BOYER Volvarina in Brazil NOVAPEX 1 (2) : 35-55, 20juin 2000 The genus Volvarina (Volutacea Marginellidae) in Brazil. : Part 1 : revision of the species described by A. Bavay, and closely related species Franck BOYER 1 10, Chemin du Marais du Souci, 93270 Sevran, France KEYWORDS. Marginellidae, Brazil, Volvarina, taxonomy, type matenal, phylogeny. ABSTRACT. Six Brazilian species of Volvarina described by A. Bavay are revised, together with a recently described related one. Two new species of Volvarina are described. The phylogeny ofBrazilian Volvarina is tackled. RESUME. Six espèces brésiliennes de Volvarina décrites parA. Bavay sont révisées, ainsi qu'une espèce apparentée récemment décrite. Deux nouvelles espèces sont décrites. Laphylogénie des Volvarina brésiliennes est abordée. INTRODUCTION In 1985, E. C. Rios tentatively performed the only gênerai faunistic inventory ofBrazilian marginellids. The Brazilian fauna of marginellids remains poorly After the study of the correspondant iconography, known, despite the great diversity presented by the and a comparison with the available material stored new phenae brought to hght by local collectors over in European public and private collections, we can many years. Most of thèse phenae are provisionally state that the Brazilian fauna of marginellids, as attributed to Caribbean taxa, depending on very presented in Rios, shows only very superficial superficial similarities. This practice tends to similarities with the fauna known from the Caribbean dissuade further investigations about taxonomic or Sea. biogeographic aspects. Concerning the genus Volvarina s.s., just one single Very few species were described from Brazil well-known species from the Caribbean Sea can be subsequently to those described by Bavay at the confirmed as belonging to the Brazilian fauna. This beginning ofthe rwentieth century : species, Volvarina avena Kiener, is represented - One large Prunum (Marginella matthewsi Van Mol throughout the whole Caribbean area by important and Tursch, 1967), one large Dentimargo populations in shallow and moderate depths, whilst {Marginella cloveri Rios and Matthews, 1972) and just known by scarce spécimens from the Brazilian one tiny Volvarina {Prunum amphorale de Souza, coasts. 1992), ail ofthem from equatorial zones, in moderate Ail the other named Caribbean species belonging to depths. the genus Volvarina s.s. and proposed by Rios as - One médium sized Volvarina {V. pontesi Rios and inhabiting Brazil, are invalidated by his figures, and Leal, 1993) frombathyal levels, offSoutheast Brazil. it would hâve been more correct to hâve considered Récent records confirm that an important number of them as non-described endémie species from Brazil. tiny to large marginellid species are still to be This applies to the following références no. 528 (as discovered along the thousands of kilometers of P. amabilis Redfield), no. 530 (as P. bella Conrad), Brazilian coasthne. However, their study cannot no. 544 (as V. albolineata d'Orbigny), no. 546 (as V. senously be undertaken without a prehminary avenacea Deshayes = junior synonym for V. avena revision of the available taxa, the comparison of Kiener), no. 547 (as V. gracilis C.B. Adams, linkable populations, and a reliable knowledge of proposed by Rios as a senior synonym ofM.joubmi both their géographie andbathymétrie distribution. Bavay = V. bahiensis Tomlin). By reviewing the taxonomic status of species The occurrence of V. gracilis C.B. Adams in Brazil belonging to the genus Volvarina sensu stricto whose remains to be verified, as explained below, and V. identity remains in doubt, ît is hoped to help to bahiensis Tomlin is not the same species (cf. clarify the documentation concerning Brazilian discussion on V. bahiensis). On the other hand, marginellids, and to encourage further research on référence no. 524 in Rios is to be confirmed really to their biogeography and phylogeny. This paper is deal with the genus Dentimargo, as the flexuous mainly devoted to the species described by A. Bavay labrum of the figured shell better suggests a from 1906 to 1922. Volvarina species. As demonstrated below, several 35 Bou k The genus Volvanna in Brazil Novapex 1 (2) : 35-55, 20juin 2000 bjdl\ knowo species of Volvanna reaching a length SYSTEMATICS of 2 ta 4 nun are living along the Brazilian coasts, and référence no. 524 just could be a non-described Genus Volvanna Hinds, 1844. one Type species Marginella nitida Hinds, 1844 Référence no 54°- îs compared to V. roberti Bavay, (subséquent désignation by Redfield, 1871), junior firom Made'ira, Northwest Afnca. However, figure synonym of Volvanna mitrella (Risso, 1826). M l> clearly represents a shell of V. serrei