6926475883

2C2

Arquivos de Zoologia

cerithioideans except Modulus. The main character is the modification of shell in a turriform shape (1); there are also the loss of glandular tissue in esophagus (91) and the simplification of hypobranchial gland (46), among other characters.

Node 3. It was a surprise to reunite Campanile with Serpulorbis and Turritella, occurred by 8 synapomorphies, being the most important the loss of determinate growth in shell (4) and loss of m2 (60). There is also the possibility of duplication of ducts to digestive gland (a reversion), but this character was not found in 7! hookeri. There are two possibilities, the preferred here is that T. hookeri lost, tertiarily, the second duet. The other possibility is that Campanile and Serpulorbis duplicated it independently with each other. Both explanations are equally parsimonious. Other morę remote possibility, but also considered, is that the single duet to digestive gland appeared in node 5, and Modulus and Turritella eonverged to this condition too. Further studies of morę species and on ontogeny could clarify this question.

Campanile symbolicum. This species presents a great reunion of autapomorphies, but comparable to other branches of the tree. The Australian Campanile was ineluded in present analysis (mainly of South American species) with the objective of clarifying its relation with the cerithioideans. Houbrick (1981a, 1988, 1989) and Haszprunar (1988) cast doubts on its systematic position. This allowed the separation in its own superfamily (Houbrick, 1989), but the present analysis reveals that Campanile can be perfectly regarded as cerithioidean, inclusive with some close relation to vermetids and turritellids. Some of the polemic characters in Campanile are also found in other groups of never questioned cerithioideans, like, e.g., the pectinate osphradium (node 12) and the double duet to digestive gland (Serpulorbis). The inclusion of Campanile within the Cerithioidea agrees also with results on paraspermatozoa (Healy, 1986) (see also Ponder & Waren, 1988). In the consensus tree obtained by Ponder & Lindberg (1997:182), Campanilidae occupies a position between the Cerithidae and remain caenogastropods. That position, in certain point of view, is compatible with the result obtained herein.

Node 4. The close relation between turritellids and vermetids was expected, and also found by Houbrick (1988). From 6, 2 important

synapomorphies unitę both groups: the food groove> (38) and the tali gili filaments (43).

Node 5. This branch unites those species with dorso-ventrally flattened snout (9) and a large central pad in stornach (93), among 7 synapomorphies.

Node 12. There is a problem in choosing C. atratum as the single representative of genus Cerithium. This taxon is a very heterogeneous group, with monophyly uncertain (Houbrick, 1992a). Surely when morę Cerithium will be known in morę detail, several speeies attributed to this genus probably will be replaced. From 4, an I important synapomorphy unites the species of this i branch: the peetinate and ridge-like (in aoverview) I osphradium (27).    I

Node 13. The miniaturized forms Bittium, Finella and Alaba are united by 171 synapomorphies, whieh indieated morę than single eonvergence due to miniaturization. Some eharacters, such as satellite osphradium fold (33), rectum narrow (48), crescentic ridge similar to een-tral pad (96) and posterior furrow of pedał sole (14), I among others, are notable synapomorphies.

Node 15. The union between planaxids and thiarids is not a surprise, already Morrison (1954) pointed out the close relationship of both groups under the light of the presence of a brood pouch. In the present analysis, the brood pouch of both groups, were not considered homologues, based on | data discussed above. Even so, both groups were maintained united due the remainder 6 characters. The genera of planaxids were analyzed phylogeneticly by Houbrick (1987a), dispensing any complementary comment. Faunus_y like other Melanopsinae, has been related to Thiaridae (Houbrick, 199 lb), at least the operculum has some similarity in being eccentric nucleate (Houbrick, 1991 b: 40-41, fig. 16). On the other hand, they are gonochoristic, lack tentacles in mantle border and lack brood pouch, which indicates that the relationship of these groups merits revaluation. Some different arrangement of the Thiaridae is presented by Starobogatov & lzzatullaev (1980), based mainly on pallial gonoduets. Simulathena papuensis Houbrick, 1992b, described in Planaxidae, appears to be better placed in Thiaridae in spite of marinę habitat, sińce presents tentacles in mantle border and dorsal placed brood pouch.