relationship of Pomatiopsis lapidaria and Oncomelania formo- sana (Prosobranchia: Hydrobiidae). Malaco- logía, 6: 1-143. DAVIS, С. M., 1968, New Tricula from Thailand. Archiv für Molluskenkunde, 98: 291-317. DAVIS, G. M., 1969a, Reproductive, neural and other anatomical aspects of Oncomelania mini- ma (Prosobranchia: Hydrobiidae). Venus, Jap- anese Journal of Malacology, 28: 1-36. DAVIS, G. M., 1969b, A taxonomic study of some species of Semisulcospira in Japan (Mesogas- города: Pleuroceridae). Malacologia, 7: 211- 294. DAVIS, G. M., 1979, The origin and evolution of the gastropod family Pomatiopsidae, with emphasis on the Mekong River Triculinae. Academy of Natural Sciences of Philadelphia, Monograph 20, 120 p. DAVIS, G. M., 1980, Snail hosts of Asian Schisto- soma infecting man: evolution and coevolution. Malacological Review, Supplement 2, The Mekong Schistosome, p. 195-238. DAVIS, G. M., 1981, Different modes of evolution and adaptive radiation in the Pomatiopsidae (Gastropoda: Mesogastropoda). Malacologia, 21: 209-262. DAVIS, G. M. & CARNEY, W. P., 1973, Description of Oncomelania hupensis lindoensis, first in- termediate host of Schistosoma japonicum in Sulawesi (Celebes). Proceedings of the Academy of Natural Sciences of Philadelphia, 125: 1-34. DAVIS, G. M. & GREER, G. J., 1980, A new genus and two new species of Triculinae (Gastropoda: Prosobranchia) and the transmission of a Malay- sian mammalian Schistosoma sp. Proceedings of the Academy of Natural Sciences of Phil- adelphia, 132: 245-276. DAVIS, G. M., KITIKOON, V. & TEMCHAROEN, P., 1976, Monograph on “Lithoglyphopsis” aper- ta, the snail host of Mekong River schistoso- miasis. Malacologia, 15: 241-287. DAVIS, G. M., MAZURKIEWICZ, M. & MAN- DRACCHIA, M., 1982, Spurwinkia: morphology, systematics, and ecology of a new genus of North American marshland Hydrobiidae (Mollus- ca: Gastropoda). Proceedings of the Academy of Natural Sciences of Philadelphia, 134: 143-177. DAVIS, G. M. & PONS DA SILVA, M. C., 1984, Potamolithus: morphology, convergence, and re- lationships among hydrobioid snails. Malaco- logia, 25: 73-108. DEACON, J. E. 8 MINCKLEY, W. L., 1974, Desert fishes. In: Desert Biology (BROWN, G. W., Jr., ed.), Vol. 2, Academic Press, New York, р. 385— 488. 116 HERSHLER FRAUENFELD, G. R. von, 1863, Vorlaufig Auf- zahlung der Arten der Gattungen Hydrobia Htm. und Amnicola Gld. Нат. in der kaiserlichen und in Cuming's Sammlung. Verhandlungen der kaiserlich-koniglichen zoologisch-botanischen Gesellschaft in Wien, 13: 1017-1032. FRETTER, V. & GRAHAM, A., 1962, British proso- branch molluscs. Ray Society, London, 755 p. FULLINGTON, R. W., 1978, The Recent and fossil freshwater gastropod fauna of Texas. Ph.D. dis- sertation, North Texas State University, Denton, 279 p. GABB, W. M., 1869, Description of fossils from the clay deposits of the Upper Amazon. American Journal of Conchology, 4: 197-200. GREGG, W. O. 8 TAYLOR, D. W., 1965, Fonte- licella (Prosobranchia: Hydrobiidae), a new genus of West American freshwater snail. Mala- cologia, 3: 103-110. GREVE, L. DE, 1938, Eine Molluskenfauna aus dem Neogen von Iquitos am Oberen Amazonas in Peru. Abhandlungen der Schweizerischen Palaeontologischen Gesellschaft, 61: 1-133. HERSHLER, R., in press, The hydrobiid snails (Gastropoda: Rissoacea) of the Cuatro Ciéne- gas Basin: systematic relationships and ecology of a unique fauna. Journal of the Arizona- Nevada Academy of Sciences. HERSHLER, R. & DAVIS, G. M., 1980, The morphology of Hydrobia truncata (Gastropoda: Hydrobiidae): relevance to systematics of Hy- drobia. Biological Bulletin, 158: 195-219. HOLSINGER, J. R., 1973, Two new species of the subterranean amphipod genus Mexiweckelia (Gammaridae) from México and Texas, with notes on the origin and distribution of the genus. In: Studies on the cavernicole fauna of Mexico and adjacent regions (REDDELL, J. R., ed.), Association for Mexican Cave Studies, Bulletin 5: 1-12. HOLSINGER, J. R. & LONGLEY, G., 1980, The subterranean amphipod fauna of an artesian well in Texas. Smithsonian Contributions to Zoology, 308, 62 p. HOLSINGER, J. В. & MINCKLEY, W. L., 1971, A new genus and two new species of subterranean amphipod crustaceans (Gammaridae) from northern Mexico. Proceedings of the Biological Society of Washington, 83: 425-444. HUBENDICK, B., 1955, The anatomy of the Gas- tropoda. In: The Percy Sladen Trust Expedition to Lake Titicaca in 1937. Transactions of the Linnean Society, ser. 3, 1(3): 309-327. HUBRICHT, L., 1940, A subterranean snail from an artesian well. Nautilus, 54: 34-35. KADOLSKY, D., 1980, On the taxonomic position, the species and the paleoecological significance of the genera Eubora, Toxosoma, and Littoridina (?) in the Pliocene Pebas Formation of the Upper Amazon Region (Gastropoda: Prosobranchia). Veliger, 22: 364-375. KARNEi, Н. S., 1978, A survey of the subterranean aquatic fauna of Bexar County, Texas. M.S. thesis, Southwest Texas State University, San Marcos, 118 p. LABOUNTY, J. F., 1974, Materials for the revision of cichlids from northern Mexico and southern Texas, USA (Perciformes: Cichlidae). Ph.D. dis- sertation, Arizona State University, Tempe, 120 p. LONGLEY, G., 1981, The Edwards Aquifer: Earth’s most diverse groundwater ecosystem? /п- ternational Journal of Speleology, 11: 123-129. LYTLE, С. L., 1972, Revision of the Notropis pros- erpinus group, subgenus Cyprinella of Notropis, from south Texas and northern Mexico (Pisces: Cyprinidae). M.S. thesis, Arizona State Univer- sity, Tempe, 74 p. MAGNIEZ, G., 1972, Deux Stenasellidae caverni- coles nouveaux de l'Amerique central: Mexi- stenasellus parzefalli п. sp. et Mexistenasellus wilkensi п. sp. (Crustacea, Isopoda, Asellota). International Journal of Speleology, 4: 19-31. MEYER, E. R., 1972, Late-Quaternary paleoecolo- gy of the Cuatro Ciénegas basin, Coahuila, Mex- ico. Ph.D. dissertation, Arizona State University, Tempe, 113 p. MEYER, E. R., 1973, Late-Quaternary paleoecolo- gy of the Cuatro Ciénegas Basin, Coahuila, Méx- ico. Ecology, 54: 982-995. MILLER, В. R., 1977 ["1974”], Composition and derivation of the native fish fauna of the Chihua- huan Desert region. Transactions of the Sym- posium on the Biological Resources of the Chi- huahuan Desert Region, United States and Méx- ico. National Park Service Transactions and pro- ceedings Series, 3: 365-382. MILLER, R. R. & MINCKLEY, W. L., 1963, Xiphophorus gordoni, a new species of platyfish from Coahuila, Mexico. Copeia, 1963: 538-546. MILSTEAD, W. W., 1960, Relict species of the Chihuahuan Desert. Southwestern Naturalist, 5: 75-88. MINCKLEY, W. L., 1969, Environments of the bol- son of Cuatro Ciénegas, Coahuila, México. Uni- versity of Texas (El Paso), Science Series, 2: 1-65. MINCKLEY, W. L., 1977 [*1974"], Endemic fishes of the Cuatro Ciénegas basin, northern Coahuila, México. Transactions of the Sym- posium on the Biological Resources of the Chi- huahuan Desert Region, United States and Mex- ico. National Park Service Transactions and Pro- ceedings Series, 3: 383-404. MINCKLEY, W. L. & COLE, G. A., 1968, Pre- liminary limnologic information on waters of the Cuatro Ciénegas Basin, Coahuila, México. Southwestern Naturalist, 13: 421-431. МОВАЕКА, О. J., 1977, A biogeographic analysis of the Chihuahuan Desert through its herpeto- fauna. Biogeographica (Junk, The Hague), 9: SSD; MORRISON, J. P. E., 1945, Durangonella, a new hydrobiine genus from México, with three new species. Nautilus, 59: 18—23. MORRISON, J. P. E., 1946, The nonmarine mol- CUATRO CIÉNEGAS HYDROBIIDS 117 lusks of San Jose Island, with notes on Pedro Gonzalez Island, Pearl Islands, Panama. Smithsonian Miscellaneous Collections, 106: 1— 49. PARODIZ, J. J., 1969, The Tertiary non-marine Mollusca of South America. Annals of the Carne- gie Museum, 40: 1-242. PILSBRY, H. A., 1944, Molluscan fossils from the Rio Pachitea and vicinity in eastern Peru. Pro- ceedings of the Academy of Natural Sciences of Philadelphia, 96: 137-153. PILSBRY, Н. A. & FERRISS, J. H., 1906, Mollusca of the Southwestern States. Il. Proceedings of the Academy of Natural Sciences of Phil- adelphia, 58: 123-175. PONDER, W. F., 1966, On a subterranean snail and a tornid from New Zealand. Journal of the Malacological Society of Australia, no. 10: 35— 40. REDDELL, J. R., 1965, A checklist of the cave fauna of Texas. |. The Invertebrata (exclusive of Insecta). Texas Journal of Science, 17: 143- 187. ROHLF, F. J., KISHPAUGH, J. & KIRK, D., 1972, NT-SYS; numerical taxonomy system of multi- variate statistical programs. Stony Brook, New York. ROHRBACH, F., 1937, Oekologische und morpho- logische Untersuchungen an Viviparus (Bel- lamya) capillatus Frauenfeld und Viviparus (Bel- lamya) unicolor Oliver, unter Berucksichtigung anderer tropischer Formen und im Hinblick auf phyletische Beziehungen. Archiv fur Mollusken- kunde, 69: 177-218. RUSSELL, R. H., 1971, Mollusca of Fish Springs, Juab County, Utah: rediscovery of Stagnicola pilsbryi (Hemphill, 1890). Great Basin Naturalist, 31: 223-236. SCHALIE, H. van DER, 1936, Ovoviparity among mollusks. Nautilus, 50: 16-19. SCHMIDT, K. P. 4 OWENS, D. W., 1944, Amphib- ians and reptiles of northern Coahuila, México. Field Museum of Natural History (Chicago), Zoology Series, 29: 97-115. SMITH, M. L., 1981, Late Cenozoic fishes in the warm deserts of North America: a reinterpreta- tion of desert adaptations. In: Fishes of North American deserts (NAIMAN, R. J. & SOLTZ, D.L., eds.), Wiley, New York, p. 11-38. SNEATH, P. 8 SOKAL, R., 1973, Numerical taxon- omy. The principles and practice of numerical classification. Freeman, San Francisco, 573 p. STANLEY, S. M., 1979, Macroevolution: pattern and process. Freeman, San Francisco, 332 p. TAYLOR, D. W., 1966, A remarkable snail fauna from Coahuila, México. Veliger, 9: 152-228. TAYLOR, D. W., 1974, The Tertiary gastropod Orygoceras found living. Archiv fúr Molluskenk- unde, 107: 93-96. TAYLOR, D. W. & MINCKLEY, W. L., 1966, New world for biologists. Pacific Discovery, 19: 18— 22: THOMPSON, F. G., 1968, The aquatic snails of the family Hydrobiidae of peninsular Florida. Univer- sity of Florida, Gainesville, 268 p. THOMPSON, F. G., 1977, The hydrobiid snail genus Marstonia. Bulletin of the Florida State Museum, Biological Science, 21: 113-158. THOMPSON, F. G., 1979, The systematic rela- tionships of the hydrobiid snail genus Nym- phophilus Taylor 1966 and the status of the subfamily Nymphophilinae. Malacological Re- view, 12: 41-50. THOMPSON, F. G. & McCALEB, J. E., 1978, A new freshwater snail from a spring in eastern Alabama. American Midland Naturalist, 100: 350-358. VAN DEVENDER, Т. R., 1976, The biota of the hot deserts of North America during the last glacia- tion: the packrat midden record. American Quaternary Association Abstracts for 1976 meet- ing, p. 62-67. VAN DEVENDER, T. R., 1977, Holocene wood- lands in Southwestern deserts. Science, 198: 189-192. VERMEIJ, С. J. 8 COVICH, A. P., 1978, Coevolu- tion of freshwater gastropods and their pred- ators. American Naturalist, 112: 833-843. WEBB, R. G. & LEGLER, J. M., 1960, A new softshell turtle (genus Trionyx) from Coahuila, México. University of Kansas, Science Bulletin, 40: 21-30. WELLS, Р. V. 1977 [“1974”], Post-glacial origin of the present Chihuahuan Desert less than 11,500 years ago. Transactions of the Symposium on the Biological Resources of the Chihuahuan Desert Region, United States and Mexico. National Park Service Transactions and Pro- ceedings Series, 3: 67-83. APPENDIX 1 Collection localities for this study. Localities 1— 100 are shown in Fig. 3. 1. Laguna Churince (southernmost of the three Posos Bonitos), 19.5 km SSW of Cuatro Ciénegas along the highway. 2. small seep feeding the middle of the three Posos Bonitos. 3. Rio Churince, 100 m downstream from Laguna Churince. 4. Rio Churince, wide pool area, 1000 m downstream from Laguna Churince. 5. Rio Churince, very large pool area due E of Laguna Grande. 6. Small pool at NW edge of (5). 7. Small pool in marshy area, N of (5). 8. Small pool in marshy area, W of (7). 9. Laguna Grande, playa lake terminus of Rio Churince. 10. Posos de la Becerra, 4 km S of the tip of Sierra de San Marcos along the highway. 11. Stream from cool springs at Los Chiceros. 12. Small spring feeding stream at Los Chiceros. 13. Stream at Los Chiceros, below warm water inflow from canal from Posos de la Becerra. 14. Small spring, 2.24 km S of tip of Sierra de San Marcos along the highway, at SE corner of the springfield. 15. Small spring, 3m W of (14), feeding same stream. 16. Small spring, 40 m N of (14). 17. Small spring, 146 m N of (16). 18. Large 118 HERSHLER spring, 130 m N of (17). 19. Large spring, just W of marsh terminus of (18). 20. Large spring, 10 m W of (19). 21. Large spring, 25 m NW of (20). 22. Juan Santos Laguna, NW of (21). 23. Small spring, 62 m N of (18). 24. Small springhole (no outflow), 370 m NNE of (23). 25. Small spring, 52 m NNW of (24). 26. Small spring, 88 m N of (25). 27. Small spring, 165 m NNW of (26). 28. Small spring at SE corner of vegetated area, 24 m N of (27). 29. Small spring (flowing N), about 300m W of (28). 30. North Spring, 960 m S of tip of Sierra de San Marcos along the highway. 31. Small spring, 57m W of (30). 32. Small spring, 27 m N of (31). 33. Small spring, 54 m N of (32). 34. Small spring, 60 m NW of (33). 35. Small spring, 328 m N of (30). 36. Small spring, 300 m NW of (22). 37. Large cold spring, 800 mS, 1 km N oftip of Sierra de San Marcos. 38. Large cold spring, about 50 m W of (37). 39. Small spring hole (no outflow), 10 т $ of (38). 40. Small spring, 800 m S, 1.6 km W of tip of Sierra de San Marcos. 41. Large spring, 800 m S, 1.25 km W of tip of Sierra de San Marcos. 42. Large spring, due W of junction of streams from (41) and (43). 43. Large spring, 800 т $, 1.22 km W of tip of Sierra de San Marcos. 44. Rio Garabatal, due W of (42). 45. Small spring, due W of (42). 46. Large spring, 230 m NE of (43). 47. Small spring, just W of marsh terminus of (46). 48. Large spring, 800 m S, 400 m W of tip of Sierra de San Marcos. 49. Large spring, 700 m S of tip of Sierra de San Marcos along the highway. 50. Rio Mesquites, 200 m upstream from junction with springs from the west. 51. Small spring, 100 m S of Rio Mesquites at house of Tierra Blanca. 52. Small spring, just N of highway, 320 m NE of tip of Sierra de San Marcos. 53. Rio Mes- quites at the highway, 9.3km SSW of Cuatro Ciénegas. 54. Rio Mesquites, where small marshy stream branches off the Mojarral area. 55. Small stream branching from (54), due E of (76). 56. Small spring, 320m S of Rio Mesquites at the highway. 57. Small spring, 370 m S of tip of Sierra de San Marcos along dirt road on the east side. 58. Small spring, due E of (59). 59. Small spring, due E of (60). 60. Small spring, 670 m S of tip of Sierra de San Marcos along dirt road. 61. Small spring, due E of (62). 62. Small spring, 150 т NE of (63). 63. Large spring, 1.12 km S of tip of the Sierra de San Marcos along dirt road. 64. Small spring, 40 m S of (63). 65. Small spring, 72m S of (64). 66. Small spring, 60 m S of (65). 67. Small spring, 1.4 km S of tip of Sierra de San Marcos along dirt road. 68. Small spring, 1.6km S of tip of Sierra de San Marcos along dirt road. 69. Small spring, due W of (70). 70. Large spring, 400 m E, 60 m S of (68). 71. Stream from (70), 130 т above mash. 72. Small spring, 2.7 km S of tip of Sierra de San Marcos along dirt road. 73. Mojarral West Laguna, about 200 m N of (76). 74. Stream draining marsh due W of (73). 75. Stream draining (73). 76. Mojarral East Laguna, 10.2 km S of Cuatro Ciénegas along the highway, 1.4km S of highway. 77. Small spring, feeding SW corner of (76). 78. Pools 30 m down- stream from (55). 79. Large spring, 2.4 km S of tip of Sierra de San Marcos along dirt road, 900 m E of road. 80. Large spring (receiving flow from 79), 100 m S of (79). 81. Los Remojos, northern spring, 500 m S of (80). 82. Los Remojos, middle spring. 83. Los Remojos, southern spring. 84. Los Re- mojos, large pool receiving inflows from (81), (82), and (83). 85. Large spring, 1.2km S of (83). 86. Large spring, 400 m S of (85). 87. Large spring, 115 m SE of (86). 88. Large spring, about 1.2 km S of (86). 89. Large spring, about 100 m S of (88). 90. Large spring, 60 m SW of (89). 91. Large spring, 200 т $ of (90). 92. Large spring, about 200 т М of (95). 93. Large spring, due E of (90). 94. Large spring, due SE of (98). 95. Laguna Escobedae, 6.72 km S of tip of Sierra de San Marcos along dirt road, 1 km E of road. 96. Large spring, 6.9 km $ of tip of Sierra de San Marcos along dirt road. 97. Laguna Tio Candido, 9.3 km S of tip of Sierra de San Marcos along dirt road. 98. Large spring, 10.7 km S of tip of Sierra de San Marcos along dirt road. 99. Laguna Anteojo. 100. Smaller spring, due W of (99). 101. Santa Tecla Laguna (= La Tecla), 22.2 km S of tip of Sierra de San Marcos along dirt road, 1.6 km SE of road. 102. Spring alongside the Rio Salado de Nadadores, 3.04km N of Sac- ramento, Coahuila. 103. Rio Salado de Nadadores, at Carino de la Montana, 3.84 km E of Sacramento, Coahuila. APPENDIX 2 The material collected by the author and ex- amined during this study is listed below by species. For each species, the lots are listed in order of locality number (1-103), with the catalog numbers and dates of collection following. All of the material is deposited in the Department of Malacology at the Academy of Natural Sciences of Philadelphia (ANSP). The initial letter A refers to specimens in alcohol. Other catalog numbers refer to dry shell lots. 1. Nymphophilus minckleyi. 1: A9879-A, 23 Mar. 1979; A9879-E, 17 Dec. 1981; A9878-l, 19 Dec. 1981. 3: A9878-G, 23 Mar. 1979. 5: A9879-B, 16 May 1980. 10: A9878-H, 4 Aug. 1979. 13: A9877-B, 2 June 1979; A9878-A, 17 Nov. 1980; A9878-E, 9 Feb. 1981. 18: A9878-D, 13 June 1979. 30: A9878- F, 9 Apr. 1979. 37: A9878-B, 11 July 1980. 38: A9877-C, 12 Apr. 1979; A9877-F, 18 July 1980. 40: A9877-A, 27 May 1981. 41: A9879-G, 17 Feb. 1981. 42: A9879-D, 10 Apr. 1981. 43: A9878-C, 8 July 1979. 53: 355196, A9877-D, 26 July 1979; A9876, 23 Apr. 1979. 64: A9879-F, 4 June 1981. 71: A9879-C, 1 May 1981; A9879-H, 12 Dec. 1981. 76: A9877-H, 7 May 1979; 355197, A9877-E, 7 Apr. 1981. 79: A9877-G, 3 Apr. 1979. 97: 355195, A9879-J, 21 May 1979. 2. Nymphophilus acarinatus. 98: holotype: 355255, 20 Dec. 1981; paratypes: 355256, 20 Dec. 1981. 101: A9929-B, 13 July 1980; A9929-C, 28 May, 1981. 3. Mexistiobia manantiali. 14: A9888-E, 25 Nov. CUATRO CIÉNEGAS HYDROBIIDS 119 1980. 16: A9887-B, 11-14 June 1979; A9886-K, 5 June 1980; A9888-J, 27 Nov. 1980. 17: A9887-L, 7 Dec. 1980. 25: A9887-G, 11 Feb. 1981. 27: A9888- |, 14 Feb. 1981. 28: A9887-L, 17 Jan. 1981. 29: A9888-B, 7 Apr. 1979. 31: A9887-A, 18 May 1981. 36: A9887-K, 20 June 1981. 38: A9886-l, 8 Jan. 1981; A9887-H, 16 Jan. 1981, A9887-F, 5 Feb. 1981. 43: A9886-M, 13 July 1979. 51: holotype: 355205, 13 July 1979; paratypes: A9887-D, 20 Apr. 1979; 355204, A9888-L, 13 July 1979. 52: A9886- D, 21 June 1981. 57: A9886-C, 15 June 1981. 59: A9886-B, 19 June 1981. 64: A9886-N, 31 May 1979; A9887-E, 30 May 1981; A9886-G, 4 June 1981; A9886-L, 21 June 1981. 65: A9886-H, 31 May 1979; 355206, A9888-F, 28 Apr. 1981. 67: A9886-F, A9887-J, 30 Apr. 1981. 68: A9889-K, 2 May 1981. 72: A9887-C, 19 May 1979. 74: A9888- С, 5 Apr. 1980. 77: A9886-J, 16 June 1981. 4. Coahuilix hubbsi. 16: A9892-A, 11-14 June 1979; A9892-M, 5 Jan. 1980; A9882-H, 27 Nov. 1980. 18: A9892-C, 19 Dec. 1981. 24: A9892-J, 14 Dec. 1981. 31: A9892D, 18 May 1981. 33: A9892- B, 21 May 1981. 35: 355210, 9 May 1981. 38: A9893-C, 8 Jan. 1981. 58: A9892-N, 17 Jan. 1981. 61: A9892-I, 18 June 1981. 64: A9893-B, 26 Apr. 1981; A9892-G, 4 June 1981; A9892-K, 21 June 1981; A9892-E, 13 Dec. 1981. 66: A9892-L, 28 Apr. 1981. 67: 355209, A9893-A, 30 Apr. 1981. 68: A9892-F, 1 May 1981. 5. Coahuilix landyei. 16: A9894-M, 27 Nov. 1980. 17: A9894-1, 10 June 1979. 24: A9894-A, 13 Feb. 1981. 28: A9894-L, 7 Apr. 1979; A9894-C, 17 Jan. 1981. 31: A9894-F, 18 May 1981. 38: A9894-E, 8 Jan. 1981. 59: A9894-H, 19 Jan. 1981. 63: A9894- K, 16 Apr. 1981. 64: holotype: A9894-N, 29 Apr. 1981; paratypes: 355211, 27 Apr. 1981. 65: A9894- B, 26 Apr. 1981. 61: A9894-J, 30 Apr. 1981. 69: A9894-G, 6 May 1981. 6. Cochliopina milleri. 36: A9884-E, 5 Feb. 1981. 38: A9884-J, 3 Sept. 1978; A9884-C, 10 Apr. 1981; 355200, A9884-K, 5 Jan. 1981. 41: A9884-D, 17 Feb. 1981. 42: A9884-C, 10 Apr. 1981. 43: A9884- В, 8 July 1979. 50: A9884-1, 1 Sept. 1978. 53: A9884-F, 23 Apr. 1979. 55: A9884-G, 7 May 1979. 97: A9884-H, 21 May 1979. 7. Cochliopina riograndensis. 101 A9885-D, 10 July 1979; A9885-E, 13 July 1980; 355202, 28 May 1981. 102: 355201, A9885-A, 25 May 1981. 103: A9885-B, A9885-C, 24 July 1979. 8. Mexithauma quadripaludium. 1: A9883, 23 Mar. 1979; 355198, A9881-F, 25 Aug. 1980. 10: A9880-E, 9 Apr. 1979; A9881-C, 4 Aug. 1979. 18: A9880-D, 13 June 1979. 20: A9880-G, 20 June 1979. 22: A9880-B, 3 Apr. 1979. 30: A9880-I, 23 Mar. 1979. 42: A9881-E, 10 Apr. 1981. 49: A9881- D, 23 Feb. 1981. 50: A9880-C, 1 Sept. 1978; A9880-F, 26 Feb. 1981. 71: A9880-A, 1 May 1981; A9881-H, 12 Dec, 1981. 76: A9880-H, 9 May 1979; A9882, 26 July 1979; A9881-A, 10 Apr. 1981. 97: 355199, 23 May 1981. 98: A9881-G, 20 Dec. 1981. 99: 355257, 10 Apr. 1981. 100: A9881-B, 10 Apr. 1981. 101: A9880-J, 28 Mar. 1981. 9. Durangonella coahuilae. 2: A9922-J, 13 Dec. 1981. 4: A9922-1, 15 May 1980. 5: A9922-G, 16 May 1980, A9924-B, 19 Nov. 1980. 6: A9922-K, 17 May 1980; 355244, A9928-K, 8 Oct. 1980. 7: A9922-H, 20 May 1980. 8: A9922-L, 27 May 1980. 9: A9922-D, 29 Mar. 1979; A9922-B, 16 May 1979; A9922-C, 30 May—10 June 1980; A9922-E, 23 Aug. 1980; A9922-F, 10 Oct. 1980; 355249, 22 Oct. 1980. 12: A9924-L, 19 Dec. 1980. 13: 355250, A9923-B, 13 Nov. 1980. 14: A9923-E, 25 Nov. 1980; 355246, A9928-M, 27 Nov. 1980. 16: A9926- M, 9 June 1979; A9928-J, 27 Nov. 1980; A9925-E, 6 Dec. 1980. 17: A9927-C, A9928-B, 10 June 1979; A9927-K, 7 Dec. 1980. 18: A9923-K, 15 Jan. 1980; A9927-L, 19 Dec. 1981. 23: A9927-A, 8 Feb. 1981; A9927-J, 11 Feb. 1981. 24: A9926-1, 8 Feb. 1981; A9924-1, 13 Feb. 1981; A9928-C, 14 Dec. 1981. 25: A9926-K, 8 Feb. 1981. 26: A9928-H, 14 Dec. 1981. 27: A9927-F, 8 Feb. 1981; A9927-B, 14 Feb. 1981. 28: A9924-N, 7 Apr. 1979; A9926-F, 13 July 1979; A9924-C, 17 Jan. 1981. 29: A9927-D, 14 Feb. 1981. 30: A9925-A, 28 Jan. 1980. 31: A9928-I, 18 May 1981. 35: A9922-A, A9924-G, 9 May 1981. 37: A9926-H, 11 July 1980; A9925-K, 7 Feb. 1981. 38: A9923-D, 15 May, 1979; A9923-F, 18 July 1980; A9924-F, 8 Jan. 1981; 355251, 20 Jan. 1981; A9926-G, 5 Feb. 1981; A9923-C, 5-6 Jan. 1981; A9927-H, 16 Jan. 1981; A9923-H, A9926-J, 6 Feb. 1981. 41: A9925-G, 17 Feb. 1981. 43: A9924-M, 7 July 1979; A9927-1, 13 July 1979; 355245, A19928- L, 17 Feb. 1981. 45: A9926-N, 10 Apr. 1981. 48: A9925-L, 23 Feb. 1981; A9925-J, 5 May 1981. 49: A9925-F, 23 Feb. 1981. 50: A9925-C, A9925-l, 1 Sept. 1978. 51: 355247, A9928-N, 20 Apr. 1979. 52: A9924-D, 21 Jan. 1981. 56: A9928-F, 18 Mar. 1981; A9925-B, 19 Mar. 1981. 58: A9924-A, 17 Jan. 1981. 59: A9923-J, 19 Jan. 1981. 61: A9927- M, 18 Jan. 1981. 62: A9927-N, 15 June 1981. 63: A9926-C, 8 Jan. 1979. 64: A9926-B, A9928-D, 31 May 1979; A9928-A, 21 Jan. 1981; A9926-L, 27 Apr. 1981; A9928-G, 30 May 1981. 65: A9923-A, 22 May 1979; A9924-H, 31 May 1979; 355248, A9923-L, 4 Apr. 1980; A9923-1, 24 July 1980; A9923-J, 1 Oct. 1980. 72: A9927-E, 19 May 1979. 73: 355252, A9926-D, 5 Apr. 1980. 74: A9926-A, 2 May 1979; A9925-H, 12 Apr. 1980. 75: A9925-M, 4 May 1979. 77: A9924-E, 16 Jan. 1981. 10. Mexipyrgus churinceanus. 1: A9909-H, 23 Маг. 1979; 355230, 15 Sept. 1980; A9907-I, 20 Sept. 1980; А9909-1, 13 Mar. 1981. 5: A9908-I, 16 Мау 1980; A9907-H, 24 Sept. 1980; 355231, 5 Oct. 1980. 10: A9909-C A9911-A, 27 Oct. 1980; 344232, 29 Oct. 1980; A9909-D, 31 Oct. 1980. 11: 355233, 2 Nov. 1980; A9901, 10 Nov. 1980. 13: A9911-F, 8 Nov. 1980. 18: 355220, A9912-1, 12 June 1980; 355216, A9912-E, 12 Aug. 1980; A9911-H, 8 Dec. 1980. 19: A9911-J, 10 Dec. 1980. 20: A9910-G, 26 June 1980; 355234, 10 Dec. 1980; A9907-E, 11 Dec. 1980. 21: 355214, A9912-C, 29 Apr. 1981. 22: A9907-B, 3 Apr. 1979; 355241, 16 Dec. 1980; A9907-C, 18 Dec. 1980. 30: A9911-C, 9 Apr. 1979; A9907-F, 20 Aug. 1980; 355222, A9914, 18 Dec. 1980. 37: 355213, A9912-B, 30 Dec. 1980. 38: 355236, 1 Jan. 1981; A9910-H, 5-8 Jan. 1981; 120 HERSHLER A9911-B, A9911-D, A9911-E, 16 Feb. 1981; A9910-J, 21 May 1981. 43: A9910-J, 7 July 1979 A9911-G, 16 Feb. 1981. 44: A9911-1, 10 Apr. 1981. 46: A9908-J, 19 Feb. 1981; 3553212, A9912-A, 20 Feb. 1981. 47: 355227, 23 Feb. 1981. 48: A9895 23 Feb. 1981; 355235, 24 Feb. 1981; A9897, 5 May 1981. 49: A9898, 23 Feb. 1981; 355242, 24 Feb. 1981; A9907-G, 5 May 1981. 50: 355218, A9920 26 Feb. 1981. 53: A9909-E, 23 Apr. 1979; A9909-F, 6 Mar. 1981. 54: A9909-G, 8 Mar. 1981. 65: A9909- A, 31 May 1979. 67: A9910-E, 4 May 1981; A9910- D, 5 May 1981. 70: A9910-F, 1 May 1981; A9909-J, 2 May 1981. 71: A9910-A, 4 May 1979; 355240, A9910-B, 25 Mar. 1981; A9905, 5 May 1981; A9910-C, 4 May 1981; A9907-A, 12 Dec. 1981. 73: 355219, A9912-H, 11 Mar. 1981. 76: A9906, 16 Mar. 1981; 355239, 17 Mar. 1981. 78: A9907-D, 9 Apr. 1981; 355215, A9912-D, 29 Apr. 1981. 79: A9900, 6 May, 1981; 355237, A9908-F, 8 May, 1981. 81: 355223, A9915, 11 May 1981. 82: 355226, A9918, 11 May 1981. 83: A9902, 9 May 1981; 355243, 11 May 1981. 84: A9908-A, 12 May 1981. 85: A9908-C, 14 May 1981. 86: A9896, 14 May 1981. 87: A9909-B, 16 May 1981. 88: 344229, A9921, 17 May 1981. 89: A9908-B, 17 May 1981. 90: 355225, A9917, 17 May 1981. 91: A9908-G, 18 May 1981. 92: A9908-H, 20 May 1981. 93: A9904 22 May 1981. 94: A9908-E, 22 May 1981. 95: 355221, A9913, 20 May 1981. 96: 355218, A9912G, 20 Dec. 1981. 97: 355224, A9916, 23 May 1981. 98: A9907-J, 20 Dec. 1981. 99: 355217 A9912-F, 10 Apr. 1981. 100: A9908-D, 10 Apr. 1981. 101: A9903, 28 May 1981. Shell 1. Shell shape: a) planispiral b) trochoid-globose с) ovate-conic d) turriform 2. Maximum shell dimension: a) <2 mm b) =2 mm <5 тт с) =5 mm 3. Shell sculpture: a) collabral ribs b) spiral cords c) absent 4. Apical whorl microsculpture: a) wrinkled, pitted b) absent 5. Shell with periostracal bands 6. Shell aperture flared External features 7. Tentacle ciliation: a) absent b) Hydrobia-like c) Spurwinkia-like 8. Animal blind, unpigmented 9. Mantle edge papillate 11. Orygoceras? sp. 67: 355254, A9929-A, 19 Dec. 1981. 12. Paludiscala caramba. 2: A9890-D, 13 Dec. 1981. 16: A9890-K, 11 June 1979; A9890-H, 5 Jan. 1980; A9889-K, 27 Nov. 1980. 17: A9890-J, 10 Jan. 1979. 18: A9891-J, 19 Dec. 1981. 23: A9889-N, A9891-D, 11 Feb. 1981. 24: A9891-H, 13 Feb. 1981; A9889-M, 14 Dec. 1981. 25: A9891-K, 11 Feb. 1981. 26: A9889-G, 14 Dec. 1981. 27: A9891- L, 11 Feb. 1981. 28: A9880-B, 17 Jan. 1981. 31: A9891-E, 11 May 1981. 32: A9890-L, 18 May 1981. 33: A9889-A, 21 May 1981. 35: A9889-l, 9 May 1981. 38: A9880-H, 16 Jan. 1981; A9890-l, 8 Jan. 1981. 56: A9889-L, 18 May 1981. 57: A9890-M, 15 Jan. 1981. 58: A9889-C, 17 Jan. 1981. 59: A9889- E, 19 June 1981. 60: A9889-B, 19 Jan. 1981. 61: A9890-C, 18 Jan. 1981. 62: A9890-A, 15 Jan. 1981. 63: 355207, A98911, 16 Apr. 1981. 64: A9889-J, 31 May 1979; A9889-F, 26 Apr. 1981; A9890-F, 13 Dec. 1981. 65: A9889-H, 31 May 1979; 355208, 24 July 1980; A9891-I, 26 Apr. 1981. 66: A9891-A, 28 Apr. 1981. 67: A9890-E, 30 Apr. 1981. 68: A9891-B, 1 May 1981. 69: A9891-G, 6 May 1981. 72: A9891-F, 19 May 1979. APPENDIX 3 Fifty-one characters and their character states are used in this study to distinguish between genera and generic groups of Hydrobiidae. To the right are references to publications in which the various character states have been figured. “H” indicates a figure within this paper. H, Figs. 15A-G, I-K H, Figs. 4, 9, 27 В, Fig 3% НВ, Fig. 33 H, Fig. 19 H, Fig. 27 i, Figs 32 H, Fig. 11; Thompson, 1977, fig. 4 H, Fig. 39A inky Tee Siz Boeters, 1974, figs. 9-11; Bole, 1970, fig. 6 H, Fig. 40A Davis, 1966, fig. 2; Hershler & Davis, 1980, fig. 1A Davis et al., 1982, fig. 5; H, Figs. 29A, B H, Fig. 30A, 31A 10. A WZ CUATRO CIENEGAS HYDROBIIDS 121 Osphradium elongate (=0.30 of the ctenidium length) Number of operculum whorls: a) <4 b) =4 Position of nucleus of operculum along its long axis: a) <30% b) =30% Digestive system . Digestive gland with reduced tubercles . Intestine with anterior loop . Caecal chamber extends posterior to stomach . Origin of basal cusps of central tooth of radula: a) from face of tooth b) from lateral angles . Number of pairs of basal cusps on central tooth of radula: a) 1-2 b) =3 . Central cusp of lateral tooth: a) massive b) small Male reproductive anatomy 19: 28. 29. Male gonad morphology: a) simple lobes b) bush-like c) single un-lobed mass . Male gonad extends onto stomach . Seminal vesicle coiling: a) on stomach b) posterior to stomach . Position of prostate: a) entirely posterior to end of mantle cavity b) overlaps mantle cavity . Anterior vas deferens: a) exits from anterior tip of prostate b) exits from posterior portion of prostate . Penis with slender penial filament . Penis ciliated . Penis with terminal eversible papilla . Penial lobe(s): a) absent b) bulb-like с) simple d) with folds Gland(s) of penis: a) absent b) large, specialized c) small, in glandular ridges Of snails having glandular ridges on the penis: a) <4 ridges b) =4 ridges Female reproductive anatomy 30. Reproductive mode: a) oviparity b) ovoviviparity Davis & Pons da Silva, 1984, fig. 6 H, Fig. 35B H, Fig. 5B H, Fig. 35B H, Fig. 5B H, Fig. 22B H, Fig. 17C; Boeters, 1974, fig. 3 H, Fig. 7A Fig. 16C; Davis & Pons da Silva, 1984, fig. 14 Fig. 28D ‚ Figs: 12B, € , Fig. 28D Fig. 6B ‚ Fig. 124 = en cn 2 Davis 8 Greer, 1980, fig. 9A H, Fig. 31A H, Figs 176 H, Fig. 17C Davis & Greer, 1980, fig. 9A H, Fig. 17C; Davis et al., 1982, fig. 14A НЕЮ. SIMA H, Fig. 31A Davis, 1967, pl. 25, fig. 4 Ai Rig? SilB H, Figs. 35D, 44B H, Figs. 35D, 44A H, Figs. 25D, 31B H, Figs. 17D, 21C H, Fig. 35D H, Fig. 8A H, Figs. 17D, 21C, 44A H, Figs. 8B, C; Boeters, 1974, figs. 6, 7; Thompson, 1968, figs. 42-47 H, Figs. 8B, C, 13D; Thompson, 1977, figs. 5A, 7A, MiGs 13 Thompson, 1968, figs. 42-47 122 36. 37. 38. 40. 42. 43. 44. 45: 46. 47. 48. 49. HERSHLER . Female gonad morphology: a) relatively large, lobed H, Fig. 41A b) relatively large, un-lobed Tay le TAS) c) relatively small, a mere thickening at end of H, Fig. 30A oviduct . Female gonad extends onto stomach H, Fig. 30A . Oviduct without coil HAE ИВ . Length of pallial oviduct/length of body a) <0.30 В Рю ИВ b) =0.30 H, Fig. 30A . Posterior pallial oviduct: a) without bend H, Fig. 17B b) with short bend H, Figs. 30A, 36A; Davis et al., 1982, fig. 10A c) with long bend, coiling in >1 plane Н, Fig. 43 Albumen gland: a) normal sized besas b) reduced in size H, Fig. 26A c) reduced to a glandular smear H, Fig. 36A Of ovoviviparous taxa, capsule gland: a) with 1 tissue region Hershler & Davis, 1980, fig. 2 b) with 2 tissue regions Ae righ УВ c) with 3 tissue regions H, Fig. 22A Length of bursa/length of pallial oviduct: a) <0.20 H, Fig. 30B b) =0.20 <0.40 KIN РО. ИВ с) =0.40 . Position of bursa: a) posterior to pallial oviduct H, Figs. 7A, B b) overlapped by pallial oviduct H, Fig. 14B Normal seminal receptacle present г: Рю. 7В . Secondary seminal receptacle present H, Fig. 22A Of snails with a normal seminal receptacle, the length of the seminal receptacle/length of bur- sa is: a) <0.30 H, Fig. 42B b) =0.30 <0.50 H, Figs. 26B, C с) 20.50 H, Fig. 30B Of snails with a normal seminal receptacle, opening of seminal receptacle into: a) oviduct H, Fig. 30B b) sperm duct H, Fig. 42C Of snails with a normal seminal receptacle, the seminal receptacle is: a) overlapped by the bursa H, Fig. 42B b) lateral to the bursa H, Fig. 30B Female with: a) ciliated ventral channel inside the pallial ovi- i, Fig; 7B duct b) spermathecal duct H, Fig. 30A Of snails with a ciliated ventral channel: a) pallial oviduct opens at anteror tip H, Fig. 14D b) pallial oviduct opens laterally НЕЮ. ИЕ Of snails with а ciliated ventral channel: a) bolster weakly developed H, Fig. 14C b) bolster well-developed мВ [5 7D Of snails with а spermathecal duct, the duct is: a) long H, Fig. 30A b) short H, Fig. 26A Of snails with a short spermathecal duct, the duct is: a) muscularized H, Fig. 42B b) non-muscularized H, Fig. 26A CUATRO CIÉNEGAS HYDROBIIDS 50. Of snails with a long spermathecal duct, the openings of the spermathecal duct and pallial oviduct are: a) separate H, Figs..S6F, G b) joined H, Fig. 22F c) separate, but with an open channel between H, Fig. 30E them 51. Of ovoviviparous snails, the anterior pallial ovi- duct has: a) a slight muscular loop H, Fig. 41B b) a well-developed muscular loop H Fig. 26E 123 MALACOLOGIA, 1985, 26(1-2): 125-135 THE PELAGIC GENUS PTEROTRACHEA (GASTROPODA: HETEROPODA) FROM HAWAIIAN WATERS: A TAXONOMIC REVIEW Roger R. Seapy Department of Biological Science, California State University Fullerton, CA 92634, U.S.A. ABSTRACT The four recognized species of pterotracheid heteropods appear to be present in Hawaiian waters. Specimens that could be assigned to Pterotrachea hippocampus Philippi, 1836 and P. minuta Bonnevie, 1920, however, were indistinguishable on the basis of visceral nucleus shape, eye morphology, fin and sucker size in males, pedal ganglion location, morphology of the male copulatory apparatus, and radular structure. The morphologies of the visceral nucleus and eye have been considered of particular importance by previous workers in separating the two species. The ratio of length to maximum width of the visceral nucleus was found to be highly variable, however, probably as a result of preservation effects and differences between individuals in fullness of the gut. Eye morphology, on the other hand, appears to be the most conservative and useful taxonomic characteristic. Regression of the ratio of eye length to retinal width against body length resulted in a size-specific relationship, with the eye shape ranging from narrowly triangular in small specimens (as in P. minuta) to broadly triangular in large individuals (as in P. hippocampus). The present results suggest that P. minuta should be considered a junior synonym of P. hippocampus. Key words: Pterotrachea; Heteropoda; taxonomy; pelagic; Hawaii. INTRODUCTION All four species of Pterotrachea that are currently recognized (Van der Spoel, 1976) appear to occur in oceanic waters off Oahu, Hawaii. Two of these species, P. coronata Niebuhr (ms. Forskal), 1775, and P. scutata Gegenbaur, 1855, possess distinctive taxo- nomic features. However, P. hippocampus Philippi, 1836, and P. minuta Bonnevie, 1920, to which the great majority of specimens col- lected off Hawaii can be assigned, could not be clearly distinguished. Large specimens could be identified as P. hippocampus, speci- mens of intermediate size were closer in appearance to P. hippocampus, whereas small individuals were more similar to P. minuta. Both species have been reported from the North Pacific Ocean off Baja Califor- nia (Dales, 1953; McGowan, 1967) and Japan (Okutani, 1957a,b). (Additional records and synonyms of P. hippocampus and P. minuta from the world’s oceans are cited in Van der Spoel (1976).) To resolve this taxonomic problem, morphometric analyses were performed to allow comparison of the Hawaiian material with descriptions of P. hippocampus and P. minuta (Bonnevie, 1920; Tesch, 1949; Rich- ter, 1968, 1974; Van der Spoel, 1972, 1976) and with the holotype of P. minuta. Pterot- rachea coronata and P. scutata collected from Hawaiian waters were examined for comparative purposes. MATERIALS AND METHODS Samples were collected in oceanic waters (bottom depth between 900 and 2,300 m) ata distance of 6.5 to 14.8 km off the western (leeward) side of the island of Oahu, Hawaii, with a 3-m Isaacs-Kidd midwater trawl (IKMT) that was continuously open and was equipped with a large cod-end bucket to minimize damage to captured animals. Speci- mens of Pterotrachea spp. were obtained from IKMT tows taken between the surface and 800 m (day), and the surface and 200 m (night) in June 1978, May—June 1981, and December 1981, during cruises of the R/V KANA KEOKI, University of Hawaii. The eyes and visceral nucleus are the prin- cipal morphological features used to separate the species of Pterotrachea (Bonnevie, 1920; Tesch, 1949; Okutani, 1957a; Richter, 1968; Van der Spoel, 1972, 1976). The shapes of these structures (Figs. 1, 2) were character- ized for series of specimens that ranged broadly in size. Within one hr. after capture, (125) 126 SEAPY FIG. 1. Natural swimming posture (lateral view) of a 50.2 mm male Pterotrachea hippocampus collected from Hawaiian waters in June 1981. Measurements: FL = fin length, SL = sucker length, VNL = visceral nucleus length, VNW = visceral nucleus width. specimens were preserved in buffered, 10% seawater-formalin solution. For the eyes, the ratio between eye length and retinal width (Fig. 2) was determined using an ocular mi- crometer scale in a dissection microscope with the eye oriented such that the dorsal surface was at a 90” angle to the plane of view through the microscope. This perspec- tive is important because the eyes are not radially symmetrical around the optical axis. In lateral view, the eye is flattened posterior to the lens and possesses a narrow, strip-like retina (Grenacher, 1886; Hesse, 1900). The ratio of length to maximal width of the visceral nucleus was obtained while viewing each individual from the right side (Fig. 1). Additionally, the fin and sucker lengths (Fig. 1) and the lens diameter (Fig. 2) of specimens that could be assigned to P. hippocampus and P. minuta were measured with the ocular micrometer. Body lengths were determined with vernier calipers (measured to the nearest 0.1 mm) in order that the length to width ratios and the lengths of the fins and suckers could each be compared with respect to the size of the animal. The holotype of Pterotrachea minuta Bon- nevie, 1920 (ZMUB 23259) was obtained from the Zoologisk Museum, University of Bergen, Norway. Body length was de- termined with vernier calipers. Lens diameter, eye length, retinal width, visceral nucleus length and width, sucker length, and fin length of the male specimen were measured with the ocular micrometer. RESULTS Among the four species of Pterotrachea that are currently recognized, two species pairs are immediately apparent on the basis of eye shape (Fig. 3). In broadest outline (dorsal view), the eyes of P. coronata and P. scutata are approximately rectangular in shape (Figs. 3F 4 G). In contrast, the eyes of the holotype of P. minuta (Fig. 3E) and of P. hippocampus/minuta from Hawaii (Figs. 3A to D) have retinas that are broader than the lens, with the result that the eye is roughly triangu- lar. When expressed in terms of the ratio between eye length and retinal width, these differences can be quantified (Fig. 4): at all sizes, the ratio exceeds 1.8 in P. coronata and P. scutata, and is less than 1.6 in P. hippocampus/minuta. Separation of P. coronata from P. scutata in the Hawaiian material is not possible on the PTEROTRACHEA TAXONOMY 127 FIG. 2. Eyes and brain of the Pterotrachea hippocampus from Fig. 1 drawn in dorsal view. Measurements: EL = eye length, LD = lens diameter, RW = retinal width. Structures: CPG = cerebropleural ganglion, DSP = distal screening pigment, OG = optic ganglion, PSP = proximal screening pigment, R = retina, S = statolith. basis of eye morphology, since the eye length to retinal width ratios overlap and range from about 1.8 to 2.5 (Fig. 4). The two species are clearly different, however, on the basis of visceral nucleus shape (Fig. 5). Length to width ratios ranged from 4.1 to 7.2 for P. coronata, while in P. scutata the ratios were less than 4.0 and ranged more narrowly from 2.0 to 3.9. The Hawaiian specimens assignable to P. hippocampus and P. minuta could not be distinguished on the basis of either visceral nucleus or eye morphology. The length to width ratio of the visceral nucleus for all sizes averaged 2.9 and decreased with increasing body size (Fig. 5). The visceral nucleus of the P. minuta holotype had a length to width ratio of 3.2, which is only slightly greater than the 2.9 average for the Hawaiian specimens and is well within the range of obtained ratios (2.1 to 3.9). The ratio of eye length to retinal width decreased sharply with increasing body length (Fig. 4), from 1.6 in small specimens to nearly 1.0 in large individuals. Variation about the regression line for P. hippocampus/ minuta in Fig. 4 was reduced substantially by using corrected body length values based on the lens diameter of each individual rather than on the body length measurements taken with vernier calipers. In another heteropod, Carinaria cristata forma japonica Okutani, 1955, lens diameter was shown (Seapy, 1980) to be an excellent predictor of body length. For P. hippocampus/minuta, then, lens diameter and body length were meas- ured for a series of specimens that did not appear to have lengthened or shortened as a result of capture damage or due to preserva- tion. A clearly-defined line of best fit resulted (Fig. 6), which was used subsequently to estimate the body lengths of the P. hippo- campus/minuta plotted in Figs. 4, 7 and 8. It is noteworthy that little sexual difference could be distinguished in the lens diameter-body length relationship (Fig. 6), which is sub- stantiated by the statistical result that the regressions for the two sexes were not signifi- cantly different (p > .05; F-test for coin- cidental regressions). The eye length to retinal width ratio of the P. minuta holotype (Fig. 3E) was 1.4. The animal was extremely stretched and deflated, probably as a result of net damage during capture, and measured 38.8 mm in length. The lens was only 0.33 mm in diameter, how- ever, which would be characteristic of a 20.5 mm animal from Hawaiian waters (Fig. 6). This latter body length is consistent with 128 SEAPY F FIG. 3. Right eyes of Pterotrachea spp. in dorsal perspective. (A) to (D) P. hippocampus/minuta: Hawaii, 1981, (A) female, 16.2 mm, (B) male, 39.7 mm, (C) male, 50.2 mm and (D) female, 68.7 mm (SBMNH 33884); (E) P. minuta: holotype, male, 38.8 mm (ZMUB 23259); (F) P. coronata: female, 78.4 mm, Hawaii, June 1981 (SBMNH 33885); (G) P. scutata: male, 60.8 mm, Hawaii, June 1981 (SBMNH 33886). The lens in the right eye of the holotype specimen of P. minuta was partially detached, as indicated by the arrow in (E). the maximal size of 25 mm for P. minuta recorded by Tesch (1949) from the “Dana” Expedition material and by Thiriot-Quiévreux (1973) for specimens presumably taken in the general region of the type-locality of the spec- ies (29° 8’ N, 25° 16’ W) off West Africa. Specimens reported from waters off Japan by Okutani (1957a) ranged from 16 to 29 mm in length. The eye length to retinal width ratio for the holotype specimen of P. minuta falls with- in the range of ratios obtained for Hawaiian individuals in the 15 to 25 mm size range (Fig. 4). Two features described by Tesch (1949) and noted later by Van der Spoel (1976) as distinctive for P. minuta were the small sizes of the sucker and fin in males. To determine whether P. hippocampus males could be dis- tinguished from P. minuta males in the Hawaiian fauna, lengths of the suckers and fins were measured for a series of 30 in- dividuals which ranged widely in size and were plotted against body length (Figs. 7 and 8, respectively). Although the relationship shows variability, sucker length increases in an approximately linear fashion as body length increases (Fig. 7). Assuming that the live holotype of P. minuta was 20.5 mm in length, the measured sucker length of 0.40 mm agrees well with those of Hawaiian individuals in this size range (Fig. 7). Fin length increases with body length, but not in a linear manner (Fig. 8), with individuals below about 30 mm in body length exhibiting much greater variability than those larger than 30 тт. As in the case of the sucker, the 5.5 mm fin length of the holotype specimen of P. minuta falls within the range of measure- PTEROTRACHEA TAXONOMY 129 235 A = ©" TD P: tat = P. scutata = o E 2.0 © a E 2 = Pp oronata Е coronat o о o” DS a) Ф > Ww qa o o = oO = с 1.0 Р. hippocampus/minuta 20 40 60 80 100 120 Body Length (mm) FIG. 4. Comparisons of body length (mm) and the ratio of eye length to retinal width for Pterotrachea spp. collected in June 1978 and May—June 1981 from Hawaiian waters. P. scutata, squares (Y = —.0007X + 2.257, r = -.07;p > .05; п = 22), where solid squares are males (Y = .0017X + 2.120; п = 10) and open squares are females (Y = — .0020X + 2.343; п = 12). P. coronata, triangles (Y = —.0058X + 2.518, г = — .79; р < .001; п = 52), where solid triangles are males (Y = — .0071X + 2.560; п = 24) and open triangles are females (Y = —.0057Х + 2.524; п = 28). Р. hippocampus/minuta, circles (Y = —.0090Х + 1.652, r = — .85, p < .001; п = 70), where solid circles are males (Y = — .0099X + 1.653; п = 32) and open circles are females (Y = — .0093X + 1.667; п = 38). The open hexagon denotes the eye length to retinal width ratio for the holotype of P. minuta (body length estimated to be 20.5 mm on the basis of lens diameter). 8.0 o = 5 Е cc = A = f= A A A = A DT 6.0 = % 2 A A a, A = А a A A O ah’ A 4 A АА = A ‘= A Aa aA N A 5 A = 40 A e Ar 2 À A A = = O о u À O 8 O O a L_ вв Р. scutata = © OR |] © O HDP A) - > Г g ES 09) =) a = 2.0 O DO O во ом Р. hippocampus/ a minuta o O D > 20 40 60 80 100 120 Body Length (mm) FIG. 5. Comparisons of body length (mm) and length to width ratios of the visceral nucleus of Pterotrachea spp. collected in June 1978 and May-June 1981 from Hawaiian waters. P. coronata, solid triangles (Y = -0006X + 5.585, r = —.02, р > .05; п = 55). P. scutata, solid squares (Y = .0023X + 2.922, r = .08, p> .05; п = 22). P. hippocampus/minuta, open circles (Y = —.0129X + 3.338, r = —.37, р < .01; n = 74) 130 SEAPY Lens Diameter (mm) T T TT T т Tl т 1 10 20 30 40 50 60 Body Length (mm) T FIG. 6. Relationship between lens diameter (mm) and body length (mm) of Pterotrachea hippo- campus/minuta (Y = .0081X + .164, r = .96, р < .001; n = 48) collected in May—June 1981 from Hawaiian waters. Males (solid circles; n = 19) and females (open circles; n = 24) are plotted sepa- rately, and the lines of best fit (Y = .0086X + .148 and Y = .0075X + .181, respectively) are indicated as long and short dashed lines. Sucker Length (mm) 10 20 30 40 50 Body Length (mm) FIG. 7. Relationship between sucker length (mm) and body length (mm) of Pterotrachea hippocampus/minuta (solid circles; Y = .0321X — 318, г = .90, р < .001; п = 30) collected in May-June 1981 from Hawaiian waters. The open circle denotes the sucker length of the holotype of P. minuta (body length estimated to be 20.5 mm on the basis of lens diameter). ments made on Hawaiian animals of similar body lengths (Fig. 8). The position of the pair of pedal ganglia relative to the insertion point of the anterior edge of the fin has been used by Tesch (1949) and Okutani (1957a) to distinguish P. minuta from P. hippocampus. According to these authors, the pedal ganglia are located anterior to the insertion point (base) of the fin Fin Length (mm) 10 20 30 40 50 Body Length (mm) FIG. 8. Relationship between fin length (mm) and body length (mm) of Pterotrachea hippocampus/ minuta (solid circles; n = 30) collected in May—June 1981 from Hawaiian waters. The open circle de- notes the fin length of the holotype of P. minuta (body length estimated to be 20.5 mm on the basis of lens diameter). in the former species, while in the latter spec- ies the ganglia are positioned just posterior to the insertion point. This difference was also illustrated (Figs. 34a and 49), but not dis- cussed by Bonnevie (1920). A series of 53 specimens of P. hippocampus/minuta from Hawaii were examined for position of the ped- al ganglia. The ganglia were situated just posterior to the insertion point in a single individual (22 mm in length); directly above the insertion point in 21 specimens (12 to 42 mm); slightly anterior in 20 animals (14 to 37 mm); and distinctly anterior in 12 speci- mens (21 to 38 mm). Clearly, there was no size-related pattern to the location of the ped- al ganglia, and the observed variability in position implies that this is not a valid taxonomic criterion. A secondary sexual characteristic that potentially distinguishes species is the copulatory apparatus in males. Van der Spoel (1976) stated that such differences could be of taxonomic utility, and he illustrated the penis and associated structures of P. hippo- campus and P. minuta, although he did not discuss how these structures differed in the species diagnoses. Detailed descriptions of the male reproductive system of P. coronata and P. hippocampus were given earlier by Gabe (1965). Examination of approximately 50 Hawaiian specimens of P. hippocampus/ minuta demonstrated considerable variability, irrespective of size, in the appearance of the copulatory apparatus. Comparisons were also made between the Hawaiian specimens PTEROTRACHEA TAXONOMY 131 and the illustrations of the copulatory appara- tus of P. hippocampus and P. minuta in Van der Spoel (1976: 400, 401) and for P. hippo- campus in Gabe (1965: 1038, 1039) with the general conclusion that the present material most closely resembles the illustrations of P. hippocampus by Gabe. The variability in structure that | observed can probably be attributed to the state of the organ at the time of death and/or to preservation effects. The morphology of the copulatory. apparatus, then, would not appear to be a usable taxonomic character. Radular differences were cited by Bonnevie (1920) to distinguish P. minuta from P. hippo- campus. The heteropods have a taenioglos- san radula (one lateral and two marginal teeth on either side of the central or rachidian tooth in each row) characteristic of the mesogastro- pods (Fretter 8 Graham, 1962). In the genus Pterotrachea, the central tooth is broad and possesses an enlarged median spine with about 5 small spines arranged on both sides (Bonnevie, 1920, textfig. B). Bonnevie consid- ered the central spine to be more pronounced in P. hippocampus than in P. minuta. Thiriot- Quiévreux (1971) also noted a strongly de- veloped central spine in P. hippocampus, but did not state that it was reduced in P. minuta. In fact, the central spine in P. minuta appeared in Thiriot-Quiévreux's scanning electron micrographs to be of comparable size in the two species. Based on the latter information, | would not consider the median spine size to be of taxonomic value. The second radular difference reported by Bonnevie (1920) was that the lateral (or in- termediate) tooth in P. minuta possesses a small secondary spine near its free end. Although Thiriot-Quiévreux (1971) did not de- scribe such a structure, her scanning electron micrograph for P. minuta indicates its pres- ence. The lateral teeth of P. hippocampus in larval, juvenile, and adult forms were illus- trated by Richter (1968: 51) and clearly show a large secondary spine in the larva, its reduc- tion in the juvenile, and its absence in the adult. If P. minuta represents the young of P. hippocampus, as the present study suggests, then the loss of this secondary spine would explain the apparent species difference de- scribed by Bonnevie. Microscopic examina- tions of the radulae of a series of 12 P. hippocampus/minuta between 15 and 40 mm in body length from the Hawaiian material are in close agreement with the observations of P. hippocampus by Richter. It is noteworthy, however, that the secondary spine in the Hawaiian specimens was not lost until a body length of 34 mm was attained. This is signifi- cant because this body length is somewhat greater than the upper size limit of approx- imately 25 to 30 mm cited for P. minuta by Tesch (1949), Okutani (1957a), and Thiriot- Quiévreux (1973). Lastly, Thiriot-Quiévreux (1971) observed a peculiar feature that is shown in her scanning electron micrograph of the radula of P. minu- ta: the fusion of the first and second spines as a bifid spine located immediately adjacent to the protruding middle spine on the central tooth. Similar bifid spines were illustrated by Buchmann (1924: 525) for the central teeth of smali P. coronata. In one specimen of P. coronata, a bifid spine was formed by the two spines immediately lateral to the central spine (as shown for P. minuta by Thiriot- Quiévreux), while for another small P. corona- ta, a bifid spine was formed by the second and third lateral spines. | did not observe any such fused spines in any of my specimens. Since Thiriot-Quiévreux did not comment on the consistency of this feature on adjacent rows or on the central teeth of other radulae, and similar bifid spines have been reported to occur irregularly in another pterotracheid by Buchmann (1924), | question the utility of this minor feature as a taxonomic characteristic. DISCUSSION The morphometric analyses of Pterot- rachea from Hawaiian waters indicate that two of the species, P. coronata and P. scuta- ta, are distinctive in possessing rectangular eyes (Fig. 3F, G) with an eye length to retinal width ratio exceeding 1.8 (Fig. 4). Separation of these two species can be made on the basis of differences in the shape of the viscer- al nucleus, with the length to width ratio ex- ceeding 4.0 in P. coronata and being less than 4.0 for P. scutata (Fig. 5). Additionally, P. scutata can be distinguished by the lateral expansions of the anterior portion of the body that form a gelatinous disc (see plate V in Tesch, 1949, and plate Il in Okutani, 1957a). Separation of the remaining two species of Pterotrachea, P. minuta and P. hippocampus, from the Hawaiian material was not possible. Bonnevie (1920) and subsequent workers (Tesch, 1949; Richter, 1968; Van der Spoel, 1976) considered P. minuta to be in- termediate between P. coronata and P. hip- 132 SEAPY pocampus in terms of visceral nucleus and eye shapes. Perhaps because the visceral nucleus is a much larger structure than the eye and can be measured more easily, both Tesch (1949) and Van der Spoel (1976) used the visceral nucleus in their keys to distin- guish P. minuta (length/width ratio = 3) from P. hippocampus (length/width ratio = 2). However, ratio values for the Hawaiian speci- mens were highly variable and ranged from 2.1 to 3.9 (Fig. 5). This variability could have resulted from differences in gut fullness, pres- ervation effects, and possibly several other factors (see Appendix for elaboration). Nonetheless, it is noteworthy that the length/ width ratio decreased with increasing body size (Fig. 5), such that smaller individuals possessed, on the average, a more elongate visceral nucleus than larger animals; e.g., based on the line of best fit in Figure 5, a 20-mm animal would have a ratio value of 3.2, while that of a 70-mm specimen would be 2.4. In view of the observed variability in shape, the visceral nucleus appears to be useful only in distinguishing species whose nuclei differ greatly, as between P. coronata and the other species of Pterotrachea (Fig. 5). The eye length to retinal width ratios for a wide size range of Hawaiian P. hippocampus, minuta revealed a steeply-sloping, size- specific relationship (Fig. 4), in which the eyes of small individuals (Fig. 3A) strongly resem- bled the holotype of P. minuta (Fig. 3E), and intermediate-to-large specimens (Figs. 3B-D) were comparable to the eyes of P. hippocam- pus photographed by Richter (1968: 370) and illustrated by Bonnevie (1920: 9) and Van der Spoel (1976: 400). Furthermore, the eye length to retinal width ratio for the holotype of P. minuta falls within the range of ratios re- corded from the Hawalian specimens (Fig. 4). Tesch (1949) considered that male P. minuta possessed a sucker and fin that were conspicuously smaller than those of P. hippo- campus. Comparisons of sucker size and fin size with body length (Figs. 7, 8) gave clear, size-specific relationships for the Hawaiian material. The sucker and fin sizes of the holotype of P. minuta proved to be well within the range of the Hawaiian animals of compar- able size, assuming that the true body length of the holotype was close to 20 mm (based оп its lens diameter) and not the measured value of 38.8 mm. The stretched condition of the holotype was almost certainly the result of net damage. In collections off southern California made with 1-m plankton nets and 3-m IKMT nets lacking cod-end devices, | have observed small P. coronata and P. scutata in a similar state. However, large individuals of these two species do not seem prone to this distortion. For the holotype specimen of P. minuta, the sucker and fin would have appeared small in relation to overall body length. Conversely, in large P. hippocampus, which may be less easily stretched, the fin and sucker would have appeared pro- portionately larger. This size difference was probably further exaggerated because the sucker and fin in undistorted males (Figs. 7, 8) appear to increase disproportionately in size with increasing body length. For ex- ample, the sucker on a 15-mm specimen is 1.1% of the body length, while it is 2.6% of the body length for a 50-mm individual (Fig. 7). Similarly, a 15-mm animal has a fin length equal to about 17% of its body length, while the fin length of a 50-mm specimen is about 25% of its body length (Fig. 8). Hypothetically, while the adult morpholog- ies of P. hippocampus and P. minuta could overlap considerably, there might be larval differences that could serve to distinguish the species. One study (Richter, 1968) in- vestigated the larvae of Mediterranean heter- opods and described three types that were distinctively different in shell morphology and which could be assigned to the genus Ptero- trachea. Richter was able to follow develop- ment of the larvae for a maximum of 10 days after metamorphosis. Since the radula, visceral nucleus, and general body morpholo- gy had not differentiated to the point that species could be recognized, Richter based his tentative identifications of the three larval types on eye morphology. The first two