Bacterial symbionts that resembled mollicutes were discovered in the marine bryozoan Watersipora arcuata in the 1980s. In this study, we used PCR and sequencing of 16S rRNA genes, specific fluorescence in situ hybridization, and phylogenetic analysis to determine that the bacterial symbionts of "W. subtorquata" and "W. arcuata" from several locations along the California coast are actually closely related ␣-Proteobacteria, not mollicutes. We propose the names "Candidatus Endowatersipora palomitas" and "Candidatus Endowatersipora rubus" for the symbionts of "W. subtorquata" and "W. arcuata," respectively.Bacterial symbionts of bryozoans in the genus Watersipora were originally reported in 1983 (38). The bacteria were consistently observed in a surface groove with an overhanging flange unique to Watersipora larvae (38). Inside adult colonies of Watersipora, the bacterial symbionts were found in clusters surrounded by host mesodermal cells (38). The bacterial symbionts were present in Watersipora larvae immediately upon their release into the water column, indicating vertical transmission of the bacteria, although transmission has not been observed directly. With their irregular shape and apparent lack of a cell wall in transmission electron micrographs, the bacterial symbionts resembled mollicutes and were called mycoplasma-like organisms (38). In 1989, an Acholeplasma sp., a type of mollicute thought to be the symbiont, was isolated from Watersipora larvae on mycoplasma agar amended with filter-sterilized adult Watersipora extract (5). However, the molecular techniques necessary to confirm that the isolated Acholeplasma sp. was actually the symbiont were unavailable at the time.Calcareous marine bryozoans in the genus Watersipora are found worldwide in the tropics and subtropics (31). Along the coast of California, they are common on floating docks, boat bottoms, and other sturdy substrates in bays and harbors and are considered successful invaders that probably arrived in the 1960s (3). The systematics of the genus Watersipora are unresolved (15, 29, 31), making classification of host samples to the species level problematic. This is illustrated by the first report of symbionts in Watersipora, which identified the bryozoan host as Watersipora cucullata (38). Later, the identification was revised to Watersipora arcuata (37). Because of the confusion in systematics, it is critical to keep thorough records of the Watersipora samples, including host gene DNA sequences, light micrographs, and in some cases, scanning electron micrographs (SEMs), in order to study this symbiosis.Examples of associations between bryozoans and bacteria are abundant (2,24,36,38). In most cases, the roles of the bacteria in the lives of their host bryozoans are unknown. The symbiosis between the ␥-proteobacterium "Candidatus Endobugula sertula" and the bryozoan Bugula neritina is an exception for which there is extensive evidence indicating that "Candidatus Endobugula sertula" is the source of the bryostatins, a family of polyketides which...