Bavay, described from Brazil, the latter species also being Volvarinaperrieri(Bavay, 1906) illustrated and correctly named under the référence Fig. 1 - Ibis. no. 550. It is noted that the name of V. lactea Kiener is Marginella (Volvarina) perrieri Bavay, 1906, pp. proposed under the référence no. 548 and illustrated 248-249, pi. 7, figs 5-6. b\ a shell of Hyalina pallida Linnaeus, which belongs to a différent non-radulate genus (even if Type material. probably a distant relative of the complex Prunum Volvanna). H. pallida is common ail around A. Four syntypes in MNHN (type collection), in 2 the Caribbean province, in rubble or sand, and it can lots : mm occur in very shallow water in sheltered - Lot 1 : 3 shells (sized 10.8 x 4.3 to 10.3 x 4.2 environments. The species may not be rare in Brazil, mm) deposited by Bavay with two labels, one of but local collectors do not employ methods for them bearing the inscription "type figuré. I. collecting microshells in shallow environments. Malouines. Expéd. de La Vénus 1839". Six spécifie taxa belonging to the genus Volvanna - Lot 2 : 1 shell (sized 10.5 x 4.2 mm) from the ss were described by Bavay : V. perrieri 1906, V. Jousseaume collection, erroneously labelled serrei 1913, V. joubini 1913 (preoccupied and "Marginellaperrieri Rochebr." renamed V. bahiensis by Tomhn, 1917), V. germaini B. One syntype in IRSNB, from the Dautzenberg 1913, V.perexilis 1922, V.pupa 1922. collection, labelled in the hand of Bavay "VolV The first species is described from material obtained perrieri Bavay. Voyage de La Vénus 1839. I. by Dupetit-Thouars during the campaign of the ship Malouines. Muséum". La Vénus in 1836-1839. The three next species were obtained by Mr P. Serre, correspondent ofA. Bavay Original description. in Salvador de Bahia. The last two species were (Translated from French) found in Jousseaume collection, dredged off Parahyba. "Shell of poor length, slender fusiform, particularly It is proposed to review hereunder thèse supposed bulging in its upper part; spire well apparent rounded endémie species, accepting ail of them in the genus conoid, with an obtuse top, made offour whorls; the Volvanna, sensu Gofas (1989, p. 160). In the course last whorl of the shell raising clearly around the of this study, some related phenae are studied, and aperture; this one triangular very slender, larger on two new species are described. the base which is obliquely rounded; nght border extemally thickened and slightly curved inwards ail Abbreviations along its médium part; columellar border bearing on its base four somewhat délicate oblique plaits. IRSNB Institut Royal des Sciences Naturelles de Colour transparent porcellaneous white but : Belgique, Brussels. suggesting however three very faint shades of fawn MCZ Muséum of Comparative Zoology, Harvard bands on the dorsum ofthe shell." : Umversity, Cambridge, USA. As noted by Bavay, the shells of V. perrieri look like MHNG Muséum d'Histoire Naturelle de Genève, the well-known Caribbean V. avena (Kiener, 1834). : Switzerland. Bavay, however, establishes the identity of V. MNHN Muséum National d'Histoire Naturelle, perrieri through slight différences with V. avena, : Paris. having a less slender shape, thinner and more oblique MNRJ Museo Nacional do Rio de Janeiro, Brazil. columellarplaits, and a subtranslucid whitish colour. : MORG Museo Oceanogrâphico do Rio Grande, The syntypes correctly match the description, even : mm Brazil. for the average size (given as 1 1 x 5 by Bavay). NNM Nationaal Natuurhistonsch Muséum, Leiden, : the Netherlands. Type locality. Coll. F.B. Collection ofthe author. Malouines Insulas (Falklands). : 36 BOYER Volvarina in Brazil NOVAPEX 1 (2) : 35-55, 20juin 2000 Remarks. In fact, the so-called new species quoted by Bavay from the French magellanic expédition are V. hahm None of the located syntypes hâve a small hole on (Mabille, 1884) and V. dozei (Mabille and the ventral side of the last whorl, just near the anal Rochebrune, 1889), both found between the Straight canal, as pictured in the type figure (Fig. la, b). ofMagellan and the Falklands, at a depth of 120 m. However, the origin of syntypes housed in MNHN V. hahni Mabille is clearly a junior synonym of V. and IRSNB îs beyond doubt. Bavay specified that warreni (Marrât, 1876) as confirmed by the study of several samples coming from the expédition of La the four syntypes stored in MNFTN - Type collection. Vénus were represented in the Pans Muséum, The high-spired V. dozei does seem really to be a without stating precisely how many. It seems différent species, and clearly belongs to the same probable that more than five shells belonged to the "subantarctic group" of V. warreni which holds