larvae possessed eyes which were shaped like those of P. hippocampus and P. minuta, re- spectively, while the third larva had a more tubular eye like that of P. coronata or P. scutata and a unique, enlarged anterior lobe of the four-lobed velum. Richter was cautious, however, in assigning species names, and indicated that he would defer to the results of further developmental studies. In her review paper, Thiriot-Quiévreux (1973) agreed with Richter that three distinctive larval types from the Mediterranean Sea could be assigned to Pterotrachea. However, the larva that Richter considered closest to P. minuta on the basis of eye shape, Thiriot-Quiévreux (1969, 1971) identified as P. coronata. She was apparently PTEROTRACHEA TAXONOMY 133 following the earlier designation of the spec- ies as P. coronata by Franc (1948) on the basis of its possession of about 30 transverse grooves on the shell. Larval differences, then, would appear to be of taxonomic utility, but additional developmental studies of the larvae following metamorphosis are obviously re- quired. Nonetheless, an important result of Richter's (1968) and Thiriot-Quiévreux's (1973) studies was that only three larval types could be distinguished in waters that re- portedly contained all four species of Ptero- trachea. The absence of a fourth larval type may be due to there being only three species of Pterotrachea. In conclusion, | was not able to find any evidence supporting the separation of P. minuta from P. hippocampus on the grounds that the former species has (1) a more elon- gate visceral nucleus, (2) a triangular eye that is narrower, (3) a conspicuously smaller fin and sucker in males, (4) pedal ganglia posi- tioned anterior to the insertion point of the fin, or (5) a radular morphology that includes a more weakly developed median spine on the central tooth and a secondary spine on the lateral tooth. Eye morphology appears to be the most useful taxonomic characteristic in pterotracheids. However, because the size- specific relationship between eye shape and body length reported here for P. hippo- campus/minuta (Fig. 4) was not previously noted, earlier investigators possibly mistook small P. hippocampus for P. minuta. Based on the Hawalian material, P. minuta does not appear to be a valid species and should be treated as a junior synonym of P. hippocam- pus. Studies of the pterotracheid faunas in other regions of the world's oceans are re- quired to substantiate the present results and permit a formal synonymy of P. minuta with P. hippocampus. A broad size range of male and female specimens of P. hippocampus/minuta from the Hawaiian study collection has been de- posited in the National Museum of Natural History (USNM 804410) and the Santa Bar- bara Museum of Natural History (SBMNH 33887). The following key incorporates the above information and represents a modification of earlier keys based on eye and visceral nu- cleus morphology (Tesch, 1949; Okutani, 1957a; Van der Spoel, 1976), with P. minuta omitted. KEY ТО ЭРЕСЕЗ ОЕ THE GENUS PTEROTRACHEA la. Eyes rectangular in dorsal view. Eye length greater than 1.8 times the width of the Tetinallbaser „rar ee 2 1b. Eyes narrowly triangular (small in- dividuals) to broadly triangular (large in- dividuals) in dorsal view. Eye length less than 1.6 times the width of the retinal Базе er P. hippocampus 2a. Visceral nucleus short; length 2 to 4 times the maximal width. Anterior portion of the body expanded laterally as a gelatinouszdiser eee P. scutata 2b. Visceral nucleus elongate; length 4 to 7 times the maximal width. Anterior portion of the body not expanded laterally as a gelatinoussdiscer nee eo P. coronata ACKNOWLEDGMENTS The specimens used in this study were collected during “Fido” cruises of the R/V KANA KEOKI, University of Hawaii. | am par- ticularly grateful to Dr. Richard E. Young, Department of Oceanography, University of Hawaii, for inviting me to participate in these cruises, for provision of laboratory space aboard ship, for stimulating discussions on the virtues of studying heteropods, and for suggestions to improve the manuscript. The text benefitted greatly from the critical reviews of Drs. Kathrina Mangold, James Carlton, and two anonymous reviewers. Measurements of specimens and preparation of the manuscript were completed at the Santa Barbara Museum of Natural History in the laboratory of Dr. F. G. Hochberg. | am most appreciative of Dr. Hochberg for providing laboratory space, providing interpretative assistance in drawing the first three figures, and critically reviewing the manuscript. | am also grateful to Ms. Laurie Marx of the Santa Barbara Museum of Natural History, who prepared most of the illustrations. Finally, | thank Mr. Audun Fos- shagen of the Zoologisk Museum, Bergen, Norway, for the loan of the holotype of Pterot- rachea minuta. LITERATURE CITED BONNEVIE, K., 1920, Heteropoda. Report on the Scientific Results of the “Michael Sars” North 134 SEAPY Atlantic Deep-Sea Expedition 1910, 3(2)(Zoolo- gy): 3-16, 5 pl. | BUCHMANN, W., 1924, Uber den Pharynx der Heteropoden. Zeitschrift fur Anatomie und En- twicklungsgeschichte, 73: 501-540. DALES, R.P., 1953, The distribution of some heter- opod molluscs off the Pacific coast of North America. Proceedings of the Zoological Society of London, 122(4): 1007-1015. FRANC, A., 1948, Véligeres et Mollusques Gastér- opodes des Baies d'Alger et de Banyuls. Journal de Conchyliologie, 88: 13-35. FRETTER, V. & GRAHAM, A., 1962, British pro- sobranch molluscs; their functional anatomy and ecology. Ray Society, London, 755 p. GABE, M., 1965, Données morphologiques et his- tologiques sur l'appareil génital male des Hétér- opodes (Gastéropodes Prosobranches). Zeits- chrift fur Morphologie und Okologie der Tiere, 55: 1024-1079. GRENACHER, H., 1886, Abhandlungen zur ver- gleichenden Anatomie des Auges. Il. Das Auge der Heteropoden, geschildert an Pterotrachea coronata Forsk. Abhandlungen der Naturfors- chenden Gesellschaft zu Halle, 17: 64 p., 2 pl. HESSE, R., 1900, Untersuchungen Uber die Organe der Lichtempfindung bei niederen Thieren. VI. Die Augen einiger Mollusken. Zeits- chrit fur Wissenschaftliche Zoologie, 68: 379- 477, pl. 25-32. McGOWAN, J. A., 1967, Distributional atlas of pelagic molluscs in the California Current region. California Cooperative Oceanic Fisheries In- vestigations, Atlas No. 6, 218 p. OKUTANI, T., 1957a, On pterotrachean fauna in Japanese waters. Bulletin of the Tokai Regional Fisheries Research Laboratory, 16: 15-21, 3 pl. OKUTANI, T., 1957b, Holoplanktonic Gastropoda in the “Kuroshio” area, south of Honshu, May 1955. Records of Oceanographic Work in Japan, New Ser., Special Number, March 1957, p. 134— 142. RICHTER, G., 1968, Heteropoden und Heter- opodenlarven im Oberfláchenplankton des Golfs von Neapel. Pubblicazioni della Stazione Zoolo- gica di Napoli, 36: 346-400. RICHTER, G., 1974, Die Heteropoden der “Meteor”-Expedition in den Indischen Ozean, 1964/65. “Meteor” Forschung-Ergebnisse, (D), 17: 55-78. SEAPY, R. R., 1980, Predation by the epipelagic heteropod mollusk Carinaria cristata forma japo- nica. Marine Biology, 60: 137-146. SPOEL, S. van DER, 1972, Notes on the identifica- tion and speciation of Heteropoda (Gastropoda). Zoologische Mededeelingen Rijksmuseum van Natuurlijke Historie te Leiden, 47: 545-560. SPOEL, S. VAN DER, 1976, Pseudothecosomata, Gymnosomata and Heteropoda (Gastropoda). Bohn, Scheltema & Holkema, Utrecht, 484 p. TESCH, J. J., 1949, Heteropoda. Dana-Report, 34: SAP эре. THIRIOT-QUIEVREUX, C., 1969, Charactéristi- ques morphologiques des véligeres planctoni- ques de Gastéropodes de la région de Banyuls- sur-Mer. Vie et Milieu, 20(2B): 333-366. THIRIOT-QUIEVREUX, C., 1971, Contribution a l'étude de Гогдаподепезе des Heteropodes (Mollusca, Prosobranchia). Zeitschrift fúr Mor- phologie und Okologie der Tiere, 69: 363-384. THIRIOT-QUIEVREUX, C., 1973, Heteropoda. Oceanography and Marine Biology, an Annual Review, 11: 237-261. APPENDIX The length to width ratios for the visceral nucleus were highly variable for the examined species of Pterotrachea (Fig. 5), and particu- larly for P. coronata. Hypothetically, this var- ¡ability is the result of: (1) differences between individual animals in the fullness of the di- gestive system portion contained within the visceral nucleus at the time of preservation, (2) changes resulting from placing individuals in preservative solution, (3) differences be- tween individuals in the degree of gonadal development, and (4) morphological changes in the basic shape of the nucleus as related to age (size) of the individual. During a cruise of the R/V KANA KEOKI in December 1981, data were obtained that ad- dressed the first two hypotheses using P. hippocampus/minuta. Before proceeding, however, several points should be made rela- tive to the third and fourth hypotheses. Although | have noted qualitatively that the width (thickness) of the visceral nucleus is greater in specimens of P. hippocampus/ minuta that have a well-developed gonad, this relationship was not quantified. Enlarged gonads are particularly evident in large speci- mens and could explain the somewhat lower length to width ratios recorded for larger P. hippocampus/minuta and, as a result, the slightly negative slope of the regression line in Fig. 5. Because of variations in the shape of the visceral nucleus that appear to result from the first three factors hypothesized above, the fourth factor, change in shape as a function of age of the individual, cannot be addressed at the present time. The effect of gut fullness on visceral nu- cleus shape was tested by isolating freshly- collected specimens in finger bowls. The selected individuals appeared healthy and undamaged by trawl capture. The length and width of the nucleus were measured through a dissecting microscope with an ocular mi- crometer at varied intervals for up to 2.5 to 7.1 PTEROTRACHEA TAXONOMY 135 = 0 E ee - CE = al NS $ 2 = ES - E . ur Chaar =3 N = € O E 4 > = => о . 6 . Е E +4 E | . vo. 0 . == . o <— e-2-—— —— = FE . = E = ae | = A . . о = -.2 e . . o el = . —e OUT = = 0 fe CES AR ee = = h : x = . о SL _ zus? = e EE . o = -3 ы o © € = o 7-4 a 5 o à . IT 177 —— 0 1 2 3 4 5 6 if Time (hrs) FIG. 9. Change in length (mm), width (mm) and length to width ratio of the visceral nucleus as a function of time in Pterotrachea hippocampus/ minuta collected from Hawaiian waters in Decem- ber 1981. hrs. Fecal strings were produced that com- monly remained attached to the animal. Thus, it was possible to determine which specimens were eliminating fecal matter. In 8 of 14 an- imals tested, fecal strings were produced which varied in thickness (not quantified) and in length (from 4 to 18mm). The results of these 8 trials are summarized in Fig. 9. In all cases, length of the nucleus decreased with time. In 6 of the trials, the width also de- creased with time, but to a much lesser de- gree. The length to width ratio clearly de- creased with time in 5 of the 8 trials, a result that can be attributed to the relatively greater decrease in nucleus length. It is noteworthy that the excretion of large amounts of wastes by P. coronata was observed on two occa- Change in Length (mm) e e a e . 1 © 8 60 . . 80 > ЕЕ ЕЕ 40 À one 20 . . | Dre р . . . .: 8 0 o 0.0 . . 1 A IE OS =:20 E . Se 20 (er . 2 0.0 ’”. os 20 . e e SS “о 3 51° e == 40 o H u a . . > o = 60 ps oo 8 S 5 7-80 р A = ke) = 2 20 30 40 50 60 Body Length (mm) FIG. 10. Change in length (mm), width (mm), and length to width ratio of the visceral nucleus after preservation in Pterotrachea hippocampus/minuta collected from Hawaiian waters in December 1981. sions to result in a dramatic shortening of the nucleus and therefore a substantial reduction in the length to width ratio. The effect of preservation on shape of the visceral nucleus was examined by measuring the length and width of the nuclei from a series of 32 freshly-captured and apparently healthy specimens (ranging in length from 22.0 to 55.6 mm) before and after preserva- tion in a buffered, 10% seawater-formalin solution (Fig. 10). In 29 of the trials, the length decreased following preservation, while the width increased in 30 cases. With only two exceptions, the length to width ratio de- creased after preservation. Preservation re- sulted in contraction of the nucleus in nearly all of the trials and, because of simultaneous increases in width, the length to width ratio decreased. These results imply that speci- mens preserved while still alive would exhibit a lower length to width ratio than those that had died before preservation. MS AN у р vos 301 ра eo 3 2 CT NN + 5 > oy р se a À | a wid) seein as ae n° Oe $ See My IM TT or à ae: LS MALACOLOGIA, 1985, 26(1-2): 137-143 TAXONOMIE EXPERIMENTALE DE BIOMPHALARIA (GASTROPODA: ‚PLANORBIDAE)—III. MOBILITES ENZYMATIQUES CONSIDEREES COMME ELEMENTS DE DIAGNOSTIC POUR LES BIOMPHALARIA ANTILLAIS. ETUDE DE SEPT SYSTEMES ENZYMATIQUES Jens Erik Jelnes'* & Jean-Pierre Pointier? RESUME Une analyse électrophorétique de sept systemes d'enzymes chez Biomphalaria glabrata, B. straminea, B. schrammi, B. havanensis et Biomphalaria sp. a nettement révélé des differences de mobilités enzymatiques entre les especes. Des caractères ont été considérés comme éléments de diagnostic pour les trois premières espèces originaires de la Guadeloupe et de la Martinique. Les informations obtenues suggèrent d'autre part, que les caracteres enzymatiques peuvent également étre utiles pour les dé- terminations spécifiques dans d'autres iles des Antilles. Mots clés: Les Antilles; Biomphalaria; caracteres diagnostiques; Hbdh; Pgi; Pgm; Got. INTRODUCTION La faune malacologique dulcaquicole des departements francais de la Guadeloupe et de la Martinique est actuellement bien con- nue. En Guadeloupe, des études détaillées ont été entreprises sur la taxonomie et la distribution des Mollusques d'eau douce (Pointier, 1974, 1976), tandis que des re- cherches similaires étaient réalisées en Marti- nique (Guyard & Pointier, 1979). Toutes ces études ont été essentiellement motivées par le fait que des espèces appartenant au genre Biomphalaria jouent le rôle d'hôte in- termédiaire de Schistosoma mansoni Sam- bon, 1907, agent de la schistosomose in- testinale dans ces îles. En Guadeloupe et Martinique, trois es- pèces seulement de Biomphalaria sont actuellement reconnues: B. glabrata (Say, 1818—localité type: Guadeloupe), B. schram- mi (Crosse, 1864—localité type: Guadeloupe) et B. straminea (Dunker, 1848—localite type: Amérique du Sud). Dans les Grandes Antilles, des inventaires malacologiques ont été établis principalement en Haiti (Robart et al., 1977) et a Porto-Rico (Ferguson & Richards, 1963; Harry & Huben- dick, 1964). Dans les Petites Antilles, la repartition de la schistosomose intestinale et de ses hótes intermédiaires a été revue récemment par Prentice (1980) et dans de nombreux cas, la présence de certaines especes de Biompha- laria п’а pu étre établie avec certitude par suite de déterminations douteuses. Ces problemes posés par l'identification des espèces de Mollusques hôtes inter- médiaires ont amené à rechercher de nouveaux critères de determination. L'utilisa- tion des méthodes électrophorétiques en- zymatiques s'inscrit parfaitement dans le cadre de ces recherches. Les résultats d'une étude préliminaire réalisée sur huit enzymes chez des especes de Biomphalaria américains, ont déja montré que les caractéristiques de ces enzymes pouvaient étre tres utiles pour l'identification et donc la révision de ce genre dans la zone Néotropicale (Jelnes, 1982). Nous pré- sentons dans cet article les résultats obtenus par l'analyse de sept loci enzymatiques con- cernant 23 échantillons de populations de Biomphalaria antillais. Les especes étudiées sont B. glabrata, B. schrammi, B. straminea, B. havanensis (Pfeiffer, 1839—localité type: La Havane, Cuba), et Biomphalaria sp. dont l'identification pose un probleme mais qui pourrait correspondre a B. albicans (Pfeiffer, 1839—localité type: Cuba), В. pallida (Adams, 1846—localité type: Jamaique) ou B. obstructa Morelet, 1849—localité type: Île de Carmen, Mexique). ‘Danish Bilharziasis Laboratory, Jaegersborg Alle 1 d, DK-2920 Charlottenlund, Danemark. “Present address: Thyboron Alle 82, DK-2720 Vanloese, Denmark. “Laboratoire de Biologie Marine et Malacologie, Ecole Pratique des Hautes Etudes, 55 Rue de Buffon, 75005 Paris, France. (137) 138 JELNES & POINTIER MATÉRIEL ET MÉTHODES L'origine des souches étudiées est donnée Tableau 1. Toutes les souches sont con- sidérées comme “sauvages” selon le critere utilisé par Jelnes (1982) exception faite de l'échantillon no. 6 qui correspond a une 2e génération de laboratoire. Les techniques utilisées pour la préparation des échantillons, l'électrophorese sur gel d'amidon, et la révélation des enzymes ont deja été décrites en detail (Henriksen & Jelnes, 1980; Jelnes, 1982). Pour chaque individu de l'échantillon étudié, les enzymes suivants ont ete analyses: 3-hydroxybutyrate déshydrogénase (EC 1.1.1.30), phosphoglu- cose isomérase (EC 5.3.1.9), isocitrate dés- hydrogénase (EC 1.1.1.41), alpha-glycéro- phosphate déshydrogénase (EC 1.1.1.8), glutamate-oxaloacetate transaminase (EC 2.6.1.1), mannose-6-phosphate isomerase (EC 5.3.1.8) et phosphoglucomutase (EC 2.7.5.1). Les esterases n'ont pas été étudiées car elles donnent des schemas extremement complexes qui se révelent par conséquent peu utilisables pour les identifications des especes. TABLEAU 1. Origine des souches analysées. RESULTATS Sept enzymes representant 7 loci gén- etiques ont été analysees par électrophorese sur un total de 269 Mollusques. Les mobilités des differentes bandes obtenues par les dif- ferentes enzymes sont presentes Tableau 2. Des analyses supplémentaires, non pré- sentées Tableau 2, ont été effectuées pour comparer la position des bandes dans les cas ou il a été nécessaire de savoir si deux en- zymes avaient une mobilité identique ou non. Des photographies de zymogrammes repré- sentatifs de Biomphalaria ont été présentés dans un précédent article (Henriksen & Jelnes, 1980). 3-hydroxybutyrate déshydrogénase (Hbdh). Un total de six mobilités differentes ont été observées pour cette enzyme. Les deux plus rapides (1,34 et 1,20) caractérisent B. glabrata tandis que 1,00 est caractéristique de B. havanensis, 0,74 de Biomphalaria sp., 0,93 de B. schrammi et 0,47 de B. straminea. Pour la population no. 2 (B. glabrata) une variation d'allozymes est observée et les phé- notypes suivants ont été trouvés: Hbdh-1,20 No Especes lles Localités No. de référence a DBL 1 В. glabrata Sainte-Lucie Marécage a Soufriere 80/257 2 В. glabrata Guadeloupe Mare des Grands Fonds 80/259 + 81/1 3 В. glabrata Guadeloupe Mare de Céligny 81/135 4 В. glabrata Guadeloupe Mare de Céligny 81/201 5 В. glabrata Guadeloupe Mare de Tombeau 81/202 6 В. glabrata Martinique Marécage du Quartier Boisneuf 81/248 7 В. glabrata Martinique Marécage de l'Anse Riviere 81/266 8 В. glabrata Martinique Riviere de Pointe La Mare 81/268 9 В. glabrata Porto Rico Humacao 81/234 10 В. glabrata Hispaniola Santo-Domingo, jardin botanique 81/249 11 В. straminea Martinique Marécage du Quartier Boisneuf 80/249 12 В. straminea Martinique Canal de Sainte-Marie 81/134 13 В. straminea Martinique Riviere Epinette a Trinité 81/170 14 В. straminea Martinique Canal de Sainte-Marie 81/171 15 В. straminea Martinique Marécage du Quartier Boisneuf 81/182 16 В. straminea Martinique Riviere de Pointe La Mare 81/267 17 В. schrammi Guadeloupe Mare de Tombeau 81/186 18 В. schrammi Guadeloupe Etang Cocoyer 81/187 19 В. schrammi Guadeloupe Mare de Céligny 81/200 20 B. schrammi Guadeloupe Mare de Tombeau 81/253 21 B. schrammi Guadeloupe Mare de Céligny 81/258 22 B. havanensis Hispaniola Haiti 81/250 23 Biomphalaria sp. Hispaniola Santo-Domingo, jardin botanique 81/251 BIOMPHALARIA ANTILLAIS: L'ELECTROPHORESE 139 TABLEAU 2. Mobilités enzymatiques observées chez des Biomphalaria antillais. Les phénotypes sont exprimés par les valeurs des mobilites enzymatiques par rapport a la souche de référence de Biomphalaria camerunensis originaire de Kinshasa, Zaire. A / la séparation des valeurs de rm indique le polymorphisme génétique. A + la séparation des valeurs de rm indique que les deux bandes sont trouvées chez tous les individus analysés. Mobilites enzymatiques observées No. de Nombre d'individus souche analyses Hbdh Ра! Idh Gpdh Got Mpi Pgm 1 9 1,34 1,18 = 1,08 0,50 0,92 071-085 2 8 1,20/1,34 1,18 0,92 1,08 050 0,92 0,71+0,85 3 14 1,34 1,18 0,97 1,08 0,50 092 0,71-0,85 4 7 1,34 118 0,97 1,08 0,50 092 0,71+0,85 5 2 1,34 1,18 0,97 (OBS ° 0,507 10192 0/71 0:85 6 8 1,34 1,18 0,97 1,08 0,50 0,92 0,71+0,85 Y 12 1,34 1,18 0,97 1,08 0,50 0,92 0,71 +0,85 8 24 1,34 1,18 0,97 1,08 0,50 0,92 0,71+0,85 9 40 1,34 1,18 0,97 1,08 0,50 0,92 0,71+0,85 10 7 1,34 1,18 1,01 1,08 0,50 0,92 0,71 +0,85 11 17 0,47 1,34 0,88 1,08 0,50 0,92 « 0,85 +1,00 12 11 0,47 1,34 0,88 1,08 050 0,92 0,85+1,00 13 19 0,47 1,34 0,88 1,08 050 092 0,85+1,00 14 16 0,47 1,34 0,88 1,08 050 0,92 -0,85+1,00 15 3 0,47 1,34 0,88 1,08 0,50 0,92 0,85+1,00 16 17 0,47 1,34 0,88/0,92 1,08 050 092 0,85+1,00 17 4 0,93 1,00 0,83 0,91 1,00. 0576. 0/71 0,85 18 3 0,93 1,00 0,83 0,91 100 0761100185 19 1 0,93 1,00 0,83 0,91 1,00 0,76 0,71+0,85 20 28 0,93 1,00 0,83 0,91 1,00 0,76 0,71+0,85 21 6 0,93 1,00 0,83 0,91 1,00 07600,35 22 5 1,00 1,18 0,92 1,08 1,00 0,92 0,85-1,00 23 8 0,74 1,18 0,97 1,08 OOO IS 1.001 (4 spécimens) et Hbdh-1,20/1,34 (4 spéci- mens). Alpha-glycérophosphate déshydrogénase (a-Gpdh). Deux mobilités seulement peuvent étre révélées: 0,91 chez B. schrammi et 1,08 ат chez les autres especes étudiées. cette enzyme, trois mobilités differentes sont observées: 1,18 chez Biomphalaria sp., B. Glutamate-oxaloacétate transaminase glabrata et B. havanensis, 1,34 chez B. stra- minea et 1,00 chez В. schrammi. || n'a été noté aucun cas de variation d'allozymes. Isocitrate déshydrogénase (Idh). Cinq mobilités differentes ont été observées: 1,01 chez une population de B. glabrata, 0,97 chez B. glabrata et Biomphalaria sp., 0,92 chez une population de chacune des trois especes B. glabrata, B. straminea et B. havanensis. La mobilité de 0,88 a été trouvée seulement chez B. straminea et 0,83 semble caractéris- tique de B. schrammi. Pour la population no. 16 (B. straminea), on peut noter une variation d'allozymes représentée par les alleles Idh- 0,88 et Idh-0,92. Treize spécimens du phé- notype Idh-0,88 et 4 spécimens du phénotype Idh-0,88/0,92 ont été observés. (Got). Pour cette enzyme deux mobilités ont également été notées: 0,50 chez B. glabrata et B. straminea et 1,00 chez Biomphalaria sp., B. schrammi et B. havanensis. Mannose-6-phosphate isomerase (Mpi). Trois mobilités ont été observées. B. schram- mi parait caractérisé par une mobilité de 0,76, tandis que chez B. glabrata, B. straminea et B. havanensis, celle-ci est de 0,92. Biompha- laria sp. est caractérisé par 0,98. Phosphoglucomutase (Pgm). A la révéla- tion enzymatique tous les individus pré- sentent deux bandes. Pour Pgm les mobilités de 0,71 et a 0,85 ont été révélées chez B. glabrata et B. schrammi, tandis qu'elles sont de 0,85 et 1,00 chez B. straminea et B. 140 havanensis et de 1,00 et 1,13 chez Biompha- laria sp. Aucun cas de variation d'allozymes n'est rapporté. DISCUSSION Jelnes (1982) a montré que les modeles enzymatiques d'autres échantillons de Biom- phalaria américains sont reproductibles et in- dépendants de l'áge, de la taille, de la nutri- tion des Mollusques et de la qualité de l'eau, si les méthodes décrites sont employées. Ceci est encore le cas pour cette étude, quoique deux enzymes (Mpi et Idh) se réve- lent si faiblement qu'il est quelque peu difficile d'établir leurs phénotypes. Pour cette raison, aucune donnée sur l'enzyme Idh n'a été rap- portée pour la population no. 1 et les valeurs de rm quelque peu variables de cette enzyme pour la population no. 10 peuvent étre dues a une incertitude dans l'établissement des phé- notypes. Une population de B. glabrata originaire de Belo Horizonte, Brésil (Jelnes, 1982), pré- sente une difference sur un seul enzyme (Hbdh-1,20) par rapport aux populations de B. glabrata antillais étudiés ici. Cette grande similitude des reponses enzymatiques d'échantillons d'origine tres differente peut être l'indice d'une faible variation géographi- JELNES & POINTIER que de l'espece. Au contraire de B. glabrata, B. straminea montre un plus grand degré de variation géographique. La population étu- diée par Jelnes (1982) et originaire d'Améri- que du Sud differe de la souche martiniquaise par des mobilités differentes des trois en- zymes: Pgi, Got et Pgm. La valeur des caractéristiques enzymatiques comme élément de diagnostic Des mobilités identiques ne furent trouvees pour aucune enzyme des cinq especes étu- diées (Tableau 2). Le Tableau 3 présente les nombres d'enzymes pour lesquelles les mo- bilités enzymatiques peuvent étre con- sidérées comme éléments de diagnostic. On y voit que le nombre minimum de bandes enzymatiques “diagnostic” est de trois. Toutes les cinq especes étudiées présentent des mobilités differentes en ce qui concerne Гепгуте Hbdh. Pour Pgi les espèces de Mar- tinique et de Guadeloupe montrent égale- ment des mobilités enzymatiques différentes. Le Tableau 4 présente quelques mobilités enzymatiques qui peuvent étre utilisées pour l'identification des espèces de Biomphalaria antillais. Comme les techniques utilisées per- mettent la révélation de deux enzymes a par- TABLEAU 3. Nombre d'enzymes utilisables comme critere d'identification spécifique pour des combinations deux par deux des especes de Biomphalaria. B. glabrata B. straminea B. schrammi B. havanensis B. straminea 4 B. schrammi 6 Y B. havanensis 3 3 6 Biomphalaria sp. 4 6 6 4 TABLEAU 4. Quelques valeurs rm, utiles pour l'identification de Biomphalaria sp. de la région Caraïbe. A / la séparation des valeurs de rm indique le polymorphisme génétique. Pour les allozymes, les mobilités rares sont mises entre parentheses. A + la séparation des valeurs de rm indique que les deux bandes sont trouvées chez tous les individus analysées. Hbdh Pgi Pgm Got Idh B. glabrata 1,34/(1,20) 1,18 0,71+0,85 0,50 (1,01)/0,97/(0,92) B. straminea 0,47 1,34 0,85 + 1,00 0,50 (0,92)/0,88 В. schrammi 0,93 1,00 0,71 +0,85 1,00 0,83 B. havanensis 1,00 1,18 0,85 + 1,00 1,00 0,92 Biomphalaria sp. 0,74 1,18 1100113 1,00 0,97 BIOMPHALARIA ANTILLAIS: LELECTROPHORESE 141 tir d'une seule électrophorese, l'utilisation des enzymes Pgi et Hbdh sera donc recomman- dée pour les déterminations spécifiques. L'enzyme Pgi est visible sur la plaque environ cing a dix minutes apres l'application du ré- vélateur tandis que pour Hbdh, la durée de révélation est de une a deux heures. Donc, si l'on considere que la mobilité de Гепгуте Ра! est un critere d'identification valable (comme c'est le cas pour les échantillons guadelou- péens et martiniquais), il est possible de véri- fier les déterminations gráce aux valeurs des rm de Hbdh. Il convient cependant d'émettre deux ré- serves si Гоп veut extrapoler les résultats a d'autres régions que la Guadeloupe et la Mar- tinique. La premiere est que cing autres es- peces de Biomphalaria sont signalées dans les Antilles dont les profiles enzymatiques ne sont pas connus (voir ci-apres). La deuxieme est qu'il existe une certaine variation géog- raphique des enzymes chez les Biomphalaria comme l'ont déjà montré les résultats ex- posés. Les mobilités enzymatiques présentées dans le Tableau 2 sont toutes exprimées par rapport a la mobilité de la méme enzyme fonctionelle chez Biomphalaria camerunensis (Boettger, 1941) originaire de Kinshasa, Zaire et maintenu en élevage au Laboratoire Danois de Bilharziose. ll est évidemment criticable d'utiliser une espece africaine de Biomphalaria comme témoin pour des études réalisées sur des Biomphalaria en Amérique. N'importe quelle souche ou population de Biomphalaria américain pourrait en principe servir de matériel de référence. ll est cepen- dant préférable d'utiliser une espece qui ne présente pas de grande variation géograph- ique et il y a un avantage supplémentaire a choisir une population qui ne révele pas de variation génétique dans les enzymes étu- diées. De toutes les especes analysées dans ce travail. il semble que B. glabrata remplisse le mieu, „es conditions. D'autre part cette espèce a l'avantage supplémentaire de pou- voir étre facilement maintenue en élevage dans les conditions du laboratoire. A partir des mobilités enzymatiques ex- primées par rapport a В. camerunensis il est possible de calculer les mobilités par rapport a toute autre souche qui serait choisie comme référence, a condition que les valeurs de la nouvelle souche de référence soient connues par rapport a celles de B. camerunensis. Pratiquement, il suffit de diviser les valeurs de rm de l'échantillon par rapport a B. camer- unensis, par les valeurs de rm de la nouvelle souche de référence par rapport a B. came- runensis. Comme les valeurs de rm varient lorsque la souche de référence change, il est evidemment tres important de préciser la souche de référence qui est utilisée. D'autre part il est souhaitable de comparer les mobil- ités enzymatiques des differentes souches de référence utilisées dans les différentes labor- atoires. En dehors des especes étudiées dans cet article, d'autres Biomphalaria ont été signalés dans les Antilles. ll s’agit de В. helophila (d'Orbigny, 1835—localité type: Callao, Pé- rou), В. peregrina (d'Orbigny, 1835—localité type: Patagonie, Argentine), B. obstructa, B. albicans et B. pallida. Une des especes analysées dans ce travail sous la dénomination de Biomphalaria sp. pose un probleme de détermination. Du point de vue morphologique, elle apparait assez proche de B. havanensis bien que la coquille presente quelques differences. Du point de vue enzymatique les deux formes présentent des differences portant sur 4 des 7 enzymes étudiées. Nous considérons donc ces deux formes comme deux especes distinctes. Trois especes déja décrites des Antilles pourraient correspondre au Biomphalaria sp. analyse dans ce travail: В. albicans, В. pallida et B. obstructa. Paraense & Ibanez (1964) ont place B. albicans en synonymie avec B. helophila tan- dis que Robart et al. (1977) considerent B. albicans comme une espece distincte. A Por- to Rico, Harry & Hubendick (1964) con- siderent également cette espece comme dis- tincte bien que les coquilles présentent des caractéristiques assez voisines de celles de B. helophila avec notamment une déviation prononcée du dernier tour de spire vers la gauche. Les spécimens analysés ici du point de vue enzymatique sous la dénomination de Biomphalaria sp. ne comportent pas cette déviation vers la gauche du dernier tour de spire bien qu'ils soient adultes. La deuxieme espece, qui nous paraít mieux correspondre a Biomphalaria sp. est B. palli- da. Décrite originellement de la Jamaique, cette espece a également été signalée a Por- to Rico par Harry et Hubendick (1964) qui la considerent comme une espece distincte de B. havanensis et B. albicans. Les caractéris- tiques de la coquille apparaissent proches de celles de l'échantillon de Biomphalaria sp. analysé dans ce travail. B. pallida pourrait donc correspondre a notre espece. 142 JELNES & POINTIER В. obstructa a été décrit par Morelet de l'île de Carmen (Mexique) et est signalé a Porto Rico et Cuba par PAHO (1968). Cette espece pourrait correspondre également du point de vue morphologique a Biomphalaria sp. B. helophila n'a été signalé jusqu'ici que dans l’île de la Barbade. Nous avons vu que ce planorbe est caractérisé par une déviation prononcée du dernier tour de spire vers la gauche comme B. albicans. Enfin, la presence de B. peregrina semble tres douteuse dans les Antilles d'apres les révisions présentées par PAHO (1968) et Prentice (1980). Cette revue des especes de Biomphalaria signalées dans les Antilles fait apparaítre les lacunes qui existent encore concernant la systématique de ce genre. Afin d'arriver a une meilleure compréhension de cette taxonomie, il sera nécessaire d'analyser beaucoup plus d'échantillons de populations d'origines tres diverses. REMERCIEMENTS Cette recherche a été financée par les organismes suivants: PNUD/Banque Mondiale/OMS—Special programme for Re- search and Training in Tropical Diseases et le Conseil National de la Recherche en Sci- ences Naturelles Danois (allocation 11-2313). Nous remercions Madame R. Herk-Hansen pour son assistance technique compétente ainsi que les differentes personnes qui ont fourni le matériel vivant: Dr. М. Vargas de Gomez (échantillons no. 10, 23, 24), Dr. E. R. Tiben (échantillon no. 9) et Dr. A. Théron (échantillon no. 6). TRAVAUX CITÉS FERGUSON, F. F. & RICHARDS, C. S., 1963, Fresh-water mollusks of Puerto Rico and the U.S. Virgin Islands. Transactions of the Amer- ican Microscopical Society, 82: 391-395. GUYARD, A. & POINTIER, J. P., 1979, Faune malacologique dulcaquicole et vecteurs de la Schistosomose intestinale en Martinique. An- nales de Parsitologie (Paris), 54: 193-205. HARRY, H. W. & HUBENDICK, B., 1964, The freshwater pulmonate Mollusca of Puerto Rico. Meddelanden frán Góteborgs Musei Zoologiska Avdeling, 136: 1-77. HENRIKSEN, U. B. 8 JELNES, J. E., 1980, Ex- perimental taxonomy of Biomphalaria (Gastro- poda: Planorbidae)—!. Methods for experimental taxonomic studies on Biomphalaria carried out by horizontal starch gel electrophoresis and staining for twelve enzymes. Journal of Chromatography, 188: 169-176. JELNES, J. E., 1982, Experimental taxonomy of Biomphalaria (Gastropoda: Pulmonata). Il. Electrophoretic observations on eight enzyme systems of the South American species: Biom- phalaria glabrata, B. straminea and B. tena- gophila. Journal of Natural History, 16: 209-217. PAN AMERICAN HEALTH ORGANIZATION, 1968, A guide for the identification of the snail intermediate hosts of schistosomiasis in the Americas. 122 p. х PARAENSE, W. L. & IBANEZ, H., 1964, “Aus- tralorbis helophilus” (Pulmonata, Planorbidae). Revista Brasileira de Biologia, 24: 249-258. POINTIER, J. P., 1974, Faune malacologique dul- caquicole de l'ile de la Guadeloupe (Antilles francaises). Bulletin du Museum National d’His- toire naturelle, Paris, 3° sér., n°235, Zool., 159: 9055933: POINTIER, J. P., 1976, Répartition locale et biogéographie des Mollusques dulgaquicoles de la Guadeloupe (Antilles francaises). Malacologi- cal Review, 9: 85-103. PRENTICE, M. A., 1980, Schistosomiasis and its intermediate hosts in the Lesser Antillean Is- lands of the Caribbean. Bulletin of the Pan Amer- ican Health Organization, 14: 258-268. ROBART, G., MANDAHL-BARTH, G. & RIPERT, C., 1977, Inventaire, repartition géographique et ecologie des mollusques dulcaquicoles d'Haiti (Caraibes). Haliotis, 8: 159-171. BIOMPHALARIA ANTILLAIS: L'ELECTROPHORESE 143 TAXONOMIA EXPERIMENTAL DE BIOMPHALARIA (GASTROPODA: PLANORIBIDAE)—!!. MOVILIDADES ENZIMATICAS COMO CARACTERES DIAGNOSTICOS DE LOS BIOMPHALARIA DE LAS ANTILLAS. UN ESTUDIO DE SIETE SISTEMAS ENZIMATICOS Jens Erik Jelnes & Jean-Pierre Pointier RESUMEN El análisis electrophoretico de siete sistemas enzimaticos que pertenecen a las especies Biomphalaria glabrata, B. straminea, B. schrammi, B. havanensis y Biomphalaria sp. reveló unas diferencias distinctas de movilidad entre las citadas especies. En cuanto a las tres primeras especies, unas propiedades enzimaticas que son útiles para la identificación de la especie, han sido propuestas en material de origen las islas Martinique y Guadeloupe. Han sido conseguidas informaciónes que demuestran la utilidad de las propiedades enzimati- cas en la identificación en otras islas del Caribe. EXPERIMENTAL TAXONOMY OF BIOMPHALARIA (GASTROPODA: PLANORBIDAE)—Ill. ENZYME MOBILITIES AS DIAGNOSTIC CHARACTERS FOR ANTILLEAN BIOMPHALARIA. A STUDY OF SEVEN ENZYME SYSTEMS Jens Erik Jelnes 4 Jean-Pierre Pointier SUMMARY Electrophoretic analysis of seven enzyme systems of Biomphalaria glabrata, B. straminea, B. schrammi, B. havanensis and Biomphalaria sp. revealed clear differences in mobility between the species. For the three first mentioned species enzyme characters useful for identification to species have been suggested for material originating from the islands of Martinique and Guadeloupe. Data have been obtained that suggest enzyme characters as useful for species identification on other Caribbean islands. MALACOLOGIA, 1985, 26(1-2): 145-151 A NEW PIGMENTATION MUTANT IN BIOMPHALARIA GLABRATA Charles $. Richards! Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20205, U.S.A. Biomedical Research Institute, 12111 Parklawn Drive, Rockville, MD 20852, U.S.A. ABSTRACT Three alleles of a gene regulating mantle pigmentation in Biomphalaria glabrata are de- scribed: pigment coalescence, discrete spotting, and absence of black mantle pigment. This gene is apparently located on a different chromosome than the gene with three alleles determining basic pigmentation (wildtype, blackeye, and albino). Combinations of these two pigmentation characters result in nine different pigment phenotypes. Key words: Biomphalaria glabrata; pigmentation; genetics; mutation. INTRODUCTION Biomphalaria glabrata (Say) is the major intermediate host for Schistosoma mansoni in the Western Hemisphere. It has been demon- strated that genetics plays an important role in the host-parasite relationship, determining variations in susceptibility of B. glabrata for infection by S. mansoni (Newton, 1953; Richards & Merritt, 1972; Richards, 1973a, 1975, 1977). Pigmentation served as a genet- ic marker in these studies. Newton (1954) demonstrated that albinism in B. glabrata is a simple recessive character. Richards (1967) reported a third allele of the same gene, “blackeye,” dominant over albino but reces- sive to black wildtype pigmentation. In wild- type pigmentation (C), snails have black eyes, black pigment in connective tissue of headfoot and mantle collar and in the mantle epithelium in the form of spots (Richards, 1969). As snails age, the mantle often be- comes all black. In blackeye pigmentation (c?), snails have black eyes and black pig- ment in the mantle epithelium in the form of spots, but are deficient in black headfoot and mantle collar pigment. Albino snails (c) lack black pigment. Black mantle pigmentation of wildtype and blackeye B. glabrata occurs in variable degrees and typically in the form of spots (Richards, 1969, 1973a). Richards (1969) developed relatively “unspotted” B. glabrata by selection for minimal spotting through several generations and concluded from crossing experiments that spotting was dominant over lack of spotting. Although additional morphological charac- ters showing single factor inheritance have been reported (Richards, 1972, 1973b, 1974a, 1974b, 1980), no linkage has been demonstrated with factors determining varia- tion in susceptibility to S. mansoni infection. Additional simple morphological characters are therefore needed. Studies on mantle pigmentation reported here add to our knowl- edge of the complex variations in mantle pig- ment and provide additional good genetic markers (Richards, 1984). MATERIALS AND METHODS Methods of maintaining snails and crossing procedures were described by Richards 8 Merritt (1972). The origins of the snail lines involved in these studies were indicated in a diagram by Richards (1975). Snails were reared in isolation, reproducing by selfing, and each digit in the numerical identification represents an individual snail in each suc- ceeding generation. Wildtype Puerto Rican B. glabrata lacking black mantle pigment spots were provided by Mr. Walter Stewart (NIAMDD, NIH). He observed snails with and without mantle spotting and demonstrated that spotting was “Current address: P.O. Box 30233, Bethesda, MD 20814, U.S.A. (145) 146 RICHARDS dominant over lack of spotting (personal com- munication). Random segregation of codomi- nant alleles at an autosomal locus of a selfed heterozygote should produce ratios of 1:2:1 among the Fs progeny. For dominant traits, a ratio of 3:1 is expected. A Х? statistic was calculated to test the fit of the data to Mendel- ian expectations and a X* test for homogene- ity was calculated to determine consistency among families. A contingency X? statistic was Calculated to test for independent assort- ment of the two loci contributing to the pig- ment phenotype. RESULTS Mantle pigment variations described in this paper are determined by three alleles of a different gene from that determining albinism. These alleles are designated as follows: 53 = coalesced mantle pigment, S = discrete man- tle spotting, s = absence of these mantle pigment expressions (Figs. 1-9). Alleles S° and S are codominant, and in heterozygous SS juvenile or young adult snails (<10 mm), this genotype is indicated by a combination of spots and coalescence. In older S°S snails, coalescence may obscure the spots. Before the interaction of the mantle pigment alleles was Clarified, both S°S° and SS snails were scored as coalesced (Figs. 10, 11). Combinations of the two pigment charac- ters (basic pigmentation: С = wildtype, с? = blackeye, с = albino; mantle pigmentation: S° = coalescence, S = discrete spotting, S°S = codominant pigmentation, s = absence of mantle pigmentation) result in nine different phenotypes (Fig. 1-9). Crossing experiments that demonstrated the inheritance follow. Offspring of spotted blackeye B. glabrata 1-3-1-4-2-10-1 (c°c°SS) by selfing included 19 (c°c°SS) with discrete mantle spotting and 6 (c°c°SS) mutants with coalesced mantle pigment (Fig. 10). The observed progenies of the 19 c®c’SS F,s by selfing totaled 155, all c°c’SS. Observed progenies of the 6 c°c?S1s F,s by selfing totaled 77 c’c°(S°S? + 99$): 3366350 = 1.1684 Pr 020) Several crosses were carried out involving the 6 mutant F;s or their descendants. Some of these are depicted in Figs. 11 and 12. Mutant snail 1-3-1-4-2-10-1-4 produced off- spring by selfing in the ratio 20 ce? (S°S? + 5:5): 8215552 = 01047 1di — 110184) (Fig. 11). This snail was mated with an albino cc(SS), resulting in postcross hybrid offspring in the ratio 12 c°cS4S : 12 cross pe = 0.0042, df = 1, P = 0.84). Eight of the c’cSS F,s, reared in isolation and reproduced by selfing, produced 134 Fos: 65 (c?c? + c°c)(S°S? + $95) : 30 (cPc® + cPc)SS : 39 cc (S°S* + 595 +155) : (02 = 3.262 dae = 0.04). Four of the c?cSS F,s produced by selfing: 55 (c°c° + c°c)SS : 21 сс (SS) (X= = 0.155, df = 1, P = 0.64). An F; albino (No. 5) from one of the с?с59$ Fıs (No. 7) was first selfed and then mated with a spotted mantle blackeye snail (c°c?SS) (Fig. 11). All the offspring of this cross, from both parents, were blackeye snails that de- veloped coalesced mantle pigment (c°cS°S), suggesting that albino No. 5 was homo- zygous for coalesced mantle pigment (сс59$9). One of the offspring of F> (No. 5) by precross selfing (7-5-1) was used in the three- way cross illustrated in Fig. 12. At about the same time the coalesced man- tle pigment mutant was observed in our lab- oratory, Mr. Walter Stewart, working with Dr. Ned Feder (NIAMDD, NIH), observed mantle pigment variation in a wildtype B. glabrata stock of Puerto Rican origin. Some snails lacked black mantle pigment, while others displayed discrete mantle spotting. By cross- es, they demonstrated that the unspotted character was simple recessive (personal communication). They kindly provided us with some of these unspotted (CCss) snails to determine the relationship between the var- ious types of mantle pigmentation. CCss snails were used in several crosses with other pigment types in paired matings. The results were consistent with those in Fig. 12, which is of interest as a promiscuous mating of three snails involving six alleles, three for each of two pigment genes. Each of the snails bred true for its pigment pattern by selfing before the matings: a wildtype snail lacking mantle pigment (CCss), a blackeye snail with spotted mantle (c°c°SS), and an albino carrying alleles for coalesced mantle pigment (ccS°S° No. 7-5-1, Fig. 11). The three snails were maintained together in a 400 ml beaker for 10 days and then re- isolated. Resulting offspring demonstrated that each snail had cross-fertilized both of the others. The wildtype CCss snail produced two types of hybrid offspring, Cc’Ss and CcS“s; the spotted blackeye produced Cc’Ss and c’cS°S hybrid offspring; and the albino сс5959 produced CcS%s and c’cS®S hybrid offspring. Hybrid F,s of all the above types were isolated and progenies by selfing were A NEW BIOMPHALARIA PIGMENTATION MUTANT 147 FIGS. 1-9. Photomicrographs of nine Biomphalaria glabrata ranging 4—7 mm in maximum shell diameter, showing nine different pigment phenotypes (36 possible genotypes): Fig. 1, wildtype with coalesced mantle pigment (CCSIS?, CCS“s, Сс?59$9, Cc®S%s, Coses”, CcS“s); Fig. 2, wildtype heterozygous for mantle pigment (CCSS, Сс?$9$, Сс$9$); Fig. 3, wildtype with spotted mantle (CCSS, CCSs, Сс?$$, Сс?$$, CcSS, CcSs); Fig. 4, blackeye with coalesced mantle pigment (c?c?S°S%, c®c?S%s, с?с$959, c®cS“s); Fig. 5, blackeye heterozygous for mantle pigment (c?c°S°S, c”cS*S)—poor lighting makes the headfoot of this snail appear dark and obscures the partially coalesced mantle spots; Fig. 6, blackeye with spotted mantle (cPePSS, c?c?Ss, cPcSS, c®cSs); Fig. 7, wildtype with unpigmented mantle (CCss, Cc’ss, Ccss); Fig. 8, blackeye with unpigmented mantle (c®c®ss, c°css); Fig. 9, albino (ccS°S%, ccS°S, ccSS, ccS“s, ccSs, ccss). 148 RICHARDS STOCK 1-13-131 STOCK 243432 дао #7 EN b A < | Ar cbcbBSS ccSS Ÿ Va [#16555 6 GENERATIONS SELFING 13142 -10-1 SE Bess: 6 cbchSdS” Q NS 74 155 cbcbSS a cbcbSS #4 SEE? FIG. COALESCED MANTLE ALBINO #4 à < PRE-CROSS cbcbSSd IX сс55 seLFins / 7 — API = = > FS (ody Or г = a 5 a] 20 cbcb(SdS4+SdS) : acbcbSS 12 cbc 595: 1266655 Е Е25 ~ a > 9 (cbcb: cbc)SS : s cc(SdSd+ SdS+SS) Gr 1 (cbc. cbc)(SIS4+SdS) : г | 2245 eo E y x TNT ccSdS4 EG | PRE-CROSS SELFING Cr dec “J ecSdSd air cbcSdS FIG. 11. Diagram of crosses to determine the method of inheritance of the mantle pigment coalescence character. Fig. 10. Diagram illustrating origin of the mantle pigment coalescence mutants. 10 (CC+Ccb\(SS+Ss): cess \ 4 &cbSS: 2 cbcbss PRE- | = 5 CcbSs Os CROSS SELFING Vues Xe 7-5) ©9255 э9(СС+Сс)(53$3+545): (ECHES) ssi 7 cc(SdSd + Sds+ss) x CR SE DUT LU N) 22 (CC+Cc)Sds+ 26(CC+ Ce)ss в cbcSdS do tro (cbcb4 cbc)SdS : 5(cbcbicbc)SSi и сс($354+595+55) "т | ig (cbcbs cbc) SdSd : 6 ccSdSd FIG. 12. Diagram showing mating results when three Biomphalaria glabrata were maintained together for ten days and then re-isolated. The three snails (CCss, c’c’SS, and albino 7-5-1 сс59$59) involved six different pigmentation alleles, and the results demonstrated the inheritance of these pigment factors. A NEW BIOMPHALARIA PIGMENTATION MUTANT 149 scored. Some snails were followed through additional generations. This and other cross- es demonstrated that the coalescence allele SY was not completely dominant over S. Althou ugh coalescence appeared complete in old S°S snails, young SIS heteroz zygotes could be distinguished from either SIS” of SS homozygotes by the combination of spots and coalescence. Scored progenies (118) of SIS heterozygotes by selfing totaled 30 SIS? : 58 SiS : 3055. DISCUSSION The 6 original mutant snails (Fig. 10) all er mixed progenies in the total ratio 77 с? ($999 + $95) : 33 cPcPSS. Backcrosses en two of the snails and albinos pro- duced hybrid ‚progenies in the total ratio 32 c°cS'S : 32 c®cSS. These results suggested that the 6 snails were all heterozygous for mantle pigmentation (сбс?$9$). In retrospect, there is no way to be certain of the phenotype and genotype of the parent 1-3-1-4-2-10-1 snail. It is probable that had it developed coalesced mantle pigment, this would have been noted. If its genotype had been S°S, the 19 SS : 6 59$ ratio of its progeny would not be expected. If the parent was spotted c°c?SS, the 6 mutant progeny, all heterozygotes, could have been the result of a mutation in one Cell in the ovotestis giving rise to a group of either sperm or egg gametes. Crosses with wildtype snails demonstrated that the coalescence allele modified mantle pigment in wildtype as well as blackeye snails, but the trait is masked in the albino snails which lack black pigmentation. Data presented in Table 1A provide a sum- mary statistical analysis for crosses shown in Figs. 11 and 12 and for additional families TABLE 1. Segregation and assortment of loci controlling basic pigmentation and mantle pigment pattern in B. glabrata. Expected values based on an assumption of Mendelian segregation are shown in parentheses. A. Genetic segregation at a locus controlling mantle pigment pattern. Offspring are the product of self-fertilization in isolated snails. Offspring Parental phenotype Genotype 55 Sse 55: x Discrete Spots SS 155 32 18 dil Coalesced Mantle SiS 16 14 23 Spots/Coalesced 95° 5(5.25) 5(10.50) 11(5.25) 9.028 7(7.75) 19(15.50) 5(7.75) 2.117 3(3.75) 9( 7.50) 3(3.75) 0.637 8(6.50) 11(13.0 ) 7(6.50) 0.692 Total 23(23.25) 44(46.50) 26(23.25) 0.451 X? hom = 8.577 DF hetero. = 6 X? N.S. at 0.05 B. Simultaneous segregation and assortment of loci controlling basic pigmentation and mantle pigmentation pattern. Phenotypic expectations based on an assumption of independent assortment are given in parentheses. A X? test of goodness of fit for observed values is also shown. Offspring Phenotype Parental snails CS CSS: CSS Cs cS c?S1S CSS 625 с Xe CcS“s(selfed) 24(19.688) 3(6.533) 6(8.75 ) 2.176 CcSs(selfed) 9(11.812) 5(3.936) 7(5.250) 1.520 Cc?Ss(selfed) 10(9.00) 4(3.00 ) 2(4.00) 1.444 c’cSS%selfed) 2(2.026) 4( 4.125) 1(2.062) 4(2.75 ) 1.098 C°cSS%(selfed) 5(6.00 ) 10(12.00 ) 6(6.00 ) 11(8.00 ) 1.625 150 RICHARDS which were not included in the illustration. Genetic control of mantle pigment pattern is amply demonstrated. Homozygous in- dividuals produce only homozygous offspring and the ratio of progeny of selfed heterozy- gotes approximates the 1:2:1 ratio expected for codominant alleles at a gene locus. The occurrence and ratios of three F; phe- notypes (including albinos) in Fig. 11 and text suggested that the mantle pigment gene was probably on a different chromosome from that determining albinism. Crosses (including self- ing) involving segregation at the two loci are summarized in Table 1B. The X? statistic was used to test the fit of the data to expectation based on independent assortment of the loci. In all cases, the data are consistent with this hypothesis, and the two pigment loci are assigned to independent linkage groups. Mating between an Fz albino (Fig. 11) and a normally spotted blackeye snail indicated one albino No. 7-5 was homozygous (S°S°) for pigment coalescence. B. glabrata has many pigmentation var- iations and inheritance is complex. In a prev- ious study on mantle pigmentation (Richards, 1969), it was concluded that spotting was dominant over lack of spotting, but that this variation was determined by multiple genetic factors. The “unspotted” snail lines were de- rived by several generations of selection from stocks with relatively few mantle spots. The spotted and unspotted stocks provided by Mr. Stewart showed single factor inheritance ap- parently being homogenic for other mantle pigment factors. Results depicted in Fig. 12 demonstrated that mantle pigment coales- cence (S°), discrete spotting (S), and lack of mantle pigment(s) are three alleles of the same gene. The S* and $ alleles are both dominant over the s allele. Young S°S heter- ozygotes can usually be distinguished from $959 or SS homozygotes. An S°S heterozy- gote reproducing by selfing should produce progeny in the ratio 1 S°S°: 2S°S : 1 SS. A total of 93 offspring of snails identified as SIS heterozygotes were scored: 23 SIS” : 44 SIS : 26 SS. Combinations of these mantle pig- ment factors and the basic pigment factors C, со, and с result in 9 phenotypes (Figs. 1-9). Coalesced mantle pigmentation, de- termined by the S“ allele, is initiated and can be recognized in very small juvenile snails. The expressivity of “diffuse” pigmentation de- scribed by Richards (1969) is apparently in- fluenced by environmental conditions as well as other genotypic factors and develops later in the snail's life. Histologic studies indicate that S* mantle coalescence involves black pigment in mantle epithelial cells, as with mantle spotting. With the range of pigment markers now available, planned experiments involving mul- tiple matings and serial matings with varying time intervals could provide information on the dynamics of cross-fertilization in pop- ulations of B. glabrata. Six characters in B. glabrata, each de- termined by a single gene pair, have been described (Richards, 1973b, 1980). Mantle pigmentation variation described here con- stitutes a seventh such character. Studies to date have failed to demonstrate linkage be- tween any of these characters, so they are tentatively considered markers for seven link- age groups. ACKNOWLEDGMENTS The cooperation of Mr. Walter Stewart in providing the B. glabrata lacking mantle pig- ment, and information gained from dis- cussions with him during these studies, are greatly appreciated. Review of the manuscript by Dr. Philip T. LoVerde, Dr. David S. Wood- ruff, and Dr. Margaret Mulvey, provision of statistical analyses by Dr. Mulvey, and the technical assistance of Mr. Paul C. Shade and Mr. Thomas A. Hallack are gratefully acknowledged. These studies were funded in part by Office of Naval Research Contract N1.N00014-78-C-0081. REFERENCES CITED NEWTON, W. L., 1953, The inheritance of sus- ceptibility to infection with Schistosoma mansoni in Australorbis glabratus. Experimental Parasitology, 2: 242-257. NEWTON, W. L., 1954, Albinism in Australorbis glabratus. Proceedings of the Helminthological Society of Washington, 21: 72-74. RICHARDS, C. S., 1967, Genetic studies on Biom- phalaria glabrata (Basommatophora: Planorbi- dae), a third pigmentation allele. Malacologia, 5: 335-340. RICHARDS, C. S., 1969, Genetic studies on Biom- phalaria glabrata mantle pigmentation. Malaco- logia, 9: 339-348. RICHARDS, C. S., 1972, Biomphalaria glabrata genetics: pearl formation. Journal of Invertebrate Pathology, 20: 37—40. RICHARDS, C. S., 1973a, Susceptibility of adult A NEW BIOMPHALARIA PIGMENTATION MUTANT 151 Biomphalaria glabrata to Schistosoma mansoni infection. American Journal of Tropical Medicine and Hygiene, 22: 748-756. RICHARDS, C. S., 1973b, Genetics of Biomphalar- ia glabrata (Gastropoda: Planorbidae). Malacological Review, 6: 199-202. RICHARDS, С. S., 1974a, Antler tentacles of Biom- phalaria glabrata: genetics studies. Journal of Invertebrate Pathology, 24: 49-54. RICHARDS, C. S., 1974b, Everted preputium and swollen tentacles in Biomphalaria glabrata: ge- netic studies. Journal of Invertebrate Pathology, 24: 159-164. RICHARDS, C. S., 1975, Variations in susceptibility of Biomphalaria glabrata for different strains of Schistosoma mansoni. Parasitology, 70: 231-— 241. RICHARDS, C. S., 1977, Variations in infectivity for Biomphalaria glabrata in strains of Schistosoma mansoni from the same geographic area. Bulle- tin of the World Health Organization, 54: 706— 707. RICHARDS, C. S., 1980, Edema-horn, an abnor- mal mutant of Biomphalaria glabrata. Journal of Invertebrate Pathology, 35: 35-37. RICHARDS, C. S., 1984, Influence of snail age on genetic variations in susceptibility of Biomphalar- ia glabrata for infection with Schistosoma man- soni. Malacologia, 25: 493-502. RICHARDS, C. S. & MERRITT, J. W., Jr., 1972, Genetic factors in the susceptibility of juvenile Biomphalaria glabrata to Schistosoma mansoni infection. American Journal of Tropical Medicine and Hygiene, 21: 425-434. д MALACOLOGIA, 1985, 26(1-2): 153-163 POPULATION ECOLOGICAL ASPECTS OF THE EULIMID GASTROPOD VITREOBALCIS TEMNOPLEURICOLA' Yoshimi Fujioka Mukaishima Marine Biological Station of Hiroshima University, Onomichi P.O., Hiroshima Pref. 722, Japan ABSTRACT The parasitic life history of the eulimid gastropod Vitreobalcis temnopleuricola, parasitic on the sea urchin Temnopleurus toreumaticus, has been studied at five subtidal stations around Mukaishima Island. Rates of infestation vary with the population size of the host and with the season. Annually, the parasites begin to settle on the hosts in the autumn and grow ex- ponentially at periodic intervals until the following summer. There was little relation between the infestation rate and the host size. A marked difference in parasite position according to parasite size was found; individuals of this species settle on the oral side near the peristome at first, subsequently migrating towards the ambitus, and lastly to the tube feet. An analysis of the aggregation pattern reveals that the parasites do not distribute at random among the hosts but in a negative binomial pattern. INTRODUCTION The superfamily Eulimacea (or Eulimoidea) is a large group of free-living and parasitic species. Consequently, it is an interesting taxon in which to study adaptive mod- ifications. Seven familial taxa have been rec- ognized in this superfamily, namely: Eu- limidae, Stiliferidae, Entoconchidae, Thy- cidae, Pelseneeriidae, Paedophoropodidae, and Asterophilidae. The three first-mentioned families are closely related (Lútzen & Nielsen, 1975; Ponder & Gooding, 1978), but the phyletic interrelations of the other families are still unclear. Grusov (1965), in his histological study of Asterophila, proposed uniting the six families, with the exception of Entoconchidae, into a single broad family, Melanellidae (= Eulimidae) s.1. Waren (1980a, b, 1981) sup- ported this concept in his comprehensive taxonomical study of this family. Lutzen and his colleagues have greatly contributed to the knowledge of the taxon- omy, anatomy, and histology of this family (Lútzen, 1972a, b, 1976; Gooding & Lútzen, 1973; Lútzen & Nielsen, 1975). Ponder & Gooding (1978) gave a good review of this family. Habe (1952, 1976) studied the classification of this group and reported about 40 species from Japanese waters. Morton (1976, 1979), Elder (1979), and Euizen (1979) contributed ecological studies on spe- cies of Balcis and Mucronalia, Thyca, and Enteroxenos, respectively. Nevertheless, compared with the many anatomical and taxonomical discussions, there are very few practical reports on the population ecology of eulimids. The probable reasons for this are that the population sizes of eulimids are usually small; also, the animals are small in size. The present species, Vitreobalcis temno- pleuricola Fujioka & Habe (1983), is found in the Seto Inland Sea of Japan exclusively parasitizing the sea urchin Temnopleurus to- reumaticus (Leske), which inhabits the shel- tered subtidal zone around the southwestern Japanese waters. The present contribution is concerned with basic knowledge of the dis- tribution and growth of this species. Dr. A. Warén (University of Goteborg), on the basis of some morphological characters of shells, considered that there are two spe- cies in my samples. However, judging from the continuity of such characters and ecologi- cal knowledge, | believe that they belong to a single species. Voucher specimens for the present study are deposited in Mukaishima Marine Biological Station, Hiroshima Univer- sity (No. 3412013). ‘Contribution from the Mukaishima Marine Biological Station. No. 218. (153) 154 FUJIOKA SURVEY AREA AND METHODS The field survey and collection were carried out at five stations around Mukaishima Marine Biological Station (34°22'N, 133°13’E), in the northwest area of Bingo- nada, central part of the Seto Inland Sea of Japan (Fig. 1). St. 1 has sunken rocks with surrounding fine gravel bottom. Water depth (below the mean tidal level) ranges from 4— 9m. St. 2 has sandy bottom at depths less than 7m and muddy bottom at greater depths. St. 3 is about 4—5 т deep and has spacious sandy to fine gravel bottom. St. 4 and St. 5 are exposed rocky shores, near which currents run strong. The host sea urchin, Temnopleurus to- reumaticus, inhabits various kinds of sub- stratum. Larger populations of this species are found on fine gravel bottom at 3-7 m depth or around rocks at 3-15 m depth, but the urchin is rarely distributed in an aggre- gated pattern. The maximum mean density of the urchins reached about 6 individuals per square meter at St. 3 in 1980. After the spring of 1981, however, population size of the urchins of this station decreased, possibly because of change in the environmental con- ditions caused by construction of a nearby artificial reef and seawall. The monthly investigations were made di- urnally using scuba for 40 minutes to 2 hours Mukaishima Island 2 5 wi shi ® FIG. 1. Asketch map of the environs of Mukaishima Marine Biological Station (MMBS) showing the locations of stations studied. per station, from the end of June 1980 to November 1981. As far as possible, a full-air scuba tank was used for each station, but the number of urchins found was influenced by the population size and the water conditions such as the tidal current, the depth of water, and the turbidity. After measuring the diameter of the test (horizontal axis) of each sea urchin, the num- ber of parasites and their positions on the host were determined and recorded. From May to September these procedures were done in situ with the naked eye or by a x5 magnifying glass. This method is important for avoiding artificial disturbance of the host population, and | believe no parasite was overlooked with this method in this season. Some parasites were removed from the host and collected in a polyethylene pouch for measurements of shell length, and the rest were left to maintain the parasite population. From October to April, however, there was a slight possibility of overlooking the parasites with this method. For this reason each urchin was collected in a separate polyethylene bag and carefully transported back to the shore laboratory and examined under a binocular microscope to locate small parasites. The longitudinal shell axis was measured to the nearest 0.1 mm with the aid of a binocular microscope. RESULTS Seasonal occurrence of parasites Seasonal fluctuation in the percentage of sea urchins parasitized at each station is shown in Fig. 2. At the end of June 1980, 459 parasites were found from 300 urchins at Si. 3 and the infestation rate reached 55%, which trend continued through the next month. There was a very large host population in this station at this time. In August the infestation showed a marked decline and the parasites were rarely found at all stations in September. In this season, a few dead shells were found on the substratum around the urchins at St. 3, but living specimens were never found away from the host throughout the present survey. In November 1980, the parasites appeared on the host again, resulting from new settle- ment in this season. From November to the next summer the infestation rate at Sts. 1, 2, 4 and 5 ranged from 5.3 to 21.1%. Mating took place on the tube feet from the end of June, VITREOBALCIS ECOLOGY 155 20 sl 10 100% 1072262153 = 38 о 132 115 aie = 100 130 55 20 A +. 