nmnmmé en nn c£n Be rl UT ttttti OUR HARVARD , ., ; UuNivER Société Belge de Malacologie association sans but lucratif VOL. 8 (1-2) MARS 1993 SOMMAIRE C. Van Osselaer Studies on Olividae. XV. Anterior notch measurements B. Tursch as taxonomic characters in the genus O/iva il B. Tursch Studies on Olividae. XVI. Fasciolar region measurements J. Bouillon as taxonomic characters in the genus Oliva 11 R.G. Moolenbeek New cones from Oman and the status of Conus boschi H.E. Coomans (Gastropoda, Conidae) 19 E. de C. Rios Volvarina pontesi, a new bathyal marginellid (Mollusca, J.H. Leal Gastropoda) from off Brazil 21 L. Bozzetti Description of a new species of the genus Favartia Jousseaume, 1880 (Gastropoda, Muricidae) from the Indian Ocean 31 R. Houart A remarkable new species of Poireria (Flexopteron) (Gastropoda, Muricidae) from the Philippine Islands 33 J.M. Lauer Description of a new species and a new subspecies of Conus (Mollusca, Prosobranchia, Conidae) from the Canary Islands 37 R. Peuchot The distribution of molluscs in beach deposits as A. Tassin identification of recent changes in the littoral 51 Périodique trimestriel Bureau de dépôt 1180 Bruxelles 18. Editeur responsable R. Duchamps | Comité d'édition Dr. Y. Finet | L. Germain | R. Houart Dr. CI. Massin | Prof. B. Tursch | Dr. J. Van Goethem | Les articles et textes présentés dans cette revue réflètent l'opinion personnelle de leur(s) auteur(s), et non pas nécessairement celle de la Société ou de l'éditeur responsable. Tous droits de reproduction, de traduction et d'adaptation des articles publiés dans ce bulletin, réservés pour tous pays. All rights of reproduction are reserved without the written permission of the board. Belgique - Belgium Etranger - Foreign [avec le service des bulletins) Abonnement aux revues APEX & ARION : Subscription to APEX & ARION Membre effectif 900 BEF 1400 BEF Membre étudiant 500 BEF Versement à effectuer par mandat postal international ou pe chèque bancaire en francs belges uniquement. (sans le service des bulletins) ayable, by international money order, or by bank check Personne appartenant à la famille d'un membre effectif in Belgian Ër ancs only. et ayant la même résidence 400 BEF au nom de at name of : Versements à effectuer au C.C.P. n° 0000974225:54 de M. J. Buyle la Société Belge de Malacologie c/o M. J. Buyle, Av. Maurice Maeterlinck, 56, bte 8 Av. M. Maeterlinck, 56, 1030 Bruxelles. B-1030 Bruxelles. CONSEIL D'ADMINISTRATION DE LA SOCIETE BELGE DE MALACOLOGIE ° Président : MR. Duchamps, Av. Mozart, 52, 1 190 Bruxelles © : (02) 344.15.47 ° Vice présidents : Dr. Y. Finet, 16 Chemin des Clochetes, CH-1206, Genève (Suisse) 7:41-22-46.77.95 | M. R. Houart, St. Jobsstraat, 8, 3400 Landen (Ezemaal) © : (016) 78.86.16 | e Secrétaire MmeJ . Masson, Rue du Merlo, 10, 1180 Bruxelles © : (02) 376.62.25 e Trésorier M. J. Buyle, Av. Maeterlinck, 56, bte 8, 1030 Bruxelles © : (02) 216.68.21 e Bibliothécaire Mme M.L. Buyle, Av. Maeterlinck, 56, bte 8, 1030 Bruxelles © : (02) 216.68.21 e Relations publiques M. G. Geeraerts, Stationsstraat, 2, 1930 Zaventem © : (02) 720.87.66 | e Administrateurs Mme M.L. Bresson, Place Guy d'Arezzo, 7, 1060 Bruxelles © : (02] 343.62.38 M. L. Germain, Rue de Linthout, 77, 1040 Bruxelles © : (02) 734.80.11 | Mme À. Langleit, Av. Cicéron, 27, bte 92, 1 140 Bruxelles M : (02) 720.41.61 | M. C. Van Osselaer, Chée de Waterloo, 512, 1060 Bruxelles & : (02) 347.36.84 | M. E. Waiengnier, Rue C. Wolles, 42, 1030 Bruxelles ® : (02) 241.51.80 | VAN OSSELAER & TURSCH Studies on Olividae. XV. APEX 8(1-2): 1-10, mars 1993. Studies on Olividae. XV. Anterior notch measurements as taxonomic characters in the genus Oliva. C. VAN OSSELAER and B. TURSCH. Laboratoire de Bio-Ecologie, Faculté des Sciences, Université Libre de Bruxelles, 50, Av. F.D. Roosevelt, 1050 Brussels KEY WORDS. Mollusca, Gastropoda, Oliva, taxonomy, morphometry, anterior notch. MOTS-CLEFS. Mollusca, Gastropoda, Oliva, morphométrie, taxonomie, échancrure antérieure. ABSTRACT. Four novel measurements of the anterior notch are defined and their potential for Oliva taxonomy evidenced. Reproducibility of measurements and causes of error have been studied. RESUME. Quatre nouvelles mesures de l'échancrure antérieure sont définies et leur potentiel pour la taxonomie du genre Oliva est mise en évidence. La reproductibilité des mesures et les causes d'erreur ont été étudiées. 