original lot. As the rediscovery of the figured type numerous phenae and possible species recorded from and potential holotype remains possible, it seems southern latitudes off South Africa and offAustralia better fornow to avoid the désignation ofa lectotype, and New Zealand. Amongst thèse species is and wait for further investigations in French and numbered Volvarina pontesi Rios and Leal, 1993, Bntish collections where spécimens from Bavay are fished off Southeast Brazil in 200 to 1,140 m, and currently found. which does présent a very high and produced spire, The original characters proposed by Bavay for a longer than the aperture, and a very narrow body distinction of V. perrieri at a spécifie level are whorl. We note that ail the species belonging to this flimsy : they ail belong to the range of the natural group V. warreni are fished in deep and somewhat vanability of V. avena (Kiener). The whitish colour cold waters, as opposed to the species belonging to ofthe shell is not a common feature for V. avena, but the group V. avena which are restricted to shallow to whitish hve spécimens are found towards the moderate depths (0-25 m for V. avena), in warm northern edge of the géographie distribution of the waters. species (Florida, Bermuda, Yucatân...), and an The remarks by Bavay concerning the French équivalent situation could occur towards the southern Expédition to Cape Horn clearly led Jousseaume to limits ofîts distribution, in Brazil. attribute erroneously V. perrieri to Rochebrune (cf. Moreover, the présent condition ofthe syntypes of V. type material références), who never desenbed such perrieri leads to consider them as worn-collected a species, and did not give any mention about the spécimens, even if they did continue to fade during taxa V. avena and V. perrieri from French Antarctic their long stay in drawers. collecting. We note also that V. avena is very scarcely recorded Furthermore, the study of the report of Dupetit- from Brazil. Rios (1985) records the species from Thouars (1840) on the "Voyage de La Vénus", from Amapâ to Rio de Janeiro, in the Atol das Rocas and which was obtained the type material of V. perrieri, in the Abrolhos Islands. Two spécimens coming leads to reconstruct the real opportunities of from this latter locality are in the collection of the sampling during the travel in the Southwest Atlantic. author (leg. E.C. Rios). Thèse two spécimens, La Vénus left Tenenfe on 10 January 1837, passing mm collected at 5 m, are 11.9 to 12.1 in size and Boavista, Fernando de Noronha and Cabo Frio. présent exactly the same structure of shell, without stopping. The expédition stayed in Rio de columellar plaits and décoration of three orange Janeiro from 4 to 16 February. No collecting is bands on the last whorl, as found in spécimens mentioned, but it is very likely that diverse collected in the Caribbean Sea. accumulations of natural objects were made, as was customary at this time and for this kind of Discussion. expédition. La Vénus started from Rio on 16 February, passed by The main trouble about the taxon V. perrieri was for Santa Catanna Island on 20 February and arrived at many times îts type locality (Falklands) which the mouth ofRio de la Plata on the 24th. Soundings attributed to it the status of a magellanic species, were made along Isla de Lobos, Isla Flores, restricted to cold waters. As a resuit, V. perrieri has Montevideo and Buenos-Aires. On 8 March. La been assimilated by modem authors (cf. Carcelles, Vénus was west of the Falklands, mid-distance 1953, p. 11) to the magellanic group of V. warreni between the continent and the archipelago, and made Marrât, 1876 (senior synonym for V. patagonica Von soundings in 100 m. Then, sédiment was obtained Martens, 1881). "for the first time in this trip", using a small rake- Bavay himselfnoted that two new marginellids were dredge. recently discovered by the French Expédition to A full bag of grey-black silty sand was brought up. Cape Hom, as a confirmation of the plausibihty of Amongst the names quoted for shells sampled from the occurrence under subantarctic latitudes of a this sédiment, some of them ("terebracules, species of Volvarina belonging to a tropical group. olives...") may actually refer to species of Volvarina. 