10 100 eS = a 100 100 ¡00 35 о > 120 102 hte (5 40 13 — 50 214 110 PERCENTAGE OF URCHINS PARASITIZED se pu É Tu aoe JJASONDJFMAMJ JASON 1980 1981 FIG. 2. Seasonal fluctuation in the infestation rate of Temnopleurus toreumaticus. The number under each plot indicates the number of urchins examined. but only five cases were observed in the field and two cases in the laboratory. At St. 3 the rate increased to about 50% in January and February 1981, but after spring the rate de- creased and was unstable due to the reduc- tion in the host population. The differences of the infestation rate from January to February 1981 between every two stations were not significant at Sts. 1, 2, 4, and 5, respectively, but were significant be- tween St. 3 and each other station (Student's t-test, P < 0.01). Growth of Vitreobalcis temnopleuricola As far as the growth of Vitreobalcis temno- pleuricola is concerned, there were no re- markable differences between the five sta- tions examined. The combined seasonal change of the size frequency distribution of this species at five stations is shown in Fig. 3. In 1980, they showed gradual growth from June to August. Freshly settled individuals smaller than 1 mm shell length were contin- ually collected in the period from November to early the next spring. They grew slowly in winter, rapidly after spring, and reached a length of 3-5 mm in July to September. The parasitic life on the urchin was terminated within about ten months. There was only one generation per year and the parasitic genera- tions never overlapped each other. Fig. 4 shows the shift of the mean sizes and its standard deviations based on the results of Fig. 3. On the whole, the relationship is ex- pressed by the following exponential formula: SL = a - exp (0.1362x) where a is the initial estimated size (Oct. 1980, 0.8661) and x is the month. The formu- la can be expressed as a straight line on a logarithmic scale (a correlation coefficient = 0.9474) and therefore the relative growth rate is constant. The shell-length frequency in each month does not show a normal distribution and shows high frequencies on some specific val- ues such as 0.9, 1.2, 1.7 mm, and so on (Fig. 156 FUJIOKA 1981 JAN. | 2 3 4 5 SHELL LENGTH (mm) FIG. 3. Seasonal shell-length frequency distribution of Vitreobalcis temnopleuricola. SHELL LENGTH 9 JJASONDJ FMAM)J JASON 1980 1981 FIG. 4. Growth in mean shell-length of Vitreobalcis temnopleuricola. Bars indicate standard deviations. Sample sizes are shown in Fig. 3. 3). However, shell-length frequency accord- ing to the whorl number of the teleoconch shows a normal distribution as in Fig. 5. Such a tendency evidently appears in the small classes less than 4, but larger ones more than 5 do not correspond with a normal distribution (e.g. no. of teleoconch whorls 5, x“-test, 0.05 > Р > 0.025). Since there is no morphologi- cally apparent sexual dimorphism in this spe- cies, this abnormality may be due to large variation of the growth pattern. The first 3-3.5 whorls of this species con- stitute the larval shell and the succeeding teleoconch whorls consist of almost 7.5 whorls in fully grown specimens. On every whorl of the teleoconch there is a microscopic longitudinal line at nearly an identical position (see Fig. 6) which corresponds with the peak of the normal distribution shown in Fig. 5. These results suggest that the body whorl of the teleoconch is not always formed at a uniform rate but at periodic intervals. In other words, the resting stage of formation is repre- sented by the longitudinal line. In fact, 93% of the specimens collected were in this resting stage. The relationship between infestation rate and host size Roughly speaking, this sea urchin breeds in the summer, recruits the population in au- tumn, reaches 20 mm or so in test diameter in the next summer, and attains adult size (more than 30 mm or so) the following summer. In Table 1, the infestation rates are com- VITREOBALCIS ECOLOGY 157. No. OF TELEOCONCH WHORLS SHELL LENGTH (mm) FIG. 5. Shell-length frequency distribution according to the number of teleoconch whorls of Vitreobalcis temnopleuricola. Teleoconch whorl O means the range from 0= to <1 (the same rule applies successively). Open circles indicate mean shell-length. Bars indicate standard deviations. TABLE 1. Comparison between the infestation rate and the test diameter of Temnopleurus toreumaticus. St. 1+2+4+5 Stas No. of No. of No. of No. of Diameter of urchins urchins Infestation Diameter of urchins urchins Infestation urchins (mm) observed _ infested rate (%) urchins (mm) observed infested rate (%) <10 6 0 0.0 <<110 1 0 0.0 10S, <20 47 5 10.6 1020 2 0 0.0 20=, <30 15% 19 12a 20=,=80 Ul 32 41.6 30S, <40 637 59 9.3 30S, <40 471 193 41.0 40= 291 29 10.0 40= 43 19 44.2 FIG. 6. An apical view of the shell. Arrows indicate the longitudinal lines of teleoconch. pared according to the test diameter of Tem- nopleurus toreumaticus based on the results excepting August to November, which show the low infestation rate. St. 3 is separated from other stations because the population structure of urchins is somewhat different from that of the other stations. All size groups of urchins were parasitized except for several juveniles up to 10mm test diameter. The smallest urchin parasitized measured 11.2 mm. There seems to be no available position for the parasites on small urchins up to 10mm. There was no tendency for the (e!) (15) (500) (9'ez) (DIS) CC A) (#8) № (9) (© С аа ($) (8 €) ease aoeuns [2}0} JO эБбе}иээ.эз (0'6/) (0:92) 0 0 0 0 0 0 0 € 0 0 | 0 0 0 2 (s'6) (ev) (8p) (12S) (8er) (0'61) 0 с 0 0 | | 0 et | 0 1 0 0 0 9 (e €) (19) (499) (een) ee) (EEL) (ee) 0 | 0 0 с 0 0 2 | y 1 0 | 0 G (29) (85) Eis) leise) VETA Ka) (SEL) (Оо (00% 0 с 0 0 | | | 8 € 1 y € € 0 1 > (Ee) (er) (ze ) (vier) (8) (ive) (ve) (891) ($15) O 0 0 с 0 v 0 € 8 8 02 8 91 92 0 € 3 (271) (SA) (yo) (232) (92) (wes) (66) aioe) (er) (#61) (6'pe) as 0 12 0 e | G 9 8 22 Ly OL Sp 18 0 с (GE) (68) ВМ (65) (6 2=) Meise) 0 0 0 0 0 с с 0 Ф Z с pl 02 0 | (0'S2) (o'Sz) (00S) 0 0 0 0 0 0 0 0 | 0 0 | с 0 0 di ZV ZI di ZV ZI di ZV ZI di ZV ZI soyiseied a a — A ee en JO SLOUM IA A N Ш Il | 424090313} JO ‘ON (sasayjuajed ul ssejo |1оцм цэва 10} sabejuaoad) UIYOIN еэ$ UO uolisod e 5 5 5 nen ‘MOI WOYOQ AU} Ul UMOUS эле Seale aeuns |210} jo aBejuaolad e se passaidxe seaje aoeung */ ‘614 JO SUOISIAIP eu} UO paseq в/оэипа/доиша} зю|едоэил Jo SUONISOA DNISPIBA ‘© 37891 158 VITREOBALCIS ECOLOGY 159 AZ: Ambulacral zone ТЕ: Tube foot 17 : Interambulacral zone FIG. 7. Diagram showing the divided areas on the urchin in order to examine the placement of Vitreobalcis temnopleuricola on its host. a: Longitudinal section of the urchin showing the six horizontal zones used in this study. b: Dorsal view of the urchin showing the longitudinal surface elements. parasites to select a specific size of urchin larger than 10 mm. Therefore, the infestation frequency is very loosely related to the size of the host. Parasitic position on host The foot is well developed and functional for crawling, but plays no role for maintaining the position of this parasite. Instead, the ex- tendible proboscis and the pseudopallium, which is developed around the base of the proboscis, are used for attaching the animal to the various parts of the host, such as spines, the tube feet, the surface of the test, and rarely the stalk of globiferous pedicel- lariae (only two cases). They were not found attached to the gill or the buccal tube foot. The proboscis of the parasite did not pene- trate the plates of the host and did little dam- age to the skeletal part of the spines. The epithelial tissue at the parasitized position was very loose, but the apparent tissue re- moval was not observed. In order to examine the parasite’s position on the host, a longitudinal section of the urchin was divided into six horizontal zones (I-VI) and each zone was divided into three longitudinal elements, the interambulacral zone (IZ), the ambulacral zone excepting the tube foot (AZ), and the tube feet (TF), as shown in Fig. 7. The results are shown in Table 2 for every teleoconch whorl number. The surface area occupied by each division was measured on 10 urchins and the mean percent shown in the bottom row of Table 2. If the parasites distribute randomly on the host, it would be expected that the parasites were collected with the approximate percentage of the surface area of the host. However, the actual distribution never agreed with these percentages. Many parasites were distributed on the oral side of the host. Moreover, the following three tendencies are recognizable: (1) the parasites were never found on the peristome and periproct, (2) as they grew, they were found to migrate radially from the oral side near the peristome toward the ambi- tus, and (3) as they grew, they migrated from the area overspread by the spine to the tube foot. To verify statistically tendencies (2) and (3) a x“-test was based on the results shown in Table 2. Every adjacent two whorl classes of parasites are combined as shown in Table 3a (for tendency (2)) and Table 4a (for tendency (3)) because this test required that the ex- pected frequencies in each cell should not be too small. In addition, zones IV and V are also combined and zones | and VI are excluded for the same reason. The values of x? were calculated for every combination of whorl classes and given in Table 3b and Table 4b, respectively. In Table 3b, lower values were obtained with neighboring whorl classes ex- cept the test between whorl classes 2-3 and 4-5, and higher values with distant whorl classes. Therefore, there is a general trend for the parasites to migrate radially on the host with growth. In Table 4b, the tendency is more clear than in Table 3b. It is evident that the parasites migrate from the ambulacral and the interambulacral zones to the tube foot with growth. 160 FUJIOKA TABLE 3. Differences of parasite position on V. temnopleuricola with snail growth. a: Comparison between the whorl classes of parasites and the horizontal zones on the host. b: Results of x”-test (DF = 2). a Whorl class Il Ш IV+V Total 0-1 39 12 4 55 2-3 186 NS 28 327 4-5 15 ЗИ 8 60 6—7 5 16 4 25 b 2-3 4.00 4-5 24:5. ZOE 6-7 188195 12.74* 0.29 Whorl class 0-1 2-3 4-5 *Significant difference (P < 0.005). TABLE 4. Differences of parasite position on V. temnopleuricola with snail growth. a: Comparison between the whorl classes of parasites and the longitudinal elements on the host. b: Results of x°-test (DF = 2). a Whorl class IZ AZ WE Total 0-1 31 22 2 55 2-3 186 98 43 327 4-5 13 8 39 60 6—7 0 2 23 25 b 2-3 5:13 4—5 И 8702 6-7 62.93** OS 7.60* Whorl class 0-1 2-3 4-5 *Significant difference (P < 0.05). **Significant difference (P < 0.001). Parasite incidence amongst the hosts Analysis of the number of individuals of Vitreobalcis temnopleuricola per host reveals that large numbers of specimens had no par- asites and a small number of specimens had large numbers of parasites, up to 25. In order to judge the parasite incidence amongst the hosts, the observed frequencies and the ex- pected frequencies calculated by negative binomial probabilities and by Poisson proba- bilities are given in Table 5. They are divided into five categories on the basis of the dif- ferences in rate of infestation and season. Bliss’ (1956) method was employed for com- puting the parameter “a common k” of the negative binomial distribution (a common k is present in all categories, P < 0.05). In each of the five categories the observed frequencies agree with the negative binomial probabilities rather than with the Poisson ones, and the y°-test shows good agreement between the observed and the negative binomial distribu- tions. Hence, the parasites are not distributed at random, but are clustered. This situation continues regardless of the season. DISCUSSION Among the five stations around Mukaishi- ma Island the infestation rate of St. 3 exhibits a higher percentage than that of the four other stations (Fig. 2). Gooding & Lützen (1973) studied Robillardia cernica parasitic in the rectum of the sea urchin Echinometra and found that the highest rate of infestation oc- curred at the mid-tide level of the intertidal zone. Elder (1979), in his study of Thyca crystallina parasitic on the blue starfish Linckia laevigata, determined that the infesta- tion rate varies directly with the degree of water movement. Hoberg et al. (1980), in their examination of the endoparasitic gastropod Asterophila japonica, treated the para- site-host spatial distribution, but gave no de- tailed information about the influence of en- vironmental factors. The present species, Vit- reobalcis temnopleuricola, inhabits the sub- tidal zone where the influence of vertical com- ponents such as tidal level and water depth are not important. Although the water move- ment seems to be high at Sts. 4 and 5, there is no direct relation between the infestation rate and water movement. At St. 3 the host urchin formed the highest density population until the winter of 1981. V. temnopleuricola may have a pelagic larval stage because the parasites are not distributed at an aggregated pattern even in the recruit season. Therefore, it is possible that the larger the population of the host becomes, the more effectively the larva settles. The effect of chemical in- VITREOBALCIS ECOLOGY 161 TABLE 5. Comparison of the observed frequencies of urchin for different numbers of parasites with the expected frequencies calculated from the negative binomial and the Poisson distributions. st 1 7214.55 St. 3 Expected Expected No. of No. of frequencies Expected No. of No. of frequencies Expected parasites hosts negative frequencies parasites hosts negative frequencies per host observed binomial Poisson per host observed binomial Poisson X1'80-11'81 (К = 0.1208) VI'80 (Кс = 0.4669) 0 121117 1225.5 1161.8 0 135 152.2 65.0 1 97 82.4 176.1 1 70 54.5 99.4 2 25 25.7 13.4 2 33 30.6 76.0 3 6 10.1 0.67 3 25 19.3 38.8 4 2 et 0.03 4 10 12.8 14.8 SE 5 2.98 0.00 5 10 8.76 4.54 6 3 6.12 evs x = 4.35 < 5.99 (P = .05, DF = 2) Y 3 4.33 0.25 8= 11 11.45) 0.06 11/81-V'81 (К = 0.3444) x? = 11.03 < 12.59 (Р = .05, DF = 6) 0 454 453.4 437.3 1 53 54.2 80.0 1'81-11'81 (Кс = 0.5487) 2 12 12.6 7.31 3 5 Saat 0.45 0 102 103.0 67.0 4= 1 1.43 0.02 1 39 36.5 67.0 2 17 18.2 33.5 Хх = 033 = 3.84 Р = 05 DE — 1) 3 9 10.4 MEA 4 7 5:95 2.79 VI'81-VI11'81 (Кс = 0.1644) 5 2 3.50 0.56 6 2 2.09 0.00 0 609 601.1 582.0 7 2 1.26 0.00 1 33 42.8 73.2 gs 2 1.09 0.11 2 13 10.8 4.61 3 2 3.38 0.19 4 0 116) 0.01 x = 0.63 < 781 (Р = 05, ВЕ = 3) БЕ 3 0.68 0.00 E (Р^— 05, ВЕ) teractions in larval settling, if any, is a subject for future study. Morton (1979) reported that Mucronalia ful- vescens and Balcis shaplandi are both ectoparasitic on the starfish Archaster typi- cus, and that they show a biannual reproduc- tive pattern and leave the host in the late summer. Thus it appears that seasonal fac- tors are necessary to explain the differences of the infestation frequencies. In the present study, parasites were rarely found on or in the vicinity of hosts in September and October. It is clear that there is one generation per year and that growth occurred from autumn to the next summer. Generally speaking, growth of most mollusks corresponds with a sigmoid curve (Wilbur 8 Owen, 1964), but the present species grows exponentially. Thus the plateau phase toward the end of the life span is not observed. It is suspected that absence of a plateau has been caused by either the death of parasites or migration from the hosts to other substrates until oviposition, although there is no reliable information about the be- havior after mating. The attachment positions of ectoparasitic eulimids are usually specific for each species (e.g. Habe, 1952, 1976; Lútzen 8 Nielsen, 1975; Lútzen, 1976; Morton, 1976; Warén, 1980a,b, 1981). However, Elder (1979) found that the position changed with growth of Thyca. He demonstrated that the juvenile Thyca is restricted to the aboral surface of Linckia and the adults to the oral surface. Elder also discussed either positive geotactic of negative phototactic behavior of the ju- 162 FUJIOKA veniles. Then the adults migrate to a more favourable site for proboscis penetration. It was found in the present study that У. temnopleuricola also migrates with increasing shell length (Table 2). The species has a tendency to settle on the oral side near the peristome at first, be radially oriented toward the ambitus, and finally to migrate to the tube foot. Since the space associated with the tube feet is large and the surface area of the ambulacral and interambulacral plates of the ambitus are larger than those of the oral side, it is believed that the parasites migrate with growth to larger and more favourable sites for parasitism. In other words, the parasite posi- tion is restricted spatially by the external appendages of the host. In the Aglossa, the radula has degenerated and a long acrembolic proboscis is developed for suctorial feeding. Examples of ectopara- site proboscises perforating the plates or skin of echinoderms have been reported by Fretter (1955) for Balcis devians and B. alba, by Bacci (1948, cited by Fretter, 1955) for Melanella comatulicola, by Baer (1952, ditto) for Mucronalia mittrei, by Bartsch (1907, cited by Lutzen, 1972b) for Eulima ptilocrinicola, by Lutzen & Nielsen (1975) for Echineulima, by Elder (1979) and Warén (1980a) for Thyca, and so on. The proboscis of Balcis per- onellicola Kuroda 8 Habe also deeply pene- trates the perivisceral cavity of the laganid, Peronella japonica Mortenson (personal observation). Fretter 8 Graham (1962: 259) suggested that the proboscis of Balcis alba passes into the body of the host, perhaps seeking the gonad. Morton (1976) mentioned that Mucronalia fulvescens and Balcis shap- landi draws the fluid contents from the tube feet and coelom of host, respectively. Pulici- cochlea (Pseudoretusa) faba also feeds on the host’s body fluids (Ponder & Gooding, 1978). On the other hand, Pelseneeria and Pulicicochlea s.s. are considered to digest the epithelium of the body of echinoids (Fretter & Graham, 1962: 255; Ponder & Gooding, 1978). The proboscis of the present species, V. temnopleuricola, does not penetrate the plates of the hosts, and many specimens were collected from the spines, attaching superficially. Therefore, it is believed that the species is a parasite of the epithelial tissue, and feeds on it slowly. Such a mode of attach- ment and feeding are considered the most primitive ectoparasitic form. The analysis of the aggregation pattern re- vealed that the parasites do not distribute at random amongst the hosts but in a negative binomial pattern (Table 5). Elder (1979) obtained similar results on Thyca and men- tioned that the pattern was caused by “the accumulation of repetitive waves of random infestation.” This same interpretation can be applied to the present results because the recruitment of parasites extends for a long period and the large number of solitary in- dividuals suggest negation of the presence of pheromone among the individuals when they settle onto the host. The parasites seldom migrate from the host to other substrata or another host. This is evident by reason of the facts that (1) the living specimen was never found off the host, (2) some specimens were collected in the growth stage of shell formation (they do not migrate even in this time), and (3) negative binomial distribution is observed throughout the parasitic life. As there are large numbers of solitary individuals, however, it is possible to migrate from one host to another only for a short period during reproduction. In the present study we dealt only with the post-larval parasitic period. Examination of the life span from breeding to settlement is a subject for future study. ACKNOWLEDGMENTS | am deeply indebted to Dr. W. K. Emerson, American Museum of Natural History, and Mr. R. B. Sigel, Hiroshima University, for reading the manuscript. Special thanks are also due to Dr. R. Robertson, Academy of Natural Sci- ences of Philadelphia, to Dr. T. Habe, Tokai University, to Dr. A. Waren, University of Goteborg, Sweden, and to Dr. A. Inaba, Hiroshima University, for their valuable sug- gestions. Mr. H. Ochi of the Mukaishima Marine Biological Station assisted me in col- lecting on a cold winter day, to whom my thanks are due. REFERENCES CITED BLISS, С. 1., 1956 ["1958”], The analysis of insect counts as negative binomial distributions. Pro- ceedings Tenth International Congress of Entomology (Montreal), 2: 1015-1031. ELDER, H. Y., 1979, Studies on the host parasite relationship between the parasitic prosobranch Thyca crystallina and the asteroid starfish Linckia laevigata. Journal of Zoology (London), 187: 369-391. VITREOBALCIS ECOLOGY 163 FRETTER, V., 1955, Observations on Balcis de- vians (Monterosato) and Balcis alba (Da Costa). Proceedings of the Malacological Society of Lon- don, 31: 137-144. FRETTER, V. 8 GRAHAM, A., 1962, British pro- sobranch molluscs; their functional anatomy and ecology. Ray Society, London, xvi + 755 p. FUJIOKA, Y. 8 HABE, T., 1983, A new species of Vitreobalcis (Prosobranchia: Eulimidae) from the Inland Sea of Japan. Venus, the Japanese Jour- nal of Malacology, 42: 13-16. GOODING, R. U. & LUTZEN, J., 1973, Studies on parasitic gastropods from echinoderms III. A de- scription of Robillardia cernica Smith 1889, parasitic in the sea urchin Echineulima Meus- chen, with notes on its biology. Kongelige Danske Videnskabernes Selskab Biologiske Skrifter, 20(4): 1-22, 4 pl. GRUSOV, E. N., 1965, The endoparasitic mollusk Asterophila japonica Randall and Heath (Pro- sobranchia: Melanellidae) and its relation to the parasitic gastropods. Malacologia, 3: 111-181. HABE, T., 1952, Parasitic gastropods found in echi- noderms from Japan. Publications of the Seto Marine Biological Laboratory, 2: 73-85, 1 pl. HABE, T., 1976, Parasitic gastropods from echi- noderms of Japan. Bulletin of the National Sci- ence Museum (Tokyo), series A (Zoology), 2: 157-168, 3 pl. НОВЕВС, М. K., FEDER, HH! М. € JEWETT, 5. C., 1980, Some aspects of the biology of the para- sitic gastropod, Asterophila japonica Randall and Heath (Prosobranchia: Melanellidae), from southeastern Chukchi Sea and northeastern Bering Sea, Alaska. Ophelia, 19: 73-77. LUTZEN, J., 1972a, Studies on parasitic gastro- pods from echinoderms II. On Stilifer Broderip, with special reference to the structure of the sexual apparatus and the reproduction. Konge- lige Danske Videnskabernes Selskab Biologiske _Skrifter, 19(6): 1-18, 1 pl. LUTZEN, J., 1972b, Records of parasitic gastro- pods from crinoids, with description of a new genus, Goodingia (Gastropoda, Prosobranchia). Steenstrupia, 2: 233-246. LUTZEN, J., 1976, On a new genus and two new species of Prosobranchia (Mollusca), parasitic on the tropical sea urchin Echinometra mathaei. Israel Journal of Zoology, 25: 38-51. LUTZEN, J., 1979, Studies on the life history of Enteroxenos Bonnevie, a gastropod endopara- sitic in aspidochirote holothurians. Ophelia, 18: 1-51. LUTZEN, J. & NIELSEN, K., 1975, Contributions to the anatomy and biology of Echineulima n.g. (Prosobranchia: Eulimidae), parasitic on sea urchins. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening, 138: 171-199. MORTON, B., 1976, Selective site segregation in Balcis shaplandi and Mucronalia fulvescens (Mollusca: Gastropoda: Aglossa) parasitic upon Archaster typicus (Echinodermata: Asteroidea). Malacological Review, 9: 55-61. MORTON, B., 1979, The population dynamics and expression of sexuality in Balcis shaplandi and Mucronalia fulvescens (Mollusca: Gastropoda: Aglossa) parasitic upon Archaster typicus (Echi- nodermata: Asteroidea). Malacologia, 18: 327- 346. PONDER, W. F. & GOODING, R. U., 1978, Four new eulimid gastropods associated with shallow- water diadematid echinoids in the western Pacif- ic. Pacific Science, 32: 157-181. WAREN, A., 1980a, Revision of the genera Thyca, Stilifer, Scalenostoma, Mucronalia and Echi- neulima (Mollusca, Prosobranchia, Eulimidae). Zoologica Scripta, 9: 187-210. WAREN, A., 1980b, Description of new taxa of Eulimidae (Mollusca, Prosobranchia), with notes on some previously described genera. Zoologica Scripta, 9: 283-306. WAREN, A., 1981, Revision of the genera Apicalia A. Adams and Stilapex Iredale and description of two new genera (Mollusca, Prosobranchia, Euli- midae). Zoologica Scripta, 10: 133-154. WILBUR, K. M. & OWEN, G., 1964, Growth. In: WILBUR, K. M. 8 YONGE, C. M., eds., Physiolo- gy of Mollusca, vol. 1. Academic Press, London and New York. 473 p. У, MALACOLOGIA, 1985, 26(1-2): 165-172 QUANTITATIVE ASPECTS OF LOCOMOTION BY THE MUD SNAIL /LYANASSA OBSOLETA! Ronald V. Dimock, Jr. Department of Biology, Wake Forest University, Winston-Salem, North Carolina 27109, U.S.A. ABSTRACT The ciliary-powered locomotion of the marine mud snail //vanassa obsoleta was quantified for animals exhibiting positive phototaxis under controlled conditions. The absolute rate of locomo- tion was approximately 2 mm/sec for snails of about 18 mm total shell length. The maximum speed attained was 3.3 mm/sec. Although the absolute rate of locomotion was positively correlated with shell length, relative speed (% shell length moved/sec) was markedly negatively correlated with snail size. Morphometric analysis of pedal morphology indicated that pedal surface area, which increases with shell length, was closely associated with the absolute speed of mud snails. Although snail speed was more than 50% faster on glass than on sand, the presence of a conspecific’s mucous trail as substratum upon which to crawl had no effect on this gastropod’s locomotion. Maintenance of //yanassa in the laboratory for as few as three weeks from the time of collection resulted in a 50% reduction in speed of locomotion and a significant decrease in this snail's overall activity. INTRODUCTION The marine mud snail //yanassa obsoleta (Say, 1822) is typical of many prosobranchs inhabiting soft substrata (Miller, 1974a; Pal- mer, 1980) in that it employs cilia as its prin- cipal mechanism of locomotion (Copeland, 1919). Although many aspects of gastropod locomotion have been examined, including classification (Miller, 1974b), functional morphology (Miller, 1974a; Gainey, 1976; Lawrenz-Miller, 1977), functional significance (Linsley, 1978; Palmer, 1980), and energetic relationships (Calow, 1974; Denny, 1980; Houlihan 4 Innes, 1982), little quantitative information on ciliary locomotion exists. In addition to being responsive to a variety of sensory stimuli (Crisp, 1969; Dimock & Parno, 1981), /. obsoleta can detect and sub- sequently pursue a trail of mucus deposited by a conspecific (Trott & Dimock, 1978). While some understanding of the mech- anisms involved in mucous trail-following has been achieved (Bretz & Dimock, 1983), the functional significance of this behavior is not yet known. It may well be that trail-following is related to quantitative or energetic parame- ters of locomotion, since ciliary locomotion also involves mucus. Thus, the present study was undertaken to quantify the locomotion of I. obsoleta. The interrelationships among Snail size, morphometry of the foot and the rate of crawling have been examined. In addi- tion to being determined for snails crawling on glass, snail speed has also been quantified for animals traversing sand, both in the pres- ence and absence of a conspecific’s mucous trail. The behavioral consequences of pro- longed maintenance of mud snails in the lab- oratory are shown to include significant changes in this gastropod’s overall activity and rate of locomotion. MATERIALS AND METHODS All I. obsoleta were collected from the New- port River marshes at Morehead City, North Carolina, and were held in aquaria (30- 32%. $, 20-22°C) with salt-marsh mud, but were not given additional food. Experiments were performed between six and eight days after the animals were collected, except as described below. The behavioral studies were conducted between 1000 hr and 1800 hr at 22-24°C with animals in 31% S artificial sea water (Instant Ocean Sea Salts). Each snail was used in only one experiment and was ‘Contribution No. 203 from the Tallahassee, Sopchoppy and Gulf Coast Marine Biological Association. (165) 166 DIMOCK measured to 0.1 mm (total length from anteri- or edge of aperture to tip of spire) with vernier calipers after it was used. Except as other- wise specified all snails were from 16.6- 18.2 mm in total shell length. The positive phototaxis of /. obsoleta (Dim- ock & Parno, 1981) was employed in an effort to standardize the stimulus conditions under which parameters of locomotion could be ex- amined. Locomotion was monitored in a chamber (130 x 15 x 15 mm) constructed of clear 3 mm lucite into which a glass slide could be placed to form a removeable floor. The chamber was illuminated horizontally from one end with a 1 ст? beam of 5 x 10 ? рЕ/т?/зес intensity of 500 nm light, following the methods of Dimock 8 Parno (1981). In a typical trial a light-adapted snail was introduced into the chamber with its anterior toward the light source. Since llyanassa is strongly positively phototactic in this light regi- men, at least 90% of the snails that moved at all moved directly toward the light. In most experiments an animal's rate of locomotion (mm/sec) was calculated by timing with a stopwatch its progression over a 4 or 5cm path mid-way in the chamber. Snails less than 10 mm were timed over а 2cm path. Any snail which failed to move or to traverse the prescribed distance within 2 min was scored as no response. The slowest snail that reached the criterion took 95 sec, but most did so in <45 sec. The absolute rate of locomotion on glass was determined for 64 /. obsoleta of X shell length = 17.5 mm. The relationships between size and rate of locomotion were determined for an additional 31 snails (shell length = 7-20 mm) which were also tested separately on detergent-cleaned glass slides. Analyses of the morphometry of the foot of llyanassa were accomplished using photo- graphic reproductions of the feet of 20 snails (shell length = 9.1- 18.3 mm). The photo- graphic technique and its effectiveness for determining parameters of pedal morphology have been described (Dimock, 1984). Briefly, the procedure involved photographing the ventral surface of each snail as it crawled vertically up the side of a glass aquarium. Identically enlarged images of the snails’ feet and of standard units of area (graph paper) were cut out of the photographs. Linear di- mensions of the foot were determined by direct measurement of the respective photo- graphs, with length being the longest ante- rior-posterior dimension and width being the maximum width exclusive of the antero-lateral horns of the propodium. The area of the foot was calculated by comparing the weight of an image of the foot to the weight of a standard unit of photographic area. The wet weight in air (including the shell) was determined for all snails used in the size-rate analyses. The weights of 20 addi- tional snails (shell length = 8- 17 mm) were measured for animals suspended both in air and in sea water (31%. S) to permit the de- termination of immersed weight. The effects of the type of substratum on mud snail locomotion were assessed under three conditions for snails of mean shell length = 17.5 mm. First, a snail was timed crawling on a glass slide in the test chamber. Second, a different animal was timed as it crawled upon the surface of about 5 mm of clean sand (particle size <0.25 mm) that had been layered over a clean slide in the cham- ber. Finally, another snail was timed as it moved toward the light on the mucous trail that had been deposited by the snail which first traversed the sand. The dimensions of these snails relative to the size of the cham- ber insured that a snail crawled directly on the surface of the mucous trail. The influence of the duration of mainte- nance in the laboratory on the locomotion and overall activity of /. obsoleta was assessed by determining the rate of crawling on clean glass for groups of snails that had been held in the laboratory for 1, 2, 3 and 4 weeks. The level of activity of these animals was de- termined by reference to their tendency to traverse the prescribed distance in the test chamber within the allotted time. The data have been analyzed by one-way analysis of variance (ANOVA) and Student- Newman-Keuls multiple range test (SNK), as appropriate. Correlation between various parameters was assessed by Pearson's product-moment correlation (Zar, 1974). The data for the correlations of shell length with area of the foot and with the ratio of snail weight to area of the foot were first log trans- formed. The weight employed in the expres- sion of snail weight/unit area of foot was the estimated immersed weight. RESULTS The absolute rate of locomotion of /. obsoleta (mean shell length = 17.5 mm) whose speed on glass was determined 6 or 7 MUD SNAIL LOCOMOTION 167 days after being collected was. 1.96= 0.49 mm/sec (X + SD; N = 64). The max- imum rate recorded for this experimental group was 3.3 mm/sec for a snail of 17.3 mm shell length. The absolute rate was positively correlated with shell length (r = 0.375, df = 29, P = 0.038; Fig. 1). However, the relative rate of locomotion (% shell length moved/sec) was significantly negatively correlated with snail size (г = —0.833, df = 29, P <0.001; Fig. 2). The analyses of the morphometry of the foot of /. obsoleta revealed several associa- tions among shell length and the size and the shape of a snail's foot (Table 1). Not sur- prisingly, foot length (FL), foot width (FW), RATE (С mm/s ) a FESTES OA AS SE 18 and, consequently, foot area (FA) were all positively correlated with shell length. There was no correlation between shell length and the ratio of FL/FW, which remained constant at approximately 2.0 over the range of snail size examined (Table 1). The rate of increase of the area of the foot as a function of shell length (Table 1) was significantly greater than that predicted from scaling relative to the square of shell length (slope of regression of log FA on log SL >2; t-test, P = 0.027). The immersed weight of llyanassa was 53.0 + 3.9% (X + SD; N = 20) of a snail's weight in air. There was no correlation be- tween shell length and the percent reduction of snail weight upon immersion. 192209 SHELL LENGTH С mm ) FIG. 1. The absolute rate of locomotion (mm/sec) of I. obsoleta as a function of shell length (mm). 168 DIMOCK RATE С % SHELL LENGTH/Ss ) Té 9 ГО SHELL LENGTH ( mm ) FIG. 2. The relative rate of locomotion (% shell length/sec) of /. obsoleta as a function of shell length (mm). The immersed snail weight (SW)/unit area of foot (A) increased significantly with shell length (r = 0.857, df = 29, P <0.001; Table 1). However, the slope of the regression of log immersed SW/log A on log SL was signifi- cantly smaller than the 3/2 ratio that would be predicted due to scaling (t-test, Р <0.001). Mud snails crawling on glass were approx- imately 57% faster than snails crawling over clean sand (Table 2). The presence of a conspecific's mucous trail had no significant LS 14 15 1S 1? 1809820 effect on the rate of a snail's locomotion over sand (Table 2). In fact, the rate of crawling of |. obsoleta was not significantly different when snails traversed from O to 4 mucous trails overlaid upon one another on sand (Dimock, unpublished observations). The speed of mud snails on glass also is not significantly affected by the presence of a conspecific's mucous trail (L. Styers, unpublished observa- tions). The rate of locomotion of mud snails de- MUD SNAIL LOCOMOTION 169 TABLE 1. The association of shell length with morphometric parameters of the foot of llyanassa obsoleta. Variables Least-squares regression equation Г af Р ЕЕ MEER VIDAS 0.925 18 <0.001 Х = SL MESES Y = 0.43X — 0.14 0.924 18 <0.001 Х = SL Y = FL/FW Y = 2.04 — 0.002X 0.021 18 n.s. X = log SL Иод RAS Y = 2.23X — 0.83 0.957 18 <0.001 X = log SL Y = log SW/log A? Y = 0.50X + 0.31 0.734 18 <0.001 SL = shell length (mm); ?FL = foot length (mm); ®?FW = foot width (mm); “РА = area of foot (mm?); °SW/A = immersed snail weight (mg)/area (mm?) of foot. TABLE 2. The effects of the type of substratum on locomotion by Ilyanassa obsoleta.' Type of substratum Clean Clean Mucous trail glass sand on sand Rate of locomotion (mm/sec) : (X + SD, N = 25) 173 = 0.49 1100516 1.722.025 ANOVA: F = 27.6 df=2,72 P <0.001 Rates linked by horizontal line are not significantly different (SNK). 'Mean shell length = 17.5 mm (range = 16.6 - 18.2 mm). The snails had been in the laboratory 6 days. TABLE 3. Effects of maintenance in the laboratory on the speed and activity of llyanassa obsoleta.' Time in laboratory (weeks) 1 2 3 4 Rate of locomotion (mm/sec, X + SD) 2.19 + 0.49 1.45 = 0.28 0:99 = 0.23 1.09 + 0.34 N 39 41 34 35 ANOVA: F = 88.9 df = 3,145 Р <0.001 Rates linked by horizontal line are not significantly different (SNK). % of snails failing to reach test criterion 9.3% 4.7% 34.6% 32.7% N 43 43 52 52 ‘Mean shell length = 17.5 mm (range = 16.8- 18.1). creased significantly with continued mainte- about 50% of the rate after one week. This nance in the laboratory (Table 3). After three reduction in crawling speed with prolonged weeks in the laboratory the absolute rate of maintenance in the laboratory was accom- locomotion of /. obsoleta had decreased to panied by an increased tendency of snails to 170 DIMOCK be unresponsive to experimental manipula- tion. Fully % of the experimental animals failed to reach the response criterion after three or more weeks in the laboratory (Table 3). This decline in locomotory rate and overall activity was not affected by including clams and shrimp, in addition to marsh mud, in the diet of this facultative-scavenger deposit- feeding mud snail (Dimock 8 Styers, un- published observations). DISCUSSION The published techniques by which the speed of gastropods has been determined do not always facilitate inter- or intraspecific comparisons of parameters of locomotion. Not only may the experimental conditions under which speed is determined be in- completely controlled or inadequately de- scribed, but such rates may also be calcu- lated from the total distance moved during time intervals that could include lengthy cessation of locomotion (Calow, 1974; Bert- ness & Schneider, 1978). The data of the present study apparently comprise the first quantitative analysis of ciliary locomotion by a single species of gastropod in response to a controlled stimulus. The range of absolute speeds recorded for llyanassa obsoleta (up to 3.3 mm/sec) are comparable to the few published reports of locomotion by sometimes unspecified spe- cies of the family Nassariidae (Copeland, 1919; Miller, 1974a; Palmer, 1980). Linsley (1978) seems to have published the only re- cent data on a close and sometimes sympat- ric relative of /. obsoleta, Nassarius vibex (Say, 1822), but one has to assume, appar- ently as Palmer (1980) did, that Linsley's data are in mm/sec, since that author never speci- fied units for the ‘average speeds’ in his paper. In any event, rates of locomotion on the order of a few mm/sec seem to be typical of many small species of gastropods which employ ciliary locomotion (Miller, 1974a). The effect of size on the speed of ciliary- powered gastropods is not yet well resolved. However, the absolute rate of locomotion is positively correlated with shell length for /. obsoleta (Fig. 1). This relationship is similar to that described by Miller (1974a) for an assort- ment of unspecified gastropods from 15— 80 mm long that also employ ciliary locomo- tion. However, Miller's data also depict similar high rates of locomotion for species (in- dividuals?) both <15mm and >80 mm in length, with lower rates occurring in in- termediate size ciliary movers (Miller, 1974a, table IV). Certainly the fact that /. obsoleta can vary the activity of its pedal ciliation and concomitantly its speed (Copeland, 1919) could complicate an assessment of the relationship between size and speed for this or other ciliary-powered gastropods. It is clear, however, from Fig. 2 and the data of Miller (1974a) that small, ciliary-powered snails attain the fastest relative rates (shell length/sec) of all types of gastropod locomo- tion except leaping. The pedal surface area of /. obsoleta increases significantly more rapidly with in- creased shell length than is predicted by simple scaling (Table 1). If the density of locomotory cilia/unit area of foot is constant, the effect of this changing pedal area would be the provision of a rapidly increasing locomotory surface as the snails get bigger. This increased locomotory surface would be partly offset by the increasing weight/unit area of foot that also occurs as /. obsoleta gets larger. However, the ratio of weight/area of foot increases less quickly with increased shell length than scaling predicts (Table 1). The effective load that pedal locomotory cilia- tion must bear could be further reduced by the rather routine occurrence of gas bubbles in the mantle cavity and/or the gut of llyanassa (Kushins & Mangum, 1971); however, the contribution of such gas bubbles to the buoyancy of large and small mud snails has not been determined. Thus, it seems likely that the greater absolute speed of large ver- sus small //уапа$5а is primarily attributable to the increased pedal surface area that accompanies larger shell size. llyanassa crawls significantly faster on glass than it does on a sandy substratum (Table 2). A similar relationship between the type of substratum and the speed of ciliary- powered gastropods was reported by Miller (1974a) who demonstrated that several spe- cies of Cassis were 45-70% faster on lucite than on sand. The presence of a conspecific’s mucous trail on sand had no detectable effect on the speed of /. obsoleta (Table 2). Thus, the significance of the mucous trail-following behavior of this mud snail is not clearly re- lated to locomotion. However, the role of trail mucus vis-a-vis locomotion might be quite different under other experimental or environ- mental conditions of the type of substratum, the exposure of snails to currents, or of the MUD SNAIL LOCOMOTION 171 relative extent of immersion of mud snails in water. It may be that trail-following is more related to the energetic costs of mucus pro- duction and of transport (Calow, 1974; Den- ny, 1980; Houlihan & Innes, 1982). The lim- ited data of Hall (1973) suggesting that Litto- rina irrorata (Say, 1822) crawls faster on than off a mucous trail were interpreted by him as implicating trail mucus as an energy-saving device. The waning speed and concomitant in- crease in the percentage of animals exhibiting no response to the stimulus condition with increased duration of maintenance in the lab- oratory (Table 3) provide quantitative evi- dence for what heretofore had been sub- jective personal observation. From the first few days to a week or two after mud snails were brought into the laboratory, they are very active and spend a lot of time on the walls of aquaria. Later, however, their overall activity seems to diminish as they increas- ingly spend time on or in the substratum, rarely climbing the aquarium walls. It is clear that llyanassa is behaviorally quite different after extended maintenance in the laboratory. Although the notion that animals become more ‘abnormal’ with prolonged laboratory maintenance is not new, quantitative be- havioral evidence to that effect among marine invertebrates is not widely available. There is a plethora of evidence that starva- tion in the laboratory results in a reduction of metabolic activity and the assumption of a standard rate of metabolism by a variety of marine invertebrates, including gastropods (Newell & Roy, 1973; Bayne & Scullard, 1978; Newell & Branch, 1980). The decline in the activity and rate of locomotion of /. obsoleta does not appear to be simply a function of the availability of food, although this animal’s nu- tritional requirements may be complex (Cur- tis, 1979). Significant, rapid changes in the behavior of /. obsoleta as a consequence of maintenance in the laboratory should be cautionary to other investigators. ACKNOWLEDGMENTS | thank E. Lynn Styers for his able assis- tance with some of the work reported here. Drs. G.W. Esch and R.E. Kuhn provided helpful comments on an earlier draft of this manuscript. | especially appreciate the helpful suggestions of Dr. A.R. Palmer. This re- search was supported by a grant from the Wake Forest University Research and Publi- cation Fund. LITERATURE CITED BAYNE, B. L. & SCULLARD, C., 1978, Rates of oxygen consumption by Thais (Nucella) lapillus (L.) Journal of Experimental Marine Biology and Ecology, 32: 97-111. BERTNESS, M. D. & SCHNEIDER, D. E., 1976, Temperature relations of Puget Sound thaids in reference to their intertidal distribution. Veliger, 19: 47-58. BRETZ, D. D. & DIMOCK, R. V., Jr., 1983, Be- haviorally important characteristics of the mucous trail of the marine gastropod //yanassa obsoleta (Say). Journal of Experimental Marine Biology and Ecology, 71: 181-191. CALOW, P., 1974, Some observations on locomo- tory strategies and their metabolic effects in two species of freshwater gastropods, Ancylus flu- viatilis Müll. and Planorbis contortus Linn. Oeco- logia, 16: 149-161. COPELAND, M., 1919, Locomotion in two species of the gastropod genus Alectrion with observa- tions on the behavior of pedal cilia. Biological Bulletin of the Marine Biological Laboratory, Woods Hole, 37: 126-138. CRISP, M., 1969, Studies on the behavior of Nas- sarius obsoletus (Say) (Mollusca, Gastropoda). Biological Bulletin, 136: 355-373. CURTIS, L. A., 1979, On the broad nutritional re- quirements of the mud snail, //vanassa obsoleta (Say), and its polytrophic role in the food web. Journal of Experimental Marine Biology and Ecology, 41: 289-297. DENNY, M., 1980, Locomotion: the cost of gastro- pod crawling. Science, 208: 1288-1290. DIMOCK, R. V., Jr., 1984, Determining the area of a gastropod's foot. Veliger, 27: 93-96. DIMOCK, R. V., Jr. & PARNO, J. R., 1981, Bi- modal sensitivity to monochromatic light by the mud snail llyanassa obsoleta. Marine Behavior and Physiology, 7: 291-296. GAINEY, L. F., Jr., 1976, Locomotion in the Gastro- poda: functional morphology of the foot in Neriti- na reclivata and Thais rustica. Malacologia, 15: 411-431. HALL, J. R., 1973, Intraspecific trail-following in the marsh periwinkle Littorina irrorata Say. Veliger, 16: 72-75. HOULIHAN, D. F. & INNES, A. J., 1982, Oxygen consumption, crawling speeds, and cost of trans- port in four Mediterranean intertidal gastropods. Journal of Comparative Physiology, 147: 113- ИР KUSHINS, L. J. & MANGUM, С. P., 1971, Ве- sponses to low oxygen conditions in two species of the mud snail Nassarius. Comparative Bio- chemistry and Physiology, 39A: 421-435. LAWRENZ-MILLER, S., 1977, Locomotion in gas- tropod molluscs and evolution of the brain. An- 172 DIMOCK nals of the New York Academy of Sciences, 299: 26-34. LINSLEY, R. M., 1978, Locomotion rates and shell form in the Gastropoda. Malacologia, 17: 193- 206. MILLER, S. L., 1974a, Adaptive design of locomo- tion and foot form in prosobranch gastropods. Journal of Experimental Marine Biology and Ecology, 14: 99-156. MILLER, $. L., 1974b, The classification, taxonom- ic distribution, and evolution of locomotor types among prosobranch gastropods. Proceedings of the Malacological Society of London, 14: 233- 272. NEWELL, R. C. & BRANCH, G. M., 1980, The influence of temperature on the maintenance of metabolic energy balance in marine inverte- brates. Advances in Marine Biology, 17: 329- 396. NEWELL, R. C. & ROY, A., 1973, A statistical model relating the oxygen consumption of a mol- lusk (Littorina littorea) to activity, body size, and environmental conditions. Physiological Zoolo- gy, 46: 253-275. PALMER, A. R., 1980, Locomotion rate and shell form in the Gastropoda: a re-evaluaton. Malaco- logia, 19: 289-296. TROTE, TE JAS DIMOCK, В. М. IS/S: traspecific trail following by the mud snail llyanassa obsoleta. Marine Behavior and Physiology, 5: 91-101. ZAR, J. H., 1974, Biostatistical analysis. Pren- tice-Hall, Englewood Cliffs, 620 p. MALACOLOGIA, 1985, 26(1-2): 173-181 FUNCTIONAL ASPECTS OF TRAIL FOLLOWING BY THE CARNIVOROUS SNAIL EUGLANDINA ROSEA Anthony Cook School of Biological and Environmental Studies, New University of Ulster, Coleraine, Northern Ireland ABSTRACT Euglandina rosea (Ferussac) follows the trails of potential mates and of its gastropod prey. Conspecific trail following is decreased after copulation. Prey trail following is decreased after feeding and recovers in about nine days. Feeding also recovers over approximately the same time span. The direction in which Euglandina follows both prey and conspecific trails was influenced by the direction of illumination. In even illumination, there is no directionality in prey trail following. Snails discriminate between trails since they do not follow their own trails, nor do they persist in following the trails of less favoured prey. The elongated lips are used extensively for sampling the substrate and their surgical removal prevents trail following. The ability of the trail to promote trail following persists if it is kept dry for 24 h but a decay in this ability is detectable after soaking the trail in water for 30 min. Whilst the values of trail following measures are higher than have been reported for other pulmonates, the underlying features of this behaviour are broadly similar, making Еид/апата a suitable subject for further work on the characterisation of the active factors involved in pulmonate trail following. Key words: gastropod; behaviour; mucus; feeding; courtship; Euglandina. INTRODUCTION Trail following in gastropods is widespread (Cook, 1977). It is, however, not a universal feature of gastropod behaviour as attempts to demonstrate it in Lymnaea and Otala, for example, have failed (Cook, 1977; Chase et al., 1978; Bousefield et al., 1981). Some species such as Limax pseudoflavus Evans (Cook, 1977), Biomphalaria glabrata Say (Townsend, 1974; Bousefield et al., 1981) and Achatina fulica Bowdich (Chase et al., 1978) tend to follow trails only infrequently and not very far. In these species no clear-cut function for trail following has been demon- strated and vet it is these same species which are being used for the further analysis of this behaviour. Established functions for trail following in- clude trail blazing through soft substrates (Hall, 1973), aggregation (Lowe & Turner, 1976), prey location (Paine, 1963; Cook, in press) and courtship (Quick, 1946; Cook, in press). Also, many gastropods home and, where this has been examined closely, trail following is frequently involved. (Limpets— Cook et al., 1969; Cook, 1979b; Onchidium— McFarland, 1980 and, to a lesser extent, Limax pseudoflavus—Cook, 1979a, 1980.) Preliminary observations on Euglandina rosea indicated that, unlike the majority of terrestrial pulmonates, it shows clear-cut re- sponses to the trails of conspecific and prey species and also appears to have well- defined anatomical adaptations for trail following. Therefore, it might be a good model for the characterisation of active factors. Be- fore further work can be undertaken it is necessary to quantify the animal's trail follow- ing responses. MATERIALS AND METHODS Euglandina rosea with shells of between 3 and 4 cm in length were collected from the Botanic Gardens at the Florida Institute of Technology in Melbourne, Florida and main- tained in a laboratory at about 20°C. All observations were conducted between March and August, 1981. Each animal was kept ina separate container with a floor of damp filter paper and was fed individually. All experiments were conducted on poly- thene sheets in an open arena approximately 80 х 40 cm. For some of the experiments the laboratory illumination was uneven since the room lights were to the right of the area within (173) 174 COOK which the experiments were conducted. Most animals tended to move away from the light (i.e. from right to left over the experimental area). Under these circumstances all marker trails (i.e. the trails laid first) were laid from right to left. In some experiments the tracker animal was placed in the centre of the polythene, pointing towards the marker trail and about 1 cm from it. In these experiments, therefore, all tracker animals made a right angle contact with the marker trail and made a choice concerning the direction in which to follow the trail. In other experiments the track- er animal was placed on the end of the mark- er trail. The experiment was stopped when the tracker left the polythene. Mucus can be made temporarily visible by breathing on it. This proved to be the best method for observing traces of mucus left by the lips. The trail itself was made visible by immersing the polythene sheet in a suspen- sion of talcum powder. Two experimental formats were used. The first used polythene sheets of 70 x 20 cm. The marker animal laid a trail along the long axis of the sheet and the tracker animal was then placed near the centre of the trail. This format allowed the measurement of the fre- quency with which marker and tracker trails were superimposed, the length of that superimposition and the direction in which the marker trail was followed relative to the direc- tion in which the trail was laid. This ex- perimental format is illustrated in Fig. 3. It was used in experiments examining directionality and the effects of feeding and copulation on trail following. Deroceras laeve (Müller) (a limacid slug) was mostly used to lay these marker prey trails, but in a short series of experiments conducted in the U.K. the very similar Deroceras caruanae (Poll.) was used. Whilst most experiments were conducted in uneven illumination a short series was con- ducted in a symmetrical light regime. This was provided by closed arenas of 70 x 50 x 40 cm with matte black walls and white floors and ceilings illuminated by a centrally placed 30 W light 40 cm above the floor. The second format utilised 20 x 20cm polythene sheets. Marker trails were laid in the normal way. These smaller trails allowed a greater number of replicates to be per- formed but at the expense of the accuracy with which the distance followed could be estimated. To examine the specificity of trail following, three prey species were used to lay 20 cm marker trails. These were Deroceras laeve, Veronicella floridana Leidy (a systellommat- ophoran slug), and Philomycus carolinianus (Bosc) (an arionid slug). The tracker Eu- glandina was placed on the end of the marker trail and a record was made of whether the trails were subsequently superimposed, the length of the superimposition, and the length of the marker trail. The effect of the removal of the head ap- pendages was examined in a similar way. The operations were performed with iri- dectomy scissors on unanaesthetised ani- mals. The two pairs of head tentacles were easily removed but complete extirpation of the lips was impossible. These were, there- fore, amputated from the point where the lip left the lateral line of the body. Only one animal was available for each of these op- erations. After a one week recovery period each animal was tested on 20 cm conspecific trails twice per day for five days. The tracker animals were placed on the end of the marker trails and the proportion of the marker trail followed was measured. 