Phylogenetic analysis reveals an evolutionary transition from internal to external brooding in Epiactis Verrill (Cnidaria: Anthozoa: Actiniaria) and rejects the validity of the genus Cnidopus Carlgren

- Epiactis is a polyphyletic group, but Epiactis sensu stricto is monophyletic.
- Cnidopus is not a valid genus.
- External brooders have evolved from internally brooding ancestors in the North Pacific.
- Internal brooding has independently evolved more times than external brooding.
- Brooding and hermaphroditism are not strongly correlated.

Reproductive behaviors in the sea anemone genus Epiactis provide an opportunity for investigating the evolution of reproductive phenomena such as brooding and sex allocation (hermaphroditic vs. gonochoric) in a group of closely related and easily accessible species. However, given its broad geographic distribution, the striking diversity in reproductive behaviors, and the lack of synapomorphy for the genus, the monophyly of Epiactis is questionable. Here we perform phylogenetic analyses to test the monophyly of Epiactis and the validity of Cnidopus, which consists entirely of species once assigned to Epiactis. We use the large number of brooding species in Epiactis to investigate evolutionary patterns in brooding modes and characteristics associated with them. We find a monophyletic group of North Pacific Epiactis species: this group includes the type species of the genus and species that brood internally or externally, and that are hermaphroditic or gonochoric. Based on the results, we reject the genus Cnidopus because its circumscription renders Epiactis sensu stricto paraphyletic. Ancestral character state reconstruction indicates that in the North Pacific, externally brooding species evolved from internally brooding ancestors and that sex allocation is highly labile. Species relationships in Epiactis and Aulactinia appear to conform to geographic patterns more strongly than to taxonomic hypotheses. Contrary to expectations based on other invertebrates, we fail to find a strong correlation between brooding and hermaphroditism.

129