37 BoM K The genus Volvarina in Brazil Novaphx 1 (2) : 35-55, 20juin 2000 In tact, tuo spécimens of the species V. warrem are French Scientific Mission to Cape Horn later stored m MNHN General collection, and labelled obtained the slender high-spired V. hahni (= V. as "M hahni Rochebrune, collected by Dupetit- warreni) and V. dozei, but no phena from the group Iliouars, Expédition of La Vénus (Malouines)". This V. avena/perrieri. Numerous collectings hâve been origin îs quite likely and could strongly support the reported during the XXth century from Magellanic conviction of-Bavay (ifthèse spécimens were known province by British, Argentinian and Uruguayan b\ hnn) about the subantarctic origin of V. perrieri. authors (Powell, 1951; Carcelles, 1946; Ureta, 1961) On 9 March, soundings were made in the with many records of V. warreni, but with no neighbourhood of Eagle Pass. On 11 March, La mention ofthe group V avena/perrieri. Venus passed along the Straight of Magellan, and This fact certainly does allow us to conclude that the amved in Valparaiso on 27 April, without stopping Falklands is an incorrect type locality for V. perrieri along the southwestern Chilean coast. Bavay, and it is highly probable that the type So, \ve can state that the expédition had only two real material of V. perrieri was dead-collected on beaches possibilities for collecting shell material in the in the bay of Rio de Janeiro (which seems to be Southwest Atlantic : during a 12 daystay in Rio de precisely the southern limit of the distribution of V. Janeiro, and during a single dredging attempt in 100 avena Kiener) and later mixed up with the material m, west ofthe Falklands, in the same area where the obtained offthe Falklands. FFiigg.. 1Ib:istypleecftigoutryepseooffMMaarrggiinneellllaa(aVvoelnvaarVianlae)npceirernineerisBinavKaiye,ner1,9061,83p4i.. VMIHI,NfGigs258-463.2 (1 1.3 x 4.5 mm). : 38 BOYER Volvarina in Brazil NOVAPEX 1 (2) : 35-55, 20juin 2000 Conclusion. (Dali, 1881)], is just a misspelling for M. avena Kiener, as clearly shown by the redescription of "M. Volvarina avena was originally described by Kiener avenacea Kiener" by Deshayes (1844, pp. 454-455). (1834) as "Marginella avena Valenciennes", and the Many more synonymous taxa hâve subsequently referred matenal is said to hâve been studied in the been described as M. varia Sowerby, 1846, M. Massena collection (included in the Delessert beyerleana Bemardi, 1853, and other names till the collection, now in Geneva Muséum). récent years, but without any évident biologie In his description, Kiener refers to "white or meaning, except that Volvarina avena has been yellowish hyahnous" shells, and in his comments, he shown to be a polymorphic species, with more or less says that young spécimens are yellowish with three the same kind of intrapopulational and ecological darker spiral bands on the last whorl. In fact, Kiener variations within the différent parts ofits widespread clearly refers to several spécimens, possibly coming distribution. from différent origins. We propose to consider Volvarina perrieri (Bavay, When quoting the taxon M. avena as a species from 1906) as a junior synonym of Volvarina avena Valenciennes, Kiener surely referred to a label (Kiener, 1834). attached to a lot, originally studied by Valenciennes and designatedby himselfas "M. avena", with an eye Volvarina serrei(Bavay, 1913) to a future description. In fact, Valenciennes never Figs 2-10, 32, 35 did publish such a species, and the nomenclatural rule in this case is that Kiener has to be considered as Marginella (Volvarina) serrei Bavay, 1913a, pp. the original author ofthe taxonM. avena. 359-360, text-fig. 3. The material which is now stored in the Delessert Type material. collection (MHNG) as supposed syntypes of M. avena Kiener is a lot ofthirteen shells removed from Fourteen supposed syntypes in Pans Muséum a card with the manuscnpt mention "M. avena Kiener (MNHN - type collection), in two lots : - varia Sow. Antilles", without any référence to - Lot 1 : Ten shells (adult, subadult and juvénile), Valenciennes or to the Massena collection. Actually, most of them deeply wom or drilled, rather slender, the labels in the Delessert collection are generally except for one médium sized full-white adult shell, mm poor, and the original labels from Kiener were rarely pictured herein (Fig. 2). Shells are 5.7 to 3.7 mm conserved. It was verified that amongst the thirteen long and 2.6 to 1.65 wide. With a label from supposed syntypes, there are full-white shells as well Bavay "Bahia, Brésil, Sables, M. Sene". On one : as yellowish shells with darker bands. Furthermore, more récent label "paratypes". : mm Kiener gave a size of 12 x 4.5 (5.5 x 2 hnes) for Lot 2 Four shells without label, in same : the species, and the supposed syntypes in the conditions of conservation. One worn adult, two mm Delessert collection range from 8.7 to 1 1.7 long. subadults and onejuvénile. So, this matenal can be accepted as the material used by Kiener for the description of V. avena, even ifthe Other material examined. label joined to this lot was apparently attributed afterwards. It is therefore