20 cm trails were also used to examine the effect of soaking trails in water. Deroceras marker trails were either left dry overnight or soaked for 30m and then dried overnight. Euglandina was placed on the centre of the polythene, pointing towards the tracker trail but about 1 cm away from it. Their behaviour on making contact with the marker trail was noted. RESULTS The head of Euglandina is highly adapted for chemoreception (Fig. 1). The posterior (optic) tentacles, which in other pulmonates are associated with distance chemoreception, possess an accessory lower lobe beneath the eye. The lips are greatly extended and are extremely mobile. Fig. 2 shows an example of the contacts made by the lips with the sub- strate, A) during traii following, B) on encountering the end of a followed trail, and C) not trail following. This track illustrates firstly the large area of substrate covered by the lips alone during searching movements and secondly the regular pattern of lip dab- bling movements during locomotion. The fre- quency of lip/substrate contacts was mea- sured during locomotion, before and during trail following, and subsequent to the search- ing movements seen after the snail had lost TRAIL FOLLOWING BY EUGLANDINA 175 A Sie, hun , OG NS ‘ ‘ DOME SHOR N > L en FIG. 1. The head of Euglandina. The posterior (optic) tentacles have an accessory lower lobe beneath the eye and the lips are greatly extended. the trail (e.g. area C in Fig. 2). The distance over which such measurements are possible varies so all data were reduced to the number of contacts per cm and are given in Table 1. Lip/substrate contacts are significantly in- creased during trail following and during peri- ods of locomotion after losing the trail. Directionality Euglandina which had neither mated nor fed for at least 7 days were tested against 70 cm trails of Deroceras laeve and other Euglandina. Some trails were rotated through 180” after being laid so that they ran from left to right. The proportion of left turns is shown in Table 2 and the experimental format in Fig. 3. lt can be seen that for both conspecific and prey trails, turns are generally made to the left rather than relative to the direction in which the trail was laid. That is, the direction of trail following is determined not by trail cues but by other environmental cues; presumably in this case by differential illumination. As a check that directional responses to the trail were not being masked by an overriding TABLE 1. The frequency of lip/substrate contacts during locomotion. Mean no. of contacts/cm s.e. N t test Normal movement 4.23 0.23 8 (control) Trail following 5.47 0.4659 24177 After end of trail 5.87 0.572.972. 2:6677 SOLO TABLE 2. The frequency of left turns taken onto the marker trails shown in uneven illumination. The total number of replicates of each experiment is indicated (N). Trail orientation Rotated Normal 180° Marker CON) eon м x“ Euglandina 84 12 75 12 0 N.S. Deroceras We 43 56 25 176 COOK FIG. 2. The use of the lips in Euglandina. A) Shows the pattern of lip contacts whilst following the trail of Deroceras, B) shows a searching pattern executed by the lips after the Deroceras trail ended, and C) shows the lip contacts as the snail moved away. response to the direction of illumination, an experiment was performed which was similar to that described above but conducted in a symmetrical light regime. Deroceras caruan- ae was used to lay the trails. The results are given in Table 3 and there is no significant indication of the ability to detect the di- rectionality of the marker trail either measured by the direction of turning or by the distance followed. Specificity The ability of Euglandina to follow trails of various prey species was tested by placing the snail on the end of a 20cm trail. The marker species were Deroceras laeve, Philo- mycus carolinianus, and Veronicella florida- na. Individuals were also tested against their own and another Euglandina trail using the 70 cm experimental format. All followers had neither fed nor mated for at least 7 days. The results are shown in Table 4. There is a significant decrease in the fre- quency with which trails are followed to their ends between Deroceras and Philomycus (p <0.001). An apparent decrease in the length of the trail followed when responses to Veron- icella trails are compared to those of Dero- ceras is not significant. An individual follows its trail significantly less than it does that of another Euglandina (p <0.001). Effect of feeding Animals were starved for a varying length of time and then their trail following ability was tested. A limited number of animals was avail- able for these observations so each animal was used more than once. 73 observations were made in all and divided into six three- day starvation periods; i.e. 1-3 days (n = 18 TRAIL FOLLOWING BY EUGLANDINA Are = _§_§ii — BD ст — H,,\—,———_© 84% + — 75%. + -— ==“ —— 16% — 25% FIG. 3. The 70 cm trail experimental format. A marker trail was laid on a polythene strip 70 cm long and tracker snails placed about halfway along this trail. In this experiment, comparison between the behaviour on normal trails (above) and that on those rotated through 180° (below) shows no significant difference in the frequency of animals turning left. TABLE 3. The lack of directionality in the following of Deroceras trails in even illumination (N = 10). Direction relative Distance followed to marker % following mean + ъ.е. Forwards 50 26:6 227401 Backwards 50 UNES 10) t= 1.16; d.f. = 8 (N.S.) N.S.—Not Significant. observations), 4-6 days (n = 14), 7-9 days (n = 12), 9-12 days (n=9), 13-15 days (n = 11), and longer than 16 days (n = 8). Since each animal contributed to more than one of these groups the experiments were not run concurrently but data were accumulated over a period of several months. All starvation periods commenced with feeding to satiation and ended with the snail being tested for its ability to follow a 70 cm Deroceras trail before being offered a slug as prey. The frequency of following, the percentage of the available trail followed and whether the snail subsequently ate a Deroceras were recorded. The resuits are summarised in Fig. 4. Both measures of trail following and feeding activity were de- TABLE 4. Specificity of trail following in Euglandina. % followed No. of Marker % followed to end replicates Deroceras 88 80 25 Philomycus 53 2055 15 Veronicella 90 70 М.5. 10 A different Euglandina 100 100 24 The same Euglandina 8 055 24 ”"—p <.001; N.S. Not Significant. creased following a meal but all animals were back to normal after about nine days. Quantitative comparisons may be made for starvation periods of up to 12 days (53 observations). Significantly fewer animals fed in the first six days than in the last six. 02-962 dt = 16 p 0%) Signiticantly more snails followed trails if they subse- quently fed than if they did not feed (x? = 6.8; d.f. = 1; р <.01). Furthermore, considering only those snails which followed trails (34 observations), significantly more snails fol- lowed to the end of the marker trail if they fed than if they did not feed (x? = 13.6; 4.1. = 1:p <0.001). 178 COOK ho] SOM o 3 £ 100 Lo= y o. po Oo E O Oo © + sol ef? = $ 5 + Е Е a Bee 3 60 IR [= © о > 40 [ а. 20 Days since last fed FIG. 4. The effect of food deprivation on trail following. Measures of trail following (percentage of the trails followed and for those that were followed, the percentage of the available trail followed) and the frequency with which snails accepted food all increase with the increasing duration of starvation. Effect of copulation Euglandina was kept in isolation and readi- ly mated when two individuals were placed together. Each animal was allowed to lay two 70 cm trails. The first of these trails was used immediately to test the trail following ability of a different snail with which it was subse- quently allowed to mate. The next day the second trail was used to test the trail following ability of the mate. In these experiments the followers were placed near the middle of the marker trail so that when they turned to follow the trail approximately half of the marker trail was left unfollowed. This remaining half was tested with an unmated snail to determine whether the trail could still elicit normal trail following behaviour from an unmated in- dividual. Individuals which had mated the previous day were also tested on fresh mark- er trails laid by Euglandina with which they had not mated. The results are shown in Table 5. There is a significant reduction (p <0.001) in the frequency of trail following after copulation as well as a significant reduc- tion (p <.001) in the distance followed by those individuals that did follow trails. The decrease in trail following is not specific to the recent copulatory partner, nor is it attributable to a decline in the effectiveness of the trail. Both conspecific trail following and courtship behaviour had returned to normal after about seven days. No animal in these experiments laid eggs. It would be expected that before courtship the trail following frequency would be as high as it is in the control (unmated) animals. The deficit in Table 5 is attributable to two animals. During courtship one snail follows the trail of the other and then mounts the shell from the rear. The two snails which did not follow trails prior to courtship eventually adopted the pas- sive role of the mounted snail during courtship (i.e. they showed no trail following during courtship either). Trail persistence Euglandina was allowed to lay trails on 20 ст square pieces of polythene. Twelve such pieces were soaked for 30 min and a further thirteen left dry. All trails were dry when tested. Test snails were placed halfway along the trail pointing towards it but not in contact with it. Movements were classified as 1) IGNORED (where no examination of the trail with the lips occurred, 2) EXAMINED (where the marker trail had been investigated by the lips of the follower but where no trail following had taken place), 3) FOLLOWED (the trails of the marker and follower were superimposed but not for the whole distance available), and TRAIL FOLLOWING BY EUGLANDINA 179 TABLE 5. The effects of copulation on trail following. Each combination of marker and tracker animal was repeated 24 times. Frequencies are compared using x” and distances using ‘t tests. All comparisons are made with the unmated control (top line). Measure of trail following Frequency of following Distance followed Marker Tracker % cm (= s.e.) future partner partner before mating 92 36.9 + 3.4 future partner partner after mating 19 132622 55/5 = unmated partner stranger after mating РУ ПО future unmated partner stranger 100 N.S. A **—р <0.01. N.S.—Not Significant. TABLE 6. The responses to trails soaked for 30 min in water. Dry trails Soaked trails Response (n = 13) (n = 12) Ignored 0% 50% Investigated with lips 16% 0% Turned onto trail 8% 25% Followed to end 76% 25% TABLE 7. The effect of removing the head appen- dages (one animal for each treatment tested ten times). Appendage % of trails Proportion of removed followed trail followed None 100 0.85 Posterior tentacles 100 0.94 Anterior tentacles 100 0.88 Lips 40* 0.29** "*—Mann-Whitney U-test; р <0.01. *—y? = 5.9; р <0.05. 4) FOLLOWED TO END (where the maximum amount of trail following had occurred). These results are shown in Table 6. A x? test per- formed on these data comparing the frequen- cy of positive responses to the marker trail with the frequency with which the trail was ignored completely showed that there was a significant reduction in these responses after the trail had been soaked (x? = 4.58; 4.1. = 1; p <0.05). The effect of tentacle and lip amputation Four animals were used in amputation ex- periments. From one the posterior (optic) tentacles were removed, from another the anterior ones and from a third the lips were amputated. The fourth was left unharmed as a control. All animals had been used in prev- ious experiments and had behaved normally. After amputation the snails were less active than before but fed when a slug was pre- sented to them. Each animal was tested ten times on fresh 20 cm trails laid by normal Euglandina. The test animals were placed on the end of the marker trail and the percentage of occasions on which the trail was followed was measured. For those that followed part of the trail, the proportion of the available trail that they followed was also measured. These results are shown in Table 7. Whilst in- terpretation must be cautious because only one animal was used for each of the treat- ments, the results suggest that the lips are essential for the successful performance of trail following. DISCUSSION Euglandina is well suited for use as an experimental animal for the further analysis of the general features of trail following in pul- 180 COOK monates. Although this snail is highly adapted for trail following, the features of its trail following behaviour conform to those seen in less specialised pulmonates. Firstly, any di- rectionality exhibited by the tracking animal is not based on the way in which the trail was laid (Tables 2 and 3). This is similar to the pattern seen in other terrestrial pulmonates, e.g. Limax and Achatina (Cook, 1977; Chase et al., 1978). Directionality has been demon- strated in the trail following of the aquatic pulmonate Biomphalaria glabrata (Say) (Townsend, 1974) but it has been suggested that this may be attributed to the decay of the trail in water rather than to any directionality inherent in the way in which the trail was laid (Cook, 1977). Secondly, the ability of the trail to promote trail following declines in water over a similar time scale to that observed in other pulmon- ates (Table 6) (e.g. Biomphalaria—Town- send, 1974; Bousefield et a/., 1981; Physa— Wells & Buckley, 1972; Limax—Cook, un- published observations). Thirdly, because of the nature of their diet the hunger and there- fore the trail following of the animals is easily controlled (Fig. 4). Finally and most impor- tantly, trail following occurs regularly and at very high frequencies. A suitable Euglandina follows trails on over 90% of the occasions that it meets them. Limax on the other hand will only follow trails on approximately 30% of such occasions (Cook, 1977). The distances followed are also very high. Limax rarely fol- lows trails for more than 20 cm, whereas Eu- glandina rarely follows a suitable trail for less than this distance. In the present experiments when a Snail followed a trail it normally fol- lowed for the full 35 cm allowed and in pre- liminary experiments snails followed trails for more than one meter. Rough comparisons can be made by es- timating ‘coincidence index’ (Townsend, 1974) for Euglandina. The data presented in Table 5 (unmated animal following conspecif- ic trails) produce an estimated index of 0.88. This compares with a maximum value of 0.19 for Biomphalaria (Townsend, 1974), 0.66 for Mariaella dussumieri (Gray) (Ushadevi & Krishnamoorthy, 1980), 0.49 for Achatina (Chase et al., 1978), and an estimate of 0.26 for Limax (from data in Cook, 1977). In the experiments involving Achatina, Biomphalaria and probably Mariaella the trackers were placed at the end of, or in contact with, the marker trail. In the experiments with Eu- glandina and Limax however, the trackers were placed away from the trail so it might be expected that for these animals the coinci- dence index has been underestimated. Eu- glandina therefore follows trails far more fre- quently and for greater distances than any other pulmonate in which this behaviour has been measured. The regularity of trail following in Eu- glandina is related to its clear function. The frequency with which it follows prey trails and the decrease in the measures of trail following of less favoured prey (Table 4) and after a meal (Fig. 4) are all clear evidence of the close relationship between trail following and feeding. Similarly the high frequency of following conspecific trails but not its own (Table 4), the decrease in the measures of trail following after copulation (Table 5), and the initial contact during courtship being from the rear (Cook, in press) all indicate the inte- gral part that trail following plays in mate finding and courtship. These changes in behaviour after feeding or copulation in Euglandina fall into the gener- al pattern of changes associated with motiva- tional phenomena. Trail following is the initial part of the appetitive behaviour for both events. Once adequate consummatory acts have been performed the snails enter a quies- cent phase in which the appetitive behaviour is difficult to elicit. Furthermore the be- haviours measured during the feeding obser- vations recover in sequence—the trail follow- ing measures recovering faster than feeding itself (Fig. 4). After copulation the decrease in the measures of the appetitive behaviour is not specific to the original partner (Table 5). The Coolidge effect (Michael & Zumpe, 1978) is therefore not apparent in these observa- tions. Trail following then is an important part of the behavioural repertoire of this snail. This is reflected in structural adaptations for its per- formance. The organs most closely involved with trail following in other pulmonates are the lower (anterior) pair of tentacles (Limax— Cook, unpublished data; Mariaella—Usha- devi & Krishnamoorthy, 1980; Acha- tina—Chase & Croll, 1981). In Euglandina, however, the lips are greatly elongated and these are used for trail detection (Figs. 1, 2; Tables 1, 7). There appears to be no residual trail sensing ability in either pair of tentacles. The lack of trail-controlled directionality in following should be a considerable handicap if a snail is to find prey or mates by this means. However, snails are influenced in TRAIL FOLLOWING BY EUGLANDINA 181 their choice of direction by the same factors that influenced the laying of the original trail (at least as far as illumination is concerned) so this need not be such a serious problem as it would first appear. ACKNOWLEDGMENTS | thank Kerry Clark for his invaluable advice and hospitality. | am also indebted to other colleagues at the Florida Institute of Technol- ogy where part of this work was conducted. This work was supported by a Scientific In- vestigations Grant from the Royal Society of London. REFERENCES CITED BOUSEFIELD, J. D., TAIT, A. 1., THOMAS, J. D. & TOWNER-JONES, D., 1981, Behavioural stud- ies on the nature of stimuli responsible for triggering mucus trail tracking by Biomphalaria glabrata. Malacological Review, 14: 49-64. CHASE, R. D. & CROLL, R. P., 1981, Tentacular function in snail olfactory orientation. Journal of Comparative Physiology 143: 357-362. CHASE, R. D., PRYER, K., BAKER, R. & MADI- SON, D., 1978, Responses to conspecific chemi- cal stimuli in the terrestrial snail, Achatina fulica. Behavioural Biology, 22: 302-315. COOK, A., 1977, Mucus trail following by the slug Limax grossui Lupu. Animal Behaviour, 25: 774— 781. COOK, A., 1979a, Homing by the slug Limax pseudoflavus. Animal Behaviour, 27: 545-552. COOK, A., 1979b, Homing in the Gastropoda. Malacologia, 18: 315-318. COOK, A., 1980, Field studies of homing in the pulmonate slug Limax pseudoflavus (Evans). Journal of Molluscan Studies, 46: 100-105. COOK, A., in press, Diet, habitat and courtship of the carnivorous snail Euglandina rosea. Journal of Molluscan Studies. COOK, A., BAMFORD, O. S., FREEMAN, J. D. B. & TEIDEMAN, D. J., 1969, A study of the homing habit of limpets. Animal Behaviour, 17: 330-339. HALL, J. R., 1973, Intraspecific trail following in the marsh periwinkle Littorina irrorata Say. Veliger, 16: 72-75. LOWE, E. F. 8 TURNER, R. L., 1976, Aggregation and trail-following in juvenile Bursatella leachii plei. Veliger, 19: 153-155, 1 pl. McFARLANE, |. D., 1980, Trail-following and trail- searching behaviour in homing of the intertidal gastropod mollusc, Onchidium verruculatum. Marine Behaviour and Physiology, 7: 75-108. MICHAEL, R. P. & ZUMPE, D., 1978, Potency in male rhesus monkeys: effects of continuously receptive females. Science, 200: 451-453. PAINE, R. T., 1963, Food recognition and predation on opisthobranchs by Navanax inermis. Veliger, 6: 1-9: QUICK, Н. E., 1946, British slugs (Pulmonata; Tes- tacellidae, Arionidae, Limacidae). Bulletin of the British Museum of Natural History, Zoology, 6(3): 103-226. TOWNSEND, С. R., 1974, Mucus trail following by the snail Biomphalaria glabrata Say. Animal Be- haviour, 22: 170-177. USHADEVI, $. V. 8 KRISHNAMOORTHY, В. V., 1980, Do slugs have silver track pheromone? Indian Journal of Experimental Biology, 18: 1502-1504. WELLS, М. J. & BUCKLEY, 5. К. L., 1972, Snails and trails. Animal Behaviour, 20: 345-355. MALACOLOGIA, 1985, 26(1-2): 183-189 THE ORGANISATION OF FEEDING IN THE CARNIVOROUS SNAIL EUGLANDINA ROSEA Anthony Cook School of Biological and Environmental Studies, New University of Ulster, Coleraine BT52 1SA, Northern Ireland ABSTRACT The types and sequences of behaviour involved in feeding on different types of prey are described. The sequences of behaviour were similar for small snails, large snails and slugs, though they were modified in response to the defensive tactics of the prey. The process of feeding can be conveniently divided into an attack, consumption of the prey and clearing up. The time allocated to each of these phases for each prey type is presented. Small snails eaten whole were consumed most rapidly (1 min). Slugs took longer (1.5 min) since time was allocated to clearing up mucus and other debris (1 min). Snails that were too large to be eaten whole but whose body weight was similar to that of the slugs took a considerable time to be consumed (15 min). Approximately two thirds of this time was taken in grazing over the empty shell after the soft parts had been eaten. Key words: Gastropoda; feeding; behaviour; Euglandina. INTRODUCTION Euglandina rosea (Férussac) is a predatory snail native to the SE U.S.A. Its general biolo- gy has been studied as part of programmes to introduce the snail into tropical islands as a control measure against large pest species such as Achatina fulica Bowdich (Chiu & Chou, 1962; Davis & Butler, 1964; Mead, 1979). Whilst there have been some de- scriptions of its behaviour (e.g. courtship— Cook, in press a; feeding—Cook, 1984, and trail following—Cook, 1985) these accounts have not included detailed analyses of its feeding behaviours when presented with dif- ferent prey. The present paper compares in detail the feeding behaviour on a range of prey. MATERIAL AND METHODS Sixteen Euglandina rosea snails with shells between 3 and 4 cm long were collected from the Botanic Gardens at the Florida Institute of Technology in Melbourne. The prey used was Deroceras laeve (Muller), a limacid slug about 2 cm long, Polygyra septemvolva Say, a snail with a flat shell about 7 mm in diameter and Succinea campestris Say, a bulbous snail with a shell about 10-12 mm long. A repre- sentative sample (6) of each prey type was taken, the shells dissolved in 1M HCI and the remaining soft parts dried to constant weight at 60°C. The mean weight of Deroceras and Succinea were similar (.03 g) but Polygyra weighed considerably less (.007 g). Observations were conducted on an open white surface approximately .4 x .8 т. This surface was covered with clear polythene which was changed between prey items so that the predator was not confused by irrele- vant trails. The euglandinas used in these experiments had been starved for at least seven days. The prey was allowed to walk a short distance and then an active Euglandina was placed on its trail. All experiments there- fore, commenced with trail following. For each observation the sequence of events was re- corded using a tape recorder. The tape was subsequently transcribed and the timing of each event noted. Observations were made on the consumption of 33 Deroceras, 30 Polygyra and 20 Succinea. RESULTS Classification of predator behaviour The behaviours recorded were as follows. 1. Trail following (TF). The predator followed the prey trail which was normally visible as a damp streak. This behaviour is accompanied by increased lip activity. (183) 184 COOK 2. Contact (Con). The predator made contact with the prey. Normally the initial contact was made by the posterior (optic) tentacles. 3. Eversion (Ev). The odontophore and upper lip of the snail were everted producing a large white balloon- like structure protruding from an area roughly bounded by the tentacles and the lips (Fig. 2C). 4. Strike (Str). The everted mouthparts were brought down rapidly onto the prey. A successful strike resulted in penetration by the radular teeth and the pinning of the prey. 5. Swallow (Sw). The prey or part of it was drawn into the protruded mouthparts. 6. Mop up (Mop). The everted mouthparts were moved over the substrate or the prey shell after the body of the prey had been consumed. 7. Inversion (Inv). The mouthparts were inverted. 8. Search (Se). The anterior of the body of the predator was raised and either waved in the air or moved over the substrate. This body activity was accompanied by increased tentacle and lip activity. 9. Moved off (Off). The predator moved away from the site at which the prey had been eaten or ceased to trail follow in cases where the prey escaped an attack. 10. Pick up (P.U.) The prey shell was lifted on the anterior portion of the foot. This action did not neces- sarily detatch the prey's foot from the sub- strate. 11. Move down (M.D.) The lifted prey was moved down the foot and the head of the predator bent over the top of the prey shell (Fig. 3C). 12. Rotate (Rot). The prey shell was rotated. This is an un- satisfactory behaviour to define since the prey shell was frequently rotated little by little in the performance of other behaviours. Occa- sionally this shell manipulation occurred by itself. 13. In and out (In, Out). In the case of larger prey the head of the predator was inserted into the shell of the prey and the flesh eaten from within. Details of the predators' behaviour within the prey shell were not consistently clear. 14. Escape (Esc). The predator lost physical contact with the prey. Prey that escaped completely did so by leaving the experimental area. Only Dero- ceras escaped once the initial contact had been made. Behavioural sequences The sequences of behaviours recorded were cast in transition tables and flow di- agrams of the sequences constructed (Fig. 1). By their definitions it is not possible for the behaviours to be randomly organised and therefore no test for non-randomness was conducted. 1. Succinea. Succinea campestris responds to being attacked by withdrawing into its shell. Its large foot allows it to adhere firmly to the substrate. These snails are, however, weak compared to Euglandina and attachment was never a successful defence. Once attacked, Succinea was always eaten. The most fre- quent sequence of events during feeding was as follows. After the approach from the rear, Euglandina moved from right to left over the surface of the prey shell (Fig. 2A). The shell was picked up on the foot and the prey was commonly exposed with its foot attached to the substrate and the shell angled upwards (Fig. 2B). This revealed the columellar muscle of the prey. In these circumstances the first strike was at the columella. The first strike results in the prey detaching from the sub- strate and withdrawing into the shell if it had not already done so (Fig. 2C). After this with- drawal the prey shell is held between the predator’s foot and the ground. Occasional rotations of the prey shell occur on the foot which result in the prey shell being better placed for the predator to enter. The normal position of the prey at first is with its spire to the right with respect to the Euglandina (Fig. 2A-C). During the consumption of the ex- posed part of the prey the prey shell is gradu- EUGLANDINA FEEDING 185 FIG. 1. A flow diagram of the behaviours involved in feeding on A) Succinea campestris, B) Polygyra septemvolva and C) Deroceras laeve. The width of the lines indicates the frequency of transition from one behaviour to another expressed as a percentage of the number of animals involved. The behaviours concerned are: TF—Trail following; Con—Contact; P.U—Pick up; M.D—Move down; Ev.—Evert; Rot— Rotate; Str—Strike; Sw.—Swallow; Mop—Mop up; Inv—Invert; Se—Searching; Off—Moving off; In—Head in prey shell; Out—Head out of prey shell; Esc—prey escapes. ally turned so that the spire is on the left. When Euglandina enters the shell to remove the remnants of the prey it therefore normally enters with its foot over the columella (Fig. 2D). In 35% of attacks an initial strike at the columellar muscle was effective, the prey ex- tracted from its shell whole and eaten outside the shell. Failure to strike at the columellar muscle was a result of the premature with- drawal of the prey into its shell or a poorly directed first strike. After eating the soft parts of the prey the mouthparts were inverted only to be re-everted immediately or after some prey shell manipulation. Re-eversion was fol- 186 COOK FIG. 2. Euglandina feeding on Succinea. The predator approaches over the body whorl (A) and lifts the shell (B) before everting its mouthparts and striking in the region of the columella (C). The predator inserts its head into the prey shell to eat the remnants of the prey (0). The Euglandina shell length is approximately 3 cm. lowed by a thorough cleaning of the prey shell both inside and out. Feeding on Succinea was normally ended by extensive searching movements prior to moving off. 2. Polygyra. The defensive response of Polygyra septemvolva is to withdraw deep into the body whorl of its shell. The aperture of the shell is protected by the small tooth. These passive defenses are ineffective, how- ever, against Euglandina of the size used and all Polygyra attacked were eaten. After the predator made contact with the prey (Fig. 3A) its shell was lifted. Because of the weak foot the whole animal was always lifted from the substrate. The Euglandina reared up with the shell attached to its foot (Fig. 3B). The prey shell was then moved down the foot and occasionally rotated (Fig. 3C) before the mouthparts were everted. The mouthparts were moved over the prey and the shell swal- lowed whole (Fig. 3D, E). There is no strike. Small shelled prey are lifted clear of the sub- strate and eaten whilst attached to the foot (Fig. 3F, G). Larger prey shells rest on the substrate whilst being eaten. The prey shell can be seen moving back in the 'neck' of the predator. It is digested whole. After swallow- ing, the mouthparts were inverted and the animal moved off with no mopping up and little searching. 3. Deroceras. Deroceras laeve is a small slug capable of rapid locomotion. After the initial contact the predator partially overtook the fleeing slug (Fig. 4A) and everted its mouthparts (Fig. 4B). The initial strike was at the tail or the mantle. This strike normally pinned the slug and it was swallowed in one or two pieces (Fig. 4C, D). After the prey had been consumed the predator mopped up the debris (Fig. 4E) before inverting the mouth- parts and moving off. When attacked, Deroceras responded with an active defence consisting of rapid locomo- tion and two specific behaviours. The first and most common was the tail flick in which the rearmost portion of the slug was lifted and EUGLANDINA FEEDING 187 FIG. 3. Euglandina feeding on Polygyra. Contact is made with the prey shell (A). It is lifted (B) and then moved down the foot (C). The mouthparts are everted (D) and the prey is consumed whole (E). Small prey are lifted clear of the ground (F) and eaten off the foot (G). The Euglandina shell length is approximately Suem: flicked rapidly from side to side in a slapping motion. The second was a flaring of the man- tle over the head. These two behaviours will be considered together as ‘Flick’ for the pur- pose of analysis. Because the slugs moved rapidly and dis- tracted the predator with these mantle and tail movements the strike sometimes missed or sliced off the rear completely, allowing the prey to escape. Escape was normally fol- lowed by the inversion of the mouthparts and the resumption of trail following. The defensive behaviour of the slug has a considerable effect on the success of the attack (Fig. 5). Forty-two percent of all strikes resulted in the prey escaping. Only 12 of the 33 prey animals flicked their tails or flared their mantles. Of these, 9 escaped once or more and 6 of these escaped completely. Only two animals escaped without flicking or flaring but both these were damaged in an attack and escaped whilst the snail was mopping up mucus and other debris. Comparison of behaviour patterns— handling times The total time taken to consume a prey item is made up of various components. Table 1 shows the time taken from the initial contact to eversion of the mouthparts (attack time); the time from that eversion to when the soft parts were completely consumed (eating time), and from that time to when the Eu- glandina moved off (clearing up time). When slugs were the prey, only those attacks in which the prey did not escape are included. These three times represent three distinct phases of an attack which are common to all prey types. 1. Attack time. The time taken to subdue a Polygyra was short because they were never 188 COOK TABLE 1. Handling times of Euglandina with different prey species. (Means + s.e.) Time (sec) Prey type Attack time Eating time Clearing up time Total (N) Succinea LOTES 266 + 39 631 + 219 933 + 216 (20) Polygyra E 28/16 LOS 67-27, (39) Deroceras 26+ 8 ЕЕ 54 = 9 94 = 12 (19) FIG. 4. Euglandina feeding on Deroceras. The slug is attacked after being overtaken (A). The initial strike is frequently made at the mantle (B) and the prey sucked into the mouth in one or two bites (C—D). After the prey has been swallowed the debris is mopped up (E). The Euglandina shell length is approximately 3 cm. firmly attached to the substrate and were raised clear of it. Subduing a slug took longer because slugs invariably moved rapidly and the predator slowed down to strike. Attacking large snails took the longest time since the predator examined the shell and manoeuvred to make a standard approach across the body whorl from right to left before lifting it. 2. Eating time. The time at which the soft parts of a Succinea were completely con- sumed within the shell was taken as the time at which the clearly visible, dark digestive gland disappeared from the shell apex. The time taken to consume this snail depends upon the site of the initial attack. In seven of the 20 attacks observed the initial strike se- vered the columellar muscle and the prey was extracted from its shell whole. For all attacks on Succinea the mean time (+ s.e.) taken to eat the exposed part of the prey was 107 + 16 sec (n = 20). In cases where the remnants of the digestive gland had to be extracted from the shell a further 213 + 37 sec (n = 13) were expended before the complete con- sumption of the prey. The time taken to eat a EUGLANDINA FEEDING 189 FIG. 5. A flow diagram of the events occurring during feeding on Deroceras showing the re- sponses of the slug and its effect on its survival. The conventions are as in Fig. 1. The additional behaviour is: Fli—prey flicks tail or flares mantle. small prey snail is all taken up by the process of extending the mouthparts over the shell. Slugs were eaten extremely rapidly. 3. Clearing up time. Euglandina took the longest time to clear up after eating Succinea. Of the final 613 sec, 464 + 139 was spent moving the everted mouthparts over the sur- face of the prey shell. The remaining time was spent in similar activity on the substrate and in searching behaviours. Polygyra was eaten whole and left no de- bris and no Euglandina ever performed mopping up behaviours after eating this spe- cies. Amean of 21 + 3 sec of the final 25 sec was spent after the inversion of the mouth- parts and before moving off. There was little searching behaviour and this static time allowed the passage of the uncomfortably large prey to the stomach. Slugs left a considerable amount of debris since the initial strike normally splits the di- gestive gland which spills out. Also, slugs produced a copious mucus as a response to being attacked. Of the final 54 sec 43 + 8 sec was therefore spent with the mouthparts everted. DISCUSSION Most pulmonates are herbivores and in some instances their feeding behaviours and their physiological bases have been de- scribed in detail (e.g. Helisoma—Kater, 1974; Lymnaea—Rose & Benjamin, 1979). In general they have cyclical feeding patterns involving repeated odontophore movements integrated with regular swings of the head from side to side (Dawkins, 1974). Some pul- monates are facultative predators consuming annelids and fellow gastropods as the oppor- tunity arises (Cooke, 1895). In these cases the regular feeding behaviour is supple- mented by other behaviours. Rollo & Welling- ton (1979) describe aggressive encounters between slugs. Some of their descriptions of the behaviour of Limax maximus L. bear a striking resemblance to the behaviour of Eu- glandina, e.g. “Following this initial ‘tasting’ the aggressor suddenly bit the victim, striking simultaneously with its everted radula and jaw.” A “rear and lunge” behaviour is also described. The structures used by slugs to bite in these encounters appear to be the same as those used by Euglandina, though the latter has much longer, slicing, radular teeth (Solem, 1974) and the everted mouth- parts are much larger. These behaviours in Limax maximus may not be primarily preda- tory since they occur seasonally, coinciding with breeding and may be better viewed as territorial behaviours (Rollo & Wellington, 1979). Nevertheless the predatory behaviour of Euglandina is clearly a specialised version of behaviours which appear in non-predatory pulmonates. There are few descriptions available of the feeding behaviour of other specialised preda- tory pulmonates with which to compare that of Euglandina. Solem (1974) states that pul- monates with specialised stabbing radulae (e.g. Testacella) ‘harpoon’ their prey, whilst a New Zealand predatory snail, Paryphanta, apparently smothers its prey by enveloping it in the folds of its foot and dragging it into the body whorl. Most gastropods are passive prey and, once caught, play no active part in determin- ing the behaviour of the predator. Some potential prey items have passive defences such as distasteful mucus or an extremely tough integument (Cook, in press a). The active defence behaviours of Deroceras laeve are those seen commonly in the limacid slugs (Rollo & Wellington, 1979). Tail flicking, man- tle flaring and rapid movement often lead to misdirected strikes and the frequent escape of the slug. In the present experiments ap- proximately 25% of slugs escaped com- pletely. In the field the proportion of escaping slugs is likely to be higher since the slug can avoid further attacks by hiding in cracks too 190 COOK small to accommodate the shell of Eu- glandina. The handling times of Succinea, which was the largest snail used in the present work (Table 1) seem out of proportion to the benefit gained from its consumption since its soft parts are about the same size as those of the Deroceras. The greatest gain in energy per unit time actually spent feeding therefore is probably derived from Deroceras but the pre- cise relationships between handling times and prey preferences are problematical. The prey of Euglandina normally lives aggregated and at high densities (Cook, 1984). It would seem to be, therefore, of no advantage to take up so much time in the complete con- sumption of prey tissues as is seen with Suc- cinea. There have been no experimental studies of prey preferences in Euglandina, however, and therefore there is no sound basis for the interpretation of the extended handling time of larger prey. ACKNOWLEDGMENTS This work was supported by a Scientific Investigations Grant from the Royal Society of London. It is a great pleasure to acknowledge the hospitality and assistance of Kerry Clark at the Florida Institute of Technology. | am also grateful to Tony Pitcher for his comments on an early draft of the manuscript. REFERENCES CITED CHIU, S. C. & CHOU, K.-C., 1962, Observations on the biology of the carnivorous snail Euglandina rosea. Bulletin of the Institute of Zoology, Academia Sinica (Taipei), 1: 17-24. COOK, A., 1985, Functional aspects of trail follow- ing by the carnivorous snail Euglandina rosea. Malacologia, 26: 173-181. COOK, A., in press a, Courtship in the carnivorous snail Euglandina rosea (Férussac). Journal of Molluscan Studies. COOK, A., 1984, Feeding by the carnivorous snail, Euglandina rosea (Ferussac). Journal of Mollus- can Studies, Suppl. 12A: 32-35. COOKE, A. H., 1895 [1959 reprint], Molluscs, In Cambridge Natural History, HARMER, S. F. & SHIPLEY, A. E., eds., 3: 1-459. Wheldon & Wesley, Codicote, England. DAVIS, С. 4. & BUTLER; G: BD Jr, 1964, №- troduced enemies of the giant African snail, Achatina fulica Bowdich, in Hawaii (Pulmonata: Achatinidae). Proceedings of the Hawaiian Entomological Society for 1963, 18: 377-389. DAWKINS, M., 1974, Behavioural analysis of coor- dinated feeding movements in the gastropod Lymnaea stagnalis (L.) Journal of Comparative Physiology, 92: 255-271. KATER, S. B., 1974, Feeding in Helisoma trivolvis: the morphological and physiological bases of a fixed action pattern. American Zoologist, 14: 1017-1036. MEAD, A. R., 1979, Economic malacology, with particular reference to Achatina fulica. In Pul- monates, FRETTER, V. & PEAKE, J., eds., vol. 2B, Academic Press. London, x + 150 p. ROLLO, C. D. & WELLINGTON, W. G., 1979, Intra- and inter-specific agonistic behaviour among terrestrial slugs (Pulmonata: Stylommatophora). Canadian Journal of Zoology, 57: 846-855. ROSE, R. M. & BENJAMIN, P. R., 1979, The relationship of the central motor pattern to the feeding cycle of Lymnaea stagnalis. Journal of Experimental Biology, 80: 137—163. SOLEM, G. A., 1974, The shell makers— introducing mollusks. Wiley, New York, xii + 289 p. MALACOLOGIA, 1985, 26(1-2): 191-200 CAUSES OF LIFE HISTORY VARIATION IN THE FRESHWATER SNAIL LYMNAEA ELODES Kenneth М. Brown’, Dennis В. DeVries? & Bonnie К. Leathers?* ABSTRACT Intraspecific life history variation in the pond snail Lymnaea elodes Say was studied in three ponds in northeastern Indiana. Population densities fluctuated more dramatically in the tempo- rary ponds due to high juvenile mortality caused by unpredictable drying times. Shell growth rates were greater in a more productive, permanent pond. In a reciprocal transfer experiment, snails in the most productive pond, regardless of origin, grew roughly twice as fast, and laid eight to nine times as many eggs as snails in the less productive, temporary ponds. However, snails originating from the most vernal pond always had slower growth rates, smaller shell lengths at maturity, and higher fecundities than other snails reared in the same pond. Although proximal factors like habitat productivity therefore explain much of the intraspecific life history variation, genetic divergence among populations is still discernible. The lower productivity and uncertain nature of vernal ponds apparently favor early maturity and high fecundity in this freshwater snail. Key words: Lymnaea; life history variation; proximal and evolutionary causes. INTRODUCTION Intraspecific life history variation is often considered as the result of natural selection producing local adaptations in life history tac- tics. Stearns (1976) summarized the pro- posed selection forces, as well as how they are predicted to produce covariation in life history traits. However, there are other expla- nations for intraspecific life history variation. Variation may be the result of environmentally induced phenotypic changes such as de- velopmental plasticity, or physiological acclimation (Stearns, 1980). Second, varia- tion (or the lack of it) may be due to phylo- genetic constraints caused by past evolution- ary history. For example, Calow (1978) sug- gests that egg size may be phylogenetically limited in fresh-water prosobranchs. Finally, intraspecific variation may indeed be due to differing selection regimes among habitats. It is therefore necessary to determine the degree of genetic basis to intraspecific life history variation before we can rule out any of these alternatives. One technique is to use quantitative genetic methods to determine the heritabilities of and genetic correlations among life history traits. However, this approach is often difficult under field con- ditions. Another technique is to perform reciprocal transplant experiments. That is, if individuals are transferred among habitats, and compared to residents, the relative im- portance of genetic and environmental factors can be discerned. For example, Berven (1982) was able to show that water tempera- ture was the most important factor explaining intraspecific life history variation in wood frogs from different elevations. However, genetic variation did occur in some life history traits, often opposing the environmentally induced variation (“countergradient selection”). Thus both proximal and evolutionary factors may be important in explaining life history varia- tion, and more studies need to be done to extend our knowledge of the relative im- portance of each. The objective of the current study is to perform such an analysis for the freshwater pulmonate snail Lymnaea elodes (= palus- tris) Say. We present sampling data on food resource levels, population dynamics, and shell growth rates in three pond populations in northeastern Indiana. We then reciprocally transplant snails among all three populations in a field rearing experiment, and compare growth rates, shell size at maturity, fecundi- ties, and egg weight and caloric content among ponds and populations. We also rear snails from all three populations under con- "Crooked Lake Biological Station, Indiana-Purdue University at Fort Wayne, 2101 Coliseum Blvd., East, Fort Wayne, IN 46805, U.S.A. “Department of Zoology, Ohio State University, Columbus, Ohio, U.S.A. “Department of Ecology, Ethology, and Evolution, University of Illinois, Champaign, Illinois, U.S.A. (191) 192 BROWN, DEVRIES 8 LEATHERS stant conditions in the laboratory to determine if contrasts remain among populations and thus have a genetic basis. Lymnaea elodes is a good candidate for such a study. Populations vary in productivity, length of life cycle, and fecundity, possibly because of differences in habitat productivity (Hunter, 1975). On the other hand, Forbes 8 Crampton (1942) reported considerable in- traspecific variation in life histories, which they considered to be genetic, since variation remained after several generations of labora- tory rearing. Finally, the density of adult snails may also determine fecundity in L. elodes (Eisenberg, 1966, 1970). The species is fairly common in ephemeral habitats in the northern United States and Canada (Harman 8 Berg, 1971; Brown, 1979). It has an annual life cycle, with breed- ing in late spring to early summer. Juveniles overwinter in temporary ponds by burrowing into the soil, and adults can survive for a second year by forming epiphragms (Eisen- berg, 1966; Jokinen, 1978). Populations are bivoltine in extremely productive habitats (Hunter, 1975) and univoltine in temporary ponds (Brown, 1979). Lymnaea elodes feeds primarily on periphyton, although it also uti- lizes carrion (Brown, 1982). METHODS Sampling of habitats The three ponds are within 30 km of the Crooked Lake Biological Station, 33 km NW of Fort Wayne, Indiana. The ponds are drawn to scale in Fig. 1. Surface areas of the tempo- rary ponds (A, B) vary, but the areas shown are representative for the breeding period of L. elodes. Pond A is the most vernal, drying by mid July. It, like the other ponds, has a muck substrate (a mixture of clay and organic detritus). Food resources for snails are both allochthonous (leaf litter, decaying grasses) and autochthonous (periphyton composed mostly of diatoms and blue-green algae). Pond B dries by late July or early August. Food resources are decaying terrestrial grasses that invade the pond during the dry season, and periphyton. Pond F 1$ per- manent, and food resources for snails are mostly autochthonous (periphyton, duck- weed, submerged plants). The ponds have similar levels for water temperature, pH, and dissolved oxygen (Brown, 1982). Pond B has AREA= .016 Ha De g E N X CROOKED STARKE STATION POND A AREA= 1.58Ha POND F AREA = .81 Ha / =—| ar FIG. 1. Location and surface area of the three ponds sampled. Ponds A and B are temporary; pond F is permanent. harder water, and lower dissolved phosphate levels (19 ug/f vs. 44 ug/t in pond A and Е). Since nutrient levels differ among ponds, we test whether periphyton biomass also varies among ponds. Periphyton biomass ас- cumulation was determined with a Wildco.. periphyton sampler. From 3 to 8 replicate, preweighed, slides were removed at each date during a study of biomass accumulation. Slides were dried overnight at 60°C in a dry- ing oven, and weighed on an analytical bal- ance (sensitivity = .1 mg). Values reported are dry biomass per slide + s.e. Ponds were quantitatively sampled at bi- weekly intervals during the field season, start- ing in 1978 in A and B, and 1979 in F. Sampling areas were allocated randomly throughout the pond, but no area was sam- pled twice on the same day. An Ekman dredge (sampling area = .05 m*) removed vegetation and the first few cm of substrate. Hauls were pooled in groups of four to form a sample. Preliminary sampling was done early in the spring to determine the number of replicate samples needed at each date. Early spring was chosen because snail densities were lowest then, and sampling variances highest. In general, enough samples are re- quired in benthic studies to reduce the ratio of LYMNAEA LIFE HISTORY VARIATION 193 PONDA e PONDB O PONDF @ 35 35 30 30 we a = NS 25 a w a 2 o = 79 3 20 2 x т ~ О 15 5 15 о 7h > = 10 < w = on a 0 0752747678710 No. OF SAMPLES 0 0—2 4 6 8) 10 Мо. OF SAMPLES FIG. 2. Sampling variation for preliminary Ekman dredge sampling of the three ponds, plotted against the number of replicate samples. the standard error to the mean to less than 20%, or to a point where the mean density ceases to fluctuate (Green, 1979). After 10 replicate samples (40 dredge hauls), Fig. 2, standard errors were less than 20% of mean densities in all 3 ponds, and mean densities were stable in ponds A and B. Ten replicate samples were therefore taken at each date. Adult snails were removed first from sam- ples, and the remaining vegetation was hand sorted, and juveniles and egg cases re- moved. Samples containing mud were washed through a series of sieves. A sample of 50 egg cases was counted at each date and the mean number of eggs per case was multiplied by the number of cases per sample to estimate egg abundances. Egg, juvenile, and adult counts were converted to a m* basis to determine changes in population dynamics. Growth in shell length was estimated by following cohorts in the ponds. The age of the cohort (symbolized X) was the time since the peak of egg production in the appropriate breeding season. Examination of shell length histograms indicated a sigmoid relationship between shell length and age: little growth in the first fall and winter, rapid growth the next spring, and slowing of growth after the onset of reproduction. Differences in shell growth rates among populations were therefore es- timated by fitting growth curves to a sigmoid function: ЕС E exps(C2—-C3 00) The parameter Cl is equivalent to the final shell length, the parameter C3 is proportional to shell growth rates, and the parameter C2 has no simple biological analogue. Values of parameters were fit with an iterative non- linear technique (Conway et al., 1970), and non-linear 90% confidence intervals were used to determine whether parameter con- stants differed among populations. Experimental methods The effect of resource abundance was de- termined by rearing snails in each of the three ponds. The efiect of snail density was de- termined by rearing snails at densities of two and four snails per container. We tested for genetic differences in tactics by rearing two populations in each pond. Snails from the most temporary pond (Pond A) were reared in each pond. Comparison populations in ponds B and F were resident snails, the pond F snails were used in pond A. The design was completely randomized with a factorial ar- rangement of treatments (3 ponds x 2 pop- ulations per pond x 2 densities per popula- tion). We used 1 { plastic containers, with op- enings covered by aluminum screens to minimize fouling (Brown, 1979, 1982). Alumi- num ions were not released into solution as the pH of the ponds seldom dropped below 6 (Brown, 1982). Containers were attached to floats to allow air space for the air breathing pulmonates. There were 15 replicate contain- ers in each of the 12 treatments. Two to four immature snails (4-7 mm) were introduced to the containers on May 9-12, 1980. Each week containers were removed and the snails measured, 5g of fresh pond vegetation added, and all egg cases removed. The ex- periment was terminated on July 17, as pond A was drying. Analysis of covariance was used to remove variation (due to shell growth before the start of the experiment) in fecundity per snail, shell length at maturity, egg weight in mg, and growth increments in shell length. Fecundity data were also log-transformed to remove a mean-variance correlation. 