LAUSR

Background The diversification process known as the Lake Tanganyika Radiation has given rise to the most speciose clade of African cichlids. Almost all cichlid species found in the lakes Tanganyika, Malawi and Victoria, comprising a total of 12–16 tribes, belong to this clade. Strikingly, all the species in the latter two lakes are members of the tribe Haplochromini, whose origin remains unclear. The ‘out of Tanganyika’ hypothesis argues that the Haplochromini emerged simultaneously with other cichlid tribes and lineages in Lake Tanganyika, presumably about 5–6 million years ago (MYA), and that their presence in the lakes Malawi and Victoria and elsewhere in Africa today is due to later migrations. In contrast, the ‘melting pot Tanganyika hypothesis’ postulates that Haplochromini emerged in Africa prior to the formation of Lake Tanganyika, and that their divergence could have begun about 17 MYA. Haplochromine fossils could potentially resolve this debate, but such fossils are extremely rare. Results Here we present a new fossil haplochromine from the upper Miocene site Waril (9–10 million years) in Central Kenya. Comparative morphology, supported by Micro-CT imaging, reveals that it bears a unique combination of characters relating to dentition, cranial bones, caudal skeleton and meristic traits. Its most prominent feature is the presence of exclusively unicuspid teeth, with canines in the outer tooth row. † Warilochromis unicuspidatus gen. et sp. nov. shares this combination of characters solely with members of the Haplochromini and its lacrimal morphology indicates a possible relation to the riverine genus Pseudocrenilabrus . Due to its fang-like dentition and non-fusiform body, † W. unicuspidatus gen. et sp. nov. might have employed either a sit-and-pursue or sit-and-wait hunting strategy, which has not been reported for any other fossil haplochromine cichlid. Conclusions The age of the fossil (9–10 MYA) is incompatible with the ‘out of Tanganyika’ hypothesis, which postulates that the divergence of the Haplochromini began only 5–6 MYA. The presence of this fossil in an upper Miocene palaeolake in the Central Kenya Rift, as well as its predatory lifestyle, indicate that Haplochromini were already an important component of freshwater drainages in East Africa at that time.