proposed to accept this lot Ail in the collection ofthe author. of thirteen shells from the Delessert collection -North Brazil (00°12'S, 47°03'W). Para State, 33 m. (MHNG, ex 993/192) as syntypes of M. avena Dead dredged, 26-04-68 by NOAS Expédition - Kiener. N.NE. IL One adult drilled shell. Length 7.85 mm. : We designate and figure one of the best conserved Leg. P. deSouza, 09-91. and coloured shells as the lectotype of M. avena - Natal State. Three white adult shells, beach Kiener (MHNG 28432, ex 993/192; 11.3 x 4.5 mm; collected. Length 4.1 to 5.2 mm. : Fig. Ibis a, b), the other twelve shells ofthe lot being - Bahia State, Salvador, Ribeira. Three spécimens, considered as paralectotypes (MHNG 28433, ex live collected by screening, at 1 m, in sand with 993/192). broken shells. Length 5.2 to 5.6 mm. : It is proposed to confirm as type locality the broad - Bahia State, Salvador. Two spécimens live géographie référence given by Kiener : "Habite les collected at 1 m, under rocks, in sand. Length : 7.9 to mers des Indes Occidentales" or : "Inhabits the seas 9.1 mm. Leg. J. Coltro, 05-90. ofthe West Indies", which means in fact, for French - Bahia State, Itapanca. One spécimen, live people from this time, ail the Main and Lesser collected. Length : 7.75 mm. Leg. de Souza, 13-03- Antilles. 91. We note that the taxon Marginella avenacea - Espirito Santo, Guarapari. Three spécimens, live Deshayes, 1844, often quoted by authors -even in collected under rocks, at 15-20 m, by A. Bodart, 10- modem times - as being distinct from the species M. 92. Length: 8.1 to 8.95 mm. avena [and sometimes used for naming V. avenella 39 BOYI K The genus Volvarina in Brazil Novapex 1 (2) : 35-55, 20juin 2000 Espirito Santo, Guarapan. Three spécimens, live It was not possible to sort out from thèse two lots any dredged at 20-25 m, on bryozoan bottom. Length : spécimen which could clearly match the type figure mm to5.55 (Fig. 3). It seems that the type Figure is a blend of - Espinto Santo, Guarapan. One spécimen, live "characteristic features" of the species, and not the dredged at 30-40-m. Length : 6.8 mm. direct picture ofone shell. Consequently, it does not Rio de Janeiro State. Six spécimens and shells, live seem possible to distinguish any holotype in the lots. and dead dredged in muddy sand, offRio de Janeiro, It should be emphasized that the mention "paratypes" at 80-100 m. Length : 6.1 to 7.9 mm. Leg. J. Coltro. attached to the first lot strongly suggests that an - Sao Paulo State, I. Ihabela. One spécimen, live holotype was at the same time marked, but that it dredged at 30-40m, in sand and broken shells. could hâve been loaned or provisionally misplaced Length : 8.4 mm. Leg. J. Coltro, 11-91. dunng the penod we examined this type material. For the same reasons as presented about V. perrieri, Original description. we prefer to refrain from the désignation of any (Translated from French). lectotype, the type material (lot 1) being sufficient for the détermination of the species, and the "Médium sized shell, slender egg-shaped, narrowed discovery of complementary lots belonging to the to the base, white and shining; four whorls for the type material remaining possible in public or private spire, which îs produced, clearly conical,just slightly collections. obtuse to the top; aperture slightly sinuous, narrowed There are several features of the shell morphology to the top, widened to the base; wide and thick and of the pattern of décoration which may be labrum in its upper and médium part, thinner to the important for the distinction at the spécifie level base, antenorly slightly margined, smooth within; (Figs4-10, 32, 35): columellar side bears four oblique subequal plaits, -The anterior columellar plait is very short and thin, the upper one being slightly more faint than the faintly apparent under the second plaît (Fig. 35). others. -The number ofcolumellar plaits varies from 4 to 5. The size, in this species, is somewhat variable, but The frequency of a well-marked fifth plaît seems to the gênerai shape is constant. be more important in large spécimens (over 7 mm) Length : 6 mm to 4 mm; width : 2.2 to 1.6 mm". living in shallow to moderate depths (1 to 40 m) : in this case, the rate ofprésence of a fifth plaît is 70%. Type locality. This rate is inverse for smaller spécimens (4 to 7 mm), amongst whose 70% bear just four plaits and Bahia (Brazil). Found in sands sent to Paris Muséum 30% bear a very faint fifth one. However, more by Mr Serre. investigation is needed, and it is to be venfied ifthis distribution has a concrète statistic worth for the Distribution. whole species V. serrei, or if it just dépends on the observed populations. Known from Para State (Northern