194 BROWN, DEVRIES 8 LEATHERS Three replicate egg samples from each treatment were combusted in a Phillipson microbomb calorimeter to determine caloric investment in eggs. Separate samples were ashed overnight at 550°C in a muffle furnace to convert caloric data into calories per mg ash free dry weight (A.F.D.W.). To test the hypothesis that caloric expenditures in eggs were independent of pond productivity, pop- ulation, and density, all data were fit to a common regression line of calories against A.F.D.W. of sample. We also tested whether population con- trasts in life history traits would remain under common conditions with a laboratory rearing experiment. Snails were reared in controlled temperature cabinets maintained at 21°C + 1°C (an average temperature during growth periods in the pond) and a 12:12 light cycle. Pairs of immature snails were placed in aer- ated 3 { aquaria, with 15 aquaria from each population. Fresh water and 10g of fresh pond vegetation were added bi-weekly. Pond vegetation in all experiments was a mixture of grasses and leaves from the ponds, dried to kill eggs and small snails. It provided both a raw food resource and a substrate for per- iphyton colonization in the aquaria and con- tainers. No supplements were added since they may artificially bias growth and fecundity results (Eisenberg, 1970). Shell length and egg laying rates were monitored weekly. RESULTS Field sampling The accumulation of periphyton biomass on submerged slides differed among the three ponds (Fig. 3). Pond B, with lower dis- solved phosphate levels, had the lowest per- iphyton biomass. Biomass accumulation was similar in the other two ponds; production was somewhat higher in pond A up until 10 days, but increased no further. Accumulations in pond F continued to increase (Fig. 3). The two temporary ponds dried at this point, but biomass in F was still greater after 30 days. Thus, in terms of standing crop of periphyton, the three ponds would be ranked F = A > B. Shell growth was related to differences in periphyton biomass among ponds (Table 1). The value of parameter 3 (proportional to growth rate) was highest in F, somewhat low- er in A, and significantly lower in pond B. Non-linear confidence intervals for parameter 3 Ln PERIPHYTON BIOMASS (MG) e POND А OPOND B @ POND F 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 DAYS FIG. 3. Natural logarithms of periphyton biomass accumulation (X + s.e., п given in each case) on submerged slides in each pond. 3 overlapped slightly in ponds A and F, but not with pond B. The model predicted snails in pond F to reach shell lengths of 20.6 mm during their first year, 19.8 mm in pond A, and only 14.1 mm in pond B. The smallest repro- ductive snails are about 14 mm in L. elodes (Brown, 1979) and so snails may not reach reproductive size in their first year in the least productive pond. However, final adult shell lengths in the three populations (parameter 1) overlap broadly (Table 1), because of a difference in life cycle length. Two year old snails could seldom be found in pond F, but cohorts could be followed in the two temporary, less pro- ductive ponds for two or three seasons. Snails in the temporary ponds therefore reach LYMNAEA LIFE HISTORY VARIATION 195 TABLE 1. Parameter estimates for fit of shell growth to a sigmoid function and 90% non-linear confidence intervals. Non-linear 90% Ratio of explained to Pond Parameter Value confidence interval total sum of squares A C1 2146 23.2-32.1 96% C3 .026 .024—.030 B C1 28.9 26.1 3128 97% C3 .005 .004—.006 Е C1 26.3 22.922918 99% C3 .031 .029-.035 10,0004 t 10,0005 4 = : t POND A oo | А EGGS N | | 4 O JUVENILES | 1,0004 POND B A MID ADULTS | W A EGGS | | k | | O JUVENILES j и | | ADULTS | 62) LP | | |. San | | | | | À Я 100! Ze I EA = | | = ike | | Е e co = | \ | x I 1 = INTA! | | : = | A | 3 \ | “Ad о, one | | | И ha \ | | A | Log | | \| | | Wiis | | | | s el Sr т Е r Jl + т 1 С 4 a 4 5 6) 7 LS 6 4 5 6 7 4 AZ 1978 1979 1980 1978 1979 1980 FIG. 4. Semilogarithmic plots of densities of adults, juveniles, and eggs through time in Pond A. Values are means = s.e. (n = 10). the same final shell lengths, but at a later age due to their slower growth rates. The abundances of eggs, juveniles, and adults fluctuated dramatically through time, especially in the two temporary ponds (Figs. 4, 5). Egg abundances were usually highest in pond A, often near 10,000 per m?. Juvenile densities were usually near 100 per т? in the most vernal pond, while adult densities were near 20 per m°. A crude estimate of survivor- ship to maturity (simply adult density divided by egg density) would be near 0.2%. Inspec- tion of seasonal trends indicates that egg FIG. 5. Semilogarithmic plots of densities of adults, juveniles, and eggs through time in Pond B. Values are means + s.e. (n = 10). laying usually peaked in early June, and that no eggs survived over the winter (Fig. 4). Juvenile and adult densities increased during the season, due both to recruitment and the shrinking of the pond. Adults apparently over- wintered much more poorly than juveniles. The 1979 field season was very short with the pond drying by mid June, and even juvenile densities were much lower in 1980, indicating that early pond drying significantly lowers ju- venile survivorship. Egg densities in pond B were somewhat lower, usually peaking around 5,000 per m? 196 BROWN, DEVRIES 8 LEATHERS 1,0007 | POND F | A EGGS || 4 | OJUVENILES | FR | ADULTS al D || a || 0 = a | \ Е | \ a 1a a = ON ES =) eh N | | Z | A | < | | | = Ой | п | | JN | | | | | | | Ne net O EN As 7 1979 1980 FIG. 6. Semilogarithmic plots of densities of adults, juveniles, and eggs through time in Pond F. Values are means + s.e. (n = 10). (Fig. 5). Juvenile densities varied dramatically from year to year, but an average would be around 10 per m”, similar to the adult density. Thus an estimate for survivorship to maturity would again be near 0.2%. Adult survivorship was fairly high over the winter of 1978-1979, but was very poor after the short field season of 1979. Again as in the most vernal pond, densities of juveniles and eggs increased dur- ing the season due to recruitment and shrink- ing of the pond. In contrast, egg densities never reached above 500 m? in the permanent pond (Fig. 6). Juvenile densities averaged around 30 per m”, and adult densities around 10 per mí. A crude estimate for survivorship in the perma- nent pond would then be approximately 2.0%. Of course, this survivorship estimate could be biased upwards if eggs or juveniles were be- ing preyed upon at a higher rate in the per- manent pond. Experiments In the field rearing experiment, snails grew rapidly in the most productive pond (Fig. 7), and growth increments decreased linearly in the other two ponds. Pond productivity had a highly significant effect on shell growth, as did snail density (Table 2). However, snails from the most vernal pond still had significantly lower growth rates than comparison pop- ulations, at each site and density (Fig. 7). The significant pond A x density interaction term (Table 2) indicates pond A snails suffered greater decreases in growth rates at the high- er densities. In summary, judging from the relative magnitude of F ratios (Table 2) habi- tat productivity had the greatest effect on growth increments, followed by population contrasts, and density. The average coeffi- cient of variation for growth increments, over all treatments, was 31.9%. Habitat productivity had even more dramat- ic effects on snail fecundity (Fig. 7). Snails in pond F grew roughly twice as large, but laid on the average eight times as many eggs. Initial shell size also had a significant effect on fecundity, unlike growth increments (Table 2). Density also had a highly significant effect on fecundity. Population contrasts in fecundities were also significant (Table 2). Snails from the TABLE 2. Table of F values from ANCOVA of 1980 field container experiment. An asterisk indicates significance at .05 level, two at .01 or more. Growth Shell length Fecundity Egg Source of variation increment (mm) at maturity (mm) per snail weight Initial size (covariate) .58 7298 64.34** 3.47 Pond 63.327” 28 1018 81165 4.38* Pond A snails vs. others 29.84** 68.63** 6.46* 3.04 Density 14.60** 13545 23.48** <1 Pond * pond A snails 2.85 2.79 <<] <1 Pond x Density <1 1.50 <1 4.73* Pond А x density 5.095 <1 il <1 Pond A x density x pond 1.84 1275 <1 <1 LYMNAEA LIFE HISTORY VARIATION GROWTH (MM) FECUNDITY PER SNAIL 197 FIG. 7. Response surfaces for shell growth increments (left) and fecundity (right) for snails reared in 3 different ponds, and at two snail densities per container. Ponds are arranged from left to right in order of declining productivity. Response surfaces for pond A snails have open circles, comparison populations are solid circles. Dotted lines indicate where response surfaces lie below each other. Data are means + s.e. for each treatment (n = 15). most vernal pond had significantly higher fecundities than comparison populations, regardless of rearing site or density. Judging from the value of F ratios, habitat productivity had the greatest effect on fecundity, followed in order by initial shell length, density, and population contrasts. Over all treatments, the average coefficient of variation for fecundity was three times higher than for growth rates, (106.5%). Shell length at maturity was also a function of habitat productivity (Fig. 8, Table 2). Snails matured at smaller sizes in the less pro- ductive ponds. Snails from the most vernal pond also reproduced at significantly smaller shell lengths. Finally, both density and initial shell length had significant effects on size at maturity. In contrast, mean dry egg weights varied extensively within treatments (Fig. 8), and only the pond main effect and the pond- density interaction were significant in the analysis of covariance (Table 2). There was a general trend of decreasing egg weight with declining habitat productivity, although not marked. The pond-density interaction oc- curred because mean egg weight decreased with density in pond A, but remained the same or increased in the other two ponds. Although in most cases pond A snails laid less massive eggs (Fig. 8), the high variation within treatments (coefficient of variation = 136%) kept contrasts from being significant (Table 2). Finally, neither habitat productivity, source population, nor snail density affected caloric content per mg A.F.D.W. of eggs (Fig. 9). Over all treatments, the coefficient of variation in calories per mg A.F.D.W. was only 12.8%. In a separate analysis of variance the effects of habitat productivity (p = .27), population contrast (p = .42), and density (p = .34) were all insignificant. Although snails from the temporary pond tended to lay more eggs in the laboratory rearing experiments, no significant differ- ences occurred due to high variances within populations (Fig. 10). A one way analysis of covariance (with initial shell length) indicated that neither shell growth increments (p = .54), 198 BROWN, DEVRIES 8 LEATHERS SHELL LENGTH (MM) FIG. 8. Response surface for mean dry egg weights (left) and shell length at maturity (right) for snails reared at 2 densities in each of the ponds. Symbols as in Fig. 7. Standard errors are not included for egg weights due to the high coefficient of variation (see text). 607 11 010 y A | 210 = 1 = 030 A2 о 1 /32 < 7: Oo у \ = -1.97 Е 4.45х 20 = Bs 8 r= .96 6 11 => SS LF See 5 10 15 MG. ASH FREE DRY WEIGHT FIG. 9. Regression of caloric content in eggs against mg A.F.D.W. of sample. Numbers 1-4 are from snails reared in pond А, 5-8 in В, and 9-12 in F. In each pond the first 2 numbers are resident snails at 2, and 4 snails per container, and the latter 2 are transferred snails. final shell length (p = .08), nor fecundity per snail (р = .33) differed significantly among populations. Again, coefficients of variation for fecundity (169%) were much higher than for growth rates (30%). DISCUSSION Habitat productivity explained a consider- able proportion of intraspecific variation in life history tactics of L. elodes. In the field rearing experiment, snails grew roughly 1.6 times larger in pond F than in pond A, and 1.8 times larger than in the least productive pond B. Fecundities were 8.1 times greater, averaged over both source populations, in pond F than in pond A, and 9.2 times greater than in pond B. Habitat productivity also increased shell size at maturity, but did not have as large an effect on mean weight of eggs, or caloric content in eggs. Shell growth data collected from field populations also indicated that snails grew faster in the more productive ponds and had shorter life cycles. Overall, resource availability is an impor- tant proximal determinant of life history varia- tion in snails. Hunter (1975) also found varia- tion in productivity and voltinism to be a func- tion of habitat poductivity in L. elodes, and Browne (1978) found the same for the freshwater prosobranch Viviparus geor- gianus. The length of time resources are available during the life cycle may also be important (Aldridge, 1982). Snail density also had significant effects on LYMNAEA LIFE HISTORY VARIATION 199 N (>) o ga INTIAL SHELL LENGTH O FINAL SHELL LENGTH (MM) o SHELL GROWTH (MM) on 100 50 FECUNDITY/SNAIL POND FIG. 10. Life history variation for snails reared in a common laboratory environment. Initial and final shell lengths are at the top, growth increments in the center, and fecundity per snail at the bottom. Data are means = s.e. (n = 15). growth rates and fecundities. Effects on fecundity were especially large, with in- crements in density reducing fecundities, averaged over both source populations, 63.2% in pond F, 82.4% in pond A, and 79.3% in pond B. Eisenberg (1966, 1970) suggested that adult densities regulate fecundity in this species. Our data indicate this to be more the case in temporary ponds, where periphyton biomasses were lower. Genetic divergence among populations ex- plained a comparatively small proportion of the variation in life histories. However, snails from the vernal pond always grew more slow- ly, matured at a smaller shell length, and had higher fecundities than other populations. These smaller differences may still be impor- tant over evolutionary time scales; little is known, for example, about the amount of gene flow among snail populations. In- terestingly, Berven (1982) also found a large proximal component to intraspecific life his- tory variation in ranid frogs, caused by tem- perature variation among ponds at different altitudes. Frogs from higher altitudes still grew more rapidly than low altitude frogs in the same pond, suggesting genetic adaptation produced by “counter-gradient selection.” Hence the pattern was similar to this study: most variation explained by proximal factors, but still a discernible genetic component. The selection forces responsible for such intraspecific genetic variation in most cases are still unknown. Calow (1981) noted that Lymnaea peregra from exposed sites ma- tured earlier and had higher reproductive effort. Differences also remained in the lab- oratory, Suggesting genetic origin. Calow con- sidered that exposed populations were r- selected due to greater density independent mortality. However, other studies of life his- tory variation in molluscs do not report traits covarying as expected from r- and K-theory (McCleod et al., 1981, Hart & Begon, 1982). In the current study, proximal factors ex- plain most of the variation in life histories among populations, although genetic differ- ences are still important. In temporary ponds, low resource levels reduce growth rates and fecundities. The combination of low productiv- ity and unpredictable drying dates have fa- vored the evolution of high reproductive rates in these vernal ponds, as well as early ages and sizes at maturity. ACKNOWLEDGMENTS This research was supported by NSF grant DEB 81-03539 to the senior author. All work was completed at the Crooked Lake Biologi- cal Station. EITERANVUREICHEB ALDRIDGE, D. W., 1982, Reproductive tactics in relation to life cycle bioenergetics in three natural populations of the fresh water snail, Leptoxis carinata. Ecology, 63: 196-208. BERVEN, K. A., 1982, The genetic basis of altitu- dinal variation in the wood frog Rana sylvatica. I. An experimental analysis of larval development. Oecologia, 52: 360-369. BROWN, K. M., 1979, The adaptive demography of four fresh water snails. Evolution, 33: 417-432. BROWN, K. M., 1982, Resource overlap and com- 200 BROWN, DEVRIES 8 LEATHERS petition in pond snails: an experimental analysis. Ecology, 63: 412-422. BROWNE, R. A., 1978, Growth, mortality, fecun- dity, biomass and productivity of four lake pop- ulations of the prosobranch snail, Viviparus geor- gianus. Ecology, 59: 742-750. CALOW, P., 1978, The evolution of life-cycle strat- egies in fresh water gastropods. Malacologia, 17: 351-364. CALOW, P., 1981, Adaptational aspects of growth and reproduction in Lymnaea peregra (Gastro- poda: Pulmonata) from exposed and sheltered aquatic habitats. Malacologia, 21: 5-13. CONWAY, С. T., GLASS, М. В. & WILCOX, J. C., 1970, Fitting non-linear models to biological data by Marquardt's algorithm. Ecology, 51: 503-507. EISENBERG, R. M., 1966, The regulation of den- sity in a natural population of the pond snail Lymnaea elodes. Ecology, 47: 889-906. EISENBERG, R. M., 1970, The role of food in the regulation of the pond snail, Lymnaea elodes. Ecology, 51: 680-684. FORBES, G. D. 4 CRAMPTON, H. C., 1942, The differentiation of geographical groups in Lym- naea palustris. Biological Bulletin, 82: 26-46. GREEN, R. H., 1979, Sampling design and statisti- cal methods for environmental biologists. Wiley, New York. HERMAN, W. N. & BERG, C. O., 1971, The fresh water snails of central New York, with illustrated keys to the genera and species. Search (Cornell Agricultural Station), 1: 1-67. HART, A. & BEGON, M., 1982, The status of general reproductive strategy theories, illus- trated in winkles. Oceologia, 52: 37-42. HUNTER, R. D., 1975, Growth, fecundity, and bioenergetics in three populations of Lymnaea palustris in upstate New York. Ecology, 56: 50— 63. JOKINEN, E. H., 1978, The aestivation pattern of a population of Lymnaea elodes (Say). American Midland Naturalist, 100: 43-53. McCLEOD, M. J., HORNBACH, D. J., GUTTMAN, S.1., WAY, С. М. & BURKY, А. J., 1981, Environ- mental heterogeneity, genetic polymorphism, and reproductive strategies. American Natural- ist, 118: 129-134. STEARNS, S. C., 1976, Life history tactics: a re- view of the ideas. Quarterly Review of Biology, 51: 347. STEARNS, S. C., 1980, A new view of life history evolution. Oikos, 35: 266-281. MALACOLOGIA, 1985, 26(1—2): 201-211 SCANNING ELECTRON MICROSCOPY OF THE BODY SURFACES OF BIOMPHALARIA GLABRATA Laila A. Aboul-Magd' & Samia A. Sabry Department of Microbiology and Immunology, School of Medicine, University of California, Los Angeles, CA 90024, U.S.A. ABSTRACT The body surfaces of small and large snails of the Puerto Rican strain of Biomphalaria glabrata were studied by scanning electron microscopy. The whole surface of the snail is covered by cilia except the epidermal surface of the anterior part of the head and the mantle surface covering the visceral mass. On large snails, the dorsal surface of the head away from the bases of the tentacles is irregularly folded and has a spongy appearance, while it is smooth with regular folds on small snails. The ventral and dorsal surfaces of the mantle collar of large and small snails are covered by dense cilia and microvilli. In addition, a basal layer on the external surface of the mantle collar, covered by microvilli, was only observed on large snails. Bulbous structures with an opening to the exterior were found on the surface of the pneumostome. Their function is unknown. Several globular structures are extended from the surface of the sole of the foot, probably representing mucous secretions from the subepidermal tissue. These structures were more pronounced in large snails. The surface of the sole of the foot is densely covered by cilia which may play a role in distributing the slime secretions over the surface of the foot. Key words: Biomphalaria glabrata; scanning electron microscopy; body surfaces. INTRODUCTION The anatomy and general histology of freshwater Pulmonata have been studied by several authors (Pan, 1958; Bolognani Fantin & Vigo, 1967a, b). Most of this research has used the light microscope, and very little has been done on the ultrastructure of these snails—especially dealing with the epidermal surfaces. It is known that the epidermis of freshwater gastropods has several functions such as respiration (Zaaijer & Wolvekamp, 1958; Jones,1961), osmoregulation (Green- away, 1970, 1971), and perception (Jager, 1971). Zylstra (1972) studied the epidermis and the associated subepidermal gland cells of the freshwater snail Biomphalaria by means of histochemical and electron micro- scope techniques. He reported that the single-cell-layered epidermis is composed of general epidermal cells, cilia, and a few scat- tered goblet cells. Sullivan et al. (1974) stud- ied the ultrastructure of the rectal ridge of Biomphalaria glabrata. The structure and function of the mantle cavity of Biomphalaria glabrata were studied by Sullivan & Cheng (1974); they found that the histology of the rectal ridge (a single layer of columnar epithe- lium) supports its role in the uptake and elimination of substances by the snail. They also reported that the pseudobranch has no respiratory function, but acts as a valve to close the chambers of the mantle cavity on chemical insult. The purpose of our study was to investigate the body surface structure by scanning elec- tron microscopy (SEM) of young and old Biomphalaria glabrata (Say), the intermediate host of Schistosoma mansoni, with the hope of adding information to some aspects of the host-parasite relationships of schistosome miracidia and their intermediate hosts. MATERIALS AND METHODS Snails of the Puerto Rican strain of Biom- phalaria glabrata were maintained in an aer- ated aquarium in spring water and were fed fresh lettuce. The diameter of the shell was "Present address: 11 Fahmy Matar Street, El Zaitoun, Cairo, Egypt. (201) 202 ABOUL-MAGD 8 SABRY measured and taken as an indicator of age. To measure the diameter, we started from the lateral edge of the upper margin of the aper- ture, across the shell to the other side, on the dorsal surface of the snail. The diameter range of snails studied was 1-11 mm. In preparation for SEM, a snail was put in a small petri dish with a little spring water, and the diameter was measured. Then, the shell was crushed by application of pressure by another, smaller petri dish. The shell frag- ments were removed and the snail was fixed in 2% buffered glutaraldehyde overnight. Then it was washed for 3 changes in spring water, 10 minutes each. The remaining steps in specimen prepara- tion for the SEM were carried out following Voge et al. (1978). Snails were scanned over their whole surface, including the exposed part of the foot, head, tentacles and the man- tle collar. The size of the snail was recorded on each photograph. For light microscopy, snails were fixed in 10% formalin or Carnoy's fixative, embedded in paraffin and sectioned at 4-6 um. Slides were stained with haematoxylin and eosin or Barbeito-Lopez trichrome stain. OBSERVATIONS As the whole surface of the snail is covered by Cilia, the distribution of these cilia, their density on different parts of the snail body, and the presence of microvilli in both small and large snails were studied. The body of Biomphalaria glabrata is di- vided into head, foot, mantle region and visceral mass. The head is not well de- marcated from the body; it bears the tenta- cles, eyes, mouth, lips and jaws. The epidermal surface covering the two lips and the jaws is smooth. The mouth has a dorsal appendage, or protrusion, which has a sur- face like the jaws, and is presumably the median jaw (Fig. 1). The dorsal surface of the head shows a variable distribution of ciliated epidermal cells. In some parts it is densely covered by long cilia, as at the edge of the head of large and small snails (Fig. 2); the cilia gradually become reduced to little patches around the base of a tentacle (Fig. 3). The epidermal surface covering the rest of the dorsum of the head of large snails is nearly devoid of cilia except for widely separated tufts of short cilia. The surface appears spongy, with irregular folds and many pores (Fig. 4). On the other hand, the dorsal surface of the head of small snails has regularly-spaced folds with a few tufts of short cilia (Fig. 5). The undersurface of the head is similar to that of the head near the base of the tentacle. A tentacle is a gradually tapering cylinder, attached to the dorsolateral surface of the head. Its whole surface is covered by epidermal cells having long cilia (Fig. 6) that are very dense at the tip and the upper third (Fig. 7), while decreasing in den- sity toward the base of the tentacle where much of the epidermal surface is devoid of cilia (Figs. 8, 9). The distribution and density of the cilia are the same in small and large snail tentacles. The mantle is composed of a single layer of epithelial cells which were seen in the section obtained for histology. It embraces the neck of the snail and covers the pallial region and visceral mass. The enlarged glandular part of this mantle, often visible along the rim of the shell aperture, is known as the mantle edge or mantle collar. An accessory structure of many folds situ- ated externally at the junction of the head with the mantle collar, known as the pseudo- branch, is Suggested to have a respiratory function (Malek & Cheng, 1974). The surface of the pseudobranch is covered by heavy, long cilia. The short siphon-like pneumostome is lo- cated near the pseudobranch. The surface of this structure is covered with short cilia and microvilli. Globular structures which possess openings suggestive of gland cells are scat- tered between the cilia (Fig. 18). In small snails (1.0-3.0 mm), the mantle collar is flat- tened over the surface of the body while in large snails, it is deflected dorsally, exposing its ventral surface. The ultrastructure of the dorsal surface of the mantle collar of small and large snails shows slight differences. ——-— FIGS. 1-5. Head of Biomphalaria glabrata. 1. SEM of head anterior of 9.5 mm snail, showing the edge of the head (EH), foot (F), jaws (J), lips (L), mouth (M), and tentacle (T). Scale bar 100 um. 2. SEM of edge of head of 11 mm snail, showing dense cilia. 3. SEM of surface of head of 5.5 mm snail around base of tentacle, showing tufts of cilia. 4. SEM of head surface of 8 mm snail away from base of tentacle, showing irregular folds with spongy appearance. 5. SEM of the same area as Fig. 4 of 2 mm snail, showing few tufts of cilia and smooth regular folds of surface. Scale bar for Figs. 2-5 3 um. SEM OF BIOMPHALARIA BODY SURFACES 203 204 ABOUL-MAGD 8 SABRY ÓN A ee a" Wa er ve ua | 4 mn ЖИМ AVION 4 о RASE Ae A ET FIGS. 6-9. Tentacle of Biomphalaria glabrata. 6. SEM of whole tentacle of 3 mm snail, showing distribution of cilia. Scale bar 50 ит. 7. SEM of tip and upper third of tentacle of 3 mm snail, showing dense long cilia. 8. SEM of middle third of tentacle of 3 mm snail, showing scattered groups of long cilia. 9. SEM of lower third of tentacle of 1.5 mm snail, showing very few cilia. Scale bar for Figs. 7-9 3 um. SEM OF BIOMPHALARIA BODY SURFACES E = +. E =23 Е IE Wo Se + 206 ABOUL-MAGD 8 SABRY SEM OF BIOMPHALARIA BODY SURFACES 207 In small snails, the dorsal surface shows three different areas with different types of epidermal cells (Fig. 10). The epidermal sur- face at the edge of the collar has a rim of heavy cilia (4 um in length), followed by a folded, bare surface which has microvilli on it (Fig. 11). The edge of the second area is bulbous, while the rest of the area is covered by evenly distributed microvilli (Fig. 12). The junction of this area with the third area shows big patches of short cilia with smooth surfaces in between. The surface of the third area is completely covered by microvilli followed by an aggregation of many blebs covering the whole surface (Fig. 13). In large snails (Fig. 14), the edge of the mantle collar is covered with very dense, long cilia. The ventral sur- face of the mantle collar near the edge has many long cilia scattered between the folded surface (Fig. 15). The basal surface of the mantle collar near its junction with the head is completely covered with microvilli and many tufts of long cilia. Globular structures, prob- ably secretions from goblet cells, are seen in this area (Fig. 16). In large snails, the dorsal surface of the mantle collar, folded backward, has struc- tures similar to those of the small ones, ex- cept for the presence of a basal layer on the external surface of the mantle collar This layer contains short microvilli arranged to show the outline of the underlying cells. Many globular secretions are observed in this area (Fig. 17). The mantle covering the rest of the body has a smooth surface and is covered by very short, thin microvilli. The whole of the ventral surface of the snail is composed of the foot, which forms the typical creeping sole. As the sole of the foot is the part of the body on which the snail de- pends for its movement, its surface is usually covered by an excess of mucous secretions appearing as a mesh of entangled filaments. In addition to the very densely ciliated epidermal cells covering this surface (Fig. 19), the sole surface of large snails also shows many widely distributed lobular struc- tures which seem to arise from deeper sites and protrude in between the cilia. These structures have different shapes and their surfaces vary from smooth to rough with re- ticular appearance (Fig. 20). They probably are secretions extruded from the underlying cells. In small snails, there are only small lobular structures probably representing sub- epidermal glands. The density of the cilia decreases toward the dorsum of the foot. On the dorsal surface of the foot of small snails, long single cilia are scattered among tufts of short cilia. The surface is slightly folded and smooth in appearance (Fig. 21). On the other hand, many irregular folds hav- ing tufts of short cilia are observed on the dorsal surface of the foot of large snails (Fig. 22). Cytological studies of the surface con- firmed the light microscope observations on the structure of the epidermis. However, not all the epithelial cells covering the surface are ciliated, particularly those cells covering the head and dorsum of the foot. The epidermis of B. glabrata is one layer thick, consisting mainly of four cell types: columnar cells, cili- ated columnar cells, pigment cells and goblet cells. The columnar and the ciliated columnar epithelial cells have oval or elongated nuclei with many chromatoidal bodies. The cyto- plasm of these cells is granular and basophil- ic. The ciliated columnar epithelium covering the sole of the foot is more densely stained with Barbeito-Lopez than any other part of the body (Fig. 23). A thick layer of cuticle covers much of the mouth epidermis and buccal cavity, which is thickened into the jaws (Fig. 24). The lips surrounding the mouth are also covered with the same cuticular layer that appears smooth by SEM. The epidermal cells of the tentacle are short columnar and ciliated. This observa- tion was also made by Pan (1958). As he also showed, the cilia are more dense at the tip and the upper third than on other parts. The surface of the mantle collar is com- posed of pseudostratified epithelial cells in some regions, and simple squamous epithe- lial cells in others, as already observed by Pan (1958). {eee FIGS. 14-18. 14. SEM of dorsum of head of 8.0 mm snail, showing edge of head (EH), foot (F), deflected mantle collar (MC), and tentacles (T). Scale bar 200 um. 15. Surface of edge of mantle collar showing dense and long cilia. 16. Basal part of mantle collar facing the head of 9 mm snail, showing microvilli and tufts of long cilia covering surface. 17. Dorsal surface of mantle collar of 9 mm snail, showing microvilli covering whole surface. Scale bar for Figs. 15-17 3 um. 18. Pneumostome surface of 5.5 mm snail, showing globular structures, suggesting goblet cells. Scale bar 2 um. 208 ABOUL-MAGD 8 SABRY р Я ' т р te MOE me IA Ah я TE . " = > FIGS. 19-22. Foot surface of Biomphalaria glabrata. 19. Sole of foot of 1 mm snail. showing dense cilia. Scale bar 2 um. 20. Sole of foot of 5.5 mm snail, showing lobular structures in between dense long cilia. Scale bar is 2 jm. 21. SEM of dorsal surface of foot of 1.5 mm snail, showing few tufts of short and long cilia. Scale bar 3 рт. 22. SEM of dorsum of foot of 5.5 mm snail, showing many irregular folds. Scale bar 2 um. SEM OF BIOMPHALARIA BODY SURFACES 209 FIGS. 23-24. Light microscope photographs of sections of B. glabrata, 1 mm in diameter. 23. Section of foot showing deeply stained, tall ciliated columnar epithelial cells (T) covering sole as well as short columnar cells covering dorsum (S). Note dense cilia on sole compared to those on dorsum (Barbeito-Lopez trichrome stain). Scale bar 10 um. Fig. 24. Section of surface of mouth, showing smooth homogeneous layer (arrow) covering epidermis (Barbeito-Lopez trichrome stain). Scale bar is 5 шт. 210 ABOUL-MAGD 8 SABRY DISCUSSION The body surfaces of the snail have many functions. In addition to being a protective boundary, it has a role in osmoregulation, respiration, locomotion, perception, and shell formation and regeneration (Hess, 1964). Therefore, a knowledge of the histology, his- tochemistry and surface structure of gastro- pods is fundamental for an appreciation of the host-parasite relationship. A number of re- ports are available on the detailed structure and function of snail tissues, including the epidermal structure (Pan, 1958, for Au- stralorbis glabratus; Hyman, 1967, for Pulmo- nata in general; Zylstra, 1972, for Biomphalar- ia pfeifferi, and Sullivan & Cheng, 1974, for Biomphalaria glabrata). In the present study, scanning electron mi- croscopy revealed that the exposed surfaces of the snail body are provided with cilia, con- firming the light microscope observations (Pan, 1958). The distribution and density of cilia vary in different parts of the body, and even within a given part. For example, while the whole tentacle is covered by cilia, these are dense at the tip and sparse at the base. Perhaps the differences in distribution of cilia are in some way related to the sensory func- tion of the tentacle. The foot is the main organ of locomotion for a snail. lts surface is modified to fulfill this function. The sole of the foot is covered by long cilia; a slimy secretion which has a re- ticular appearance is spread over the surface of the sole of the foot. The presence of the long cilia and mucous secretions facilitate the movement of the foot (Zylstra, 1972) in water and on the surface of containers. Hyman (1967) described the slimy secretion as Originating from the mucocytes which are gland cells located in the subepidermal tis- sue. Also, Zylstra (1972) mentioned one type of subepidermal gland cell, located in the ventral region of the foot, extending up to 400 um from the surface. These cells are grouped together and their necks form a bun- dle while the cell body has a reticular appear- ance. This is suggested by the present study, where we observed groups of secretions ex- tending from the ventral surface of the foot of large snails. Zylstra (1972) stated that the dorsal surface of the foot of Biomphalaria pfeifferi is similar to the head epidermis where cilia were only found scattered between the epidermal cells. Our observations have confirmed this. The presence of the few tufts of cilia may bring the slime secretions from the pedal gland cells to the front and ventral surfaces of the foot (Zyl- stra, 1972). It is obvious from the present work that the surface of the mantle collar differs from the mantle covering the part of the body hidden by the shell. The mantle under the shell is smooth, while in the mantle collar, it is glandu- lar in Pulmonata (Hyman, 1967). Zylstra (1972) added that the surface of the mantle collar of Biomphalaria pfeifferi is nearly covered by short microvilli. Similar microvilli were observed in our work on Biomphalaria glabrata. Hubendick (1958) suggested that these microvilli are important for adhesion to the shell, although Zylstra (1972) reported that there is very little structural basis for adhesion to the shell, unless adhesion occurs by suction. The ultrastructure of the surface of the rec- tal ridge epithelium of Biomphalaria glabrata described by Sullivan et al. (1974) is similar to what we found on the ventra! surface of the mantle collar which bears microvilli and bun- dles of cilia. Also, Malek (1980) mentioned that the mantle collar of Pulmonata is covered by irregular, occasionally branched microvilli, protruding into the fluid-filled spaces between the mantle and the shell. As to the light microscope observations, the cuticle lining the mouth which appeared as a smooth surface by SEM, and also as a homogeneous layer by light microscopy, was also seen in Lymnaea stagnalis by Zylstra (1972). Wilson et al. (1971), working with Lymnaea truncatula, found that the ground cytoplasm of the columnar epithelial cells contains many mitochondria and vesicles. Malek (1980) mentioned that the brush border of the epithe- lial cells seen with the light microscope is revealed by the electron microscope to be microvilli on the surface of each epithelial cell covering the surface of Lymnaea truncatula. In the present study, microvilli and cilia were found on the surface of B. glabrata by SEM. The ultrastructure of the body of small and large snails revealed some differences. For example, the cells covering the head in the area between the bases of the tentacles had different shapes in the two groups of snails. Also, in small snails, the glands on the sole of the foot are less developed. From all these observations, it is clear that the snail surface has a complex structure and that it provides a basis for further study on the SEM OF BIOMPHALARIA BODY SURFACES 2 functions of the snail epidermis, as well as on the site and snail age preferences by schisto- some miracidia. ACKNOWLEDGMENTS We are grateful to Professor M. Voge, De- partment of Microbiology and Immunology, UCLA, for help with the manuscript. This work was done under the tenure of an Egyptian Government Scholarship by the senior au- thor, and a Peace Fellowship by the junior author. REFERENCES CITED BOLOGNANI FANTIN, A. M. & VIGO, E., 1967a, Dati histochimici sui tipi cellulari dell'epitelio tegumentale del piede di gasteropodi acquatici. Rendiconti Istituto Lombardo Accademia di Sci- enze e Lettere, [Sezione B], 101: 99-116. BOLOGNANI FANTIN, А. М. 8 VIGO, E., 1967b, La mucinogenesi nei Molluschi, IV. Caratteristiche istochimiche dei tipi cellulari presenti nel piede e nel mantello di alcune specie di Gasteropodi. Riv. Istoch. Norm. Pat., 13: 1-18. GREENAWAY, P., 1970, Sodium regulation in the freshwater mollusc Limnaea stagnalis (L.) (Gas- tropoda: Pulmonata). Journal of Experimental Biology, 53: 147-163. GREENAWAY, P., 1971, Calcium regulation in the freshwater mollusc Limnaea stagnalis (L.) (Gas- tropoda: Pulmonata). 1. The effect of internal and external calcium concentration. Journal of Experimental Biology, 54: 199-214. HESS, O., 1964, Die Haut der Mollusken. Studium Generale, 17: 161-176. HUBENDICK, B., 1958, On the molluscan adhesive epithelium. Arkiv fór Zoologi, 11: 31-36. HYMAN, L. H., 1967, The Invertebrates, Vol. VI. Mollusca I. New York, McGraw Hill, р. 548-562. JAGER, J. C., 1971, A quantitative study of a chemoresponse to sugars in Lymnaea stagnalis (L.) Netherlands Journal of Zoology, 21: 1-59. JONES, |. D., 1961, Aspects of respiration in Pla- norbis corneus and Lymnaea stagnalis (L.) (Gas- tropoda: Pulmonata). Comparative Biochemistry and Physiology, 4: 1-29. MALEK, E. A., 1980, Snall-transmitted parasitic diseases. CRC Press, Boca Raton, Florida, 7: 105-116. MALEK, Е. A. & CHENG, Т. C., 1974, Classification and structure of the Gastropoda. Medical and Economic Malacology. Academic Press, New York and London, p. 18-26. PAN, C. T., 1958, The general histology and topo- graphic microanatomy of Australorbis glabratus. Bulletin of the Museum of Comparative Zoology, Harvard University, 119: 235-299. SULLIVAN, J. Т. & CHENG, Т. C., 1974, Structure and function of the mantle cavity of Biomphalaria glabrata (Mollusca: Pulmonata). Transactions of the American Microscopical Society, 93: 416— 420. SULLIVAN, J. T., RODRICK, G. E. & CHENG, T. C., 1974, A transmission and scanning electron microscopical study of the rectal ridge of Biom- phalaria glabrata (Mollusca: Pulmonata). Cell and Tissue Research, 154: 29-38. VOGE, M., PRICE, Z. & JANSMA, W. B., 1978, Observations of the surface of different strains of adult Schistosoma japonicum. Journal of Parasitology, 64: 368-372. WILSON, В. A., PULLIN, В. & DENISON, J., 1971, An investigation of the mechanism of infection by digenetic trematodes: the penetration of the miracidium of Fasciola hepatica into its snail host Lymnaea truncatula. Parasitology, 63: 491. ZAAIJER, J. J. P. & WOLVEKAMP, H. D., 1958. Some experiments on the haemoglobin-oxygen equilibrium in the blood of the ramshorn (Planor- bis corneus L.). Acta Physiologica Pharmacolo- gica Neerlandica, 7: 56-77. ZYLSTRA, U., 1972, Histochemistry and ul- trastructure of the epidermis and the sub- epidermal gland cells of the freshwater snails Lymnaea stagnalis and Biomphalaria pfeifferi. Zeitschrift für Zellforschung und mikroskopische Anatomie, 130: 93-134. MALACOLOGIA, 1985, 26(1-2): 213-223 ECOLOGY OF THE TERRESTRIAL SNAIL BREPHULOPSIS BIDENS (PULMONATA: ENIDAE): MORTALITY, BURROWING AND MIGRATORY ACTIVITY Gregory M. Livshits Sackler Faculty of Medicine, Dept. of Anatomy and Anthropology, Tel Aviv University, Ramat Aviv, Israel ABSTRACT This paper summarizes studies on the mortality, burrowing and migratory behaviour of the pulmonate snail Brephulopsis bidens performed at various sites within the population area (Crimea, USSR) during 1975-1977. The annual survival rate of adult snails was about 0.222— 0.252. During the active period mortality of the snails reached a maximum in July. This was followed by the onset of burrowing, which led to a decrease of mortality. Intercolonial migration was extremely limited during the reproductive period (1.4%-2.4% per 10-day period), but increased considerably towards the end of summer, reaching about 20% per 10-day period. The activity radius of individual snails was approximately 3 m. Statistically significant correlations were found between all of the following ecological parameters: mortality, burrowing, migration and population density. Key words: population; mortality; burrowing; migratory activity; correlations. INTRODUCTION Various studies have shown that physical features of the environment (Peake, 1978), or the quantity and accessibility of food (Butler, 1976) may exert a considerable influence on land snail population density. These factors alone, however, could hardly explain the marked variations observed by us (Livshits, 1983) in the spatial and temporal density of populations of the snail Brephulopsis bidens. Oosterhoff (1977), in her study of the ecology of the snail Cepaea nemoralis, suggested that under constant abiotic conditions population density may be regulated by changes in growth and emigration of the snails. In the present work, the mortality, burrow- ing and migration patterns of B. bidens were investigated for their correlation with, and in- fluence on, the population density of this snail within naturally changeable environments. MATERIALS AND METHODS Description of the Investigated Snail Brephulopsis bidens (Krynicki, 1833) (syn- onym Chondrus bidens (Kryn., 1833); Bulimi- nus bidens (Kryn., 1833)) (Pulmonata: Enidae = Buliminidae) is an endemic mollusc of the Crimean (USSR) fauna and is found in (213) steppes and open glades of foothills (Puzanov, 1925, 1926; Likharev & Ram- mel'meier, 1952). The shell of this moderate- ly-sized snail (height 15-20 mm and width 4—6 тт) is elongate-ovate, and is white often patterned with black radial bands. The life- span of B. bidens is about two years with an active period of 7-8 months annually, from April to November (Livshits & Shileiko, 1978). For the remainder of the year the snails hiber- nate beneath the surface, clumped in groups of 3 to 15 individuals. During the active period the snails often climb on grass and aggregate in more or less discrete groups or colonies of different numbers (mean size of a colony in 1975 was 46.6 + 14.4 individuals) (Livshits, 1983). The space occupied by a colony ranged between 0.039 m? and 0.500 m?; the mean distance between colonies was 0.36 m. The mean number of colonies per random 100 п? area in 1975 was 291.4 = 34.7. Dur- ing exceptionally hot and dry weather (July— August) the snails descend and burrow into the ground to a depth of 1-5 cm, remaining there for days or even weeks. Copulation and oviposition take place generally in April-May. The investigation was carried out over a period of three years (1975-1977) on the Internal Cuesta of Crimea (USSR). The stud- ied snail population occupied an area approx- imately 360 т x 70 м. This area, throughout which the snails were encountered in numer- 214 LIVSHITS ous discrete colonies, was arbitrarily divided into 12 sites of roughly 30 m in length, desig- nated by the letters A through N. Livshits (1981, 1983) described the location of the study site in detail. Mortality and Burrowing To determine the winter mortality, 9 ply- wood boxes were used (12 x 20 x6 cm) with floors overlaid with soil and leaf litter. The sexually mature snails, removed from the nat- ural habitat in October, were placed in these boxes (3 boxes in 1975/1976, n = 192 in- dividuals; 6 in 1976/1977, n = 500 in- dividuals) which were then covered with gauze and left for the winter (until April) under natural conditions. Each box was fitted into a slight depression in the ground so that its floor was on a level with the surrounding terrain. The determination of snail mortality and burrowing during May—September was car- ried out in enclosures. Sections of 1.2 x 3m within the population area were Cleared of live and dead snails of all age groups and were made inaccessible to migrants by a high 75 cm gauze fence. After 12-14 days, ensur- ing that no snails had appeared in them, test snails were introduced into each enclosure. The values of their mortality and burrowing rates were calculated from the equations: Mo = n/N; Br = N — п- m/N where Мо represents mortality, Br the proportion of burrowing individuals, n the number of snails dying per unit time, m the number of molluscs living on the grass, and N the total number of snails in the enclosure. Four such enclosures located at sites B, E, K and M were used in 1975 (duration of experiment June- September, initial number of snails in enclo- sures were 439, 361, 733 and 496 individuals respectively). In 1976 a similar experiment in a single enclosure continued from May to October. At this time the initial number of snails was 546 individuals. Enclosures like these (4 x 1 т) were also used to study the comparative mortality on all 12 sites of the population area. The sections, cleared in ad- vance of all snails, were juxtaposed to the areas where the population density was measured (Livshits, 1983). Three hundred adult molluscs were placed in each enclosure in April. Monthly readings were taken in these sections of the number of dead snails. After each observation in all described ex- periments, the dead specimens were re- moved from the enclosures. Зоо FIG. 1. Distribution pattern of 15 colonies selected for the study of B. bidens migration in 1975. Investigation of the Snail Migration Patterns The aim of our study of snail migration was threefold: 1. To estimate the radius of in- dividual snail mobility; 2. To determine the direction and extent of snail migration; 3. To determine the migration intensity of the sex- ually immature individuals. For these purposes we used the technique of Sheppard (1951) and Cain & Currey (1968) whereby the snails are marked with spots of indelible nitroenamel. Marking was carried out in the field to leave intact the native struc- ture of the colonies. All adult snails were marked in 12 colonies located at sites A, B, D, G, K, and M (2 colonies per site) in 1975 and in 12 colonies at all 12 sites (1 colony per site) in 1976. Subsequently, the decreased num- bers of marked snails in the colony between observations were used to estimate emigra- tion. During this period the burrowing and mortality of snails were also considered. Dig- ging out over a radius of 0.5 m from the center of the colony enabled us to count the number of buried individuals. After each observation, all adult snails in the colony (both marked and unmarked) were re-marked by a colour corre- sponding to a given colony. To estimate the range of individual activity and emigration directions of the snails, fifteen colonies extending over an area of about 15 m? at site О (Fig. 1) in 1975 and twelve colonies distributed over an area 2.5 x 2.5m at site F in 1977, were marked and assigned different colours. Afterwards, the emigrated snails were searched for in concentric circles having radii 1, 2, 3 and 4 m from the center of each colony. Both in 1975 and 1977 it was impossible to estimate the number of snails returning to their original colonies (“homing”) because nonmigrating and returning speci- BREPHULOPSIS ECOLOGY 215 mens could not be distinguished from one another. A similar experiment was carried out in 1976 with sexually immature snails, using select members of 32 colonies distributed over an area of 4 x 8m. All statistical procedures were after Sokal & Rohlf (1969). RESULTS Snail Mortality and Burrowing Adult mortality was determined in enclo- sures under natural conditions during June— September of 1975 and May-September of 1976. The overall mortality rate was 0.591 (N = 535) in 1975 and 0.631 (N = 776) in 1976; the difference between these two values is not significant because the duration of observations was longer by a month in 1976 than in 1975. Actually the death rate per 10-day period during the active season was 0.048 and 0.047 respectively (t = 0.05, p >>0.05, test on equality of two percentages). In plywood boxes with nearly natural con- ditions the adult snail mortality during hibernation (October—April) was 0.366 (N = 194) in 1975/1976 and 0.312 (N = 500) in 1976/1977. Per 10-day period, the winter mortality was 0.024 and 0.021 respectively, and this difference is also not statistically significant (t = 0.57, p >0.05). Also, there were no significant differences between banded and unbanded shell morphs in winter У VI VII VIII IX x Months FIG. 2. Relationship between seasonal changes in population density (D), mortality (d) and burrowing (B) of B. bidens during 1976. Snails used for mortal- ity and burrowing determination numbered 635 in- dividuals. D is expressed in number of snails per m?, d and B in percentages. mortality, albeit in summer months the differ- ences were highly significant (Livshits, 1978). Fig. 2 presents curves of seasonal changes in mortality and burrowing in relation to changes in the population density during 1976. Between May and June, while the pop- ulation density increased, the mortality re- mained at a comparatively low level (0.014— 0.048 per 10-day period) and burrowing activ- ity was nil. Subsequently, however, there was a sharp increase in mortality (0.148 per 10- days) which led to a sudden decline in the population density. Following this, burrowing commenced and there was a decrease in the population density to a constant low level. Population dwindling was due not only to mortality, which in fact at that time (August— September) diminished to a mean of 0.066 per 10 days, but also to the burrowing of adolescent snails. Our findings thus suggest that the diminished snail mortality is associ- ated with considerable increase in the propor- tion of buried snails. A similar pattern was observed during the summer of 1975 when, along with a decrease in the population density (mid-July), there was a substantial rise in mortality (from 0.025 to 0.185 within 10 days) and a sharp increase in the number of buried snails (from O to 0.491). Subsequently (from July 16 to August 20), there was considerable diminution of snail mortality (0.064 per 10 days) and the propor- tion of buried snails was 0.400. The observed decrease in the population density on the surface was not in fact reflected in the popula- tion size, which was actually much larger than apparent when buried snails were taken into account. Indeed, the steep increase in the population density observed for about two weeks each year in the fall (Fig. 2 and see also Livshits, 1983) supports this conclusion. Mortality varied considerably not only dur- ing different seasons, but also at different sites of the population area—along a density gradient (Table 1). Data showing this were obtained by investigating the snail mortality in enclosures of 1 x 4m which were placed in each site during May—October of 1975, and the results indicate that as the number of individuals рег m? decreased from site В to site N, the mortality increased in the same direction. There was a good inverse correla- tion between the mortality and density at the various sites (г = — 0.66, р <0.01, data from Table 1). Burrowing also showed a significant spatial fluctuation. Observations on burrowing and LIVSHITS 216 390 2120 8610 2910 9510 e800 750`0 vLOo'O SJue1Biu ДЮ UO} -10d014 bec 296 069 96/ 916 Etc €S¢ 0573 SIIEUS psyuew JO Jaquinn 605 — 681 - 802 - 8:01 - PELOS OCA BIO VAS 9/qeueA цоцеллэ$ао jo sajeg 'SZ6! Ul pouad Áep-01 sad циоцелбиш suapıq 'g jo solmeuAp ¡euoseas ‘с 3719gv1 — 6/1 — 9bz — — +09 — 758 — 695 26p 59215 ajdwes — 779`0 — peso — — 015`0 — 2810 — c9 0 8620 pouad Аер-0г sad uoneudiyy cOLO 5010 5500 1200 9800 cv00 7500 1500 7500 9200 61010’ 1500 poued Аер-0г iad Анерои 126 OFSOL ga 509 950: BEL 21/1 eet 919: 486 pez sell -W/Sjenpiarpui AIsusq N W 7 У H 5 4 3 а С) Я у $1а}эшелез Bale uonejndod jo says ‘SS YOR Je sjenpinipul 00€ ралэашпи UOHeUILa}Ep Ашерош 10, pasn SIIBUS “Jaquia}das—jsnBny 10, uone1Biu ‘saquiaydas—Aeyy 10; sabesane эле sanıen Ашенош BREPHULOPSIS ECOLOGY 217 TABLE 3. The movement of marked B. bidens over various distances during selected periods of 1975. Number of immigrant snails Date От 1m 2m 3m 4m 21.5-10.6 181 41 1174 6 0 10.6-21.6 20 44 13 9 0 27.6-10.7 26 50 16 A 0 10.7-22.8 127 74 0 0 0 22.8-30.9 55 12 il 11 0 Average and standard error 81.8 + 34.9 AAD 11.1 Е 334 6.6 + 2.1 0550 mortality were carried out in four separate enclosures at sites B, E, K and M during 1975. During May, snails collected at the four sites were placed in enclosures as follows: 526 individuals at site B, 684 individuals at site E, 651 individuals at site K and 267 individuals at site M. Two months later the number of dead and buried snails at each site was determined. As illustrated in Fig. 3, there was a gradual increase in burrowing activity proceeding from site B to site M and this was coupled with an increase in snail mortality and a decrease in population density. There were significant differences between the sites with respect to mortality (x? = 100.2, р <0.001) and burrowing (x° = 64.4, р <0.001). Snail Migratory Behaviour and Pattern Migration of B. bidens was studied in the field by observing individuals marked with D O 20 Frequency ¿u/s reus 'Ayısu Sites FIG. 3. Relationship between the spatial variability in population density (D), mortality (d) and burrow- ing (B) of B. bidens in July 1975. Snails used for the mortality and burrowing determination numbered 2128 individuals. B and d are given as frequency of buried and dead snails in the enclosure. different nitroenamel colours. In the course of the study, 2919 adult and 1795 juvenile snails were thus marked. Data on migrations for the entire field study period of 1975 are summa- rized in Table 2. As can be seen from the Table, a rather low level of migration (1.4% per 10 days) was observed in April when mating took place, but from early May there was a rapid increase in migration activities which reached a peak (21% per 10 days) between late August and mid-September. Fig. 4 and Table 3 present data on the radius of activity of individual snails. This radius was 3m and although snails can traverse such a distance within 20 days there were hardly any migrations into the zone ex- tending 4 m from the center. During the entire period of investigation, May-September 22.08.75 10.06.75 FIG. 4. Radius of individual migration activity of B. bidens. |, Il, Ш and IV are concentric zones of distances of 1, 2, 3 and 4 m, respectively, from a marked colony. Numbers in each zone sector shown as percentages of marked individuals found. Total number of snails marked on 21.5.75 was 612 individuals. 218 LIVSHITS TABLE 4. Directions of adult B. bidens migrations in 1975 and 1977. Numbers of snails migrating during the following periods 1975 Wil- 1977 Wil- Direc- coxon's coxon's tions 21:5— 106 177 = 2608 309 test 18.5— 20.6 - 20.8 -149 test North 12 15 2 3 44 116 13 South 4 8 3 5 | 25 11 12 West 51 67 50 22 23 39 23 East 28 32 10 27 | 9 16 39 Var 6.61 8.5 26.6 0.34 24.5 25.7 22.3 B <0.01 <0.01 <0.001 <0.5 -0.001 0.001 < 0.001 Detected in original colony 37 165 35 15 196 125 120 Total detected 132 287 100 712 297 264 207 Total marked 632 632 632 632 350 350 350 ‘In 1975 only data for migrations in West-East directions were tested by x”. 