[Original adaptive characters of intestinal Digenea of Sarpa salpa (Teleostei, Sparidae) and their interpretation in terms of evolution].

In the family Sparidae, the genus Sarpa is distinguished by a few characteristics: monospecificity, vegetarian diet and wide geographical distribution. The helminth fauna of Sarpa salpa is also very original. Indeed, the digenean parasites of this Teleostean fish are essentially classified into two families restricted to this fish. In the present paper, the author redescribes Mesometra orbicularis, M. brachycoelia, Centroderma spinosissima, Elstia stossichianum, Wardula capitellata (family Mesometridae) together with Robphildollfusium fractum (family Robphildollfusidae). Various original and yet unknown features are pointed out. Among these unusual structures, several correspond to adaptive characteristics favouring the settlement of the Digenean on the peculiar digestive gut wall of this herbivorous fish. Indeed, the intestinal mucous membrane of Sarpa salpa exhibits very few villi giving it an unusual smooth aspect. Therefore, the Mesometridae which always have just a single sucker (monostomatous) have selected a new kind of compensatory adhesive structure. Sometimes, the anterior end of the body becomes a sucker due to the particular distribution of the muscle strings; in other examples, the whole body becomes a sucker and its edges become considerably thinner to improve the tightness of the adhesive system. Other original anatomical features have been selected to allow survival in a medium rich in plant detritus. So, in the oral sucker crests ornemented by numerous sclerous denticles seem to act as a microfilter for the intestinal chyme in which plant fibres predominate. The original pharynx seems to act as a suction-force pump. The excretory system, which is of a reticular type, penetrates the whole parenchyma and this could be a response to huge intestinal fermentations. The Digenea of Sarpa salpa are not interpreted by the author as true parasites but as endocommensal symbionts. These inquiline species are not immunogenic, or at least only slightly so, since they do not feed upon the host itself but upon its intestinal chyme. In most cases this results in a high parasite density (post larvae and adults) together with a cohabitation of the various species along the various intestinal segments. Coexistence of several species, systematically very close, evidently raises the question of their reproductive isolation. The author proposes an answer founded upon data of allopatric speciation.