AbstractWhether speciation can happen in the absence of geographical barriers and if so, under which conditions, is a fundamental question in our understanding of the evolution of new species. Among candidates for sympatric speciation, Cameroon crater lake cichlid radiations have been considered the most compelling. However, it was recently shown that a more complex scenario than a single colonization followed by isolation underlies these radiations. Here, we perform a detailed investigation of the speciation history of a radiation of Coptodon cichlids from Lake Ejagham using whole-genome sequencing data. The existence of the Lake Ejagham Coptodon radiation is remarkable since this 0.5 km2 lake offers limited scope for divergence across a shallow depth gradient, disruptive selection is weak, the species are sexually monochromatic, yet assortative mating is strong. We infer that Lake Ejagham was colonized by Coptodon cichlids almost as soon as it came into existence 9,000 years ago, yet speciation events occurred only in the last 1,000-2,000 years. We show that secondary gene flow from a nearby riverine species has been ongoing, into ancestral as well as extant Lake Ejagham lineages, and we identify and date river-to-lake admixture blocks. One of these contains a cluster of olfactory receptor genes that introgressed close to the time of the first speciation event and coincides with a higher overall rate of admixture into the recipient lineages. Olfactory signaling is a key component of mate choice and species recognition in cichlids. A functional role for this introgression event is consistent with previous findings that assortative mating appears much stronger than ecological divergence in Ejagham Coptodon. We conclude that speciation in this radiation took place in sympatry, yet may have benefited from ongoing riverine gene flow.Author SummaryDespite an active search for empirical examples and much theoretical work, sympatric speciation remains one of the most controversial ideas in evolutionary biology. While a host of examples have been described in the last few decades, more recent results have shown that several of the most convincing systems have not evolved in complete isolation from allopatric populations after all. By itself, documenting the occurrence of secondary gene flow is not sufficient to reject the hypothesis of sympatric speciation, since speciation can still be considered sympatric if gene flow did not contribute significantly to the build-up of reproductive isolation. One way forward is to use genomic data to infer where, when and into which lineages gene flow occurred, and identify the regions of the genome that experienced admixture. In this study, we use whole-genome sequencing to examine one of the cichlid radiations from a small isolated Cameroon lake, which have long been the flagship example of sympatric speciation. We show that gene flow from a riverine species into the lake has been ongoing during the history of the radiation. In line with this, we infer that the lake was colonized very soon after it was formed, and argue that Lake Ejagham is not as isolated as previously assumed. The magnitude of secondary gene flow was relatively even across Lake Ejagham lineages, yet with some evidence for differential admixture, most notably before the first speciation event into the C. deckerti and C. ejagham lineage. Among the sequences that were introgressed into this lineage is a cluster of olfactory receptor genes, which may have facilitated speciation by promoting sexual isolation between incipient species, consistent with previous findings that sexual isolation appears to be stronger than ecological isolation in Ejagham Coptodon. We conclude that speciation in this radiation took place in sympatry, yet may have benefited from ongoing riverine gene flow.
Complex histories of repeated gene flow in Cameroon crater lake cichlids cast doubt on one of the clearest examples of sympatric speciation