1975, the average number of snails moving into a 0, 1, 2, 3 and 4 meter range was, respectively, 81.8, 44.2, 11.4, 6.6 and 0.0 individuals (see Table 3). The significant dif- ferences of these values were examined by two-way analyses of variance without replica- tion (Sokal 8 Rohlf, 1969). Since the main concern in this study was the distance factor, only data on this aspect are presented here. The variance (V) of the number of individuals reflecting the influence of distance was 5839.0 (df = 4), and V error was 1086.0 (df = 16). Hence, differences in mean snail num- bers according to distance are highly signifi- cant, yielding a value of F = 5.37, р <0.01. It seems that despite the ability of individual B. bidens snails to traverse distances greater than 3m, most of the adult snails displayed limited mobility and rarely travelled beyond a radius of 1 m. Data were collected also on the migratory direction of the adults as well as juveniles. To determine the direction of snail migration dur- ing the summer of 1976 (15 colonies) and 1977 (12 colonies), all the molluscs in a given area were marked with different colours in accordance with their colony of origin. Subse- quently the colonies were observed for migra- tion into areas of individuals marked with “foreign” colours. The observational results are given in Table 4, and relate to an area extending between sites D and E. Significant preferential movement of snails to the W was already observed in the first 20 days. This tendency persisted through the summer months to the end of August. Begin- ning in September, however, the picture abruptly reversed, with about half of the mi- grant snails moving to the E, and 15.4% mov- ing N or $. It should be noted that the sample colonies were mainly located in a W-E direc- tion (Fig. 1) so that information on migration in other directions was not meaningful. A changing migration pattern was also noticed in 1977 when the analysed colonies were uniformly distributed spatially. In June a tendency for movement to the north was clearly evident, but towards the end of the summer this changed to a westerly direction; in September, the snails commenced moving eastwards again (Table 4). Wilcoxon's signed-rank paired-sample test (Sokal 4 Rohlf, 1969) has been used to com- pare the extent of migration in different direc- tions during the completed seasons of in- vestigation (i.e., May-September 1975 and also May-September 1977). By this test, each two directions of snail migration con- stitutes a pair, and the various proportions of migrants that migrate in a single direction comprise a series of pairs. During the active seasons of 1975 and 1977 there were no significant overall differences between the dif- ferent directions of migration (р >0.05). Thus, during our different study periods or seasons of the year, the incidence of migra- tions in various directions differed signifi- cantly. However, the overall migration in- BREPHULOPSIS ECOLOGY 219 cidence for all the active periods as a whole is more or less equal in all directions. It seems that the snails move only around or near their own colony. In fact, near the end of the ex- periment in August, there was a sudden in- crease in marked individuals within the start- ing circle (0) and circle No. 1 (Fig. 4). This phenomenon reflects the apparent re- migration of snails to their original colony site. To assess migration of the preadolescent snails, 1795 individuals with 5—7 shell whorls, deriving from 32 colonies оп an 8 x 4 m area, were marked in July 1976. Only 212 (11.7%) of these could be detected in May 1977, most of them (179 individuals or 84.1%) still within their respective colonies and only 15.9% hav- ing migrated out of their colonies (Table 5), albeit in groups. Phenological observations indicated sever- al seasonal cycles of migration and burrowing activity, with the only difference between the two being the somewhat later commence- ment (at about the end of June) and earlier termination of burrowing. In fact, a correlation was observed between adult snail emigration and the proportion of snails burrowing during the seasonal field observations of 1975 and 176 Eig 5, г = 0:75. р <0:01): Migration as well as burrowing varied соп- siderably in space, and also correlated in- versely with the adult snail density at the sites (Fig. 6, r = 0.56, p <0.001). This was es- tablished by investigating snail emigration during July-September 1975 from colonies at sites A, B, D, K and M, and again in 1976 at each site. A very high coefficient of correlation was found also between mean rates of > Burrowing N 0 .08 16 .24 32 Emigration per 10-day FIG. 5. Burrowing of B. bidens as a function of migration. TABLE 5. Migration of preadolescent B. bidens between 26.7.1976 and 19.5.1977. Total Total Colony marked detected no. 29.7.76 19.5.77 1 54 12 2 62 7 $) 34 0 4 97 16 5 82 0 6 112 27 i 70 10 8 40 0 9 60 6 10 32 0 in 50 0 12 70 14 13 30 0 14 50 0 15 30 7 16 40 0 1174 60 0 18 60 Al 19 40 0 20 67 24 21 48 0 22 40 0 23 26 0 24 54 0 25 70 8 26 30 4 27 80 17 28 41 0 29 60 0 30 103 19 Si 62 8 32 41 0 Total in colonies 1795 179 Between colonies = 33 Sum 1795 212 mortality and migration at the same sites dur- ing the summer months of 1975 (r = 0.93, p <0.05, data from Table 1). DISCUSSION Ecological Variables Different ecological variables were studied separately on discrete colonies of B. bidens at various sites within their population area. The annual adult snail survival rate was about 0.222-0.252, which was substantially lower than for other terrestrial snails, e.g. 0.50-0.75 for Cepaea nemoralis (Oosterhoff, 1977; Wil- 220 LIVSHITS Emigration 0 -4 -2 0 2 4 6 log, Density FIG. 6. Correlation between migration of B. bidens and the adult population density. © and @ August and September of 1975, /\ and A July and August of 1976. liamson et al., 1977) or 0.70 for Sphinc- terochila zonata (Shachak et al., 1975). How- ever, the latter species have a life span of 4-6 years (Cain 8 Currey, 1968; Shachak et al., 1975; Williamson, 1976) and reproduce re- peatedly, whereas the life span of B. bidens is only two years and the snails reproduce only once (Livshits and Shileiko, 1978). It is possi- ble, therefore, that biological differences be- tween B. bidens and the other two species account for the different annual survival. Mortality rates of B. bidens varied signifi- cantly within the population area and also during different months of the active season. In the case of C. nemoralis such spatial (Cain & Currey, 1968) or seasonal (Richardson, 1975) variability was not observed. Seasonal burrowing is typical for many spe- cies of terrestrial pulmonates and has an im- portant adaptive value (Wolda, 1963; Shileiko, 1978). For species dwelling in arid zones this behaviour is crucial for survival during the hot and dry season (Yom-Tov, 1971; Shachak & Chaptan, 1976; Smith, 1976). The present study shows that burrowing can be a means for correcting the population density by lowering the mortality rate during adverse or deteriorating ecological con- ditions. Previously it was found that the differ- ent activities displayed by different B. bidens shell morphs was mainly responsible for the maintenance of mean morph frequencies in the populations during the summer (Livshits, 1978). As for the chronology of the migration pat- tern of B. bidens, this is briefly as follows: during the reproductive period (April-May), the adult snails aggregate and mate in dis- crete colonies, with snail migrations during this stage ranging between 1.4%-2.4% per 10-day period. This rather low migratory rate agrees with the obtained maximal value of the ratio of the variance S* to the mean value of the population density D (Livshits, 1983) as well as with the maximal genetic heterogene- ity observed during the reproductive period (Livshits, 1978). The young snails emerging from the eggs remain largely in place but some (approximately 15.9% per year) mi- grate. Sizeable migrations occur only after the snails mature and mate, at which time the migration attains 20% per 10-day period (July-August). However, even in these months the distance traversed by the snails may be no greater than 3m, the dispersion intensity diminishing sharply with each addi- tional meter (Fig. 4). In other snail species the rate of migration decreases with increasing age (Cain & Currey, 1968; Pollard, 1975). The activity radius of individual C. nemoralis is up to 10 m/year (Greenwood, 1974; Cameron & Williamson, 1977), while for Helix pomatia it can be about 4 m/week (Pollard, 1975). Activ- ity of adult H. aspersa, H. pomatia and C. nemoralis reaches maximum levels in late spring and early summer (Bailey, 1975; Pol- lard, 1975; Cameron & Williamson, 1977), which according to the last-mentioned au- thors is not unexpected considering that this is the height of the mating season. Correlations between Ecological Parameters and Population Density There are numerous publications on terres- trial pulmonates relating variations in popula- tion density to other ecologic parameters. Some of these correlations are summarized in Table 6. As is evident from this table, negative correlations between density and growth rate, density and level of reproduction were discerned in various snail species. Correlation between density and migration, however, may be positive or negative, if any. On the basis of data obtained in the labora- tory as well as under field conditions, Oos- terhoff (1977) proposed a causal scheme to explain the regulation of molluscan population density. Her scheme proposes the existence of a negative correlation between snail pop- ulation density and the growth rate and posi- tive dependence of emigration on density. BREPHULOPSIS ECOLOGY 221 TABLE 6. Correlations between population density and other ecological variables in several terrestrial snails. + and — are positive and negative correlations respectively. O—unaffected by population density. Variables Migration Migration Migration Mean adult size Growth rate of shell Growth rate of shell Growth rate of shell Production of eggs or young Production of eggs or young Production of eggs or young Survival Nature of correlation Species Source +t Cepaea nemoralis Cain & Currey, 1968 ! Oosterhoff, 1977 0 Cepaea nemoralis Cameron & Williamson, 1977 Helicella virgata C. nemoralis C. nemoralis H. virgata Trochoidea seetzeni C. nemoralis T. seetzeni H. virgata H. virgata Butler, 1976 Wolda, 1969 Oosterhoff, 1977 Pomeroy, 1969 Yom-Tov, 1972 Wolda & Kreulen, 1973 Yom-Tov, 1972 Butler, 1976 Butler, 1976 TABLE 7. Correlations between various ecological parameters in the studied B. bidens population. Density Migrations (D) (m) — 0.56 — — 0.66 0.93 — 0.82 0.75 Frequency of banded morph — 0.56 i Mortality Burrowing (d) (B) Illumination Source — — — Present study — — — Present study 0.88 — — Present study 0.88 0.73 — 0.85 Livshits (1981) “It was found that migration activity of the banded morphs was significantly higher than in unbanded ones. However, these correlations were obtained in snails maintained on a limited food supply and under laboratory conditions. Т During spring-summer Correlations between different ecological parameters were also observed for the pop- ulation studied (Table 7). Our own data sug- gest the following scheme of density regu- = Increase of illumination I During autumn lation: Increase of Increase of mortality migration (Decrease of banded morph frequency) ns ] Increase of » burrowin | g Decrease of density eur Decrease of — + Decrease of — + Decrease of ———> Increase of density illumination mortality (Increase of banded morph frequency) burrowing Maturation of preadolescent snails 222 LIVSHITS In the previous investigation (Livshits, 1981) it was shown that during the hot and dry period of summer the mean frequency of buried snails among the banded morphs was significantly higher than among the unbanded snails (p <0.001). Burrowing of the snails was concomitant with a considerable decline of the total mortality of animals and particularly of the banded morphs. For example, during July 1976 (before burrowing) mortality of banded snails was 13% vs. 4.2% per 10 days during August 1976. Simultaneously the rela- tive mortality of unbanded morphs in the pop- ulation increased (4.6% vs. 8.4%). As a result of this in September—October, the increase in frequency of the banded morph was parallel to a decrease in mortality. Analysis of the thermotolerance of shell pattern morphs showed that the resistance of the unbanded morph was significantly higher than that of the banded one (Livshits, 1981). Reversible correlation between banded morph frequency and illumination was also discerned, which suggests that illumination (and/or temperature) may be an important determinant of snail mortality. Indeed, maxim- al fluctuations of density (= maximal coeffi- cient of variation) were observed at sites M and N, where the frequency of the banded morph was also maximal. Increase of illumination led to the increase not only of mortality but also of migratory activity. Addi- tional experiments revealed that the banded morph preferred shaded areas (Livshits, 1981) and that the morphs may migrate in different directions actively searching for appropriate microhabitats. However, the paucity of suitable microbiotopes and the lim- ited activity radius of the individual snails lead to mass burrowing during the summer. There is consequently a decrease in the population density on the surface, in spite of the con- current decrease in snail mortality (Fig. 2). In the autumn, the diminution of insolation leads to decreased mortality and the reemergence of buried snails. Restoration of the adult pop- ulation density is effected also by maturation of pubescent snails. A histogram of age struc- ture (Livshits, 1983) clearly indicates an in- crease in the proportion of adult snails within the population by autumn. The data presented herein enable the for- mulations of several possible mechanisms to explain the seasonal and spatial fluctuation in the population density of B. bidens. ACKNOWLEDGMENTS It is a great pleasure to acknowledge the advice given to me by Professor L. Fishelson during the writing of this work. | am deeply grateful to Professor J. Lange for correcting my English. Thanks are also due to Mrs. R. Suzin, who prepared the illustrations. REFERENCES CITED BAILEY, S. E. R., 1975, The seasonal and daily patterns of locomotor activity in the snail Helix aspersa Muller, and their relation to environmen- tal variables. Proceedings of the Malacological Society of London, 41: 415-428. BUTLER, A. J., 1976, A shortage of food for the terrestrial snail Helicella virgata in South Aus- tralia. Oecologia, 25: 349-371. CAIN, A. J. & CURREY, J.D., 1968, Studies on Cepaea. Ш. Ecogenetics of a population of Cepaea nemoralis (L.) subject to strong area effects. Philosophical Transactions of the Royal Society of London, ser. B, 253: 447-482, pl. 33. CAMERON, R. A. D. £ WILLIAMSON, P., 1977, Estimating migration and the effects of dis- turbance in mark-recapture studies of the snail Cepaea nemoralis (L.) Journal of Animal Ecolo- gy, 46: 173-179. GREENWOOD, J. J. D., 1974, Effective population numbers in the snail Cepaea nemoralis. Evolu- tion, 28: 513-526. LIKHAREV, I. М. & RAMMEL’MEIER, E. J., 1952, Terrestrial molluscan fauna of the USSR. Akademiia Nauk SSSR, Moscow and Leningrad. In Russian, 512 p. LIVSHITS, G. M., 1978, Adaptive behaviour as a factor in the maintenance of the genetic stability of an isolated population of the land mollusc Chondrus bidens (Kryn.). Soviet Genetics, 14: 449-455. LIVSHITS, G. M., 1981, Survival, behaviour and spatial distribution of shell morphs in a popula- tion of the snail Brephulopsis bidens (Pulmo- nata). Oecologia, 51: 220-226. LIVSHITS, G. M., 1983, Ecology of the terrestrial snail Brephulopsis bidens: age composition, population density and spatial distribution of in- dividuals. Journal of Zoology, London, 199: 433— 446. LIVSHITS, G. M. & SHILEIKO, A. A., 1978, Life cycle of Brephulopsis bidens. Ecologia (USSR), 5: 77-83. OOSTERHOFF, L. M., 1977, Variation in growth rate as an ecological factor in the land snail Cepaea nemoralis (L.). Netherlands Journal of Zoology, 27: 1-132. PEAKE, J., 1978, Distribution and ecology of the Stylommatophora. In FRETTER, V. & PEAKE, BREPHULOPSIS ECOLOGY 223 J., eds., Pulmonata, 2: 429-526. Academic Press, London, New York, San Francisco. POLLARD, E., 1975, Aspects of the ecology of Helix pomatia. Journal of Animal Ecology, 44: 305-329. POMEROY, D. E., 1969, Some aspects of the ecology of the land snail, Helicella virgata, in South Australia. Australian Journal of Zoology, 17: 495-514. PUZANOV, I. I., 1925, Materials to study Crimean molluscs. |. Mountain molluscs. Biulleten Mos- kovskogo Obshchestva Ispitatelei Prirodi, 34(1— 2): 41-62. In Russian. PUZANOV, I. I., 1926, Materials to study Crimean molluscs. |. Steppe molluscs. Biulleten Mos- kovskogo Obshchestva Ispitatelei Prirodi, 35(1— 2): 84-101. In Russian. RICHARDSON, A. M. M., 1975, Energy flux in a natural population of the land snail, Cepaea nemoralis L. Oecologia, 19: 141-164. SHACHAK, M. 8 CHAPTAN, E. A., 1976, Some aspects of the ecology of the desert snail Sphinc- terochila boissieri in relation to water and energy flow. Israel Journal of Medical Sciences, 12: 887-891. SHACHAK, M., ORR, Y. & STEINBERGER, Y., 1975, Field observations on the natural history of Sphincterochila (S.) zonata (Bourguignat, 1853) (= S. boissieri Charpentier, 1847). Argamon; Israel Journal of Malacology, 5: 20—46. SHEPPARD, P. M., 1951, Fluctuations in the selec- tive value of certain phenotypes in the poly- morphic land snail Cepaea nemoralis (L.). Heredity, 5: 125-134. SHILEIKO, A. A., 1978, Molluscs, Ш, 6. Nauka, Leningrad. In Russian. SMITH, B. S., 1976, Life history and biology of a snail. |. Aestivation and reproduction. Victorian Naturalist, 93: 128-130. SOKAL, R. R. & ROHLF, F. J., 1969, Biometry. Freeman, San Francisco. WILLIAMSON, P., 1976, Size-weight relationships and field-growth rates of the land snail Cepaea nemoralis (L.) Journal of Animal Ecology, 45: 875-885. WILLIAMSON, P., CAMERON, R. A. & CARTER, M. A., 1977, Population dynamics of the land snail Cepaea nemoralis (L.): a six year study. Journal of Animal Ecology, 46: 181-194. WOLDA, H., 1963, Natural populations of the polymorphic land snail Cepaea nemoralis. Ar- chives Néerlandaises de Zoologie, 15: 381-471. WOLDA, H., 1969, Fine distribution of morph frequencies in the snail Cepaea nemoralis near Groningen. Journal of Animal Ecology, 38: 305— 327. WOLDA, H. & KREULEN, D. A., 1973, Ecology of some experimental populations of the land snail Cepaea nemoralis (L.). И. Production and sur- vival of eggs and juveniles. Netherlands Journal of Zoology, 23: 168-188. YOM-TOV, Y., 1971, The biology of two desert snails Trochoidea (Xerocrassa) seetzeni and Sphincterochila boissieri. Israel Journal of Zoolo- gy, 20: 231-248. YOM-TOV, Y., 1972, Field experiments on the effect of population density and slope direction on the reproduction of the desert snail Tro- choidea (Xerocrassa) seetzeni. Journal of An- imal Ecology, 41: 17-22. | À, vd 1 | o i A р у vai ' и | Al ih № | ay Е =. i a | pas 1 | | | , en | 7 a Y ' D 17 3 u | 1 1 e ar у $ | | | A | | 2 i y er ve | > 1 in ae. Re AY A A in i = i | т { ; y a MALACOLOGIA, 1985, 26(1-2): 225-239 SEASONAL CHANGES IN THE REPRODUCTIVE GROSS ANATOMY OF THE LAND SNAIL TRIODOPSIS TRIDENTATA TRIDENTATA (PULMONATA: POLYGYRIDAE) Kenneth C. Emberton Committee on Evolutionary Biology, University of Chicago and Field Museum of Natural History, Chicago, Illinois, U.S.A. ABSTRACT Triodopsis tridentata tridentata (Say, 1816) is a seasonally protandric hermaphrodite. Three Statistically independent stages in its reproductive cycle, named mating readiness, egg produc- tion, and allosperm absence, were detected by principal components analysis of the relative volumes of six reproductive organs from March through July. The range and temporal sequence of volume changes in each of these organs are shown in graphs and in illustrations of the entire reproductive systems of individuals with extreme principal component scores. The illustrations indicate especially well that lengths, widths, and volumes of some reproductive organs should be considered suspect as phylogenetic or taxonomic characters in this and probably many other land snail groups. Key words: snail; Pulmonata; reproductive anatomy; principal components; protandry; her- maphrodite. INTRODUCTION The hermaphroditic reproductive system of pulmonate gastropods is commonly used as a source of characters for phylogenetic in- ference and systematics. Despite this, in- traspecific variation in reproductive charac- ters is poorly understood. Most studies on intraspecific variation in the pulmonate repro- ductive system (e.g., Krahelska, 1912-1913; Holm, 1946; Laviolette, 1950; Lusis, 1961, 1966; Rigby, 1963, 1965; Galangau, 1964; Kugler, 1965; Luchtel, 1972 [a review]; Dun- can, 1975 [a review]) are histological in approach, are restricted to one organ or a limited set of organs, and ignore gross morphology. Several papers (e.g., McLauch- lan, 1951; Walter, 1968, 1969; Webb, 1970) contain illustrations of some intraspecific var- iation, but do not put the variation into its seasonal context and do not attempt to depict the full range of variation. A few papers dis- cuss complete seasonal cycles in the size or weight of pulmonate reproductive organs (e.g., Berrie, 1966; Smith, 1966; Runham & Laryea, 1968) but give no illustrations of gross morphology. To my knowledge, only one paper to date (Solem, 1981, fig. 53) actu- ally illustrates the full range of seasonal varia- tion in a pulmonate's reproductive organs. Such baseline data are essential for the phy- logeneticist and systematist, who must com- pare specimens collected at different seasons or in different climatic regimes. This paper illustrates the range of temporal variation in the gross morphology of repro- ductive organs in the polygyrid land snail Triodopsis tridentata tridentata (Say, 1816). The Polygyridae are an endemic and domi- nant land snail family of North America (Pils- bry, 1940). Polygyridae are increasingly used for studies in physiology (Reeder & Rogers, 1983), ecology (Solem, 1955; Blinn, 1963; Randolph, 1973; McCracken, 1976, 1980; Vail, 1978; Emberton, 1981), and ecological genetics (Fairbanks, 1979; McCracken, 1980; McCracken & Brussard, 1980). Some polygy- rids may soon be of considerable economic importance as a source of anti-A agglutinin for typing human blood (Miles & Beck, 1983). Triodopsis t. tridentata (Fig. 1) belongs to the subfamily Triodopsinae (Pilsbry, 1940; Webb, 1959) and to the subgenus Triodopsis, sensu stricto (Pilsbry, 1940; Webb, 1954; Vagvolgyi, 1968); it is the nominate member of the tridentata species complex (Vagvolgyi, 1968). Triodopsis t. tridentata is a common snail of woodlands and waste ground, ranging from southern Ontario, Michigan, and New England to middle Alabama and Georgia (Hubricht, unpublished range map). This snail lives from near sea level in New York and (225) 226 EMBERTON New Jersey to between 1200 and 1500 m in the Roan Mountains of Tennessee. The shell varies considerably in size and, to a lesser extent, in shape over its geographical range (Vagvolgyi, 1968). Triodopsis t. tridentata lives under leaf litter, logs, stones, and trash. It hibernates during the winter and is active through the spring, summer, and fall (Grimm, 1975). During the active season, the snail may enter a short- term quiescent state, with the formation of the same type of thin epiphragm of dried mucus as used in hibernation, when the weather is unseasonably cold (Ingram, 1941) or dry (per- sonal observation). Courtship is brief. Intromission is either reciprocal or one-sided and lasts 5 to 15 minutes (Webb, 1947, 1959). Small clutches of eggs are laid in loose soil under some cover, usually in early spring. The eggs are 2.0 to 2.1mm in diameter and are deeply indented when first laid (Ingram, 1944), but later appear “leathery . .. bounded by an out- er clear viscid membrane beneath which is a white crystalline layer’ (Kingston, 1966). In the laboratory the eggs hatch in two to three Height Diameter spermatheca FR > ’ prostate ‘Ge weeks and the hatchlings reach maturity in six to eight months. Some laboratory hatchlings mature by fall and others overwinter as young to complete growth the next spring. Adults live two to four years and lay eggs every three weeks to six months in the laboratory. They have not been observed to self-fertilize, and allegedly do not oviposit unless inseminated by another individual, no matter whether of the same or different species (Grimm, 1975). Self-fertilization does occur very rarely and with low fertility in the congeneric Triodopsis albolabris (McCracken, 1980). The reproductive system of Triodopsis t. tridentata is shown diagrammatically in Fig. 1. The functions of its organs, as inferred from the literature on other pulmonates, are as follows. Italicized names in the following text are as labeled in Fig. 1. The ovotestis first produces sperm, then ova (Pennypacker, 1930; Lusis, 1961; Runham & Laryea, 1968; but see Rigby, 1963, 1965). Sperm are stored in the hermaphroditic duct (Duncan, 1975; Solem, 1981). During mating, the penis is everted through the genital opening, and sperm travel from the hermaphroditic duct, TREE N 6 T A .5 Whorls (=5.4) albumen gland EE ISA hermaphroditic ( € duct EY ovotestis DD) 10 mm AAA FIG. 1. Triodopsis t. tridentata (Say): shell and diagrammatic reproductive system. Shell measurements are indicated. Rank measurements were taken of the spermatheca, prostate, uterus, talon, albumen gland, and hermaphroditic duct. TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 227 down a ciliated channel adjacent to the pros- 1979, at six different sites around Dow Lake, tate (which adds secretions), then out the tip Strouds Run State Park, Athens County, of the everted penis. The mate's penis inserts Ohio, U.S.A. (Fig. 2). The six sites were cho- through the partially everted vagina and dis- sen as relatively undisturbed, second-growth charges a sperm mass into the duct of the deciduous hill slopes with occasional sand- spermatheca, which is actually more thick- stone outcrops. For each of the 16 col- walled than depicted in Fig. 1 (Webb, 1948, lections, the site was chosen by random num- 1959; Duncan, 1975). Much of the received ber table such that all six sites would be sperm mass is probably digested by proteo- sampled before repeating any site. Collecting lytic enzymes in the spermatheca (= bursa was done each afternoon for approximately copulatrix) (e.g., Nemeth & Kovacs, 1972; three hours. All collections are summarized in Reeder & Rogers, 1979; Rogers et a/., 1980), Table 1. Collections were designed for the but a few sperm escape down the sperma- dual purpose of detecting seasonal changes thecal duct to swim up through the oviduct in reproductive structures of Triodopsis t. and the uterus to reach the talon (= recepta- tridentata and of assessing microgeographi- culum seminis = fertilization pouch = ciliated cal variation in land snail communities. hood), where they are stored and perhaps Each sample of Triodopsis t. tridentata with nurtured (Rigby, 1965; Bayne, 1973; Lind, reflected shell lips was split into two groups. 1973; Reeder & Rogers, 1983). One group was drowned overnight in tap Ova move through the hermaphroditic duct, water laced with chloral hydrate, a relaxant, are fertilized by allosperm from the talon, fixed in 95% ethanol, then stored in 70% receive a yolk from the a/bumen gland, then ethanol for gross anatomy. The second group pass into the uterus, where the egg shell is was likewise drowned in tap water with chloral added. Clutches of completed eggs travel hydrate, but was then placed directly into down the oviduct and out the genital opening Bouins solution for later histological examina- (Duncan, 1975). tion. Only the 57 specimens preserved in ethanol are considered here (Table 1, column METHODS AND MATERIALS 6) The following data were recorded for each Snails were collected at approximately specimen: shell diameter and height in mm weekly intervals from 24 March to 29 July, (Fig. 1); whorl count to the nearest 0.1 whorl TABLE 1. Collections of Triodopsis t. tridentata from Strouds Run State Park, Athens, Ohio. Number live Site Hours collected Number FMNH Date (see spent ane eS + il adults catalogue (1979) Fig. 2) collecting Adult Juvenile dissected number Notes 24 March IV 3:50 10 0 6 209209 1 April Ш 4:30 3 0 2 209232 8 April Il 3:00 6 1 3 209237 14 April | 2:30 Y 2 4 209279 22 April V 3:00 4 2 2 209294 28 April VI 3:00 8 2 4 209333 5 May IV 3:00 5 0 3 209348 14 May Ш 3:00 8 4 3 209375 21 May Il 4:00 12 2 6 209387 26 May | 3:00 4 1 4 209420 eggs first seen 3 June V 3:00 2 1 2 209453 eggs in uterus of specimen 16 June IV 3:00 9 0 5 209502 27 June Ш 3:10 5 2 3 209536 juveniles are hatchlings 2 July Il 3:00 6 0 209554 25 July VI 3:00 1 8 1 209571 juveniles are hatchlings 29 July IV 2:30 6 0 6 209584 228 EMBERTON 2 er FIG. 2. Six collecting sites (I-VI) at Strouds Run State Park, Athens, Athens County, Ohio. Dow Lake is in black. Contour interval is 60 ft. (Fig. 1); degree of protrusion of the animal from its shell (ranks O to 4); albumen gland relative volume (ranks 1 to 9); relative thick- ness of edge (ranks 1 to 6), and darkness of color (ranks 1 to 6), cream to dark brown); hermaphroditic duct volume (ranks 1 to 9); talon volume (ranks 1 to 6); uterus volume (ranks 1 to 6); spermatheca volume (ranks 1 to 7); and prostate volume (ranks 1 to 6). Rank measurements were taken as fol- lows. All 57 vials, each containing the com- plete reproductive system of one snail, were placed in a shallow, ethanol-filled dissecting tray, and the specimens were rank-ordered for each of the 9 variables in turn. For the albumen gland, for example, the specimens were ordered from small to large glands, with the number of size categories (ranks) de- termined by my ability to distinguish these from pairwise comparisons. For all organs, the sorting criterion was estimated volume rather than length, width, or area. The penis could not be ranked in like manner because there were varying degrees of extrusion and coverage by the sheath and the retentor mus- cle. The ovotestis could not be ranked be- TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 229 cause it was encased by the digestive gland and thus could not be observed directly. During the rankings, an effort was made to maintain approximately normal distributions of ranks. This was in no way difficult or unnat- ural, as the size distributions of the organs had strong central tendencies. Ranks can be used in parametric uni- and multivariate an- alyses if their distributions are approximately normal (Paul Sampson, personal com- munication). In fact, the term “rank” is some- what misleading in this case because the difference between adjacent ranks is close to a constant (the least detectable difference). Thus, the ranks approach being metric var- iables. For all statistical analyses | used BMDP-79 programs (Dixon & Brown, 1979) on the Amdahl computer at the University of Chica- go. Histograms, as well as normal and de- trended normal plots (P5D), were used to test assumptions of normality of all measured var- iables. Stepwise linear regression (P2R) and canonical correlation (P6M) were used to de- termine whether shell size or degree of an- imal extrusion from its shell could explain the size of the prostate, the uterus, or the albu- men gland (the three largest and most de- formable reproductive organs). Polynomial regression (P5R) with a two degree maximum was used to detect significant linear or uni- modal changes in each of the measured var- iables over time. For multivariate analysis, only one measure was used for the albumen gland, namely volume. This was done to restrict variables to six because of the small number of snails (57). Clustering algorithms were used to de- tect natural groupings into blocks of affinity (P3M) by the 57 individual snails (P1M), by the six reproductive organs (P3M), and by both snails and reproductive organs. Principal components analysis (P4M) detected groups of the six reproductive organs that varied together as independent units. For deriving principal components, the covariance matrix was used in order to weight each reproductive Organ according to its detectable variation. Snails having extreme values for each prin- cipal component were chosen to show the biological significance of the components and to illustrate the full range of size variation of each reproductive organ. Finally, each snail was Classified according to its stage in the reproductive cycle, with the stages de- termined by the aforementioned analysis. RESULTS The distributions of all measured variables were not significantly different from normal. Significant seasonal variation was found in 7 of the 9 measurements of reproductive organs (Fig. 3). The prostate (that is to say a generalized prostate for all dissected snails from all collections) steadily decreased in volume from March through July. Likewise, the hermaphroditic duct steadily decreased in volume from its peak in March. The uterus volume rose to a peak in late May, then decreased. The albumen gland increased slightly to a peak volume in late May, then decreased. The albumen gland increased in granularity of texture to a peak also in late May, then decreased. The albumen gland became darker in color until late April; thereafter, it became lighter. Neither the talon volume nor the edge thickness of the albu- men gland underwent any significant season- al change. Considerable variability occurred in all re- productive characters at every collection, as is obvious from the scatter of points in Fig. 3. Thus, for any given date and site, individual snails of the same population (on a single hill slope) differed widely in their reproductive development. Some of the great variation in uterus and albumen gland volumes was apparently a result of preservation: stepwise regression indicated that 25% and 18%, respectively, of their variation was explained by how far the drowning animal had retracted into its shell. These relationships were also evident in the canonical correlation analysis (not shown here for brevity’s sake.) The effect of animal extrusion on organ size was not statistically removed from the data prior to subsequent analyses; doing so would not seem to have affected the results. Clustering the reproductive organs by a standard, minimum distance, single linkage algorithm based on the absolute correlation matrix yielded the dendrogram shown in Fig. 4. Two major clusters are readily apparent. The upper cluster represents the tendency for snails with a large uterus to also have a large albumen gland and a small talon. The lower cluster represents the tendency for snails with a large spermatheca to also have a large prostate and, to a lesser extent, a large hermaphroditic duct. Clustering the individual snails yielded a finely dissected, indecipher- 230 EMBERTON 71 KO y . 0) Qi +. . fine S 7 к. у 5] © ... ооо cove a a o > а o ..... 0... is NS a 3 5 eee ee ~ e. 000 © e008 © . > о K a = Е еее oo . Е о 3 31 ee . oe . 2 = 3 . o a SS = г . . > . e. . [© 21 . . ... o. . Wy ij 1 a coarse T T T T T T TE T 1 y = eae . = Mar Apr May Jun Jul Mar Apr May Jun Jul 6] © во ° e 61 . ee { . O . thick 1 NG eo ee Я eee ee eee © . o] | .. . . . 2 < ape > 4! o ee © ..o.o © 4 № / оф see > . 3 4 eee ... ee e о | a о | 0 oa с m 2 4 x © 4 ... > . D -| ооо во © 5 Ne © 1 . . ee ооо ee . = SS > \ = > 24 . ee . 2! ое . . A 21e . . . . . thin 1 T T T - T ”, 1 T Tp T * *— : T - =" г > T Mar Apr May Jun! Jul | 61 ee + { ee хо о al . dark ee ee | ® > { оо ee e... 7- ee 0 ef . . e _ 4- ® © eee .o.oo > . o 1 $ 00010100 © oe . © 4- ss оо . 5 a 5 5 5} eve eo e... . De < fe 4 ee eo ооо . » . 5 liado 0 ie 4 0e .. ee . . «3° * 8 21° о ee + © . 27 ee о. o . 4 0 | 1: o light : E J | T T Ti ; = T T т T T Mar Apr May Jun Jul T T T al Mar Apr May Jun Jul FIG. 3. Ranked sizes and characteristics of reproductive organs plotted against time. Least squares regression lines, either linear or quadratic, are superimposed. Significance levels of the regressions are as follows: hermaphroditic duct, p <.005; prostate, p <.005; talon, p = .47; spermatheca, p <.005; uterus, p <.005; albumen gland, p = .08; granularity of albumen gland, p <.005; color of albumen gland, p = .01 (linear) or .06 (quadratic, as shown). able dendrogram. Simultaneously clustering both organs and snails produced an equally confusing array of small blocks of affinity. Results of principal components analysis are given in Table 2. Three independent com- ponents (factors) accounted for 84% of the total size variation in the six reproductive organs. These factors point out the same pattern that appeared in the cluster analysis, only in finer detail. The first factor is a linear function of the tendency for the hermaphro- ditic duct, the prostate, and the spermatheca to simultaneously increase and decrease in volume. This factor is a new, combined vari- able which may be called mating readiness: the hermaphroditic duct is large and charged with sperm, the prostate is large and full of stored seminal secretion, and the sperma- TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 231 Albumen Gland 23 63 65 Uterus Talon 59 Hermaphroditic Duct 75 Spermatheca Prostate FIG. 4. Cluster tree of reproductive organs. Num- bers at branch-points indicate the distance or sim- ilarity when the link was formed. Algorithm was a minimum distance (single linkage) method based on the absolute correlation matrix. TABLE 2. Structures of three independent factors extracted from rank-measurements of six reproduc- tive organs. Names of the factors are based on their strongest contributors, which are underscored. The total explained variance is 83.6%. Mating Egg Allo- readi- produc- sperm ness tion absence Hermaphroditic duct 1.69 0.00 0.40 Spermatheca ES 0.17 —0.37 Prostate Eve 0.17 0.36 Albumen gland 0.14 1.53 zo Uterus 0.10 115 —0.22 Talon 0.14 19 —1.00 Percentage of variance explained 45.4% 28.3% 9.9% theca is enlarged either preparatory to or as a result of receiving the sperm mass of a copulatory partner. Mating readiness com- prises about half the total variation of the six organs. The second factor is a linear function of the tendency for the uterus and albumen gland volumes to covary. This factor should be called egg production: the albumen gland is enlarged and full of yolk, and the uterus is also engorged with egg shell-producing secretions. Egg production comprises about one-fouth of the total variation of the six rank- measured organs. The variable egg produc- tion is independent of the variable mating readiness because they were derived as orthogonal principal components. The third factor is predominantly a function of talon size. This factor may be called allo- sperm absence, because large values of the factor correspond to small talons presumably having little or no stored foreign sperm. Allo- sperm absence comprises about one-tenth of the total variation of the organs; it 1$ т- dependent of both mating readiness and egg production, because of the orthogonality of principal components. In order to check the biological relevance of the three factors, as well as to view the ex- treme sizes of each organ, | examined the three or four snails with highest and lowest values for each factor. One obvious juvenile (undeveloped genitalia despite the reflected lip of its shell) had the lowest value for both mating readiness and egg production. An- other snail which was transitional between juvenile and adult had the second smallest value for mating readiness and the third smallest value for egg production. Excluding those two juveniles, | prepared drawings, all to the same scale, of the reproductive sys- tems having extreme values of each of the three factors (Figs. 5 to 7). The snail with the highest value for mating readiness (Fig. 5, top) had a hermaphroditic duct fully charged with sperm, a prostate swollen with secretion, and a huge sper- matheca with an enclosed mass, presumably sperm received in a recent mating. This snail was from the first collection in March, and had probably recently emerged from hibernation. The snail with the lowest value for mating readiness (Fig. 5, bottom) had a minute hermaphroditic duct, prostate, and sper- matheca. The snail with the highest value for egg production (Fig. 6, top) had a large, plump albumen gland and uterus for egg-making. Also, its hermaphroditic duct appeared emptied at its distal end. The snail with the lowest value for egg production (Fig. 6, bot- tom) was apparently a subadult, with all organs small and incompletely developed. This animal could not have been manufactur- ing eggs. The snail with the highest value for allo- sperm absence (Fig. 7, top) had a very small talon. This snail seems to have had both male and female systems charged and active. The snail with the lowest value for allosperm ab- sence (Fig. 7, bottom) had the largest talon of all the dissected snails. Because this snail was collected at the end of July and had an extremely eroded shell, it could be interpreted 232 EMBERTON FIG. 5. Triodopsis t. tridentata: reproductive anatomies having extreme values for the factor mating readiness. AG = albumen gland, HD = hermaphroditic duct, Ov = ovotestis, Pe = penis, Pr = prostate, Sp = spermatheca, SM = sperm mass, Ta = talon, Ut = uterus, VD = vas deferens. The arrow indicates the position of the cross section figured in the upper left. The organs with circled labels are those which make up the factor (see Table 2). “High” was collected 14 April (specimen D) and “low” was collected 29 July (specimen A). as post-reproductive. Thus, the uterus appeared as though folded down to a com- pact size, and the albumen gland appeared spent and flaccid. | next classified each snail according to its stage in the reproductive cycle by studying the 57 reproductive systems, removed from their vials, in random order. There were only two reproductive systems which totally con- fused me and which | could not place into any category. Repeating the classification proc- ess the next day, again in random order, | got a repeatability of 88%. The major difficulty was deciding between early egg production and late post-reproduction. With final de- cisions made, sometimes aided by reexamin- TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 233 FIG. 6. Triodopsis t. tridentata: reproductive anatomies having extreme values for the factor egg production. Labels are as in Fig. 5. The organs with circled labels are those which make up the factor (see Table 2). “High” was collected 14 April (specimen C) and “low” was collected 29 July (specimen C). ing the shell, the results were graphed against time (Fig. 8). Included in the graph were the juveniles and hatchlings which had not been dissected. From Fig. 8 it is evident that two age groups overwintered: adults began mating upon emergence in March, started laying eggs in May, and their offspring were crawling at least by the end of June; juveniles that had over- wintered were approaching sexual maturity by the end of July. It can be seen from Fig. 8 that the dissected 234 EMBERTON A L Ne US E 4 FIG. 7. Triodopsis t. tridentata: reproductive anatomies having extreme values for the factor allsoperm absence. Labels are as in Fig. 5. The organ with the circled label (the talon) makes up the bulk of the factor, and the organs with cross-marked labels are secondary contributors to the factor (see Table 2). “High” was collected 24 March (specimen A) and “low” was collected 16 June (specimen C). snails included a cohort of late-maturing ju- DISCUSSION AND CONCLUSIONS veniles. This cohort, then, is present in the Methods used seasonal graphs of organ volumes (Fig. 3) as a “contaminant” of the later stage of each Rank-ordering reproductive organs by visu- graph. al estimation of volume is inferior in many TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 235 Post- Reproductive Egg Producing Mating- Ready Neoadult Completing Shell Juvenile Hatchling April May June eggs first ри July FIG. 8. Reproductive state of individual snails plotted against time. The cross represents an individual with a stunted reproductive system due to a massive parasitic infestation of the digestive gland. ways to calculating volume from serial sec- tions (Lusis, 1961) or cutting apart and weigh- ing the individual organs (Smith, 1966; Runham & Laryea, 1968). Points in favor of my ranking method are that it requires mini- mal time and equipment, leaves the reproduc- tive system intact, and permits separate measurements of the uterus, prostate, and talon. Sample sites were dispersed over several square kilometers (Fig. 2) and included the range of variation in slope, aspect, depth of litter layer, and other parameters of forest microhabitat that were available in the region. Because so many sources of variation were included in the study, the seasonal trends which appeared in the data (Figs. 3 and 8) can parsimoniously be attributed to underly- ing biological properties of the snails rather than to biased sampling. Protandric cycle The regression lines in Fig. 3 and the curve in Fig. 8 represent the anatomical changes of an individual snail from March through mid- June (beyond mid-June the data are “con- taminated” by the addition of a cohort of neoadults: see Fig. 8). The adult snail comes out of hibernation in late March ready for copulation, with stores of sperm and prostatic secretion, and a spermatheca enlarged or capable of enlarging. The female organs for egg production develop soon after sperm ex- change. Egg-laying begins in early June. The cause of spermathecal enlargement remains undetermined. The spermatheca may remain relatively small until it is ex- panded by copulation and receipt of a sperm mass. Alternatively, digestion of excess auto- sperm and other internal wastes (R. Reeder, personal communication), or response to hor- monal changes, may enlarge the spermathe- ca prior to copulation. Support for the hypoth- esis of pre-copulatory enlargement is afforded by two apparently hibernating Triodopsis t. tridentata (FMNH 209162) col- lected 3 March near Site |, Strouds Run State Park (Fig. 2). These two snails had enlarged spermathecae much as in the upper anatomy 236 EMBERTON of Fig. 5. The two were deep under an icy decayed log, had well-formed epiphragms, and presumably had been dormant all winter. During the previous two days, however, mini- mum and maximum air temperatures at a nearby station had risen to 40 and 54 degrees Fahrenheit, so it is possible that these snails had reactivated and copulated before being collected. Significant color changes in the albumen gland (Fig. 3) may have been caused either by histological changes associated with yolk production and release, or by staining, under preservation, by the closely adjacent stom- ach, the color of which may have matched seasonal changes in diet. Therefore, changes in albumen gland color cannot be firmly in- dicated as a part of the protandric reproduc- tive cycle. Significant changes in the granularity of the albumen gland (Fig. 3) are readily explained by the gland’s development during the repro- ductive cycle. Overwintered adults had albu- men glands spent and granular-looking from last year’s egg laying. As these glands began producing yolk, their cells filled, and they looked less granular. After egg production, the albumen glands once again had empty cells, and appeared granular. The wide variation in reproductive state found in each collection (Figs. 3 and 8) prob- ably resulted from a mixture of three effects. First, each hillside from which a collection was taken doubtless contained a number of different microclimates which would have dif- ferentially affected the growth rates of resi- dent snails. Second, with as many as four year classes coexisting (overlapping genera- tions), and with two cohorts (early spring and late summer) per year class, a wide variation in reproductive state was to be expected. Third, it is highly likely that hatchlings from the same clutch of eggs grow at markedly differ- ent rates. This phenomenon has been documented in field populations of the con- familial Mesodon thyroidus (Blinn, 1961) and of the congeneric Triodopsis albolabris (McCracken, 1976), as well as in controlled laboratory rearings of T. albolabris (McCrack- en, 1976; Vail, 1978). In view of these three of possibly many sources of variation in repro- ductive state, it is not surprising to see the scatter of points underlying the seasonal trends in Figs. 3 and 8. This scatter can be predicted to increase in more and more southerly populations from increasingly milder climates. It would also be interesting to determine how the variance in reproductive state affects the effective population size (see, e.g., Roughgarden, 1979), which is im- portant in ecological genetics. Factors and life stages An important point needs to be made about the three factors extracted from the data (Table 2; Figs. 5-7). These factors corre- spond very closely to Smith's (1966) three phases of the underlying ontogenetic se- quence in the arionid slug Arion ater. Smith determined these three phases from histo- logical series, and found them, by elegant experimentation, to be independent of both environmental change and body weight. The first phase, “differentiation of male gonads early in the spring,” is the equivalent of mating readiness. Smith's second phase, “copulation and differentiation of the female glands,” corresponds to egg production. Smith called the transition between his first and second phases the “critical point,” which occurs at or near the act of copulation and sperm ex- change. Finally, Smith's third phase, “fertiliza- tion, egg-laying, and the onset of atrophy,” is associated in a loose way with allosperm absence. Fertilization uses allosperm and therefore increases the value of allosperm absence. No snails in this study were un- equivocally in a state of atrophy, so the cor- respondence on that aspect is unclear, unless Smith’s (1966) term “atrophy” is equivalent to my term “sexual dormancy.” The correspondence between this study and Smith’s (1966) suggests that a general mechanism underlies the reproductive cycles of both an arionid slug and a polygyrid snail. Clearly, the results of both studies are con- sistent with the hypothesis of a sequential release of separate male and female hor- mones (see Boer & Joosse, 1975), with the release of the latter triggered by some event around the time of copulation. This event is Smith's (1966) “critical point,” which Runham & Laryea (1968: 104) equated with Lavio- lette's (1950) “transitional stage.” Along the same line, Solem (1981) hypothesized that development of the female system in the camaenid land snail Amplirhagada bur- nerensis burnerensis was triggered by sperm exchange, because unmated snails, which had been kept quiescent for several months past normal mating time, when killed and dissected, had a hypertrophied male system with an undeveloped female system. TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 237 Comparable studies Male-acting and female-acting anatomies have been illustrated for the camaenid land pulmonates Meridolum jervisense (McLauch- lan, 1951) and Polydontes lima (Webb, 1970). Both these species show suites of mating organs and of egg-producing organs enlarged as in the “high” anatomies in Figs. 5 and 6 of this paper. Solem (1981, fig. 53) presented a chart showing the gross morphologies of the ovotestis, hermaphroditic duct, prostate, spermathecal head and contents, and coiled section of the vas deferens, for each of six to ten collections over a year from dry season to wet to dry for the Australian camaenid Am- plirhagada b. burnerensis. The prostate and hermaphroditic duct volumes were greatest at the beginning of the wet season and de- creased remarkably during the wet season, exactly paralleling the prostate and hermaph- roditic duct volumes of Triodopsis t. tridentata from early through late spring. Likewise, the uterus volume of Amplirhagada b. bur- nerensis was smallest in the early wet sea- son, increasing to a maximum in mid-season, then decreasing through the remaining wet and dry seasons. Seasonal changes in the spermathecal volume of Amplirhagada b. bur- nerensis were slight compared to those of Triodopsis t. tridentata, but the greatest volume apparently occurred early in the wet season. The configuration of the distal vas deferens, which was not looked at in this study, proved a useful indicator of mating state in Amplirhagada b. burnerensis: un- mated snails from the dry season had highly convoluted vasa deferentia, whereas mated snails had relatively straightened vasa de- ferentia. In summary, the Australian camaenid Amplirhagada b. burnerensis showed the same seasonal pattern of changes in the reproductive system as the American ¿ ,gyrid Triodopsis t. tridentata, with the start of the Australian wet season corresponding to the North American spring. Significance for phylogenetics Caution must be exercised when using size differences among reproductive organs as taxonomic criteria. When Lutz (1950), for ex- ample, described the new subspecies Triodopsis hopetonensis claibornensis, he used as a distinguishing character a very long, enlarged spermatheca. Clearly, this character is capable of extreme intraspecific variation (see Fig. 5), and hence is of little value as a taxonomic character. Likewise, the relative size or shape of any pulmonate repro- ductive organ should be used as a taxonomic character only with caution and a full knowl- edge of its range of intraspecific variation. Even with this restriction, the pulmonate re- productive system provides a rich source of phylogenetically and taxonomically useful characters, including the internal (or everted) structure of the penis (e.g., Webb, 1947; Solem, 1976, 1981) and vagina (e.g., Solem, 1981), and the presence and disposition of various muscles (e.g., Pilsbry, 1940) and glands (e.g., Shileyko, 1978). ACKNOWLEDGMENTS This paper was first presented at the 1982 annual meeting of the American Malacologi- cal Union, New Orleans, Louisiana. The study was supported by a Graduate Research Grant from the Graduate Student Senate, Ohio University, and by Public Health Service Genetics Training Grant GM07197-07. Col- lecting and research equipment were pro- vided by the Field Museum of Natural History (FMNH), Chicago, where the specimens are catalogued and housed. Computational funds were provided by the University of Chicago. Linnea Lahlum gave artistic advice. Alice Gough graciously improved and inked most of the drawings and served as a sounding board for my ideas. Richard Reeder offered helpful discussion and encouragement. Ellen Ember- ton entered the first draft into the computer. Alan Solem suggested and encouraged the collections and dissections, and provided in- cisive criticism of various drafts. LITERATURE CITED BAYNE, C. J., Physiology of the pulmonate repro- ductive tract: location of spermatozoa in isolated, self-fertilizing succineid snails (with a discussion of pulmonate tract terminology). Veliger, 16: 169-175, fig. 1A, 1B. BERRIE, A. D., 1966, Growth and seasonal changes in the reproductive organs of Lymnaea stagnalis (L.). Proceedings of the Malacological Society of London, 37: 83-92. BLINN, W., 1961, Aspects of ecology, behavior, and physiology of land snails, particularly of Mesodon thyroidus (Say) and Allogona profunda (Say). Ph.D. Dissertation, Northwestern Univer- sity, 95 p. 238 EMBERTON BLINN, W., 1963, Ecology of a land snail. Ecology, 44: 498-505. BOER, H. H. & JOOSSE, J., 1975, Endocrinology. In: Pulmonates, ed. FRETTER, V. & PEAKE, J., Volume 1, Functional anatomy and physiology, Academic Press, London, 1: 245-307. DIXON, W. J. & BROWN, M. B., 1979, BMDP-79, Biomedical computer programs, P-series. Uni- versity of California, Berkeley, 880 p. DUNCAN, C. J., 1975, Reproduction. In Pulmo- nates, ed. FRETTER, V. & PEAKE, J., Function- al anatomy and physiology, Academic Press, London, 1: 309-365. EMBERTON, K. C., 1981, Ecological notes on two sympatric, conchologically convergent polygyrid land snails in Ohio. Bulletin of the American Malacological Union, “1980”: 27-30. FAIRBANKS, H. L., 1979, Enzyme variation in Ashmunella levettei (Bland) (Gastropoda: Polygyridae). Ph.D. dissertation, University of Arizona, 70 p. GALANGALU, V., 1964, Le cycle sexuel annuel de Milax gagates (Drap.) et ses deux pontes. Bulle- tin de la Société Zoologique de France, 89: 510-593. GRIMM, F. W., 1975, Speciation within the Triodop- sis fallax group (Pulmonata: Polygyridae)—a preliminary report. Bulletin of the American Malacological Union, “1974”: 23-29. HOLM, L. W., 1964, Histological and functional studies on the genital tract of Lymnaea stagnalis appressa Say. Transactions of the American Microscopical Society, 65: 45-68. INGRAM, W. M., 1941, Habits of land Mollusca at Rensselaerville, Albany County, New York. American Midland Naturalist, 25: 644-651. INGRAM, W. M., 1944, Observations of egg-laying habits, eggs, and young of land mollusks on the Edmund Niles Huyck Preserve, Rensselaerville, New York. American Midland Naturalist, 32: 91— 97. KINGSTON, N., 1966, Observations on the labora- tory rearing of terrestrial mollusks. American Midland Naturalist, 76: 528-532. KRAHELSKA, M., 1912-1913, Reduktions-Er- scheinungen in der Eiweissdrúse der Schneck- en. Polska Akademia Umiejetnosci, Krakow, Wyaziat Matematyczno-Przyrodniczy, sB: Scien- ces Naturelles, “1912”: 606-621. KUGLER, O., 1965, A morphological and histo- chemical study of the reproductive system of the slug, Philomycus carolinianus (Bosc). Journal of Morphology, 116: 117-132. LAVIOLETTE, P., 1950, L'evolution de la glande hermaphrodite d’Arion rufus et ses rapports avec le croissance. Comptes Rendus de la Société de Biologie, Paris, 144: 135-136. LIND, H., 1973, The functional significance of the spermatophore and the fate of spermatozoa in the genital tract of Helix pomatia (Gastropoda: Stylommatophora). Journal of Zoology, London, 169: 39-64. LUCHTEL, D., 1972, Gonadal development and sex determination in pulmonate molluscs. |. Ar- ion cirumscriptus. Zeitschrift für Zellforschung und mikroskopische Anatomie, Berlin, 130: 279- 301. LUSIS, O., 1961, Postembryonic changes in the reproductive system of the slug Arion ater rufus L. Proceedings of the Zoological Society of Lon- don, 137: 433-468. LUSIS, O., 1966, Changes induced in the reproduc- tive system of Arion ater rufus L. by varying environmental conditions. Proceedings of the Malacological Society of London, 37: 19-26. LUTZ, L., 1950, A list of the land Mollusca of Claiborne County, Tennessee, with a description of a new subspecies of Triodopsis. Nautilus, 63: 99-105, 121-1238. McCRACKEN, G. F., 1976, The population biology of the white-lipped land snail, Triodopsis albolab- ris. Ph.D. dissertation, Cornell University, 136 p. McCRACKEN, G. F., 1980, Self fertilization in the white-lipped land snail Triodopsis albolabris. Bio- logical Journal of the Linnean Society, 14: 429— 434. McCRACKEN, С. Е. 4 BRUSSARD, P. F., 1980, The population biology of the white-lipped land snail Triodopsis albolabris: genetic variability. Evolution, 34: 92-104. McLAUCHLAN, C. F., 1951, Basic work on the life cycle of some Australian snails. Proceedings of the Zoological Society of New South Wales, “1949-1950”: 26-36. MILES, C. D. & BECK, M.L., 1983, Land snails (Polygyridae) as a source of anti-A agglutinin for typing human blood. American Malacological Bulletin, 1: 97-98. Abstract. NEMETH, A. 8 KOVACS, J., 1972, The