Abstract:
The Protobranchia, with a stomach of type I (Gastroproteia), comprise one of only two subclasses of Bivalvia. The stomachs of members of the deposit-feeding Nucu acea and Nuculanacea are basically comparable and differ from those of all filter-feeding bivalves in some important respects. There are only three slender ducts from the digestive diverticula (Oligosyringia). There is no clear crystalline style, but the wide proximal expansion of the mid gut houses an amorphous, viscous secretion in lieu. The intestinal groove does not obviously extend forwards from the mid gut into the stomach. A gastric shield and a dorsal hood are present, but there are no other stomach caeca. Any sorting areas present differ in disposition from those in filter-feeding bivalves. In the Solemyacea the alimentary canal including the stomach is much reduced and simplified, and in one species is wholly lacking. In the subclass of filter-feeding bivalves there are many ducts from the digestive diverticula into the stomach (Polysyringia) and the intestinal groove extends forwards from the mid gut into the stomach, protected throughout its course by the overhanging major typhlosole. Six of the superfamilies in the Pteriomorpha have stomachs of type III, in which the diagnostic feature is a long slender tongue of the major typhlosole which penetrates the food-sorting caecum to its apex, accompanied throughout by the intestinal groove. This feature is so remarkable that it cannot have evolved more than once and it must indicate that these six superfamilies have a common source. Other possible criteria for the clustering of superfamilies in the Pteriomorpha have been considered individually and all have been rejected for reasons given. The superfamilies Arcacea, Limopsacea, Mytilacea, Pinnacea, Pteriacea and Ostreacea, all of which possess a stomach of type III, arejudged to compose a natural assemblage, the Gastrotriteia. The review of the Gastrotriteia reports upon 27 genera assembled in ten families and six superfamilies. Two species, Trisidos semitorta and Pteria penguin are described in detail. A type B sorting mechanism, which links the apex of the typhlosolar tongue with the apex of the dorsal hood, has been found in two families, the Pteriidae and Mytilidae, and this implies phyletic affinity between these families. Tabulation of variation in the internal structure of the stomach in this group fails to reveal any other subordinate clustering. Extensive sorting areas may be present or absent, and the duct orifices may be arranged in an extended series or may form clusters, and such variation is best interpreted as random. Stomach type IV is characterized by a major typhlosole and intestinal groove which emerge from the mid gut and curve evenly to the left across the stomach floor, without production of a typhlosolar tongue and without entering into an embayment on the right anterior side of the stomach. Thirteen superfamilies exhibit this type of stomach, and analysis of structural variation shows that these may be arranged in three groups. (i) In the Pectinacea, Limacea and Anomiacea there are many duct orifices, either scattered or clustered, the major typhlosole passes towards the left pouch, and there is a conspicuous sorting area on the anterior floor of the stomach. (ii) In the Pandoracea and Clavagellacea the duct orifices are concentrated into a few embayments of the stomach wall, the major typhlosole and intestinal groove are short and either very posterior in position or pass towards the left caecum (Laternulidae). There is a conspicuous anterior sorting area on the stomach floor between the oesophageal orifice and the intestinal groove. (iii) In the remaining eight superfamilies there is not an extensive anterior sorting area on the stomach floor. The major typhlosole and intestinal groove pass towards the left caecum (Trigonacea, Crassatellacea, Carditacea, Hiatellacea and Pholadomyacea), or penetrate the left caecum (Unionacea, Leptonacea and Gastrochaenacea). The last three ol these superfamilies are intermediate in this respect between stomach types IV and V, and it is possible that one or more of these superfamilies may have reverted from stomach type V to type IV. Finally there are some taxa which have most certainly reverted from stomach type V to type IV. This has been clearly demonstrated in the Lucinacea (in the families Lucinidae and Thyasiridae), the Corbiculacea (for example in Sphaerium corneum ), the Chamacea (in Chama multisquamosa )and the Tellinacea (in Donax faba and D. semigranosus ). These changes can be variously explained as simplification of the stomach permitting opportunism in feeding in nutritionally poor habitats (Lucinidae) ; as the outcome of adoption of small size as a strategy for survival ( S. corneum ) ; as paedomorphosis ( D. semigranosus ). Consideration should be given to the possibility of other cases of reversion from stomach type V to type IV, especially among taxa with greatly simplified stomachs, and with a small body size. Of the 13 superfamilies that exhibit a stomach of type IV, the eight that are listed in group (iii), above, should be set on one side for the present time in view of the possibility that further examples of reversion from a stomach of type V may yet be