Brazil) to Sào - The outer margin of the labrum is very wide, Paulo State (Southeastern Brazil), from 1 to 100 m. narrowed in the upper part, and widely surrounding Not recorded from dredgings ofOCPS campaigns off the siphonal canal. In well-coloured spécimens, the Surinam (Leiden Muséum - Netherlands) neither margin is decorated with brown horizontal rays or from dredgings offRio Grande do Sul (E.C. Rios and stains on an orange-brown background (Figs 6, 9). J. Coltro, pers. com.) - The background colour is very variable, from milky-white in spécimens from deep levels to Habitat. orange-brown in some spécimens from moderate depths. The most coloured spécimens présent three In sand, mud or détritus; open environments of soft faint brownish bands on the last whorl, the upper one bottoms, bryozoan hard bottoms, or under rocks. being the thinner, but also the more accurately defined one (Figs 5, 7, 8, 10). The central and the Remarks. basai bands are wider, but faintly marked or indiscernible in most cases. In some spécimens, just The first lot in the type-collection of Pans Muséum the upper dark part of the central and anterior bands clearly belongs to the type material, while the status are visible, looking Hke one or two thin brown lines of the second lot is less évident, because it lacks an around the whorl. Numerous spécimens hâve a light a.tached label. Presenting however the same state of tan to yellowish semitransparent shell, without any conservation and same proportions, this second lot décoration. has probably the same ongin as the first one, and - As noted by Bavay, the size of the shell is very deserves to be kept in the type material. variable. We recorded lengths from 4.1 to 9.1 mm for shells belonging to the phena V. serrei (cf. below the 40 BOYER Volvarina in Brazil NOVAPEX 1 (2) : 35-55, 20juin 2000 comparison with the phena V. germaini), that means Further investigations could reveal complementary a ratio from 1 to 2.25, a very large one within the type material in public orprivate collections. genus Volvarina. For the closely related V. exilis Gmelin from Sénégal, we note length ratios from 1 to Other material examined. 2.15. Ail in the collection ofthe author. Discussion. - Bahia State, Salvador. Eleven dead collected adult shells. Length 3.1 to 3.7 mm. : V. serrei Bavay appears to be a valid species, widely - Pernambuco State. One dead collected adult shell. distributed along the Brazilian coasts, and closely Length : 3.35 mm. Leg. J. Coltro, 02-98. related to V. mitrella (Risso, 1826), which is the type species of the genus Volvarina. The Mediterranean Original description. V. mitrella has several close relatives along (Translated from French). Northwestern Africa V. exilis (Gmelin, 1791) from : Sénégal, V. attenuata (Reeve, 1865) from Mauritania "Small shell, subcylindncal, narrowed to the base; and Sénégal, and V. roberti Bavay, 1917 from conical spire, slightly more obtuse to the top, with a Madeira and the Canary Islands. We note that V. noticeable height (it measures slightly less than a attenuata, as V. serrei, currently présents a fifth quarter ofthe total length ofthe shell), made of four columellar plait, an uncommon feature within the whorls. genus Volvarina. V. eumorpha (Melvill, 1906), from Slender triangular aperture, narrowed to the top and the Gulf of Oman, could belong to the same regularly widening to the base; left border with four "complex V. mitrella". There is not any évident columellar plaits, of which the three upper ones are représentative of this complex in the Caribbean Sea. equal, the last one, or lower, being slightly sharp, ail As far as gênerai shell morphology is concerned, V. oblique; labrum almost straight, thickened in his avena could be a possible relative of V. serrei, but upper half-part. anatomical and biochemical data hâve to verify ifthe Colour ofthe shell pale fawn; the last whorl is omate matter holds true. However, some Brazilian with two lines : the upper one is linear, placed Volvarina seem more or less closely linked to V. slightly above the middle of this whorl; the lower serrei, and to be referable to the same complex of one, wider but less dark, is near to the base; the species, as examined in the course ofthis paper. sutures are more or less tinged with fawn. Length 3.5 mm; breadth 1.8 mm; spire height 0.8 : : Volvarinagermaini(Bavay, 1913) mm". Figs 11-14 Type locality. Marginella (Volvarina) germaini Bavay, 1913b, p. 483, pi. XX, figs 1,2. Bahia, Brazil. In sands. Type material. Distribution. Not found in MNHN (typothèque and gênerai Known from Pernambuco State to Bahia State, collection). However, a set ofthree shells is stored in Northeastern Brazil. the Dautzenberg