ul- trastructure of the epithelial cells of the seminal receptacle in the snail Helix pomatia with special reference to the lysosomal system. Acta Biologi- ca Academiae Scientarum Hungaricae, 23: 299- 308, 10 pl. PENNYPACKER, М. I., 1930, The germ-cells in the hermaphroditic gland of Polygyra appressa. Journal of Morphology, 49: 415—453. PILSBRY, H. A., 1940, Land Mollusca of North America (north of Mexico). Academy of Natural Sciences of Philadelphia Monograph 3, 1(2): [vi], 575-994, ix. RANDOLPH, P. A., 1973, Influence of environmen- tal variability on land snail population properties. Ecology, 54: 933-955. REEDER, R. №. & ROGERS; SAP MIS TS MINE histochemistry of the spermatheca in four spe- cies of Sonorella (Gastropoda: Pulmonata). Transactions of the American Microscopical Society, 98: 267. REEDER, R. L. 4 ROGERS, S. H., 1983, Histology of the seminal receptacle complex in Mesodon zaletus. American Malacological Bulletin, 1: 98. Abstract. RIGBY, J. E., 1963, Alimentary and reproductive systems of Oxychilus cellarius (Múller) (Stylom- matophora). Proceedings of the Zoological Soci- ety of London, 141: 311-359. RIGBY, J. E., 1965, Succinea putris: a terrestrial TRIODOPSIS REPRODUCTIVE ANATOMY: SEASONAL CHANGES 239 opisthobranch mollusc. Proceedings of the Zoological Society of London, 144: 445-486. ROGERS, S. H., REEDER, R. L. & SHANNON, W. A., 1980, Ultrastructural analysis of the mor- phology and function of the spermatheca of the pulmonate snail, Sonorella santaritana. Journal of Morphology, 163: 319-329. ROUGHGARDEN, J., 1979, Theory of population genetics and evolutionary ecology: an introduc- tion. Macmillan, New York, 634 p. RUNHAM, N. W. 8 LARYEA, A. A., 1968, Studies on the maturation of the reproductive system of Agriolimax reticulatus (Pulmonata, Limacidae). Malacologia, 7: 93-108. SHILEYKO, A. A., 1978, On the systematics of Trichia s. lat. (Pulmonata: Helicoidea: Hygro- miidae). Malacologia, 17: 1-56. SMITH, B. J., 1966, Maturation of the reproductive tract of Arion ater (Pulmonata: Arionidae). Mala- cologia, 4: 325-349. SOLEM, A., 1955, Studies on Mesodon ferrissi (Gastropoda, Pulmonata) 1. General ecology and biometric analysis. Ecology, 36: 83-89. SOLEM, A., 1976, Comments on eastern North American Polygyridae. Nautilus, 90: 25-36. SOLEM, A., 1981, Camaenid land snails from west- ern and central Australia (Mollusca: Pulmonata: Camaenidae) Il. Taxa from the Kimberley, Am- plirhagada lredale, 1933. Records of the West- ern Australian Museum, Supplement 11: 147- 320. VAGVOLGYI, J., 1968, Systematics and evolution of the genus Triodopsis (Mollusca: Pulmonata). Bulletin of the Museum of Comparative Zoology, 136: 145-254, pl. 1-6. VAIL, V. A., 1978, Laboratory observation on the eggs and young of Triodopsis albolabris major (Pulmonata: Polygyridae). Malacological Re- view, 11: 39-46. WALTER, H. J., 1968, Morphological features of Liberian Bulinus and B. truncatus of Egypt: a pictorial essay on snails of three subgenera (Pla- norbidae: Basommatophora). Malacological Re- view, 1: 35-89. WALTER, H. J., 1969, Illustrated biomorphology of the “angulata” lake form of the basommatophor- an snail Lymnaea catascopium Say. Malacologi- cal Review, 2: 1-102. WEBB, G. R., 1947, The mating-anatomy tech- nique as applied to polygyrid landsnails. Amer- ican Naturalist, 81: 134-147. WEBB, G. R., 1948, Comparative observations on the mating of certain Triodopsinae. Nautilus, 61: 97-103. WEBB, G. R., 1954, The life-history and sexual anatomy data on Ashmunella with a revision of the triodopsin snails. Gastropodia, 1: 13-18. WEBB, G.R., 1959, Pulmonata, Polygryidae: notes on the sexology of Triodopsis, a new sub-genus, Haroldorbis, and a new section, Shelfordorbis. Gastropodia, 1: 23-25. WEBB, С. R., 1970, Pulmonata, Camaenidae: comparative sexology and genital development of Caracolus carocolla (L.), C. marginella (Gmel- in), and Polydontes lima (Férussac). Gastro- podia, 1: 79-84, pl. 36-37. MALACOLOGIA, 1985, 26(1-2): 241-251 FUNCTIONAL MORPHOLOGY OF “EYESPOTS” OF MANTLE FLAPS OF LAMPSILIS (BIVALVIA: UNIONACEA): EVIDENCE FOR THEIR ROLE AS EFFECTORS, AND BASIS FOR HYPOTHESIS REGARDING PIGMENT DISTRIBUTION IN BIVALVE MANTLE TISSUES Louise Russert Kraemer & Charles M. Swanson Department of Zoology, University of Arkansas, Fayetteville, AR 72701, U.S.A. ABSTRACT Serially sectioned tissues of Lampsilis ventricosa “eyespots” reveal that this structure consists of (1) an epithelium with elongated, heavily pigmented cells resting on a consistent basement membrane and thrown into distinct folds; (2) an underlying series of muscles coursing parallel to the postero-anterior length of the flap; (3) connective tissue trabeculae which extend from the basement membrane down into the flap interior, separating the aforementioned muscles into bundles; (4) pigment granule clusters which accompany the connective tissue trabeculae and are distributed from the basement membrane of the “eyespot’ epithelial cells into the central flap tissues; (5) a prominent muscle in the center of the flap which is longitudinally displayed in transverse sections and which has a “vertical” orientation in the flapping mussel; and (6) an extensive nerve supply apparently associated almost exclusively with the “eyespot” muscle tissues. The role of the “eyespot” tissues in the mantle flap during the mussel's flapping behavior makes it seem unlikely that the “eyespots” have a sensory function. This conclusion comports well with findings from previous extensive behavior studies. Evaluation of (1) the nerve and muscle distribution in the “eyespot” region of the flap; (2) the obvious relationship of the pigment granule clusters to the pigmented “eyespot” epithelial cells; and (3) the pigment clusters’ evident sequential path of distribution through the flap tissues indicate that the “eyespots” are more likely effectors than sensors. Rather than “eyespots,” they should more correctly be termed “pig- mented effector spots.” Indeed it seems possible that localized mantle movements (e.g., of Lampsilis mantle flaps) or even more generalized mantle movements may serve to produce patterned distribution of pigment in bivalve mantle tissues. Key words: “eyespot” function; mantle pigment; pigment pattern. INTRODUCTION The present study is an examination of the structure and an evaluation of the putative function(s) of the “eyespots” of the mantle flaps of Lampsilis (Bivalvia: Unionacea). Man- tle flaps of this genus have recently figured prominently in systematic treatments of freshwater mussels (Burch, 1975; Johnson, 1980), and in the analysis of systematic ge- netics (Davis & Fuller, 1981). Behavioral aspects of mantle flap movements and Lamp- silis mantle flap movements themselves have been involved in reproduction and spawning (Kraemer, 1970). In recent years the molluscan bivalve man- tle has been the subject of physiological stud- ies (e.g., Sick & Siegfried, 1980; Sorenson et al., 1980; Zaba, 1981; Zaba & Davies, 1981). Some histochemistry has been carried out by Mane & Patil (1980), Wheeler (1979), Gil- loteaux (1979) and Counts & Prezant (1979). Sensitivity and control of the scallop mantle edge and of the file clam mantle edge were studied by Stephens (1978a, 1978b). In the literature reviewed for this study, however, there were no reports of investigation of man- tle structures known to be involved in repro- ductive behavior of bivalved mollusks; and no further published studies of the structure and function of the mantle flaps of the fresh-water mussel Lampsilis. BACKGROUND An earlier study (Kraemer, 1970) revealed the presence of complex spawning behavior associated with mantle flapping, which causes some species of Lampsilis to resem- ble small swimming fish. Kraemer (1970) found that mantle flaps develop only in ma- (241) 242 KRAEMER & SWANSON BS te, N 5 E / : Е Se ee ¿aio ade (a Va 0. АА МАЕ x | th on Pu ge 7 spill AAA RT / A \ ) N ©. | FIG. 1. Sequence of paired pulsing movements of mantle flaps in Spawning female Lampsilis ventricosa, viewed from the left side (modified from Kraemer, 1970). Pulsing movements may be as rapid as 3/sec. a, end of recovery phase, (“tails” out, horizontally): b, beginning of pulse (initiated at base of mantle flap “tails”); с, pulse moves along flap, causing lateral bulge; d, pulse nears “eyespot” ends of mantle flaps; e, pulse is at flap ends, which are moved down and out, horizontally, exposing the glochidia-charged water tubes of the marsupial gill(s). Insert: sketch showing characteristic “headstand” position of L. ventricosa during flapping behavior. A, anterior: AS, excurrent siphon: BS, incurrent siphon: D, dorsal: E, “eyespot” of mantle Нар; М, gravid outer gill, serving as marsupium: Р. posterior; T, “tail” of mantle flap. LAMPSILIS MANTLE FLAP “EYESPOTS” 243 ture females, and that each flap is an exten- sion of the inner lobe of the mantle edge, immediately anteroventrad to the branchial siphon. Each flap is a permanent part of the mantle, and characteristically possesses an “eyespot” (a raised, pigmented patch of epithelium) at its posterior end, and a “tail” anteriorly. The flaps function as part of a flapping behavior complex of gravid females before and during spawning, when the glo- chidia (parasitic larvae which must effect con- tact with a specific fish host) are shed into the water. The flaps move in rhythmic, paired pulses which are initiated near the “tail” ends of the flaps and move to the “eyespot” ends (Fig. 1). Near the tails of the flaps where the pulses originate, accessory mantle ganglia have been found, and may be involved as pacemakers for the flap movements (Kraem- er, 1968, 1969). A previous behavior study (Kraemer, 1970) did not imply that the “eyespots” on the flaps are photosensors, although Kraemer (ibid) produced experimental evidence indicating that the rate of mantle flap movements will increase or decrease in response to in- crements and decrements of light, respective- ly, at low light intensities. These findings con- siderably modified and amplified the earlier conclusion of Welsh (1933) that there is “pho- tic stimulation of the rhythmic contractions of the mantle flaps.” An earlier anatomical study (Kraemer, 1970) showed that there is considerable in- terspecific variation in the appearance and location of “eyespots” on the mantle flaps. For example, in Lampsilis ventricosa, a con- spicuous “eyespot” is typically visible only on the outer surface of the flap, while in L. fas- ciola, an “eyespot” is evident on both outer and inner flap surfaces. The present study was carried out on the “eyespots” of L. ventri- cosa to determine the histological character- istics of the structure, and to explain the possible role of the “eyespot” in the mantle flap movements and flapping behavior com- plex of the spawning mussel. MATERIALS AND METHODS Specimens of Lampsilis ventricosa (Barnes) used in the histological study were all gravid females collected in July, 1965 from the Raisin River in Washtenaw County, Michi- gan, from Lee Creek in Crawford County, Arkansas in June, 1964, and from War Eagle Creek in Madison County, Arkansas in July, 1976. Five “eyespots” were removed from the mantle flaps of relaxed specimens, em- bedded in paraffin blocks, sectioned trans- versely or frontally at 6-10 рт, and stained with an aniline blue variation of Mallory’s triple stain (Schmitz, 1967). One complete series of transverse sections was especially helpful. Living specimens for SEM preparation were collected from King’s River, Arkansas in February, 1982. Excision of “eyespot” tissue from the mantle flap of a relaxed specimen had to be done swiftly, since the membranous mantle flap retains its contractility even in a heavily sedated mussel. Once removed, the tissue was fixed in 2.0% glutaraldehyde, and processed for SEM examination. Samples were viewed with an ISI-60 Scanning Electron Microscope at 30KV with a working distance ОГ TS mm: RESULTS When viewed in cross-section under low magnification (Fig. 2), the “eyespot” region of the mantle flap shows a conspicuous pig- mented epithelial layer on the outer surface. Corresponding to the “eyespot” region on the outer surface of the mantle flap, the inner surface epithelium is not pigmented or other- wise modified. Underlying the inner epithelial surface one sees only a few muscles in- terspersed with loose connective tissue. In the following paragraphs, histological de- scription of the “eyespot” region will focus on the tissues occupying the outer surface (where the “eyespot” region is visible), and extending into the mid-interior of the mantle flap (Fig. 2). Muscle tissue in and near the “eyespot” region of the mantle flap Beneath the basement membrane of the outer surface (“eyespot”) epithelium of the mantle flap are bundles of muscles which course the length of the mantle flap. These muscles are separated into groups of four or more by connective tissue trabeculae which extend at right angles from the basement membrane under the surface epithelium to the interior of the flap (Fig. 2, CT). The bun- dles of transversely-sectioned outer muscles crowd against the outer epithelial surface of the flap, especially in the areas distally and proximally adjacent to the pigmented epithe- 244 KRAEMER 8 SWANSON Que DA D FIG. 2. Semi-diagrammatic cross-section of the Lampsilis ventricosa mantle flap “eyespot.” Histological boundary of the “eyespot” is indicated by the bracket. Insert: orientation of mantle flap showing plane in which tissue was sectioned. BM, basement membrane; CT, connective tissue trabecula; D, distal; EE, “eyespot” epithelium on outer surface of mantle flap; IE, epithelium on inner surface of mantle flap; LM, transversely sectioned, longitudinally arranged muscle; OE, outer epithelial layer (other than “eyespot” epithelium) of mantle flap; P, proximal; PGC, pigment granule cluster; VM, longitudinally sectioned, vertically arranged muscle. lial cells of the “eyespot” itself (Fig. 3, LM). Internal to the bundles of transversely sec- tioned outer muscles are several conspicuous longitudinally-sectioned, inner “vertical” mus- cles located near the center of the mantle flap interior (Fig. 2, VM). Small nerves course along the inner muscles, thread through the outer muscle bundles, and extend from one muscle to another (Fig. 4A, N). Small nerves often accompany the muscle-separating trabeculae of connective tissue. Occasionally one can see a nerve terminating bluntly in connective tissue (Fig. 4B, N). Basement membrane and connective tissues in and adjacent to the “eyespot” A distinct basement membrane occurs be- tween the base of the flap’s outer surface epithelium and the underlying connective tis- sue. In the area adjacent and distal to the “eyespot” epithelium, the basement тет- brane is greatly thickened, often lying over large, pale-staining, rounded cells (Fig. 5, PC). In the area adjacent and proximal to the “eyespot,” the basement membrane is distinct but not thickened (Fig. 3, BM), and is seldom associated with underlying pale, round cells. Layers of the distinct basement membrane upon which the “eyespot” epithelium stands may separate (1) to form trabeculae which dip between the underlying muscular tissues (de- scribed above); or (2) simply to surround small cavities containing pigment granules (Fig. 6A, PG). Epithelium in and adjacent to the “eyespot” Proximal to the “eyespot” epithelium, the outer epithelial layer of the mantle flap is comprised of simple, low columnar to cuboid- al epithelial cells. Distal to the “eyespot” LAMPSILIS MANTLE FLAP “EYESPOTS” 245 FIG. 3. Photomicrograph of transverse section of outer surface of Lampsilis ventricosa mantle flap, immediately adjacent and proximal to the “eyespot” region. Note low columnar to cuboidal epithelium, and absence of pigment granules. BM, basement membrane; CT, connective tissue trabecula; LM, transversely sectioned muscle; OE, outer epithelial layer (other than “eyespot” epithelium) of mantle flap. epithelium, the outer epithelial layer of the mantle flap contains cells of similar shape and smaller size. The “eyespot” epithelium itself manifests large, very elongate cells which taper to an attenuated base and often appear to be attached to the basement membrane by a threadlike stalk (Fig. 6, EE). The exposed surface of the “eyespot” epithelium shows an evident brush border of microvilli (Fig. 6, 7A). Viewed with the SEM, the “eyespot” epithelial cell surfaces are devoid of cilia (Fig. 7A), and show apical borders with well-developed microvilli (Fig. 7B). Petit et al. (1978) did not note comparable features on the mantle of Amblema (Unionidae). It is only the great length of the epithelial cells themselves (60—70 рт) which raises the “eyespot” above the surrounding surface of the flap (Fig. 2). “Eyespot” epithelial cells are uninucleate, and the nuclei of neighboring cells form a row, in register, along the base of the cells. Distal to the nucleus, more than a third of most “eyespot” epithelial cells are crowded with brown, granular pigment. In some cells the pigment fills the whole cell above the nucleus. In other “eyespot” epithe- lial cells the distal cell half is devoid of pig- ment and tapers to its free surface, some- times to form what appears to be a pore. In yet other instances, the entire cell distal to the “eyespot” epithelial nucleus may be ex- panded, without pigment granules. Unlike the epithelium on the inner surface of the flap, or much of the outer flap surface, the “eyespot” epithelium is thrown into distinct folds (Fig. 2). In the series of transverse sec- tions examined for this study, sections near the outer edges of the “eyespot” typically showed three or four folds, and sections of the center of the “eyespot” displayed as many as eight. Almost invariably, the base of the epithelial folds was associated with one or more clusters of extracellular pigment gran- ules, as described below. Extracellular pigment granules In the hundreds of serial sections examined for this study, extracellular pigment granules were found frequently, and were always 246 KRAEMER 8 SWANSON FIG. 4. Nerves associated with “eyespot” region of mantle flap. A, photomicrograph showing nerves associated with longitudinal and vertical muscles and with connective tissue of flap; B, photomicrograph showing nerve abutting connective tissues; CT, connective tissue trabecula; LM, transversely sectioned, longitudinally arranged muscle; N, nerve; VM, longitudinally sectioned, vertically arranged muscle in the mantle flap. LAMPSILIS MANTLE FLAP “EYESPOTS” 247 FIG. 5. Photomicrograph of transverse section of region adjacent and distal to mantle flap “eyespot” with ciliated epithelium and underlying pale cells. C, cilia; CT, connective tissue trabecula; LM, transversely sectioned, longitudinally arranged muscle in the mantle flap; OE, outer epithelium (other than “eyespot” epithelium) of the mantle flap; PC, pale cells. associated with the “eyespot” epithelial cells as follows: 1) At the base of two or three epithelial cells, typically within an epithelial fold, clus- ters of 3-12 large, round pigment granules occurred. 2) Small clusters of pigment granules were frequently seen surrounded by a fold of base- ment membrane, adjacent to the base of the epithelium, and often accompanied by a nu- cleus or other apparent remnants of an epithelial cell (Fig. 6). 3) Pigment granule clusters, encased in a connective tissue membrane, occurred just beneath the epithelial basement membrane (Fig. 6). 4) Large encased clusters, formed from apparent fusion of several smaller clusters, were frequently seen attached to connective tissue trabeculae, the latter extending at right angles from the basement membrane to the interior of the flap (Figs. 2, 6B). 5) The largest pigment clusters, sur- rounded by connective tissue membranes, were invariably found butted up against the vertical muscles (shown longitudinally sec- tioned in Figs. 1, 4A) near the center of the flap. 6) Not infrequently, very small longitudinal- ly arranged muscles (shown transversely sec- tioned) were found enclosed within the pig- ment clusters (Fig. 6B). 7) Rarely, extracellular pigment granules were seen between or distal to the long axis of the “eyespot” epithelial cells. We conclude from extended examination of tissues that pigment production by the “eye- spot” epithelial cells is followed by packaging of pigment into the conspicuous granules. The pigment granules then appear to undergo extracellular transport and distribution among the tissues in the interior of the mantle flap. Pigment transport seems to occur in the se- quence indicated by items 1-5 above. К also seems likely that pigment transport is effected during the rapid, rhythmic move- ments of the mantle flaps (Fig. 1). The flap movements are apparently produced by the well-innervated longitudinally-arranged mus- cles (accounting for the moving “pulse”) and 248 KRAEMER 8 SWANSON р О 6B FIG. 6. Photomicrograph of transverse section of Lampsilis ventricosa mantle flap “eyespot” region detailing pigmented epithelium, pigment granules and pigment clusters. A, note separation of basement membrane laminae at base of fold of “eyespot” epithelium; B, formation of pigment granula cluster in association with connective tissue trabecula; BB, brush border of epithelium; BM, basement membrane; CT, connective tissue trabecula; CTL, connective tissue lamina, separated from basement membrane: EE, “eyespot” epithelium on outer surface of mantle flap; PG, pigment granule; PGC, pigment granule cluster. LAMPSILIS MANTLE FLAP “EYESPOTS” 249 FIG. 7. SEM micrographs of “eyespot” epithelial surface. A, showing columnar structure of epithelial cells visible along a crack which was made during the preparation of the tissue; B, detail showing distinct microvilli on surface of epithelial cells; C, cilia; DEE, distal surface of “eyespot” epithelial cell; EE, “eyespot” epithelium on outer surface of mantle flap; MV, microvilli. by the vertically-arranged muscles (account- ing for the downward and outward jerk of the flaps at the end of each pulse, Fig. 1e). Di- rectional movement of pigment granule clus- ters to the interior of the flap could be aided by the connective tissue trabeculae in the “eye- spot” region of the flap. SUMMARY AND DISCUSSION From evidence examined for the present study, the “eyespot” of the Lampsilis mantle flap apparently does function in pigment pro- duction and transport. It is also clear that the “eyespot’ does not manifest the histological characteristics of a specialized sensor found elsewhere in freshwater bivalves: (1) there is no conspicuous thinning of the basement membrane underlying the “eyespot” epithe- lium; (2) there is no evidence of extensive innervation of the “eyespot” epithelium. Both of these phenomena are characteristic of the osphradial epithelia in certain freshwater bivalves (Kraemer, 1981). One might argue that the microvillar surface on the “eyespot” epithelial cells constitutes evidence of a sen- sory function, as certainly microvilli are com- ponents of rhabdomes which constitute the photoreceptor organelles in many mollusks. However, in this case we do not think so, for the foregoing reasons and because the mi- crovilli do not have the appearance of rhabdomeric membranes. Of course, neither the techniques of TEM or histochemistry were used in this study. Hence, it is possible to argue that TEM might turn up rhabdomes, or that histochemical experiments might de- lineate the kind of secretory activity occurring in the “eyespot” epithelial cells. However, in addition to the evidence pre- sented here, extensive studies of the be- havior of the mantle flaps (Kraemer, 1969, 1970) do not indicate a sensory function for the “eyespots.” To the contrary, in the present study, the “eyespot” epithelium and the com- plex, extensively innervated muscle com- ponents of the “eyespot” region of the mantle flap, are more likely to be effectors than sen- sors. Indeed, an effector function comports well with the whole function of mantle flap 250 KRAEMER 8 SWANSON movements in the spawning behavior com- plex of Lampsilis (Fig. 1). As mentioned above, earlier experimental evidence (Welsh, 1933; Kraemer, 1970) did indicate that in certain but not all species of Lampsilis the rate of mantle flap movement may be affected by changes in light intensity. The present study was undertaken on one of the light-sensitive species, Lampsilis ventri- cosa, in part to determine whether mantle flap “eyespots” are sensors. The results indicate that L. ventricosa “eyespots” are probably not sense organs. How, then, does one account for the appar- ent photic response of the mantle flaps men- tioned above? Specialized photoreceptors have not been identified in fresh-water mus- sels (Unionidae), despite the siphons of many species being “light” or more accurately “shadow” sensitive (skioptic). In at least one gastropod, Lymnaea, Stoll (1972) demon- strated a non-ocular light sensor. Kennedy (1960) demonstrated photoreceptive activity for the pallial nerve of the marine bivalve Spisula, although he was unable to identify pertinent photoreceptor pigments there. Conly-Dillon (1965), in his work on the spec- tral sensitivity of Pecten, notes that his find- ings do not exclude the possibility that there may be light-sensitive structures within the bivalve nervous system itself. The present study indicates that for Lampsilis, too, photo- sensors will probably have to be sought else- where than the “eyespot” of the mantle flap. As noted earlier (Kraemer, 1970: 241), “be- cause no photoreceptive function has yet been demonstrated for the ‘eyespot,’ the term is inappropriate.” “Eyespot” was used throughout that study, however, because the term was established in the literature, and because many lampsilids possess numerous other (1.е., non-raised) pigment spots. The present study provides abundant histological evidence of a “non-eye” function for the “eye- spot.” Perhaps Lampsilis “eyespots” could more appropriately be termed “pigment effectors” or “pigment effector” spots. There is compell- ing evidence from this study of (1) the produc- tion of pigment granules by the “eyespot” epithelial cells; and (2) intercellular transport of pigment granule clusters from the base of the epithelial cells to the interior of the flap. Is this phenomenon a direct or indirect effect of the pulsing movements of the mantle flaps? Whatever the answer, it seems likely that a function of the “eyespot” and associated mus- cles, nerves, and connective tissue trabecu- lae described here in the mantle flaps of Lampsilis is the patterned distribution of pig- ment within the tissues of the mantle flaps. While the pulsing movements of Lampsilis are confined to specific regions of the mature female mantle, spontaneous rhythmic move- ments of the bivalve mantle have been known for years (e.g., Redfield, 1917; Barnes, 1955; Bullock & Horridge, 1965). In the context of the present study, it seems logical to hypothesize that one effect of either localized mantle movements (e.g., in Lampsilis mantle flaps) or of more generalized, spontaneous mantle movements might be the distribution of pigment in characteristic patterns through the mantle tissue. ACKNOWLEDGMENTS It is a pleasure to thank Dr. George M. Davis as well as three anonymous reviewers for their critical examination of the manu- script, and their very helpful suggestions. It is also a pleasure to thank Roxanne Rackerby for her careful rendering of Fig. 2, and of the insert for Fig. 1, and to thank Sarah R. Orr for her skillful typing of the manuscript. LITERATURE CITED BARNES, G. E., 1955, The behaviour of Anodonta cygnea L. and its neurophysiological basis. Jour- nal of Experimental Biology, 32: 158—174. BULLOCK, T. G. & HORRIDGE, G. A., 1965, Struc- ture and function in the nervous systems of invertebrates, |. Freeman, San Francisco, 1611 p. BURCH, J. B., 1975, Freshwater unionacean clams (Mollusca: Pelecypoda) of North America. Re- vised ed., Malacological Publications, Hamburg, Michigan, 204 p. CONLY-DILLON, J. R., 1965, Spectral sensitivity of the scallop Pecten maximus. Science, 151: 345— 346. COUNTS, С. L., Ш 8 PREZANT, R.S., 1979, Shell structure and histochemistry of the mantle of Corbicula leana (Bivalvia: Sphaeriacea). Amer- ican Zoologist, 19: 1007. DAVIS, G. M. & FULLER, S. L. H., 1981, Genetic relationships among Recent Unionacea (Bival- via) of North America. Malacologia, 20: 217—253. GILLOTEAUX, J., 1979, Histochemical detection of monamine oxidase activity In smooth muscle and epithelial tissues of Mytilus edulis and Myti- lus galloprovincialis. Acta Histochemica, 65: 15— 24 JOHNSON, В. |. 1980, Zoogeography of North LAMPSILIS MANTLE FLAP “EYESPOTS” 251 American Unionacea (Mollusca: Bivalvia) north of the maximum Pleistocene glaciation. Bulletin of the Museum of Comparative Zoology, 149: 77-189. KENNEDY, D., 1960, Neural photoreception in a lamellibranch mollusc. Journal of General Physiology, 44: 277-299. KRAEMER, L. R., 1968, A comparative morphological study of mantle innervation, of statocysts, and of eyespots in the genus Lamp- silis (Pelecypoda). American Zoologist, 8: 802— 803. KRAEMER, L. R., 1969, The functional bilateral symmetry of the Lampsilis mantle: some pro- blems. Annual Reports for 1969 of the American Malacological Union, р. 28-30. KRAEMER, L. R., 1970, The mantle flap in three species of Lampsilis (Pelecypoda: Unionidae). Malacologia, 10: 225-282. KRAEMER, L. R., 1981, The osphradial complex of two freshwater bivalves: histological evaluation and functional context. Malacologia, 20: 205— 216. MANE, S. Y. & PATIL, V. Y., 1980, Histochemical analysis of muco-substances of the ventral mar- ginal folds of the mantle in Lamellidens con- sobrinus. Folia Histochemistry, Cytochemistry, 18: 47-52. РЕТИТ, 5. DAVIS: МЕ. & JONES; В. Е., 1978; Morphological studies оп the mantle of the fresh- water mussel Amblema (Unionidae): scanning electron microscopy. Tissue & Cell, 10: 619- 628. REDFIELD, E. S. P., 1917, The rhythmic con- tractions in the mantle of lamellibranchs. Journal of Experimental Zoology, 22: 231-239. SCHMITZ, E. H., 1967, Visceral anatomy of Gam- marus lacustris Sars (Crustacea: Amphipoda). American Midland Naturalist, 78: 1-54. SICK, L. V. & SIEGFRIED, C. A., 1980, Calcium and amino acid fluxes in Crassostrea virginica mantle tissue in response to changes in ambient salinity concentrations. American Zoologist, 20: 137. SORENSON, A. L., WOOD, D. S. & KIRSCHNER, L. B., 1980, Electrophysiological properties of resting secretory membranes of lamellibranch mantles: interaction between calcium and potas- sium. Journal of General Physiology, 75: 21-38. STEPHENS, P. J., 1978a, The sensitivity and con- trol of the scallop mantle edge. Journal of Ex- perimental Biology, 75: 203-222. STEPHENS, P. J., 1978b, Mechanical and chemi- cal sensitivity at the mantle edge of the file clam Lima scabra. Marine Behaviour and Physiology, 5: 79-90. STOLL, С. J., 1972, Sensory systems involved in the shadow response of Lymnaea stagnalis (L.) as studied with the use of habituation phe- nomena. Proceedings of the Koninklijke Neder- landse Akademie van Wetenschappen, Ser. C (Biol. Med. Sci.), 75: 342-351. WELSH, J. H., 1933, Photic stimulations and rhythmical contractions of the mantle flaps of a lamellibranch. Proceedings of the National Academy of Sciences, 19: 755-757. WHEELER, A. P., 1979, Oyster mantle carbonic anhydrase: evidence for plasma membrane- bound activity. American Zoologist, 19: 995. Ab- stract. ZABA, В. N., 1981, Lycogenolytic pathways in the mantle tissue of Mytilus edulis. Marine Biology Letters, 2: 67-74. ZABA, В. М. & DAVIES, J. |., 1981, Carbohydrate metabolism in isolated mantle tissue of Mytilus edulis: isotopic studies on the activities of the Embden-Myerhoff and pentose phosphate path- ways. Molluscan Physiology, 1: 97-112. i) Du il | ai 14 YN Ni nf oe L 1 nl = N 11 ls ay mM) | > 1 |' | A a vi [al | 1 } Y Hy Г | it U i ih Ш i i E m nr | 4 р f ni i | | u Ñ LA т \ . Ú fy à \ ni п В : . Su = N, | DD Mm а —— : ur т if iy fl i р у | ut 1 у | y 1 AK UN р | | | р wein у 7 ] | Pal 1 } n° Wa m das Г 1 ] ur N m ih 1 й у Is у A ra en Dr № ie ir | : | 5 he и | ous № ny | И. { 1 u Ш О SAT PET ANTON MR, ee er 1, ET Yeh wy № at: ua) wa PAT ARTE | A if io. р р A > er О | р м в ИТ | И MALACOLOGIA, 1985, 26(1-2): 253-271 A NEW MUSSEL (BIVALVIA, MYTILIDAE) FROM HYDROTHERMAL VENTS IN THE GALAPAGOS RIFT ZONE Vida Carmen Kenk' 8 Barry В. Wilson? ABSTRACT A new subfamily, Bathymodiolinae, and new genus and species, Bathymodiolus thermophilus, are described from material collected by the 1977 and 1979 expeditions to the hydrothermal vents in the Galapagos Rift Zone. This large modioliform mussel has very unusual anatomy, exhibiting extreme mantle fusion which restricts the incurrent aperture to a short byssal-pedal gape in the ventral midregion. The gills lack food grooves ventrally; the free edges of the gills fit axial ridges on the visceral mass and mantle lobes, thereby isolating the dorsal excurrent chambers from the rest of the mantle cavity. The gut is short and different from that of other mytilids in lacking a recurrent loop, the stomach is simple and lacks a deep sorting caecum, dorsal hood and left pouch, and there are but three pairs of digestive ducts opening into the stomach. The auricles of the heart have a broad connection to the longitudinal vein laterally between the branches of the divided posterior retractor muscles in addition to the normal connection anterior to these muscles. The kidney is very small. Feeding is discussed in light of high densities of chemoautotrophic sulphur-oxidizing bacteria in the environment and the possibility of a symbiotic relationship between the mussels and bacteria. INTRODUCTION The discovery in 1977 of biological com- munities surrounding hydrothermal vents in the Galapagos Rift Zone at latitude 00.47°N (Corliss & Ballard, 1977; Lonsdale, 1977; Corliss et al., 1979; Enright et al., 1981; Edmond, 1982) led to the Galapagos Rift Biology Expedition in 1979 (Ballard 8 Gras- sle, 1979; Galápagos Biology Expedition Par- ticipants, 1979). Since the initial discovery, additional submarine hydrothermal com- munities have been described at 21°N (Rise Project Group, 1980) and 11-13°N (Des- bruyeres et al., 1982). The majority of speci- mens collected on these expeditions are un- usual organisms differing from known rela- tives at generic or higher levels (Newman, 1979; Williams, 1980; Burreson, 1981; Fretter et al., 1981; Jones, 1981; Krantz, 1981; McLean, 1981; Desbruyeres & Laubier,1982; Williams & Chace, 1982). One of the most abundant and conspicuous organisms collected at some of these hydro- thermal vents is a large modioliform mussel. Although the shell form is like that of the mytilid genus Modiolus, anatomical study of preserved specimens has revealed many dis- tinctive features. This animal is described here as a new genus and species and a new subfamily is erected for it. The mussels were abundant at several vent sites in the Galapa- gos Rift Zone. The species is also present, though apparently less abundantly, at the 11— 13°N site, but was not collected or observed at the vents at 21°N. MATERIALS AND METHODS All of the specimens examined in this study were from the Galapagos Rift Zone vents, viz.: a) 79 specimens preserved in ethanol (size range 0.3 to 14.38 cm in length) collected during the 1977 expedition at Clambake 1, Oyster Bed, and Garden of Eden vent sites (dive stations 713, 723, 727, 728 and 733) and forwarded to the authors by Dr. Jack Corliss. b) 11 specimens preserved in ethanol (size range 7.9 to 16.2 cm) collected during the 1979 expedition at Rose Garden and Mus- sel Bed vent sites (dive stations 879, 880, 894 and 896) and forwarded to the authors by Dr. Fred Grassle, and 153 juveniles "Department of Biological Sciences, San Jose State University, San Jose, CA 95192 U.S.A. “Museum of Victoria, Division of Natural History and Anthropology, 285-321 Russell St., Melbourne, Victoria 3000, Australia. New address: Dept. Conservation & Land Management, Matilda Bay, Nedlands, Western Australia 6009, Australia. (253) 254 ati H FIG. 1. Diagram indicating the measurement taken for length and height. Width is greatest dimension through both valves. (size range 0.3 to 9.8 mm) collected from washing of mussels from dives 880 and 884, loaned by Dr. Howard Sanders. c) 75 dried shells deposited at the U.S.N.M., lot registration number 81331600-3380- P30000. Seven preserved specimens from sample a) were dissected under a binocular micro- scope. Anatomical drawings were done free- hand; shells were drawn with the aid of a camera lucida. Measurements taken are illus- trated in Fig. 1. The holotype and a large series of para- types are lodged at the U.S.N.M. Paratypes are also lodged at the following museums: California Academy of Sciences, San Fran- cisco; Museum of Comparative Zoology, Har- vard University; Los Angeles County Mu- seum; British Museum of Natural History, London; Museum National d'Histoire Naturelle, Paris; Museum of Victoria, Mel- bourne; Zoologisk Museum, University of Copenhagen; Academy of Natural Sciences, Philadelphia; Scripps Institute of Oceanogra- phy. KEY TO ABBREVIATIONS IN FIGURES AA, aa anterior adductor muscle a art anterior artery aff v afferent vein a | ascending lamella an anus AR, ar anterior retractor muscle au auricle b byssus bg byssal gland KENK & WILSON PA, pa pbr(a), APR pbr(p), PPR per PL pm ppr r r ap rg f DIE $ ЭВ: Sh st sta st p s v-au p V vsm byssal-pedal gape branchial septum connecting bar dorsal cul de sac digestive duct descending lamella efferent vein excurrent siphon foot food groove gill axis gastric shield position genital aperture anterior genital duct intestine incurrent chamber internal diaphragm inner demibranch inner excurrent chamber intestinal groove inner mantle fold kidney labial palp labial palp muscle longitudinal vein major typhlosole mouth oesophagus outer demibranch outer excurrent chamber oral groove outer mantle fold papilla posterior adductor muscle posterior byssal retractor mus- cle (anterior) posterior byssal retractor mus- cle (posterior) pericardium pallial line pallial muscles posterior pedal retractor mus- cle rectum renal aperture ridge for gill attachment reno-pericardial channel septum of principal filament siphonal retractor muscles stomach stomach anterior chamber stomach posterior chamber secondary venous-auricular passage ventricle valvular siphonal membrane BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 255 TAXONOMY Family MYTILIDAE BATHYMODIOLINAE Kenk & Wilson, subfam. nov. Type-genus: Bathymodiolus Kenk 8 Wilson, gen. nov. BATHYMODIOLUS Kenk 8 Wilson gen. nov.? Type-species: Bathymodiolus thermophilus Kenk 8 Wilson, sp. nov.? Diagnosis of the subfamily and genus: Shell smooth, modioliform, with sub- terminal umbones; hinge edentulous; perios- tracum hirsute; posterior retractor muscle di- vided, retractor scars separate; pallial mus- cles including siphonal retractors strong; ex- current siphon short, extensible, with internal diaphragm; inner folds of the mantle lobes enlarged, extensible postero-ventrally, fused in the mid-line antero-ventrally and postero- ventrally; auricles enlarged, fused posteriorly; gills heterorhabdic, eleutherorhabdic, with short, fleshy filaments, lacking food grooves at ventral edges of demibranchs; with tubular connections present posteriorly between free edges of ascending lamellae and gill axes; labial palps small; stomach without a deep sorting caecum or left pouch; intestine short, lacking a recurrent loop. Bathymodiolus thermophilus Kenk & Wilson, sp. nov. Type-locality: lat. 00°47'.89"N; long. 086°09'.21”W. Depth 2495 m. R/V Alvin Dive 879, at “Mussel Bed” geothermal vent, Gala- pagos Rift. Holotype (Fig. 2.): USNM 803661, preserved in 70% ethanol. Collected 20 Jan- uary 1979 by Ellis and Ballard on Alvin. Measurements: length 14.95 cm, height 6.30 cm, width 5.83 cm (Fig. 2). ®In a paper on a similar or identical mussel from the East Pacific Rise at 11°-13°N, Le Pennec et al. (1984: 70) introduced Bathymodiolus as a nomen nudum. Also, the generic name has been used repeatedly in a popular article by Laubier & Desbruyeres (1984); the manuscript species name B. thermophilis [sic] appeared on p.1510. Information in the former paper unfortunately was not considered by the authors of the present paper. ED. DESCRIPTION Shell morphology. Modioliform, solid, ellipti- cal in juveniles and subadults, arcuate in old specimens, equivalve. Anterior end rounded; dorsal margin slightly convex; postero-dorsal corner rounded in adults, angular in juveniles; posterior end rounded; ventral margin nearly straight in specimens less than 10 cm length, slightly concave in larger specimens (Fig. 3). Umbones subterminal, prosogyrate. External surface smooth, sculpture lacking, dull white beneath periostracum. Interior white, nacreous. Periostracum straw-yellow, yellow-brown antero-ventrally, often stained dark brown in large specimens. In young specimens less than 0.8 cm in length perios- tracum smooth, larger juveniles develop peri- ostracal hairs (of byssal origin, see Bottjer & Carter, 1980; Ockelmann, 1983) on posterior slope; specimens more than 2 cm long have hairs on most of shell exterior; hairs broad, flat. In addition to their own hairs many shells bear byssal end-plates of other mussels which had been attached to them, distinguish- able by oval shape and central slender strand. Ligament opisthodetic, parivincular, strong, extending most of length of dorsal margin; resilial ridge (as defined by Soot-Ryen, 1955: 7) deep, chalky and rather soft, not pitted; sub-ligamental shell ridge strong and angular, with deep groove between it and ligament anteriorly, becoming obsolete below mid- point of ligament. Hinge edentulous except for strong backward pointing projection of anteri- or hinge margin beneath anterior end of liga- ment; post-ligamental denticles lacking. Muscle scars (Fig. 4). Anterior adductor muscle scar half-moon shaped, located below umbo, distant from anterior margin (Fig. 4); young specimens may show small round scar of labial palp support muscles just behind anterior adductor; anterior byssal-pedal re- tractor scar oval, located high within umbonal cavity behind umbo; posterior adductor scar rounded-rectangular; posterior byssal-pedal retractor muscles form two separate scars with large gap between them, anterior one elongate-elliptical, located high, and close to posterior end of ligament, second one ellipti- cal and located antero-dorsally to but con- tiguous with posterior adductor scar to form a joint comma-shaped scar. Pallial line distinct, extending ventrally from anterior adductor to posterior adductor, curving upwards to form 256 KENK 8 WILSON FIG. 2. Bathymodiolus thermophilus, holotype, USNM 803661. A, anterior view; B, posterior view; C, dorsal view: D, ventral view; E, lateral view, right valve; F, lateral view, left valve. BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 257 5 ст FIG. 3. Growth series of shells illustrating change in form (paratypes, USNM 813316). FIG. 4. Diagram of shell illustrating muscle scars. 258 KENK 8 WILSON concavity in byssal region about Ys of the distance from anterior end; small siphonal retractor scar present adjacent to posterior adductor scar at posterior end of pallial line. Measurements. Rhoads et al. (1982) re- ported the largest shell in their series as 18.4cm long. Maximum shell length observed in this study 16.3cm. Length, height and width proportions of preserved specimens (paratype series, N = 79): mean height/length 0.568; range 0.514 to 0.604 mean width/length 0.362; range 0.323 to 0.438 Anatomy Musculature (Fig. 5). Main features of mus- culature evident from previous description of muscle scars. Posterior byssal retractors in two roughly equal main bundles arising together at base of byssus but diverge and attach separately to shell. Posterior pedal re- tractors thick, arising from base of foot anteri- or to origin of posterior byssal retractors, passing dorsally lateral to anterior retractors and inserting dorsally on both inner and outer ar Ppr sides of most anterior bundles of posterior retractors. Pallial muscles unusually well developed; strong siphonal retractors present, formed of amalgamated strands originating in inner mantle folds in region of excurrent siphon. Slender strand of anterior pedal retractor muscle extends anteriorly and attaches to shell behind anterior adductor, providing sup- port for labial palps. Posterior adductor large and divided into “quick” and “catch” parts; anterior adductor elongate, half-moon shaped. Foot and byssus (Fig. 5). Foot thick, flat- tened, terminally swollen; byssal groove run- ning along ventral surface almost to tip. Bys- sus profuse, usually emerging as separate strands from orifice, strands thick and strong; byssal gland yellow, extends down centre of foot behind groove, without extension dorsal to anterior retractor muscles. Mantle. Mantle lobes thin dorsally but be- come unusually thickened and muscular near posterior and ventral edges (Fig. 6). Free edges of mantle lobes have three folds as in other mytilids (Yonge, 1957); inner folds fuse Pbr(a) pbr(p) FIG. 5. Musculature; left valve, mantle lobe and ctenidia removed. BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 259 cb iexch rg +9? ¡dem im f aff у oexch A ieee A e У — pm FIG. 6. Transverse section of the right gills, through primary filaments and with one of several connecting bars between the free edge of the inner ascending lamella and the gill axis. mesially along the entire postero-dorsal slope and ventrally enclosing the mantle cavity to an unusual extent. Excurrent siphon formed posteriorly be- tween inner folds, capable of moderate exten- sion but shown in retracted position in Fig. 7; thin internal diaphragm with narrow horizontal aperture partly occludes inner end of ex- Current siphon. Fusion of inner mantle folds immediately below excurrent siphon forms horizontal shelf, the branchial septum, with an inner part reaching forward to ventral side of posterior adductor (Fig. 7). Branchial septum separates incurrent and excurrent chambers posteriorly; posterior ends of gill axes attach to its ventral surface. Ventral development of branchial septum forms an oblique, transverse partition, the valvular siphonal membrane (terminology of Yonge, 1955), joining left and right lobes postero-ventrally thus enclosing incurrent mantle cavity in that region; inner folds also fused in mid-line antero-ventrally; incurrent aperture thus confined to a short ventral pedal-byssal gape (Figs. 7, 8). Rim of gape bordered by another muscular fold which may regulate aperture size by muscular contraction. A small papilla present at posterior end of gape (Fig. 8). Free edges of inner folds form wide exten- sible frills postero-ventrally shown partly ex- tended in Fig. 7. Fig. 12 shows them fully 260 KENK 8 WILSON st al 5 v-aup ап ех $ | FIG. 7. Vascular and alimentary systems and siphonal structure; left valve and mantle lobe removed; fused inner mantle folds cut down the mid-line (in sagittal section). Outer demibranch deformed. extended in life. This structure forms function- al, though not tubular, incurrent siphon by apposition of edges ventrally. Excurrent and incurrent siphons separate as in Botula (Wilson & Tate, 1984). Mantle cavity. Mantle cavity divided by ctenidia laterally and branchial septum poste- riorly into ventral incurrent and dorsal ex- current chambers. Edges of ascending lamel- lae flanged and fitted to muscular longitudinal ridges on surfaces of mantle lobes and viscer- al mass thus completely separating incurrent and excurrent chambers in life (Fig. 6). In this way four tunnel-like, longitudinal excurrent chambers are formed along roof of mantle cavity, two on each side; chambers meet posteriorly at entrance of excurrent siphon above branchial septum. A cul de sac of excurrent chamber passes posterodorsally above rectum and posterior adductor, reaching forward as far as posterior wall of pericardium (Fig. 7); thin pericardial wall separates pericardial fluids from sea water in excurrent mantle cavity. Ctenidia. Paired ctenidia consist of inner and outer demibranchs each with descending and ascending lamellae forming W-shaped gill typical of mytilids (Fig. 6); demibranchs approximately equal-sized, inner demi- branchs extend slightly further anteriorly, out- er demibranchs slightly deeper; both demi- branchs end abruptly anteriorly (Fig. 9). Cte- nidia filibranchiate, heterorhabdic and eleutherorhabdic; interlamellar junctions lack- ing but every third to seventh filament is “prin- cipal filament” (see type B(1b) of Atkins, 1937, text fig. 4) with septum or “baffle” rising to more than half the height of gill (Fig. 6). Demibranchs rather short, filaments wide and fleshy; ventral edges lack food grooves though minute indentations present. Deep folds on outer surface of ascending lamellae just below free edges might function as food grooves; anteriorly folds continue in a loop as grooves on mantle wall and terminate in deep oral groove between labial palps leading into mouth. Inter-lamellar tissue junctions lacking; series of about four large tubular connections present in posterior area between free edges and gill axes (Figs. 6, 10) appearing to con- nect efferent veins with either afferent or BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 261 1 MES FF SRE г. AY | A Il АЯ Æ || = ра Fig. 8. External view from the ventral side (shell valves removed) showing extensive mid-line fusion of the inner mantle folds, the extended valvular siphonal membrane of the branchial septum and the small byssal-pedal gape. longitudinal veins. Filaments sometimes thickened, shortened, deformed, particularly posteriorly, possibly due to activities of poly- chaetes. Labial palps. Paired labial palps short, broad, flat, triangular; usually strongly plicate on their inner surfaces (Fig. 9); outer palps larger than inner palps and placed farther posteriorly, markedly so in some large speci- mens (e.g. Fig. 5). In very large specimens palps sometimes smooth, lacking plications on either surface. Alimentary system. Digestive tract short, more or less straight, direct. Mouth trans- verse, slit-like; esophagus enters anterior end of stomach which lies superficially in visceral mass below ligament. Stomach (Fig. 11; nomenclature of parts follows Reid, 1965) small elongate, divided into anterior and posterior chambers; back- ward-pointing pouch on left dorsal side of anterior chamber; posterior chamber swollen on right side, left pouch lacking; gastric shield small, located in normal position on antero- dorsal wall on left side of posterior chamber. Three pairs of digestive ducts enter stomach laterally (Fig. 11) one pair on left and right sides of anterior chamber and two pairs on left and right side of posterior chamber; on right side two ducts open into posterior cham- ber close together; on left side two openings spaced apart with posterior one much the larger. Major typhlosole straight except for an elbow close to its entry into intestine, passes along floor of posterior chamber in mid-line and terminates in centre of anterior chamber; minor typhlosole branches on right side at elbow and passes up right side of posterior chamber (not traced further). Intestinal groove originates in opening of posterior di- gestive duct of left side and passes forward along floor of stomach to left of major typhlo- sole, passes around tip of that typhlosole in the anterior chamber, and returns down right side to enter intestine posteriorly. Surface of major typhlosole transversely plicate. Minor intestinal groove runs along anterior side of minor typhlosole on right side of posterior chamber but its origin not located. Hood groove not observed but this may have been a consequence of the poor preservation. Style sac and intestine conjoined; crystal- line style present in some preserved speci- mens. Intestine leaves posterior end of stom- ach and traverses short distance posteriorly down mid-line; rectum turns upwards to enter pericardium and ventricle from below; thence passes posteriorly through ventricle and di- rectly down the mid-line to anus on posterior side of posterior adductor muscle; recurrent loop of intestine lacking. Vascular system. Pericardium in usual po- sition dorsally between posterior retractor muscles (Figs. 7, 12); broad reno-pericardial canal on each side passes laterally around most anterior of posterior retractor muscles, then ventrally to gill axis; canals superficial and easily seen when shell removed. Heart three-chambered (Fig. 12); medial ventricle thick, muscular, rhomboidal, traversed for much of its length by rectum; anteriorly ven- tral surface of ventricle fused to floor of peri- cardium. Paired anterior arteries arise from aortic bulb and pass forward over visceral mass; large ventral artery leads downwards through pericardial floor. Two auricles unusually large, fused together posteriorly (Fig. 12); each has an anterior arm 262 KENK 8 WILSON mth og — | \ Y aa ie A k 13 Z idem 1 FIG. 9. Ventral view of the labial palps with the mantle and anterior adductor cut away; the inner palp of the left side is pinned back to expose the oral grooves. curving laterally and downward within reno- pericardial canal but connection via oblique vein to longitudinal vein (see White, 1937, for description in Mytilus) not observed. Each auricle has wide latero-ventral flap pro- truding between bundles of posterior retractor muscles, with a wide foramen opening directly into longitudinal vein; valvular mechanism in that opening appears to be lacking. Efferent veins in free edges of demi- branchs, and afferent and longitudinal veins immediately above gill axis readily observable in hand-cut sections; longitudinal veins large and spacious in zone between reno- pericardial canal and posterior adductor. In dissections it appeared that there are several foramina between afferent and longitudinal veins in this zone; largest of these located directly below wide space connecting longitu- dinal veins and latero-ventral flaps of auricles. Plicate membranes lacking (see White, 1937 for details of these structures in Mytilus between visceral mass and gill axes and mantle lobes and gill axes). Tubular junctions between free edges of demibranches and gill axes already noted; whether these are vascular connections needs to be determined. Nervous system. Paired ganglia situated in normal positions, cerebral ganglia between anterior retractor muscles near attachment of inner labial palps; paired pedal ganglia medially just above region where anterior and posterior retractor muscles meet foot; paired visceral ganglia on ventral surface of poste- rior adductor muscle. BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL é a g ax E O NV = aes : = E aff v = gen ap LE NCA NCS FIG. 10. Location of the genital and renal apertures, and showing the connecting bars between the gill axis and the free edge of the ascending lamella of the inner demibranch. Reproductive system. In all large speci- mens examined gonad tubules were in re- gressed condition; confined to mesosoma and visceral mass over and behind digestive gland and below pericardium, lacking in man- tle lobes. Genital apertures located at tips of very small conical papillae in roof of inner excurrent Chambers at a point adjacent to byssus (Fig. 10). Excretory system. п transverse hand-cut section taken through body below pericar- dium, a small axial duct closely associated with longitudinal vein was tentatively identi- fied as kidney; duct very thin-walled and im- possible to dissect out under the microscope, longitudinal extent of it, and whether or not it is recurved, could not be determined. Renal apertures extremely small slits on slight protuberances in roof of inner excurrent chambers, just behind genital apertures (Fig. 10). BIOLOGY Life history and dispersal. The relatively ephemeral nature of a given active vent site and the distance between sites require an effective mechanism for dispersal if a species is to survive after a vent site becomes in- active. Lutz et al. (1980) examined the larval shell by scanning electron microscopy and found that prodissoconch | is small relative to the size of prodissoconch Il. Comparisons they made with the larval shells of other myti- lids such as Mytilus edulis and Modiolus mod- iolus suggest that Bathymodiolus thermophi- lus has long-lived planktonic larvae which could be transported from one vent site to another. The small size of prodissoconch | suggests very high fecundity. Lutz et al. (1980) proposed that these larvae may be induced to settle and undergo metamorphosis by encountering elevated temperatures at vent areas and might even delay metamor- phosis in the absence of this stimulus. Growth rates. Rhoads et al. (1981, 1982) derived growth rates from marked and recov- ered mussels at the “Mussel Bed” and “Rose Garden” sites and from recently settled young, which are among the highest recorded for deep-sea species. Mature mussels have mean growth rates of about 1 cm per year. 264 KENK 8 WILSON FIG. 11. Dorsal view of stomach with the dorsal wall removed, showing the openings of digestive ducts and the major typhlosole and intestinal groove. Juveniles may reach a length of 27 mm in 294 days or less. Growth rates appear to be in- fluenced mainly by food concentration. Mus- sels in dense populations close to the vents where there was a high density of microbial food grew two or three times faster than those in peripheral locations where density of micro- bial food was less. Growth rings in different individuals could be tentatively correlated at the “Rose Garden” site, indicating synchronous change in mus- sel growth, possibly in response to change in temperature and nutrient conditions. If such thermal pulsing occurs, it may serve as a cue for gonad development or spawning. As