Hydrobiologia 498: 1–83, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands. 1 The late Cenozoic Thiaridae (Mollusca, Gastropoda, Cerithioidea) of the Albertine Rift Valley (Uganda-Congo) and their bearing on the origin and evolution of the Tanganyikan thalassoid malacofauna Dirk Van Damme 1 & Martin Pickford 2 1 Laboratorium voor Paleontologie, Geologische Instituut, Krijgslaan 281, S8; B-9000 Ghent, Belgium E-mail: dirkvandamme2001@yahoo.com 2 Coll` ege de France, Chaire de Pal´ eoanthropologie et de Pr´ ehistoire, 11 place M. Berthelot, F-75005, Paris, and Laboratoire de Pal´ eontologie, UMR 8569 du CNRS, 8 rue Buffon, 75005, Paris, France E-mail: pickford@mnhn.fr Received 28 June 2002; in revised form 28 March 2003; accepted 28 March 2003 Key words: relict faunas, escalation, prey-predator coevolution, convergence, thalassoidism, iterative evolution, punctuated equilibrium, ancient rift lakes, bottle-necking, Melanoides, Potadoma, Potadomoides, Cleopatra, Pseudocleopatra, Heynderycxia gen. nov., Ellinoria gen. nov., Platymelania, Uganda, Congo, Lake Tanganyika Abstract The recent cerithioid malacofauna of meromictic Lake Tanganyika is unique in its degree of thalassoid convergence with marine molluscs. This is generally considered the result of a long-lasting intensive escalation and cladogenesis caused by a coevolutionary prey–predator interaction in a freshwater ecosystem with sea-like characteristics, i.e. exceptional longevity (ca. 7–12 Ma) and vast dimensions (present surface area: 32 900 km 2 , present maximum depth: 1470 m). In the Albertine Basin, ca 300 km north of the Tanganyika Trough, Palaeolake Obweruka existed during Mio-Pliocene times. In many aspects, it can be considered as a ’sister lake’ of Lake Tanganyika, being also long-lived (from ca 7.5 to 2.5 Ma), extensive (surface: 27 000 km 2 ) and meromictic. Like Lake Tanganyika it belonged to the Congo catchment. Although thalassoid molluscs have been known from the Albertine deposits since the beginning of the 20th Century, previous researchers recognised only four polymorphic thiarid species, two of which were thalassoid. Detailed in situ collecting shows that the perceived low species diversity and high variability of the Obwerukan thalassoids is a phenomenon of the museum drawer. In the field the thiarid species and other molluscs are not extremely polymorphic, as formerly thought, but tend to occur in discrete morphometric packages according to stratigraphic level and geographic area. The species richness of the Palaeolake Obweruka thiarids has thus been severely underestimated, as is also the case for the Recent Tanganyikan thiarids. The present authors discern 35 species and 7 genera in the Albertine deposits (2 of which are new to science), and consider this to be a conservative estimate. In their degree of diversification and morphological escalation, the Obwerukan thiarids are thus comparable to the Tanganyika thalassoids. Arguments are presented that the thalassoids of these lakes are polyphyletic, that they are derived from the same genera (mainly Potadoma, Potadomoides and Pseudocleopatra) and that the tempo and mode of their intralacustrine evolution must have been largely similar. As for the tempo of evolution, the Albertine fossil record shows that escalation in shell morphology and ornamentation in the Thiaridae was extremely rapid, as in the Ampullariidae and Viviparidae (Van Damme & Pickford, 1995, 1999). In the Thiaridae heavily ornate shells evolved from an unornamented ancestor in two consecutive bursts, each lasting less than 0.1 million years over a total period of ca. 1 million years. The punctuated escalation in the thiarids was initiated millions of years after the lake was formed. This implies that the degree of escalation is not related to age of a lake and that the Tanganyikan thalassoids or part of them may not be particularly old, even though Lake Tanganyika itself is. As to the mode of evolution, it appears that escalation is a coevolutionary reaction to predation but that this selective pressure only triggers a reaction under specific conditions and in synergy with other environmental factors. The Obwerukan thiarid fossil record also indicates that iterative evolution of molluscan shells may be of frequent occurrence. What is considered to be a single polymorphic chronospecies, static over a great length of time, may actually be a sequence of several species that branched at different moments from the same ancestral phenotype. Finally, there is the inference that bottle-necking may be, geologically speaking, so common that it forms an essential force in macroevolution.