1. INTRODUCTION obvious pointers and repeated attempts at direct measurements on the shell (TURSCH & Shell morphometry appears to be the most GERMAIN, unpublished results) have been practical, objective approach to the taxonomy shown to lack both precision and reproducibil- of the genus Oliva. Sets of protoconch meas- ity. Indirect measurements are far more urements (TURSCH & GERMAIN, 1985 and convenient and we wish to report here that ac- 1986), teleoconch measurements (TURSCH & curate observations can be made on imprints GERMAIN, 1985) and measurements of the of the anterior canal. The present paper aims subsutural groove (TURSCH & VAN OSSE- solely at defining these measurements and LAER, 1987; VAN OSSELAER & TURSCH, testing their taxonomic potential. 1988) have therefore been defined and tested. These measurements have repeatedly been shown to be useful and reliable taxonomic characters (TURSCH, GERMAIN & GREIFENEDER, 1986a and 1986b; TURSCH & HUART, 1988; TURSCH & GREIFENEDER, 1989; TURSCH & GREIFENEDER, 1989; TURSCH & HUART, 1990; TURSCH, MISSA & BOUILLON, 1992). Our search for additional shell characters has led us to test the possibilities offered by À À measurements of the anterior notch, which is conspicuous in all Olividae (see Fig. 1). The anterior notch does not possess any Fig. 1. Ventral and dorsal view of an Oliva sharp discontinuity that could be utilized as shell. The arrows point at the anterior notch. APEX 8(1-2): 1-10, mars 1993. Studies on Olividae. XV. VAN OSSELAER & TURSCH REF En] ANUA vi Fig. 3. Typical imprint of the anterior notch of an Oliva shell. REF is 1 mm reference segment. The trace of the outer lip is at the Fig. 2. Making a modelling clay imprint of the left, that of the columellar lip at the right. The anterior notch of an Ofiva shell. deepest part of each trace is also represented (see text, section 2.3). 2. METHODS 2.1. Making an imprint 2.3. Geometrical construction on the The shell to be measured is presented ver- drawing tically (apex up) and then carefully lowered The deepest part of each trace is quasi-lin- upon a flat, horizontal surface of modelling car. It is carefully drawn for each trace (see clay (plasticine) (Fig. 2). One should ensure Fig. 3), then graphically extrapolated by simultaneous contact of the lowest points of tracing the lines a and b (see Fig. 4 ). Line a is both the columellar lip and the outer lip with the direction of the lower edge of the outer lip the clay surface. A very slight vertical pres- and line b that of the lower edge of the co- sure on the shell then vields an accurate lumellar lip. Points A, B, C and D are defined imprint, consisting of two distinct traces of as the intersections of lines a and b with the comparable size. contours of the traces. Lines a and b intersect at point G. 2.2. Drawing the imprint Point E and F are defined as the midpoints The imprint is then carefuly drawn, using of segments AB and CD, respectively (Fig. 5). the camera lucida attachment of a binocular lens. The size of the drawing is controlled by 2.4. Measurements adjusting the magnification to obtain a length Let us define (see Fig. 5) the linear meas- of 2 to 5 cm for each trace. A segment of 1 urement DN as EF/REF (where REF stands mm is also drawn as an internal length refer- for the length of the 1 mm internal reference ence, using a precalibrated ocular reticulum. segment described hereabove). DN is thus the The length of this reference segment on the length of the segment EF, expressed in milli- drawing will be called REF. meters. VAN OSSELAER & TURSCH Studies on Olividae. XV. APEX 8(1-2): 1-10, mars 1993. REF \ b - 2 - - PL Fig. 4. First step of the geometrical Fig. 5. Second step of the geometrical construction on the drawing of the imprint construction on the drawing of the imprint (see text, section 2.3). (see text, section 2.3). Let us also define (see Fig. 5) the angular The two methods (direct protractor meas- measurements @ as the angle of the lines a and urements and calculation from the lengths of b; 8 as the angle of the lines EF and EG: 6 as sides of the triangle EFG) have been compared the third angle of the triangle AFG. AIl angles by performing both types of measurements on are expressed in degrees. Determination of two ten photocopies of the same drawing. This was of any of these angles automatically defines done both for a large (Oliva porphyria, speci- the third one, as their sum is equal to 180°. men BT-0345, H [height]: 114 mm) and a The angles can either be measured with a small species (Oliva hilli, specimen BT-6206, protractor, or calculated from the lengths of H: 12 mm). the segments EF, EG and FG. For the latter case, a little, simplistic computer program The results are given in Table 1. It can be based upon relations such as seen that angles «a, 8 and 8 are more accurate cosB = (EF2+EG2-FG2)/2+EF+EG when