noted above, all of the mussels examined in our study had gonads in regressed condition, which also implies synchrony of the gameto- genic cycle. Interactions with other species. TV tape recorded the brachyuran crab, Bathograea thermydron (Williams, 1980) crawling over mussels and occasionally probing them. Rhoads et al. (1982) considered these crabs to be the most likely predators on the mus- sels. They found that shells often show re- paired damage, especially in the region of the byssal notch. They suggested that mussels shorter than 2.0 cm are usually consumed if attacked. Unsuccessful predator attacks appear to be most frequent in mussels of shell lengths between 2.0 and 5.5 cm while larger mussels appear to be ignored. About one third of the preserved mussels examined contained the polynoid polychaete Branchipolynoe symmytilida (Pettibone, 1984). The worms occurred in the mantle Cavity, usually in the posterior region. The gills of specimens with these polychaetes were often thickened and uneven, possibly due to disturbance by the worms. Not all mussels with deformed gills had worms at the time of collections, nor were the gills de- formed in all specimens with worms inside. /n situ TV tape photography shows a live red polychaete leaving a mussel and swimming out of view as the mussel was being collected by R/V Alvin’s mechanical arm. Other poly- chaetes of apparently the same kind are seen swimming freely and crawling over the exteri- or of the mussels. Krantz (1981) has described a new and unusual species of predatory mite recovered from detritus associated with a sample of mussels. DISCUSSION Shell form and structure of Bathymodiolus thermophilus are typical of the Mytilidae, most closely resembling Modiolus. The anatomy also conforms generally with that of the Mytili- dae but there are several features, though common within the family, which are not found in Modiolus, and others that are unique. In the former category are the internal dia- phragm within the excurrent siphon (as in Mytilus White, 1937; Lithophaga and Leiosolenus Wilson, 1979, Botula Wilson & Tait, 1984) and the divided incurrent and ex- current siphons (as in Botula Wilson & Tait, 1984). The enlarged auricles with a second open- ing into the longitudinal vein between the two bundles of the posterior byssal retractor mus- cles and the very small, tubular kidney are unique features of the anatomy. The most obvious and remarkable feature BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 265 a art per au АА IOs. LEER gen d rpc pbr(a) s V-au p FIG. 12. Dorsal view of the heart showing the narrow anterior connection of the auricles with the longitudinal vein through the reno-pericardial channel, and the wide secondary connection between the anterior and posterior bundles of the posterior byssal retractor muscles. of this mussel is the extent of ventral fusion of the mantle lobes. It is achieved by normal fusion of the inner folds anteriorly, but poste- riorly it is achieved by extraordinary develop- ment of the valvular siphonal membrane. This structure is represented in some other myti- lids by a thin, transverse, oblique partition from the antero-ventral edge of the branchial septum, partly occluding the incurrent aper- ture (see Yonge, 1955; Botula Wilson & Tate, 1984, fig. 3). In this case the postero-ventral part of the mantle cavity is enclosed by the valvular siphonal membrane except for the short central byssal-pedal gape. Neverthe- 266 KENK & WILSON FIG. 13. Photograph of living mussels in situ; R-V Alvin Dive 885, courtesy of Dr. Robert Hessler. BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 267 less the free edges of the inner folds postero- ventrally are extensible and, judging from photographs of the animal in life (Fig. 13), must channel water along the ventral side of the body to that gape. The gills of Bathymodiolus, also, are anom- alous among mytilids for which gill structure is known. Although they have the typical ‘W’ shape, the gills are unusually thick and lack food grooves at the ventral edges of the demi- branchs. Perhaps the deep grooves at the free edges of the ascending lamellae are functional food grooves but even in that case the conclusion is inescapable that this mussel does not filter suspended particles in the usual mytilid way. Further evidence for a different feeding mechanism is found in the alimentary tract of Bathymodiolus. The short, direct gut lacking a recurrent loop and the internal structure of the stomach are quite unlike those of any other mytilid so far described. In Mytilus (Graham, 1949; Owen, 1955), Modiolus (Nelson, 1918; Reid,1965; Morton, 1977), Leiosolenus (Pur- chon, 1957, Adula (Fankboner, 1971) and Musculista (Morton, 1974) the stomach is a tumid, two-chambered organ with a deep antero-ventral sorting caecum and a promi- nent left pouch and dorsal hood posteriorly. But in Bathymodiolus the stomach is elon- gate, there is a small lateral pouch on the left side of the anterior chamber but no deep sorting caecum, and a left pouch of the poste- rior chamber is lacking. In those genera the major typhlosole originates in the caecum but in Bathymodiolus it originates on the ventral floor of the anterior chamber and runs straight down the mid-line to the entrance of the in- testine. In other genera the intestinal groove originates in the posterior left pouch and curves around the posterior chamber into the sorting caecum of the anterior chamber on the left side of the major typhlosole. In Bathy- modiolus it originates in the posterior di- gestive gland duct on the left side and runs straight forward along the left side of the major typhlosole, curves around its anterior end and returns down its right side to the entrance of the intestine. The transverse plications on the floor of the stomach between the two arms of the intestinal groove are not matched by any similar structures in other mytilids. Finally, in the other mytilid genera specified above the ducts of the digestive gland are numerous and asymmetrically grouped, while in Bathymodiolus there are three distinct pairs which enter the stomach laterally. Although the presence of a crystalline style, gastric shield and plications on the stomach floor confirm that the stomach of Bathymo- diolus is a particle-sorting chamber, it has a much more simple structure and organization than the stomachs of other mytilids which are extremely complex. If simplicity indicates primitiveness then it can be concluded that Bathymodiolus has a primitive gut. The dis- position of the paired and separate digestive ducts also may be regarded as primitive (Pur- chon, 1957). From these observations it is evident that the water circulation system within the mantle cavity, the capture and carriage of particles on the gills, and the sorting processes within the gut of Bathymodiolus thermophilus are atypical of the Mytilidae and that feeding in this species must differ from the usual. High concentrations of chemoautotrophic, sulphur-oxidizing bacteria occur in the sulphur-rich water in the vicinity of the hydro- thermal vents (Corliss et al., 1979; Galapagos Biology Expedition Participants, 1979). These bacteria could be food for the filter-feeding animals living there (Jannasch & Wirsen, 1979; Rau & Hedges, 1979; Karl et al., 1980; Williams et al., 1981). Suspensory feeding on such high concentrations of suspended bac- teria might indeed involve mechanisms differ- ent from the usual ones. There is also evidence that Bathymodiolus may obtain some or all of its nutrients through symbiosis with sulphur-oxidizing bacteria in the gills. Cavanaugh et al. (1981) discovered prokaryotic cells in the trophosome of the vestimentiferan tubeworm Riftia pachyptila Jones. Felbeck (1981) demonstrated the presence of sulphur-oxidizing and Calvin- Benson cycle enzymes in that organism, suggesting that sulphur-oxidizing bacteria ex- ist in a symbiotic relationship with it. Subse- quently Felbeck et al. (1981) have shown that the vent clams Calyptogena pacifica Dall and C. magnifica Boss & Turner (1980), and the mussel described here, also show evidence of sulphur-oxidizing enzyme activity in the gill tissues. They suggested that these bivalves inhabiting the sulphide-rich environments of the hydrothermal vents “are not only able to tolerate these toxic habitats, but in addition are capable of exploiting the energy of sul- phide to drive net CO; fixation and, thereby, reduce their dependence on ingestion of 268 KENK 8 WILSON photosynthetically fixed carbon.” The simple structure of the gills, labial palps and alimen- tary tract in Bathymodiolus is quite consistent with this possibility. Nevertheless, the an- atomical evidence indicates some degree of ciliary feeding on suspended particles, prob- ably bacteria. Nutrition based on symbiotic sulphur- oxidizing bacteria is now described for sever- al bivalves that live in sulphide-rich environ- ments (Reid & Bernard, 1980; Cavanaugh, 1983; Felbeck, personal communication). In discussing such symbioses in marine in- vertebrates, Reid & Bernard (1980) com- mented on the need for a burrow or tube in pogonophorans “to contain and confine the related organisms and prevent the dissipation of useful solutes.” The extreme degree of ventral mantle fusion in Bathymodiolus may also serve this function. In several of the preserved specimens, small particles of bright yellow particulate matter, presumed to be elemental sulphur, were observed trapped in mucus on the gills. Jones (1981) made a similar observation in Riftia. One might expect that an animal living in such a toxic environment would have a specially efficient excretory system. It is sur- prising, therefore, to find that the kidney in Bathymodiolus is so small. The vascular sys- tem, on the other hand, is unusually well developed. The clue to interpretation of these unusual structures may lie in the physiology of nutrition based on suspended or symbiotic sulphur-oxidizing bacteria, or in the mussels’ physiological tolerance to the sulphur-rich en- vironment. The phylogenetic affinities of Bathymo- diolus remain problematical. The similarity of the modioliform shell to that of Modiolus is clearly a case of parallelism for the anatomi- cal characters are very different. The extent of mantle fusion, the simple gut, the lack of ventral food grooves on the gills, the small kidney and the large auricles with second connections to the longitudinal veins, are all characters which have not been previously described in the Mytilidae. The physiological implications of these features indicate a differ- ent life-style and evolutionary origins to the Mytilinae, Modiolinae, Lithophaginae, Crenellinae or Musculininae. We confidently introduce the new subfamily Bathymodiolinae for this new genus and species. There are several small, modioliform myti- lids in the Pacific region. Knudsen (1970) described Modiolus abyssicola (Fig. 14B) 4 lmm FIG. 14. A. Small specimen of Bathymodiolus ther- mophilus. B. Holotype of Modiolus aybssicola Knudsen. from 3670-3270 m in the Gulf of Panama (5°49’N, 78°52'W). This small mussel (max. length recorded 17.2 mm) has an arcuate mo- dioliform shell similar to that of a medium to large-sized Bathymodiolus thermophilus although juvenile specimens of the latter are not arcuate (Fig. 14A). Because of this sim- ilarity and the proximity of the localities, early consideration was given to the possibility that the large mussels from the hydrothermal vents might be adults and the Panama ones juveniles of one species. Independently we have examined the anatomy of preserved specimens in the type-series of M. abyssico- la, through the courtesy of Dr. Knudsen. The posterior retractors of this species are divided but there is no siphonal development, the branchial septum is short and simple, there is a large incurrent aperture (gape) from the ventral side of the branchial septum to the anterior adductor, and the intestine is short but looped as in Modiolus. Gonads, gono- ducts and a prominent gonad aperture were observed in the larger specimens, dispelling any suggestion that these are juveniles. For these reasons, we conclude that M. abyssico- la may be a true Modiolus or at least a modio- linid, and the possibility of a relationship with the species described here can be rejected. Modiolus projectus Verco, 1908, from 200 BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 269 fathoms off South Australia is another small species (holotype length 10.9 mm) of similar form. It is characterized by a conspicuous “projecting lamina” below the ligament. The generic affinities of this species remain un- determined but this unusual character makes a relationship with Bathymodiolus improb- able. Adipicola Dautzenberg, 1927, is a genus containing several small deep water modioli- form mytilids. The prosogyrate umbos are situated well back from the anterior end and there is an anterior (lunular) keel as well as a posterior one. A sub-ligamental ridge is lack- ing, and the ligamental plate, though more or less vertical posteriorly, curves under the margin near the umbos. In A. simpsoni (Mar- shall, 1900) and A. argenteus (Jeffreys, 1876) there are pseudo-taxodont teeth or denticles on the dorsal margin beneath and behind the ligament; these are lacking in the type-spe- cies A. pelagica (Woodward, 1854). The anat- omy of these species is unknown to us although the posterior retractor muscle scars are not divided in A. simpsoni at least. The Pacific genus Terua Dall, Bartsch & Rehder, 1938, appears to be very close to Adipicola. The type-species T. pacifica Dall, Bartsch & Rehder, 1938, and T. japonica (Habe, 1971) also have hinge denticles. Dried specimens of the latter species in the USNM (204525) show a long ventral gape, no excurrent siphon, and undivided posterior retractor muscles. This complex of small modioliform mytilids is in need of revision. In the meantime, even in the absence of much information about their an- atomy, we are confident that they are not closely related to the new species from the Galapagos hydrothermal vents. Neverthe- less, an affinity for Bathymodiolus might eventually be established in this direction. ACKNOWLEDGMENTS We are most grateful to the participants in the two expeditions to the Galapagos Rift Zone for providing the material for study. This article is contribution number 18 of the Gala- pagos Rift Biology Expedition, supported by the (U.S.) National Science Foundation. We appreciate the advice and information freely provided to us by many scientists involved in the study of the hydrothermal vent fauna, in particular Dr. Jack Corliss, Dr. Fred Grassle, Ms. Linda Morse-Porteous, Dr. Horst Fel- beck, Dr. Robert Hessler, Dr. William New- man, and Dr. Ruth Turner. Dr. Jorgen Knud- sen lent us specimens of Modiolus abyssicola for comparative study. Dr. Ruth Turner, Dr. Kenneth Boss, Dr. Kurt Ockelmann and Dr. William Newman kindly read and criticized early drafts of this manuscript. Our thanks go to Ms. Tina Baird, Mrs. Lyn Anderson and Ms. Mary Alice Tatarian for help in photography and manuscript prepara- tion. We are happy to acknowledge the sup- port and assistance provided by William Minkel and by the Director and staff of the California Academy of Sciences during this study, which was conducted largely at that institution. LITERATURE CITED ATKINS, D., 1937, On the ciliary mechanisms and interrelationships of the lamellibranchs. Part Ill. Types of lamellibranch gills and their food cur- rents. Quarterly Journal of Microscopical Sci- ence, 79: 375—421. BALLARD, R. D. & GRASSLE, J. F., 1979, Return to oases of the deep. National Geographic, 156: 689—705. BOSS, K. J. & TURNER, R. D., 1980, The giant white clam from the Galapagos Rift, Calyp- togena magnifica species novum. Malacologia, 20: 161-194. BOTTJER, D.J. 8 CARTER, J. G., 1980, Functional and phylogenetic significance of projecting perio- stracal structures in the Bivalvia (Mollusca). Journal of Paleontology, 54: 200-216. BURRESON, E. M., 1981, A new deep-sea leech, Bathybdella sawyeri, n. gen., n. sp., from thermal vent areas on the Galapagos Rift. Proceedings of the Biological Society of Washington, 94: 483— 491. CAVANAUGH, C. M., 1983, Symbiotic chemo- trophic bacteria in marine invertebrates from sulphur-rich habitats. Nature, 302: 58—61. CAVANAUGH, C. M., GARDINER, S. L., JONES, M., JANNASCH, H. W. 8 WATERBURY, J., 1981, Prokaryotic cells in the hydrothermal vent tube worm Riftia pachyptila Jones: possible chemoautotrophic symbionts. Science, 213: 340-342. CORLISS, J. В. 8 BALLARD, В. D., 1977, Oases of life in the cold abyss. National Geographic, 152: 441—453. CORLISS, J. В., DYMOND, J., GORDON, L. |., EDMOND, J. M., VON HERZEN, R. P., BAL- LARD, R. D., GREEN, K., WILLIAMS, D., BAIN- BRIDGE, A., CRANE, K. & VAN ANDEL, T. H., 1979, Submarine thermal springs on the Galapa- gos Rift. Science, 203: 1073-1083. DESBRUYERES, D., CRASSOUS, P., GRASSLE, J., KHRIPOUNOFF, A., REYSS, D., RIO, J. & 270 KENK 8 WILSON VAN PRAET, M., 1982, Données ecologiques sur un nouveau site d'hydrothermalisme actif de la ride du Pacifique oriental. Comptes Rendus des Séances de l'Académie des Sciences [Paris], ser. Ш, Sciences de la Vie 295: 489—494. DESBRUYERES, D. 8 LAUBIER, L., 1982, Paralvi- nella grasslei, new genus, new species of Alvinellinae (Polychaeta: Ampharetidae) from the Galápagos Rift geothermal vents. Pro- ceedings of the Biological Society of Washing- ton, 95: 484-494. EDMOND, J. M., 1982, Ocean hot springs: a status report. Oceanus, 25: 22-27. ENRIGHT, J. T., NEWMAN, W. A., HESSLER, R.R. & MCGOWAN, J. A., 1981, Deep-ocean hydrothermal vent communities. Nature, 289: 219-220. FANKBONER, P. V., 1971, The ciliary currents associated with feeding, digestion, and sediment removal in Adula (Botula) falcata Gould, 1851. Biological Bulletin, 140: 28—45. FELBECK, H., 1981, Chemoautotrophic potential of the hydrothermal vent tube worm, Riftia pachyp- tila Jones (Vestimentifera). Science, 213: 336- 338. FELBECK, H., CHILDRESS, J. J. & SOMERO, С. М., 1981, Calvin-Benson cycle and sulphide oxidation in enzymes in animals from sulphide- rich habitats. Nature, 293: 291-293. FRETTER, V., GRAHAM, A. & McLEAN, J. H., 1981, The anatomy of the Galapagos Rift limpet, Neomphalus fretterae. Malacologia, 21: 337- 361. GALÁPAGOS BIOLOGY EXPEDITION PARTICI- PANTS, 1979, Galapagos '79: initial findings in a deep-sea biological quest. Oceanus, 22(2): 2-10. GRAHAM, A., 1949, The molluscan stomach. Transactions of the Royal Society of Edinburgh, 61: 737—776. JANNASCH, H. W. & WIRSEN, C. O., 1979, Chemosynthetic primary production at East Pacific sea floor spreading centers. Bioscience, 29: 592-598. JONES, M., 1981, Riftia pachyptila Jones: observa- tions on the vestimentiferal worm from the Gala- pagos Rift. Science, 213: 333-336. KARL, D. M., WIRSEN, C. O. & JANNASCH, H. W., 1980, Deep-sea primary production at the Gala- pagos hydrothermal vents. Science, 207: 1345— 1347. KNUDSEN, J., 1970, The systematics and biology of abyssal and hadal Bivalvia. Galathea Report 11: 1-241. KRANTZ, G. W., 1981, Copidognathus papillatus, a new species (Acari: Actinedida: Halacaridae) from the Galapagos Rift, Pacific Ocean. Canad- ian Journal of Zoology, 60: 1728-1731. LAUBIER, L. & DESBRUYERES, D., 1984, Les oasis du fond des oceans. La Recherche, 15: 1506-1517. LE PENNEC, M., LUCAS, A. 8 PETIT, H., 1984 [41983”], Etudes préliminaires sur un Mytilidae des sources hydrothermales du Pacifique. Haliotis, 13: 69-82. LONSDALE, P., 1977, Clustering of suspension- feeding macrobenthos near abyssal hydrother- mal vents at oceanic spreading centers. Deep- Sea Research, 24: 857-863. LUTZ, В. A., JABLONSKI, D., RHOADS, D. С. 4 TURNER, R. D., 1980, Larval dispersal of deep- sea hydrothermal vent bivalve from the Galapa- gos Rift. Marine Biology, 51: 127-133. McLEAN, J. H., 1981, The Galapagos Rift limpet Neomphalus: relevance to understanding the evolution of a major Paleozoic-Mesozoic radia- tion. Malacologia, 21: 291-336. MORTON, B., 1974, Some aspects of the biology, population dynamics, and functional morphology of Musculista senhausia Benson (Bivalvia, Mytili- dae). Pacific Science, 28: 19-33. MORTON, B., 1977, The biology and functional morphology of Modiolus metcalfei (Bivalvia: Mytilacea) from the Singapore mangrove. Mala- cologia, 16: 501-517. NELSON, T. C., 1918, On the origin, nature and function of the crystalline style of lamellibranchs. Journal of Morphology, 31: 53-111. NEWMAN, W. A., 1979, A new scalpellid (Cir- ripedia); a Mesozoic relic living near an abyssal hydrothermal spring. Transactions of the San Diego Society of Natural History, 19: 153-167. OCKELMANN, К. W., 1983, Descriptions of mytilid species and definition of the Dacrydiinae n. sub- fam. (Mytilacea-Bivalvia). Ophelia, 22: 81-123. OWEN, G., 1955, Observations on the stomach and digestive diverticulae of the Lamellibranchia |. The Anisomyaria and Eulamellibranchia. Quar- terly Journal of Microscopical Science, 96: 517- 537. PETTIBONE, M. W., 1984, A new scale-worm com- mensal with deep-sea mussels on the Galapa- gos hydrothermal vent (Polychaeta: Polynoidae). Proceedings of the Biological Society of Wash- ington, 97: 226-239. PURCHON, R. D., 1957, The stomach in the Fili- branchia and Pseudolamellibranchia. Proceed- ings of the Zoological Society of London, 129: 27-60. RAU, С. H. & HEDGES, J. 1., 1979, Carbon-13 depletion in a hydrothermal vent mussel: sug- gestion of a chemosynthetic food source. Sci- ence 203: 648-649. REID, R. G. B., 1965, The structure and function of the stomach in bivalved molluscs. Journal of Zoology, London. 147: 156-184. REID, R. G. B. 8 BERNARD, F. R., 1980, Gutless bivalves. Science, 208: 609-610. RHOADS, D. C., LUTZ, В. A., CERRATO, В. М. 4 REVELAS, E. C., 1982, Growth and predation activity at deep-sea hydrothermal vents along the Galapagos Rift. Journal of Marine Research, 40: 503-516. RHOADS, D. С., LUTZ, R. A., REVELAS, Е. С. & CERRATO, R., 1981, Growth of bivalves at deep-sea hydrothemal vents along the Galapa- gos Rift. Science, 214: 911-913. RISE PROJECT GROUP: SPIES, F. N., MacDON- ALD, K. C., ATWATER, T., BALLARD, T., CAR- RANZA, A., CORDOBA, D., COX, C., DIAZ BATHYMODIOLUS: A NEW GALAPAGOS RIFT MUSSEL 271 GARCIA, V. M., FRANCHETEAU, J., GUER- RERO, J., HAWKINS, J., HAYMON, R., HES- SLER, R., JUTEAU, T., KASTNER, M., LAR- SON, R., LUYENDYK, B., MACDOUGALL, J. D., MILLER, S., NORMARK, W., ORCUTT, J. & RANGIN, C., 1980, East Pacific Rise: hot springs and geophysical experiments. Science, 207: 1421-1433. SOOT-RYEN, T., 1955, A report on the family Mytilidae (Pelecypoda). Allan Hancock Pacific Expeditions, 20(1): 1-175. WHITE, K., 1937, Mytilus. Liverpool Marine Biology Committee Memoirs, 31: 1-117, 10 pl. WILLIAMS, A. B., 1980, A new crab family from the vicinity of submarine thermal vents on the Gala- pagos rift (Crustacea: Decapoda: Brachyura). Proceedings of the Biological Society of Wash- ington, 93: 443-472. WILLIAMS, А. В. 4 СНАСЕ, Е. A., 1982, A new caridean shrimp of the family Bresiliidae from thermal vents of the Galapagos Rift Zone, East Pacific. Journal of Crustacean Biology, 2: 136— 147. WILLIAMS, P. M., SMITH, K. L., DRUFFEL, E.M. 8 LINICK, T. W., 1981, Dietary carbon sources of mussels and tubeworms from Galapagos hydro- thermal vents determined from tissue ,,C activ- ity. Nature, 292: 448-449. WILSON, B. R., 1979, A revision of Queensland lithophagine mussels (Bivalvia: Mytilidae: Lithophaginae). Records of the Australian Museum, 32: 435-489. WILSON, B. R. 8 TAIT, R., 1984, Systematics, anatomy and boring mechanisms of the rock- boring mytilid bivalve Botula. Proceedings of the Royal Society of Victoria, 96: 113-125. YONGE, C. M., 1955, Adaptation to rock-boring Botula and Lithophaga (Lamellibranchia, Mytili- dae) with a discussion of the evolution of this habit. Quarterly Journal of Microscopical Sci- ence, 96: 383—410. YONGE, C. M., 1957, Mantie fusion in the Lamelli- branchia. Pubblicazioni della Stazione Zoologica di Napoli, 29: 15-171. NOTE The following appeared while this paper was in press (see also p. 255, footnote 3): FIALA-MEDIONI, A., 1984, Mise en évidence par microscopie électronique a transmission de Габопдапсе de bactéries symbiotiques dans la branchie de Mollusques bivalves de sources hydrothermales profondes. Comptes Rendus de l'Académie des Sciences [Paris], Ill, 298: 487— 492, 2 pl. LE PENNEC, M. & PRIEUR, D., 1984, Observa- tions sur la nutrition d'un Mytilidae d'un site hydrothermal actif de la dorsale du Pacifique. Ibid., 298: 493-498, 1 pl. ED! MALACOLOGIA, 1985, 26(1-2): 273-275 LEGER TO: THE+EDITORS DERIVATIONS OF ARENOPHILIC MANTLE GLANDS IN THE ANOMALODESMATA Club-shaped, multicellular glands lining at least a portion of the mantle edge, have been found in various members of the bivalve sub- class Anomalodesmata. These organs, first described from the verticordiids by Allen & Turner (1974) and named by these authors “radial mantle glands,” have since been found in members of the Lyonsiidae (Prezant, 1979a, 1981a), Periplomatidae (Morton, 1981), Parilimyidae (Morton, 1982), and Cla- vagellidae (Morton, 1984a,b). At least in the Lyonsiidae, these glands secrete an adhesive mucin (bi-layered, weakly acidic mucopoly- saccharide and glycoprotein) atop the perio- stracum that functions in the attachment of extraneous material (often sand) to the out- side of the shell (Prezant, 1979a,b, 1981a,b). Based on this function, Prezant (1981a) termed these organs “arenophilic radial man- tle glands” or “arenophilic glands.” This ex- ternal coating may serve several functions including: protection of relatively thin shelled specimens; dissuasion of boring predators (especialy naticids); camouflage; weighting and increased frictional resistance of rela- tively light, smooth shells in animals with weak byssi (e.g. Lyonsia) for increased stabil- ity in potentially shifting substrata (Prezant, 1979a,b, 1981a,b,c). In Morton’s many studies on the Anoma- lodesmata, he has described a variety of members with “radial mantle glands” as occurring in the middle mantle fold (1981, 1982, 1984a,b). Morton (1984a), for instance, states that for Clavagella australis (Sowerby), “... the tips of the siphonal crown possesses specialized subepithelial, basiphilic, glands, termed by Prezant (1979) for the Lyonsiidae, “radial mantle glands”... Such glands are developed in the middle folds, not the outer folds as suggested by Prezant....” | found that the glands in members of the lyonsiid Entodesma open distal to the periostracum-secreting cells (i.e. ventral to the periostracal groove on the inner surface of the outer mantle fold) (Prezant, 1981a). The glands in members of the genus Lyonsia are found proximal to the periostracum-secreting cells and thus open directly into the periostra- cal groove. In both genera a secretory sheath encircles the gland. In Lyonsia this sheath is derived from the outer epithelium of the mid- dle mantle fold, and in Entodesma from the inner epithelium of the outer fold. In both genera, however, the central gland (compos- ing the bulk of the organ) is derived from the outer mantle fold epithelium. Adhesive mucins from the arenophilic glands of both Lyonsia and Entodesma serve similar functions and are both released upon the periostracum. In the latter the glands are more abundant in thinner shelled juveniles where they may still be active in their ascribed functions (adult members of Entodesma may be thick-shelled crevice dwellers with strong byssi) (for a discussion of this see Prezant, 1981a). In specimens of Entodesma the secretion is deposited upon the periostracum presumably by the action of some proteolytic enzyme (probably incorporated within the gly- coprotein layer) that allows the adhesive to periodically penetrate through the overlying periostracum (secretion deposited as radial tufts in Entodesma as opposed to continuous radial lines in Lyonsia). The third genus within the Lyonsiidae, Mytilimeria, lacks arenophilic glands but has a heavy concentration of mar- ginal mucocytes dispersed along the mantle edge. M. nuttalli Conrad is typically an en- dosymbiont of compound ascidians and has consequently lost arenophilic glands with attainment of this protective and secure habitat. The arenophilic glands of lyonsiids are de- rived as invaginations of the outer mantle fold epithelium (with a surrounding middle fold sheath). If, as suggested by Morton (1981, 1982, 1984a,b), the glands proper were of middle fold origin, it would necessitate an ontogenetic migration of these organs through the periostracal groove in Entodes- ma. This is almost certainly not the case. Based on the limited evidence available, a derivation of these glands from the middle mantle fold in anomalodesmatans outside the Lyonsiidae might be hypothesized. The evi- dence for this, however, is not strong as the only detailed work thus far in print on these organs is that of Prezant (1979a, 1981a) for the lyonsiids. Also, as Morton states (1981), it is very likely that these are “specialized glands evolved in some comon ancestor of (273) 274 LETTER TO THE EDITORS the Anomalodesmata and retained in a di- verse series of descendants.” It is indeed plausible that these organs can serve as evoutionary markers but we must first expand our knowledge of them and also clear the literature of discrepancies. In a report on tube construction in Brechites, Morton (1984a) reports that “Sim- ilar glands occur in members of the Verti- cordiidae, Lyonsiidae, Pholadomyidae, Peri- plomatidae, Poromyidae and Parilimyidae (Allen & Turner, 1974; Prezant, 1979[a]; Mor- ton, 1980, 1981a,b, 1982a)....” There is, however, no indication in any of these cited references that either poromyids or pholado- myids possess arenophilic mantle glands. In fact Morton (1982; table 1) lists both the Pho- ladomyidae and Poromyacea as lacking ra- dial mantle glands. In his 1980 publication on Pholadomya candida Sowerby, Morton notes that “Radial mantle glands . . . do not occur in P. candida.” Among the Anomalodesmata additionally reported not to possess arenophilic mantle glands are the Pandoridae (Prezant, 1981a), Thraciidae (Prezant, 1981a), Laternulidae (Prezant, 1981a), Cleidothaeridae (Morton, 1984) and Cuspidariidae (Morton, 1982). The Myochamidae have yet to be adequately ex- amined for these organs. While arenophilic glands have been described by Morton (1981) from the periplomatid Periploma (Offadesma) angasi Crosse & Fischer, no such glands have been found in serially sec- tioned mantle of adult P. fragile (Totten) nor Cochlodesma praetenue (Pulteney) (Prezant, 1981 a). Arenophilic mantle glands have been found by Morton (1982) in Parilimya fragilis (Grieg) and apparently also function in adhesion of sand grains to the periostracum. Here, Mor- ton states that “Clearly the glands have an uneven distribution in Anomalodesmata. This may be a case, however, so often encoun- tered in the Anomalodesmata, of a primitive feature, possibly evolved during the Palaeo- zoic period of radiation in shallow water de- posits, subsequently retained in a wide but dis- jointed assemblage of descendants.” | concur that the glands are of an “uneven” distribution among members of the subclass. | also sug- gest that they may be of “uneven” structure and derivation in recent Anomalodesmata. It is unlikely that such specialized glands arose separately in any major group of anomalo- desmatans. The Middle Ordovician stock of this subclass likely possessed an “anlage” of arenophilic glands in the form of widely dis- persed and nonconsolidated mucocytes. The secretions from these simple glands perhaps helped support the walls of the bivalves’ bur- rows. The glands diversified early in their evolution into the structured organs we see today. Extant members of the subclass that possess arenophilic glands may reflect a bifurcation from this early stock with the struc- ture developing, phylogenetically, either in- side or outside the periostracal groove. The location of the glands in specimens of Entodesma precludes the possibility that these organs are derived from the middle mantle fold in the Lyonsiidae. Fig. 17 in Pre- zant’s (1981a) article on arenophilic glands clearly shows such a gland opening inside (Le. ventral to) the periostracal groove. Fig. 8 of the same paper shows the secretion from an arenophilic gland penetrating the perio- stracum. The lack of adequate longitudinal sections in other reports demonstrating the occurrence of similar glands prohibits the ex- act determination of specific locations and mantle origins. Cross-sectional representa- tions, as exhibited in most reports (/.e. Mor- ton, 1981, 1982, 1984a,b), strongly suggest the similar origin within the Anomalodesmata of arenophilic glands but do little to support the contention that the glands are middle mantle fold derivatives. Allen & Turner (1974) found that the radial mantle glands of the verticordiid Policordia densicostata Locard open “on to the edge of the sensory lobe of the mantle.” The latter authors also report “radial mantle glands within the sensory lobe” of Verticordia triangularis Locard. Morton (1981), however, found the “description by Allen & Turner (1974) of similar glands in the mantle of members of the Verticor- diidae . . . insufficiently detailed to facilitate comparison...” with the arenophilic glands of Lyonsia (as described by Prezant, 1979a). In addition, Morton (1981) found “No trace of retractor muscles for withdrawing the gland...” in Offadesma angasi. . . “as occur in Lyonsia (Prezant, 1979)” and suggests, that their inability to protrude may be “. . . be- cause of their length... .” Though no total length measurements are offered in the latter paper (í.e. Morton, 1981), diagrammatic representation show glands in O. angasi that are likely equal to or shorter than protrusible glands along the siphonal mantle edge of some lyonsiids. While the glands of O. angasi may not be protrusible, it is not a reflection of length. This difference as well as other differ- LETTER ТО THE EDITORS 275 ences already mentioned may offer valuable clues to the taxonomy and phylogeny of the Anomalodesmata. Because of our limited knowledge of arenophilic glands, it is presently difficult to account for intertaxon differences in mantle gland derivations in what is most likely a well established mantle feature. The loss of these glands in the lyonsiid Mytilimeria, the location of the glands deep in the periostracal groove of Lyonsia, and the establishment of the glands distal to the groove in Entodesma with the apparent concurrent development of some proteolytic enzyme involved in perio- stracal penetration, all point to the evolution- ary plasticity of this organ in one family alone. The plasticity within the subclass is certainly greater. Since arenophilic glands are present in several superfamilies of anomalodesmatans (i.e. Pandoracea, Verticordiacea, Clava- gellacea, Thraciacea), it is likely that the glands, sensu stricto, evolved early, perhaps even in primitive pholadomyacean stock of the Paleozoic. Present day pholadomyids may lack these organs but this does not pre- clude the possibility of gland loss from earlier stock. The presence of similar mantle organs in numerous anomalodesmatan families argues strongly for plesiomorphy. Since man- tle glands of members of the lyonsiid genus Entodesma are scarce in large adult speci- mens, it is still possible that similar organs will be found in other anomalodesmatan families once complete serial sections are taken of mantle edges from complete growth series. Unfortunately, Morton (1980) had only a sing- le specimen of Pholadomya candida available for histological analyses and was thus unable to serial section the entire mantle edge. A large portion of the mantle edge of this primi- tive bivalve was left unexamined. Whether or not arenophilic glands are present in P. candi- da (and it is unlikely that they are present) does not discount the possibility of similar glands in stock pholadomyaceans or at least a densely packed series of mucocytes lining the mantle edge. This might well have been the incipient evolutionary stage of present day arenophilic glands. | suggest we acknowledge the possibility that the possession of arenophilic glands per se is symplesiomorphic, but presently vari- able in form and ontogenetic origin. These subtle differences may offer strong clues to the phylogenetic progression of the Anomalo- desmata. The solutions to the presumed dis- crepancies await ontogenetic evidence of the ultimate origin of arenophilic radial mantle glands in each group in which they occur. REFERENCES CITED ALLEN, J. A. & TURNER, J. F., 1974, On the functional morphology of the family Verti- cordiidae (Bivalvia) with descriptions of new species from the abyssal Atlantic. Philosophical Transactions of the Royal Society of London, ser. B, 268: 401-536. MORTON, B. S., 1974, Some aspects of the biol- ogy and functional morphology of Cleidothaerus maorianus Finlay (Bivalvia: Anomalodesmata: Pandoracea). Proceedings of the Malacological Society of London, 41: 201-222. MORTON, B. S., 1980, The anatomy of the “living fossil” Pholadomya candida Sowerby 1823 (Bivalvia: Anomalodesmata: Pholadomyacea). Videnskabelige Meddelelser fra Dansk naturhis- torisk Forening, 142: 7-102. MORTON, B. S., 1981, The biology and functional morphology of Periploma (Offadesma) angasai [sic]. (Bivalvia: Anomalodesmata: Periplomati- dae). Journal of Zoology, 193: 39-70. MORTON, B. S., 1982, The functional morphology of Parilimya fragilis (Bivalvia: Parilimyidae nov. fam.) with a discussion on the origin and evolu- tion of the carnivorous septibranchs and a reclassification of the Anomalodesmata. Trans- actions of the Zoological Society of London, 36: 153-216. MORTON, B. S., 1984a, The biology and functional morphology of Clavagella australis (Bivalvia: Anomalodesmata). Journal of Zoology, 202: 489-511. MORTON, B. S., 1984b, Adventitious tube con- struction in Brechites vaginiferus (Bivalvia: An- omalodesmata: Clavagellacea) with an in- vestigation of the juvenile of “Humphreyia strangei.” Journal of Zoology, 204: 461-484. PREZANT, R. S., 1979a, The structure and func- tion of the radial mantle glands of Lysonia hyali- na (Bivalvia: Anomalodesmata). Journal of Zool- ogy, 187: 505-516. PREZANT, R. S., 1979b, Shell spinules of the bivalve Lyonsia hyalina. Nautilus, 93: 93-95. PREZANT, R. S., 1981a, The arenophilic radial mantle glands of the Lyonsiidae (Bivalvia: Anom- alodesmata) with notes on lyonsiid evolution. Malacologia, 20: 267-289. PREZANT, R. S., 1981b, Comparative shell ultra- structure of lyonsiid bivalves. Veliger, 23: 289- 299. PREZANT, R. S., 1981c, Taxonomic re-evaluation of the bivalve family Lyonsiidae. Nautilus, 95: 58-72. Robert S. Prezant Department of Biological Sciences University of Southern Mississippi Hattiesburg, Mississippi 39406, U.S.A. р 1 À ‘ ’ , ve Fi 2 a “ À | В À 5 Te | i | AS 5 1 Г a i ” EN \ y Lt E eer L | | | ein Bu р | i р 7 11 i" (hays | 4 Ron RU cr i | и | i | LT y | Uk > 1% na | = Mi" à {> | i aa YA ed | р N Я Fo x р | 5 r р Ира y | sd т al tou | i 2 st 4 } LA A 4! т y y | | | lay at a | | > N ae dt MN Ñ 1 «2 Г q ar e] y р 4 [2 de ry an у FU y м à = É Кон ey us ’ | т и | à A iy i PL Len LAS | hos A | | y? 4 6 ohn в * al er И и Нар are a 2 x u = AA € 5 i E MEA 1 | à i ne yey {8 | Pr Yu if A 0 т f ' Ke ol het i N N Ss,‘ vail DN сми ” ин a ar р LS on Y р N во à $ ба м = Ele ae й ha y en ¿En in 77 FOIS 14} i № я \ ( В | Y A e do i iy EN anys is Pale? Te q Sag) EN oe pare iu ió ВТ | alee Aig teen Oe TNT ioe 1" gine: a si oye mi LL A INDEX TO TAXA IN VOLUME 26 An asterisk (*) denotes a new taxon abyssicola, Modiolus, 268, 269 Acari, 264 *acarinatus, Nymphophilus, 38, 43, *45, 46, 118 Achatina, 173, 180, 183 Achatinellidae, 3, 5, 9, 30 Adipicola, 269 Adula, 267 alba, Balcis, 162 albicans, Biomphalaria, 137, 141, 142 albolabris, Triodopsis, 226, 236 Allodiscus, 1-30 Amblema, 245 Amplirhagada, 236, 237 angasi, Periploma (Offadesma), 274 Anomalodesmata, 273-275 Archaster, 161 argenteus, Adipicola, 269 ariel, Phrixgnathus, 10, 12, 13, 18, 22 Arion, 236 Ascidia, 273 aspersa, Helix, 5, 10, 22, 24, 220 Asterophila, 153, 160 Asterophilidae, 153 ater, Arion, 236 Athoracophoridae, 2 australis, Clavagella, 273 Australorbis, 210 Bacteria, 267, 268 Balcis, 153, 155, 161, 162 banksii, Freycinetia, 19 Bathograea, 264 *Bathymodiolinae, 253-*255-271 *Bathymodiolus, 253—*255-271 bella, Eubora, 113 bidens, Brephulopsis, 213-223 bidens, Buliminus, 213-223 bidens, Chondrus, 213-223 Biomphalaria, 137-143, 145-151, 173, 180, 201— Za Bivalvia, 241-251, 253-271, 273-275 Botula, 26, 264, 265 Branchipolynoe, 264 Brechites, 274 Brephulopsis, 213-223 buccinella, Cavellia, 9, 18, 23 Buliminidae, 213 Buliminus, 213-223 burnerensis, Amplirhagada, 236, 237 busbyi, Paryphanta, 2 Calyptogena, 267 Camaenidae, 3, 237 camerunensis, Biomphalaria, 139, 141 campestris, Succinea, 184, 185 candida, Pholadomya, 274, 275 caputspinulae, Paralaoma, 10, 13, 18 caramba, Paludiscala, 37, 47, 54, 58-63, 120 Carinaria, 127 carinella, Liarea, 13, 18 carolinianus, Philomycus, 174, 176 carranzae, Mexipyrgus, 87, 101, 104 caruanae, Deroceras, 174, 176 Cassis, 170 Cavellia, 9, 12, 18, 22, 23 celinde, Therasiella, 9, 14, 15 cellarius, Oxychilus, 10 Cepaea, 213, 219, 221 cernica, Robillardia, 160 Chara, 33, 34, 40, 47, 68, 74, 81 Charopa, 2, 6, 9, 11, 12, 14, 15, 17, 19, 22, 23 Charopidae, 1-30 chiron, Flammulina, 9, 15, 18 Chondrus, 213-223 chrysaugeia, Charopa, 6, 9, 17, 23 churinceanus, Mexipyrgus, 31, 37, 87-104, 119 Cichlasoma, 71 Cionella, 5, 10, 22 Cionellidae, 3 claibornensis, Triodopsis, 237 Clausiliidae, 12 Clavagella, 273 Clavagellacea, 275 Clavagellidae, 273 Cleidothaeridae, 274 coahuilae, Durangonella, 37, 47, 60, 78-87, 119 Coahuilix, 31-123 Cochliopina, 31-123 Cochliopinae, 38 Cochlodesma, 274 coma, Charopa, 6, 9, 12, 14, 17 conella, “Phrixgnathus,” 7, 10, 13, 17, 22 Cookeconcha, 14 cookiana, Geminoropa, 9, 17, 25 coresia, Delos, 9 cornucopiae, Orygoceras, 98 coronata, Pterotrachea, 125-135 coronatus, Pyrgophorus, 97 costulata, Charopa, 6, 9, 11, 15, 19 Coxia, 12 Crenullinae, 268 cristata, Carinaria, 127 Crustacea, 31 crystallina, Thyca, 160 Cuspidariidae, 274 cyanoguttatum, Cichlasoma, 71 Cytora, 4, 5, 9, 13-15, 18, 22, 24, 25 cytora, Cytora, 5, 9, 14, 15 decidua, Therasia, 9, 13, 15, 24 decussatulus, Cookeconcha, 14 Delos, 5, 9, 18, 23 densicostata, Policordia, 274 Deroceras, 174-177, 183-186, 188-190 devians, Balcis, 162 dimidiata, Otoconcha, 2 dimorphus, Allodiscus, 5, 6, 9, 12, 14, 17, 24 Durangonella, 31-123 dussumieri, Mariaella, 180 Echineulima, 162 Echinoidea, 162 (277) 278 INDEX Echinometra, 160 edulis, Mytilus, 263 egea, Liarea, 9, 14 egesta, Egestula, 2 Egestula, 2 elaiodes, Phrixgnathus, 10, 18, 19, 23 elodes, Lymnaea, 191-200 Endodonta, 28 Endodontidae, 3, 12, 15, 17 Enidae, 213-223 Enteroxenos, 153 Entoconchidae, 153 Entodesma, 273-275 erigone, Phrixgnathus, 7, 10, 12, 13, 18, 22 escobedae, Mexipyrgus, 87, 101 eta, Mocella, 9, 23, 29, 30 Eubora, 112, 113 Euglandina, 173-181, 183-190 Eulima, 162 Eulimacea, 153-163 Eulimidae, 153 Eulimoidea, 153-163 faba, Pseudoretusa, 162 faba, Pulicicochlea, 162 fasciola, Lampsilis, 243 Fectola, 6, 9, 15, 17, 19, 28 feredayi, Flammulina, 9, 15 Flammulina, 6, 9, 11, 15, 18, 19 floridana, Veronicella, 174 Fontelicella, 111 fragile, Periploma, 276 fragilis, Parilimya, 274 francesci, Paralaoma, 10, 12-15, 22 fratercula, Libera, 12 Freycinetia, 19 fulica, Achatina, 173, 183 fulvescens, Mucronalia, 161, 162 fuscosa, Charopa, 6, 9, 11, 17, 22 Gambiodonta, 28 Gastropoda, 31-143, 153-163, 165, 170, 173-189, 201-211, 225-239 Geminoropa, 9, 17, 25 georgianus, Viviparus, 198 giveni, Phenacohelix, 9 glabrata, Biomphalaria, 137-143, 145-151, 173, 180, 201-211 glabratus, Australorbis, 210 glabriusculus, Phrixgnathus, 10, 13, 18 granum, Allodiscus, 6, 9, 10, 18 greenwoodi, Rhytida, 5, 9, 13, 18, 24 Hauffenia, 58 havanensis, Biomphalaria, 137-143 hectori, Huonodon, 6, 9, 17, 18 hedleyi, Cytora, 5, 9, 14, 25 Helicarionidae, 12 Helicella, 221 Helicidae, 3 Helisoma, 189 Helix, 5, 10, 22, 24, 220 helophila, Biomphalaria, 141, 142 Hendersoniella, 12 Heteropoda, 125-135 hippocampus, Pterotrachea, 125-135 hochstetteri, Liarea, 9, 13, 18 hopetonensis, Triodopsis, 237 Horatia, 112 hubbsi, Coahuilix, 37, 53-59, 112 Huonodon, 6, 9, 17, 18 Hyalella, 49 Hydrobia, 78, 105, 108-110 Hydrobiidae, 31, 123 Hydrobiinae, 108 Hydrocenidae, 3, 5, 9 ide, Suteria, 9, 14, 15, 17, 18, 22 llyanassa, 165-172 infecta, Fectola, 9, 17 irrorata, Littorina, 171 Japonica, Asterophila, 160 japonica, Carinaria, 127 japonica, Peronella, 162 japonica, Terua, 269 jeffreysiana, Delos, 9 jervisense, Meridolum, 237 jungermanniae, Pasmaditta, 10, 18 kivi, Serpho, 9, 10, 12, 13, 15, 18, 24, 25 laeve, Deroceras, 174-176, 183-186, 189 laevigata, Linckia, 160 Lamellidae, 5, 9, 18, 19, 22, 24 Lampsilis, 241-251 *landyei, Coahuilix, 53, 54, *57-59, 119 Laoma, 4, 5, 7, 10, 12, 13, 18, 22, 29, 30 Laternulidae, 274 lateumbilicata, Paralaoma, 8, 10, 17, 18 leimonias, Laoma, 5, 7, 10, 12, 13, 18, 22, 27 Leiosolenus, 264, 267 levis, Phrixgnathus, 7, 10, 13, 18, 22 Liarea, 5, 9, 13, 14, 18, 22, 24, 25 Liareidae, 3, 5, 9, 24 Libera, 12, 28 Lima, 241 lima, Polydontes, 237 Limax, 173, 180, 188, 189 Linckia, 160, 161 Lithoglyphinae, 108 Lithophaga, 264 Lithophaginae, 268 Littoridininae, 31, 38, 53, 54, 59, 106, 108, 110, 113 Littoridinops, 11, 112 Littorina, 171 lubrica, Cionella, 10, 22 lugoi, Mexipyrgus, 87, 101, 103, 104 Lymnaea, 173, 189, 191-200, 210, 250 Lyonsia, 273-275 Lyonsiidae, 273-275 macgregori, Coxia, 12 maculata, Mocella, 6, 9, 23 magnifica, Calyptogena, 267 manantiali, Mexistiobia, 37, 46-*47-51 INDEX 279 manawatawhila, Mocella, 9, 13 mansoni, Schistosoma, 137, 145, 201 mariae, Laoma, 7, 10, 13, 18, 22 Mariaella, 180 marina, Laoma, 7,10, 13, 18, 22, 29, 30 Marstonia, 39, 40, 47 marsupialis, Fectola, 28 maximus, Limax, 188, 189 Melanellidae, 153 Meridolum, 237 Mesodon, 236 Mexipyrgus, 31-123 Mexistenasellus, 111 *Mexistiobia, 36-38, *46-51, 81, 105, 111, 118 Mexithauma, 31-123 Mexithaumatinae, 38 Mexiweckelia, 111 micra, Horatia, 112 milleri, Cochliopina, 37, 40, 64-72, 119 minckleyi, Nymphophilus, 37-45, 68, 74, 118 minuta, Pterotrachea, 125-135 mira, Fectola, 6, 9, 17, 19 mittrei, Mucronalia, 162 Mocella, 6, 9, 13, 23, 29, 30 Modiolinae, 268 Modiolus, 253, 263, 264, 267, 268 modiolus, Modiolus, 263 moellendorfi, Phrixgnathus, 7, 10, 13, 18, 22 mojarralis, Mexipyrgus, 87, 98, 101 Mollusca, 137-143, 213 Mucronalia, 153, 161, 162 multilineatus, Mexipyrgus, 87, 98, 101 Musculininae, 268 Musculista, 267 Myochamidae, 274 Mytilidae, 253-271 Mytilimeria, 273, 275 Mytilinae, 268 Mytilus, 262-264, 267 nana, Stiobia, 49 Nassariidae, 170 Nassarius, 170 Naticidae, 273 nemoralis, Cepaea, 213, 219, 221 neozelanica, Therasiella, 6, 9,14, 15, 22, 23, 25 novoseelandica, Lamellidea, 5, 9, 18, 19, 22, 24, 29, 30 nuttalli, Mytilimeria, 273 Nymphaea, 33, 40, 74 Nymphophilinae, 31, 37, 38, 108 Nymphophilus, 37—45, 47, 68, 74, 105, 110-112, 118 Obanella, 10, 13, 18, 27 obsoleta, llyanassa, 165-172 obstructa, Biomphalaria, 141, 142 ochra, Charopa, 6, 9 Offadesma, 274 Omphalorissa, 9, 18, 22, 24 Onchidium, 173 Orygoceras, 38, 53, 54, 98, 99, 112, 120 Otala, 173 Otoconcha, 2, 4 Oxychilus, 3, 5, 10, 11 pachyptila, Riftia, 267 pacifica, Calyptogena, 267 pacifica, Terua, 269 Paedophoropodidae, 153 pallida, Biomphalaria, 137, 141 pallida, Cytora, 5, 9, 13, 14, 18 palmeri, Hendersoniella, 12 Paludiscala, 37, 38, 47, 54, 58—63, 105-109, 111- 113, 120 Paludiscalinae, 38 palustris, Lymnaea, 191-200 Pandoracea, 275 Pandoridae, 274 Paralaoma, 8, 10, 12-15, 17-19, 22 Parilimya, 274 Parilimyidae, 273, 274 Paryphanta, 2, 189 Pasmaditta, 10, 18 Pecten, 241, 250 pelagica, Adipicola, 269 Pelseneeria, 162 Pelseneeriidae, 153 perdita, Flammulina, 6, 9, 15, 19 peregrina, Biomphalaria, 141, 142 Periploma, 274 Periplomatidae, 273, 274 Peronella, 162 peronellicola, Balcis, 162 pfeifferi, Biomphalaria, 210 Phenacohelix, 9, 10, 12, 15, 22, 29, 30 Philomycus, 174, 176, 177 Pholadomya, 274, 275 Pholadomyacea, 275 Pholadomyidae, 274, 275 Phrixgnathus, 2, 4, 7, 10, 12, 13, 17-19, 22, 23 Physa, 180 pilsbryi, Charopa, 6, 9, 11, 15 pilula, Phenacohelix, 9 perongiaensis, Phrixgnathus, 7, 10, 13, 17, 29, 30 Planorbidae, 137-143 planulatus, Allodiscus, 6, 9, 14 poecilosticta, Phrixgnathus, 7, 10, 12, 13, 17, 22 Policordia, 274 Polychaeta, 261, 265 Polydontes, 237 Polygyra, 183, 184-189 polygyrata, Polygyratia, 12 Polygyratia, 12 Polygyridae, 225-239 pomatia, Helix, 220 Pomatiasidae, 12 Pomatiopsidae, 37, 110 Pomatiopsis, 110 ponsonbyi, Phenacohelix, 9, 29, 30 Poromyidae, 274 Potamopyrgus, 110, 112 praetenue, Cochlodesma, 274 projectus, Modiolus, 268 Prosobranchia, 22, 165 pseudanguicula, Charopa, 6, 9, 12, 17, 19, 23 280 pseudoflavus, Limax, 173 pseudoleiodon, Huonodon, 9, 17 Pseudolibera, 28 Pseudoretusa, 162 Pterotrachea, 125-135 ptilocrinicola, Eulima, 162 Pulicicochlea, 162 Pulmonata, 173, 188, 189, 201-223, 235-239 Punctidae, 1-30 purchasi, Omphalorissa, 5, 9, 18, 22, 24 Pyrgophorus, 97, 111 quadripaludium, Mexithauma, 37, 40, 71-77 Rhytida, 5, 9, 13, 18, 23, 24 Rhytididae, 3, 5, 9 Riftia, 267, 268 rimutaka, Obanella, 10, 13, 18, 27 riograndensis, Cochliopina, 64, 65, 68, 71,111, 119 Rissoacea, 31, 123 Robillardia, 160 rosea, Euglandina, 173-181, 183-189 roseveari, Cavellia, 9, 12, 22 Rumina, 12 Schistosoma, 137, 145, 201 Schizoglossa, 2, 18 schrammi, Biomphalaria, 137-143 scutata, Pterotrachea, 125-135 seemani, Durangonella, 78, 83, 86 seetzeni, Trochoidea, 221 semimarsupialis, Taipidon, 28 Semisulcospira, 114 septemvolva, Polygyra, 183-186 Serpno 9 10, 112.113: 15, 18, 24.125 serrata, Therasiella, 9, 14, 15, 22 serratocostata, Paralaoma, 10, 13-15 shaplandi, Balcis, 161, 162 simpsoni, Adipicola, 269 Sphincterochila, 220 Spisula, 250 Spurwinkia, 72, 73, 108-110, 112, 113 stagnalis, Lymnaea, 210 Stiliferidae, 153 Stiobia, 46, 49 straminea, Biomphalaria, 137-143 Succinea, 183-186, 188, 189, 190 Suteria, 9, 14, 15, 17, 18, 22 symmytilida, Branchipolynoe, 264 Taipidon, 28 tamora, Therasiella, 9, 14, 15 temnopleuricola, Vitreobalcis, 153-163 INDEX Temnopleurus, 153-155 Terua, 269 tessullatus, Allodiscus, 9, 14, 18 Testacella, 189 Texadina, 111 thalassohelix, 9, 14, 15, 17, 22, 29, 30 Therasia, 9, 13, 15, 24 Therasiella, 6, 9, 13-15, 22, 23 thermophilis [sic], Bathymodiolus, 255 *thermophilus, Bathymodiolus, 253—255—271 thermydron, Bathograea, 264 Thraciacea, 275 Thraciidae, 274 Thyca, 153, 160-162 Thycidae, 153 thyroidus, Mesodon, 236 toreumaticus, Temnopleurus, 153-155 torquilla, Cytora, 5, 9, 14, 22, 25 triangularis, Verticordia, 274 Triculinae, 37 tridentata, Triodopsis, 225-239 Triodopsinae, 225-239 Triodopsis, 225-239 Trochoidea, 221 Tropidebora, 112 truncatula, Lymnaea, 210 typicus, Archaster, 161 unidentata, Fectola, 9 Unionacea, 241-251 Unionidae, 245, 250 Urocoptidae, 12, 25 urquharti, Allodiscus, 6, 9, 12, 22 Utricularia, 33, 40, 68 Vallonia, 5, 10, 12, 19 Valloniidae, 3, 10 ventricosa, Lampsilis, 241—251 Veronicella, 174, 176, 177 Verticordia, 274 Verticordiacea, 275 Verticordiidae, 273, 274 Vestimentifera, 267 vibex, Nassarius, 170 virgata, Helicella, 221 Vitreobalcis, 153-163 Viviparidae, 114 Viviparus, 198 worthyae, Schizoglossa, 2, 18 ziczac, Thalassohelix, 9, 14, 15, 17, 22, 29, 30 zonata, Sphincterochila, 220 AWARDS FOR STUDY AT The Academy of Natural Sciences of Philadelphia The Academy of Natural Sciences of Philadelphia, through its Jessup and McHenry funds, makes available each year a limited number of awards to support students pursuing natural history studies at the Academy. These awards are primarly intended to assist predoctoral and immediate postdoctoral students. Awards usually include a stipend to help defray living expenses, and support for travel to and from the Academy. Application deadlines are 1 March and 1 October each year. Further information may be obtained by writing to: Chairman, Jessup-McHenry Award Committee, Academy of Natural Sciences of Philadelphia, 19th and the Parkway, Philadelphia, Pennsylvania 19103, U.S.A. 26(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. Papers of local geographical interest should be submitted elsewhere, as should papers whose primary thrust is physi- ology 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, Be- thesda, 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. 8. Provide a concise and informative Ab- MALACOLOGIA 1985 stract summarizing not only contents but re- sults. A separate summary generally is superfluous. 9. Supply between five and eight key (topic) words to go at the end of the Ab- stract. 10. Use the metric system throughout. Mi- cron should be abbreviated um. 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 caption 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 2 or ‘3. 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 se- quentially as figures (not grouped as plates or as lettered subseries), and are to be arranged 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 conven- ient lengths (e.g. “200 jm,” not “163 um”). Magnifications in captions are not acceptable. 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- specifications 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. REPRINTS AND PAGE COSTS 22. 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. 33.00 or more. 23. When an article is 10 or more printed pages long, MALACOLOGIA requests that an author pay part of the publication costs. SUBSCRIPTION COSTS 24. For Vol. 26, personal subscriptions are U.S. $17.00 and institutional subscriptions are U.S. $27.00. For information on Vol. 27, address inquiries to the Subscription Office. VOL. 26, NO. 1-2 MALACOLOGIA CONTENTS . A. SOLEM & F. M. CLIMO Structure and habitat correlation of sympatric New Zealand land snail SPECIES ЛЬ AN ha TR ное R. HERSHLER Systematic revision of the Hydrobiidae (Gastropoda: Rissoacea) of the Cuatro Ciénegas Basin, Coahuila, México .......................,... HR SEAPY : The pelagic genus Pterotrachea (Gastropoda: Heteropoda) from Hawaiian Waters: a taxOnOmic review аа. a a IT ie A PEER . E. JELNES & J.-P. POINTIER Taxonomie expérimentale de Biomphalaria (Gastropoda: Planorbidae)—III. Mobilités enzymatiques considérées comme éléments de diagnostic pour les Biomphalaria Antillais. Etude de sept systemes enzymatiques....... C. S. RICHARDS A new pigmentation mutant in Biomphalaria glabrata ............... aie Y. FUJIOKA Population ecological aspects of the eulimid gastropod Vitreobalcis TENIMODIGUIICOIE, EE RER er) Coat bie Hele the daa dial al AN . У. DIMOCK, Jr. ИЕ Quantitative aspects of Е by the mud Se Ilyanassa obsoleta . A. COOK aig ha № ee é Functional aspects of trail following byähen carnivaraus snail Euglandina = D A. COOK The organisation of feeding in the Sl Euglena rosea. К. М. BROWN, D. В. DEVRIES 8 В. К. LEATHERS os : wort Causes of life history variation in the freshwater: Sait ‘Lymnaea elodes. . L. A. ABOUL-MAGD & S. A. SABRY | Scanning electron microscopy of the body surfaces of Biomphalaria Po rel EEN LE ANS A II E A ВВ . М. LIVSHITS Ecology of the terrestrial snail Brephulopsis bidens (Pulmonata: Enidae): mortality, burrowing and migratory activity .................... K. C. EMBERTON Seasonal changes in the reproductive gross anatomy of the land snail Triodopsis tridentata tridentata (Pulmonata: Polygyridae) ............... . В. KRAEMER & С. М. SWANSON Functional morphology of “eyespots” of mantle flaps of Lampsilis (Bivalvia: — ———_ Unionacea): evidence for their role as effectors, and basis for dl > ued regarding pigment distribution in bivalve mantle tissue ................. \. С. KENK & В. В. WILSON А new mussel (Bivalvia, Mytilidae) from hydrothermal vents in the Galapa- o IRE AE A A ESS LT AA Е LETTER TO THE EDITORS R. S. Prezant. Derivations of arenophilic mantle glands in the ¡AÑOMAIOIESMALA a a e a Ao SE abs р Er De INDEX TO VOL: 28; NOV 172 2 0. A vc UU AAC RARES Loh 7. 0 M Hie o г- LEN MAS Le AR NA a RUN AE tr IRD 4 y i | yh | F A МУ ba en yn AJ А RU 4 + { ' ' t , 0 2‘ Bar eth ad у | ууу" 4 fi , ‘ ER И Ur mus у ‘ } vit DATE Pe Гл Tait РР # fe Mal set d'en и, I | CRD Ph à Kan en к J м СК 7 A; } ASAE ght 28 9 200 HOME i oyu fade ччин bane ! lu » CHA Mah ‘ de ‘ t ' hos ‘ ROUE PAT ENT MEN? ‘ ee HAL aye obs ' Picea ryt 0 y р у avery a) mie nh | и AA M 2 ci НИ уе PAS ye) tte Meier, hu : ны pe i rhs y A ; Г : tale pis и аа ель ‘ } ia une d'à Mist garer + ‚ Int et A " APTE 0 e h " й rar ГУ x eae on Ен РНИИ A I ney N Е Kar 7 RAT EE ' ad thet adie add hier 7 iu ! AO AN PA РКУ ; CM : Н r р ro ay dde ны , » v lun г , N OM r ary piss Bn arg an ‘ ка Heer ук ren TE AN r ph PAPER Е bes y pd pret oh Meet we TEEN LC р tite i pu . me pa ts 4 , "| ae ие р % 1 PENN FA Aa ; Air Fi En ma fi . N р uta) a EES ' ri is i 4 Laure : Br Ba ея : hype MAT TN . , LR : or у A KR} ар i reine PURE AlN ot Het ‘ see wh : MEANS . ; DL nam mes. : .o d Ne EL уч я НИ а * 2 Hr Г DE TEE DR Ane ее os ‘ . ‘ ры M ame thn o A > f en My e 1 « NE, О A, , . RUN , ts 0 ‘ re ne) г. fete uo toe , ñ PAL ov shee te hare ES avn tay, и N ER ee ; nee on , + es of peepee ence dy q £ ot АН быв y rhume re rte it à р | 3 A 96 8 EN ASA } ; 4 ts - ye de А A OPERA roe Sf bade a 4 ATA IA e Le Won de Ki | vers FOTO PAPA TITTEN Ce tt hs Le ot ni A | i dine be oe em rd mAb poet ! ed ar ! y Der de es ‘ i ив : ИРУ £ > } TS в, . X ! RR ne N НИ , ve ATA у M2 ди j р Рен г oe г | ‚ * . у ‘ . , у Им t у 2 p EMA à + : N 3 eins y Zr 4 ' y . ; ; i . y PWI р } } ; 1 . | PS + A еж, 2» PESA ire te ea В 131) HEN , y , MÉTÉO y ; PAPE ‘ j 3 ‘ \ ! | Ум u 4 PS ULA 9 , | Нч AO IO EN JUN i | АИ | т JUNE UE) USA { ми ' bi Nahezu 4 EA , 6 QAR м wre ud { иен. Sen ` ‘ USER \ ‘ eurer \ AAC NA mi AST ER EN + WA + 1 ys A , . Kun С Fare aime ETA CIÓN > \ ‘ . ï 3 hr \ RME pen re ! 5 ARME MP Ura aus Py, : Anti, A A uy А} URS NENA Ue IA ‘ м +2943) ke 4 1 er ny edd yi nr м as m2 ve Pe УР IIA AA ЛИ Na lo wat x REN Karen | 4 (4 x | 4 ur k А Во v 5 H зая А АХ ЧИНА ay, { 4 {ин u EDER мн Ан IE EEE Lor ate } ry: eae Ау АНИ