Looks can be deceiving: discordances in phylogeny and morphology within loricate choanoflagellates

Choanoflagellates are heterotrophic holozoans that are classified into two groups based on their morphology: loricates, which possess a silica-based extracellular structure, and craspedids, which do not. Although the craspedid versus loricate morphological separation is currently supported by their phylogenetic relationship, recent evidence has suggested inconsistencies between morphology and phylogeny within each group. Loricate choanoflagellate taxonomy has historically been based on selected aspects of their lorica morphology, and on their mode of cell division, in which tectiform daughter cells emerge into a lorica synthesized by their mother cell following division, and nudiform daughter cells do not. Here, we characterize two new loricate strains that display unexpected morphological features when compared to their nearest genetic relatives. The strain BEAP0094 very closely matched the 18S ribosomal gene of the tectiform Pseudostephanoeca paucicostata , but its morphology clearly differed, due to the absence of the characteristic anterior ring found in all Stephanoeca species. Instead, its features resembled more closely those of the Acanthocorbis genus, raising the possibility of the existence of either multiple lorica morphologies within the same or very closely related species, or multiple morphological species sharing the same 18S ribosomal gene. The second strain we investigated, BEAP0360, presented a morphological match to Stephanoeca cauliculata , but its 18S ribosomal sequence did not, suggesting that different species could share the same lorica architecture. BEAP0360, here described as Cepoeca plumat a (n. gen. n. sp.), possesses a key phylogenetic placement, potentially as the earliest branching member within nudiform loricates, which would be informative for investigating the evolution of the nudiform lifestyle. Our findings are inconsistent with a strict classification based on currently defined aspects of lorica morphology and support the usage of genetic data as primary criterion for genus-level taxonomic assignment.