calculated than when directly measured, instantly vields the desired angles. especially when the angles are small. APEX 8(1-2): 1-10, mars 1993. Studies on Olividae. XV. VAN OSSELAER & TURSCH 2.5. Practical tips 2.6. Variability of imprints Best results have been obtained by using The expressions "a very slight vertical dark colored modelling clay (the better con- pressure" (section 2.1) vielding "rather shallow trast makes the drawing easier). The quasi-lin- imprints" (section 2.5) are very vague. There car, deepest part of the trace is located nearly is no easy, practical way of measuring the in the middle of the groove and 1s easier to pressure exerted on the shell or depth of observe on rather shallow imprints (where imprint 1n soft material. In order to be repro- shade 1s no problem). ducible, the process clearly needs a more accu- À minimum of practice 1s strongly advised rate description. before attempting actual taxonomic work; For a given shell, the length of the traces is some training greatly decreases the dispersion of course positively correlated with the depth of the measurements. For two series of ten of the imprint. Furthermore, as the Oliva shell measurements, separated by 30 other trials, the is not a regular surface of revolution, it is to be performance of a naïve observer (expressed in expected that the shape of the traces will also CV) improved from 4.80 to 2.20 for the vary with the depth of the imprint. The shape measurement of DN, from 14.68 to 7.14 for a, will also be affected with deviations to verti- from 3.61 to 2.70 for B and from 4.12 to 2.25 cality. This 1s indeed the case as evidenced in for 8 , all angles being calculated as decribed Fig. 6 where the trace of the outer lip is above. represented on the left. 2.7. Choice of suitable imprints If one produces a quantity of imprints of a given specimen on a slab of modelling clay, it will be seen at a glance that these imprints are very variable. The problem of which particular imprint to choose for measurements immedi- ately arises. Imprints can be characterized by the absolute size of one trace (roughly proportional to the vertical pressure exerted on the shell) and by the relative size of the two traces (roughly dependent upon deviations from verticality). This will allow an empirical optimization of experimental conditions. The graph of Fig. 7 shows the values of DN obtained for a series of purposely variable imprints obtained from the same shell (Oiva sayana BT-5315, H: 70.9 mm). Each imprint is characterized by the length (the largest diameter, not to be confused with AB or CD) of the trace of its columellar lip (CTR) and the Fig. 6. Four imprints of the same shell (Oliva length of the trace of its outer lip (OTR). sayana BT-5315, H: 70.9 mm). The traces of One sees that consistent values are ob- the outer lip (length: OTR) are represented on tained in a region where the lengths of both the left, the traces of the columellar lip traces do not differ by more than 10 % and lie (length: CTR) are on the right. DN is 39 mm. between 0.5 and 0.75 times the length of DN. Imprints vary both in size and in shape (see As an example, if we apply the above text, section 2.6). Imprints À and B are not recommended for measurements but imprints guidelines to the imprints depicted in Fig. 6, C and D do meet the requirements (see text, where DN is 39 mm, we can see that imprint A section 2.7). (OTR: 17 mm, CTR: 24 mm) should be VAN OSSELAER & TURSCH Studies on Olividae. XV. APEX 8(1-2): 1-10, mars 1993. rejected because the lengths of its traces differ 3. REPRODUCIBILITY and PRECISION too much: CTR does not lie within the limits of OTR plus or minus 10%. Imprint B (OTR: 40 Before any taxonomic application, one has mm, CTR: 38.5mm) should be rejected be- to assess the limits of confidence of these new cause the traces are too large (more than 0.75 measurements and to evaluate the relative DN). Imprints C (OTR: 28 mm, CTR: 28 mm) contribution of the various possible sources of and D (OTR: 19,5 mm, CTR: 18 mm) are error. adequate. The three consecutive phases of the proc- All these precautions might appear quite ess described hereabove can each lead to a intricate but were needed mainly to avoid gross different type of error: errors. In practice, the method 1s quite simple A. error due to the inaccuracy of the geo- and very fast. It takes only seconds to produce metrical construction and the measurements on a quantity of imprints of a given specimen and the drawing. with a bit of experience, a suitable imprint will B. error due to the inaccuracy of the be recognized at first sight. drawing of the imprint. C. error due to the non-reproducibility of the imprint process. 