identified. The remaining five superfamilies, the Pectinacea, Limacea, Anomiacea, Pandoracea and Clavagellacea compose a distinct phylogeny of bivalves, the Gastrotetartika. Within this restricted group the review reports on 14 genera assembled in nine families and five superfamilies. One species, Chlamys crassicostata is described in detail. Among the remaining eight superfamilies the review treats 20 genera assigned to 15 families, while two species, Velesunio ambiguus and Aspatharia brumpti , are described in detail. Within the Pholadomyacea, the genus Parilimya has a stomach which is in a state of transition between stomach type IV and type II. Phis provides valuable support to the proposition that the septibranchs have evolved from anomalodesmatan ancestors taxonomically close to the Verticordiidae. The septibranchs are carnivorous and have stomachs of type II (G astrodeutera); they include the families Verticordiidae, Poromyidae and Cuspidariidae. In each of the verticordian genera Halicardia , Policordia and Lyonsiella individual species show different degrees of specialization towards the carnivorous habit, those least specialized having stomachs that still possess a dorsal hood, a small sorting area of ciliated folds, a small major typhlosole and an intestinal groove. In the more highly specialized carnivorous taxa all these structures are lacking, and the stomach forms a powerful crushing gizzard with a thicker outer muscular coat and a thicker inner lining of scleroprotein. There are only two slender ducts from the digestive diverticula into the stomach (Oligosyringia). These data support the concept of transition from a delicate ciliary microphagous habit to a convulsive muscular carnivorous habit without loss of efficiency during an intermediate opportunistic omnivorous phase. Although the septibranchs are polyphyletic as indicated by the very different feeding methods of the Verticordiidae and Cuspidariidae, they may nevertheless have originated from an even earlier common ancestry in the Anomalodesmata. Although this would place the septibranch families as components of the Anomalodesmata, their extreme structural and functional specialisms mayjustify the award of independent taxonomic status at a level equivalent to each of the three phylogenies within the filter-feeding bivalves. Review ol the Gastrodeutera reports upon seven genera of septibranchs, assigned to three families placed in either one, or two, superfamilies. Recognition of stomach type V is simple, being dependent upon the presence of a semicircular or elongated flange of the major typhlosole, accompanied by an extension of the intestinal groove, which projects into the right caecum. As regards the large assemblage of bivalves which possess this type of stomach (Gastropempta), tabulation of details of internal structure has failed to indicate any natural subdivisions. This paper reports upon the stomachs of 76 genera, assigned to 24 families and to 14 superfamilies. To this phylogeny should be added three further families, the Lucinidae and Thyasiridae (Lucinacea) and the Pisidiidae (Corbiculacea), and a few species in the genera Donax , Tellina and Chama in which the stomach has reverted from type V to type IV. The internal structure of all stomachs of types IV and V was therefore examined in detail to define more thoroughly the differences between these two types of stomach, with a view to identifying any further taxa which may have reverted from type V to type IV. It was found that the ultimate destination of the major typhlosole and intestinal groove in the stomach was within the left caecum in all examples of stomach type V, and also in the Unionacea, Kellndae (Leptonacea) and Gastrochaenacea, but in no other taxa with a stomach of type IV. 1 his suggests that the Unionacea, Leptonacea and Gastrochaenacea are either transitional between types IV and V, or may have partly reverted from type V to type IV. Stomach types IV and V differ as regards the distribution of the sorting areas. In stomachs of type IV sorting areas 1, 3 and 8 are common, numbers 6 and 7 are occasional, numbers 4 and 5 are rare. In stomachs of type V sorting areas numbers 3 and 6 are almost universal, numbers 5, 7 and 8 are common, number 1 is rare and number 4 is absent. Other variations in internal structure of the stomac are without taxonomic significance. These concern the size of the oesophageal orifice, t e size and location of the gastric shield, the ciliated pathway from the oesophageal orifice to the dorsal hood, and the relationship between the style sac and the mid gut. Advantage is taken of this further information to supply more extensive definitions of stomach types III, IV and V. Attention is drawn to a curved flange that lies on the posterior floor of the stomach in certain members of the Leptonacea and of the Tellinacea, and that might indicate a phyletic affinity between these superfamilies. There are a number of similarities between the stomachs of Glauconme (Veneracea) and Geloina (Corbiculacea), and also between Petricola (Veneracea) and Trapezium (Arcticacea). A conical mound on the floor of the stomach appears to be diagnostic of families in the Unionacea. Attention is drawn to the need for further studies on the stomach in species of Chama .