collection (IRSNB) with the label : "M. (Vol")germaini Bavay. Bahia. Sables". Habitat. Depending on the close relationship prevailing between Bavay and Dautzenberg, there is no doubt Unknown. that this lot came from the sands sent by Mr Serre to Bavay, and has to be considered as belonging to the Remarks. type material. We do not know if Bavay descnbed this species V. germaini is not separable from V. serrei on the referring to a larger number of shells, and the status strength of its shell characters. The phena V. oftypes is not specified on Dautzenberg's label. We germaini is just distinguished by its smaller size, a are no more able to distinguish if any one spécimen more transparent shell, a slightly more oval outline from the Dautzenberg collection is represented by the with tendency to a shorter blunt spire and a slightly type figures (Figs 13, 14), which seem to be a more widened antenor part of the aperture, a more synthesis ofseveral shells. conspicuous spiral décoration, and a generally more However, depending on the origin of Dautzenberg's apparent first columellar plait. The principal features lot, we consider thèse three shells as syntypes. characterizing the species V. serrei are however found in the phena V. germaini : similarity of the 41 Boï i K The genus Volvarina in Brazil NOVAPEX 1 (2) : 35-55, 20juin 2000 Structure of the spire, protoconch and teleoconch, of thc gênerai morphology and proportions ofthe body Type material. whorl, organisation ofcolumellar plans, outer hp and extemal varix I'he absence ofa fifth columellar plaît Holotype of M. joubini Bavay in MNHN. Length : is also dominant in small représentatives ofthe phena 4.5 mm, breadth : 2.2 mm. However, the labrum is i serrei broken, and the initial breadth was probably 2.4 mm. The main différence conceming the phena V. Two original labels, from the hand ofBavay "M. P. : germaini is in the lightness of îts shell, associated Serre, 1913. Marginella joubini Bavay. Bahia. wîth îts transparency, which gives more contrast to Type." and "Marginella joubini Bavay. Type brisé. the spiral pattem ofthe décoration. Thèse characters, Bahia. Sables." as well as the tendency to a squatter outline, are commonly found in dwarf populations of testaceous Other material examined. marine gastropods, notwithstanding simple physical constraints of architecture and development. The Ail spécimens in the collection ofthe author. same reasons could explain the tendency to an oval -Bahia State, Salvador. Two adult shells (Length/ outline and the more protruding first columellar plaît breadth 4.2 x 1.5 mm, and 3.6 x 1.2 mm), and one : in V. germaini. juvénile (Length/breadth 2.85 x 1.1 mm). : It is noted that the size-range of the phena V. germaini is perfectly complementary of the size- Original description. range of the phena V. serrei, and that their common (Translated from French). size-range ofvariability is from 1 to 3. "Fusiform slender shell, with an irregular conical- Discussion. truncated spire, slightly narrowed at the level of the sutures, and rounded to the top; the last whorl ofthe The phena V. germaini could only represent the shell is slightly narrowed to the base, its right border dwarfspécimens within the populations belonging to goes up rather suddenly and rather high along the the species V. serrei. Another possibility could be aperture. that thèse dwarf spécimens hâve a population status Narrow aperture in the upper part, then irregularly as an ecological form expressing a local adaptation inflated to the base; labrum rather straight, thickened of the biologie species V. serrei to spécial in its médium part which extends forwards, environmental conditions. externally marginated; left border with four very However, this kind ofsituation is also favourable for subequal oblique plaits linked by a light deposit of speciation if reproductive isolation occurs. In the enamel spread over ail the left border. considered case, confirmation could corne from field Colour white. The remains of two or three fawn observations of live animais and populations. Such bands are indistinctiy seen on the last whorl." observations will allow to verify the suspected conspecificity ofboth phenae and the possiblejunior Type locality. synonymy of V. germaini Bavay (1913b) with the former V. serrei Bavay (1913a). Bahia, Brazil, in sands. Volvarina bahiensis (Tomlin, 1917) Distribution. Figs 15-17 Known from Bahia State (Eastern Brazil). Marginella joubini Bavay, 1913b, pp. 482-483, pi. XX, figs 3, 4 (not M. joubini Dautzenberg and Habitat. Fisher, 1906). Marginella bahiensis Tomlin, 1917, p. 252. Unknown. Replacement name for M. joubini Bavay, 1913, not M.joubini Dautzenberg and Fisher, 1906. Figures 2-10. 