3.1. We have first compared the measure- ments performed by two independent observers CTR on ten photocopies of the same drawing. In this case we deal only with error A. This was done both for a large (O. porphyria, specimen BT-345, height: 114 mm) and a small species 80 7.8 (O. hilli, specimen BT-6206, height: 12 mm). 19 . 84 81 esLe 7.9 79 11 The angles were obtained by calculation from the lengths of the sides of the triangle EFG. RAnTé ANS LA The results are given in Table 2, where the BB” a EEE1 284 0 82 na 1 8.3 dispersion of the data can be evaluated by the standard deviation S or the coefficient of 0 6 2 7 0R E 84 84 y 8.3 01 s4 80 85 variability CV, which 1s the standard deviation 8.3 Lui as percentage of the mean (MAYR, 1969). It 48.0 can be seen that error À is very small and that 0] the two observers obtained practically the 0.75 DN OTR same values. 45 mm 3.2. We have then compared the measure- ments performed by two independent observers on ten different drawings of the same imprint. In this case we cumulate two types of error (A and B). This was done both for a large (O. porphyria, specimen BT-345, height: Fig. 7. Values of DN measured on a series of 114 mm) and a small species (O0. hilli, speci- imprints obtained from the same shell (Oliva men BT-6206, height: 12 mm). The results are sayana BT-5315, H: 70.9 mm). To represent given in Table 3. the variability, the values of DN are plotted 3.3. Finally, we have compared the meas- for different lengths of the traces of the co- urements performed by two independent ob- lumellar lip (CTR) and the outer lip (OTR). The recommended working zone is deline- servers on ten different imprints of the same ated by broken lines (see text, section 2.7) shell. In this case we cumulate the three types APEX 8(1-2): 1-10, mars 1993 Studies on Olividae. XV. VAN OSSELAER & TURSCH of error (A.,B and C). This was done both for a large (O. porphyria, specimen BT-345, height: degrees DN 114 mm) and a small species (Q. hilli, speci- men BT-6206, height: 12 mm). The results are given in Table 4. 3.4. In conclusion, the mean values ob- tained by separate observers are quite compat- ible: they differ by less than 0.1 mm on distances and 2° on angles. Their precision 1s also quite similar. The contribution of the various types of error can be very roughly estimated by observ- ing the evolution of the average values of the coefficients of variability (CV) obtained by two independent observers on the same shell during the three steps of the process. This was done both for a large (O. porphyria, specimen BT-345, height: 114 mm) and a small species (O. hilli, specimen BT-6206, height: 12 mm). Fig. 8. Anterior notch measurements on a The results are given in Table 5. growth series of Oliva sayana. DN (black circles) increases linearly with the heigth of Errors of type A (geometrical construction the shell (H). The angular measurements and measurements on the drawing) are very (open triangles), R (open circles) and 8 (open small and probably negligible for practical squares) do not significantly vary with H. purposes. Excepted for @&, the overall dispersions seem roughly independent of the size of the shell. The greatest dispersion is observed for & , which is a much smaller angle than 8 or 0: the same angular error will result in a much larger relative error. Errors of type B (inaccuracy of the draw- ing of the imprint) are by far the largest The linear measurement DN (the width of contributor to the total error on &. They are the anterior notch) could be expected to be largely due to the extrapolation error when size-dependent. Indeed, DN increases linearly tracing lines a and b. with the height of the shell (H) as shown in As only two of the three angles need to be Fig. 8. DN is also correlated (not 1llustrated) determined (see section 2.4) it is preferable to with other linear shell measurements (the select the larger angles B and 8. width D, the length of the lip L, etc ..). On the contrary, the angles «, B and 6 appear not to vary significantly with shell size. As Oliva shells show considerable vari- 4. VARIATION WITH SIZE ations in size, it follows that DN values are better utilized under a reduced form (expressed Before attempting actual taxonomic work, as a ratio to some other linear shell measure- one should first establish whether the measure- ment) such as DN/H or expressed as DN/pnw ments defined hereabove are dependent upon (pnw being the number of postnuclear whorls). the size of the shell or not. For a given species, the angles «, 8 and 8