2. type figure ofMarginella (Volvarina) serrei Bavay, 1913a, p. 360, fig. 3. 3. syntype ofMarginella (Volvarina)serrei Bavay, 1913. MNHN (5 x 2.1 mm). 4. Volvarinaserrei Bavay. OffRio de Janeiro, 80-100 m. Coll. F.B. (7.9 x 4.1 mm). 5-7. Volvarina serrei Bavay. Ribeira, Salvador, Bahia, 1 m. Coll. F.B. (5.9 x 2.8 mm). 8-10. Volvarina serrei Bavay. OffGuarapari, Espirito Santo, 15-20 m. Coll. F.B. 42 BOYER Volvanna in Brazil NOVAPEX 1 (2) : 35-55, 20juin 2000 HoU K 1 lu- genus Volvarina in Brazil Novaphx 1 (2) : 35-55, 20juin 2000 Remaries. Volvarinaperexilis (Bavay, 1922) Figs 19,20 rhe t\pe figure (Figs 16, 17) is slightly squatter than ihc holotype, which is slightly more slender. The Marginella perexilis Bavay, 1922, pp. 78-79, text- accurate original description perfectly matches the fig. 3. fresh shells stored in the author's collection (Fig. 15). Type material. Discussion. MNHN Six syntypes in : 1 destroyedjuvénile shell, 4 The spécifie status of V. bahiensis is undoubted. The adult shells very damaged, partially broken and synonymy with V. gracilis (C.B. Adams, 1851) corroded, 1 adult shell in a good state of proposed by Rios (1985, p. 122) is incorrect. The conservation, hère illustrated (Fig. 19) and shell of the North Canbbean V gracilis is much designated as lectotype (length : 2.05 mm, breadth : larger (Fig. 18) and présents a deeply différent shape 0.75 mm). The other shells were apparently smaller of the anterior part, with a bulging left external in their original condition. They hâve to be antenor part and a less opened aperture. Columellar considered as paralectotypes. One label "Dr : plans of V. gracilis are more oblique and the third Jousseaume 1921. Marginella exilis Jouss. Brésil." one is shorter than in V. bahiensis. The dimensions of the lectotype of V. gracilis (MCZ 186119, type Original description. locality Jamaica) are length = 6.35 mm, breadth = (Translated from French). : : 2.25 mm. The dimensions of the two shells from Gran Cayman Island, in author's collection, are "Tiny shell, light and fragile, rather thick and with a : lengthbreadth 7.35 x 2.7 and 7.25 x 2.6 mm. rounded top; proportionally moderate last whorl, Furthermore, the illustration of V. gracilis as high and egg-shaped. The aperture is oblique, wide, presented in Rios (1985, pi. 42, fig. 547) seems to be with a slight sinus at its upper part. Length : 3.4 mm; a reproduction of the figure of the lectotype of V. breadth : 1.3 mm". gracilis as figured in Clench and Turner (1950, pi. 32, fig. 14), and not the représentation of a shell Type locality. collected in Brazil. No spécimen of V. gracilis was observed in Brazilian samples and collections Soundings Bank, offParahyba, northeastern coast of examined by the author. Therefore, the real Brazil, in coral stone. occurrence of V. gracilis in Brazil is to be considered as doubtful. Distribution. The similar structure oftheir spires and labrums, and pattern of décoration could signify that V. bahiensis Only known from the type locality. belongs to the same group than the North Caribbean V. gracilis. Habitat. Such a "group V. gracilis" could however be represented by more species in Brazilian waters. One Possibly hard coral bottoms. fresh shell collected in Guarapari, Espirito Santo State, was examined from A. Wakefield collection Remarks. (Great Bntain), closely matching V. bahiensis in many aspects. Nevertheless, this shell was pure milky The selected lectotype perfectly matches the white, squatter and presented a wide subcylindrical description. However, the type figure (Fig. 20) does spire with an important bulging and produced not match so clearly, rather representing a biconical protoconch. This shell was unfortunately lost, and slender outline. The rounded top and egg-shaped last detailed investigations were stopped. whorl displayed in the description are not apparent in In fact, the range of variation of V. bahiensis is the type figure, but they are perfectly apparent in the unknown, and ît remains to verify with more lectotype, as figured (Fig. 19). On the other hand, the extensive matenal if this spécimen from Guarapari type figure clearly shows the six distinct columellar represented a géographie variant (or a subspecies) of plaits, which are présent on the lectotype and not a widespread species V. bahiensis, or a common quoted in the description. individual variation within this species, or a separate In his introduction, Bavay announced that shells and unnamed species with a more southern collected off Parahyba were figured as a distribution. reconstruction from a synthesis between spécimens and fragments. Yet, we note that ail the paralectotypes suggest a very high spire, rather 44