Phylogenetic position of Sipuncula derived from multi-gene and phylogenomic data and its implication for the evolution of segmentation

The phylogenetic position of Sipuncula, a group of unsegmented marine worms, has been controversial for several decades: Especially based on morphological data, closer relationships to Mollusca or Annelida were among the most favoured hypotheses. Increasing amounts of molecular data in recent years have consistently placed Sipuncula either in close affinity to or even within Annelida, the segmented worms, and rejected a close relationship to Mollusca. Yet, it remained uncertain whether Sipuncula is the sister group of Annelida or an annelid subtaxon. Therefore, herein we gathered data for five nuclear genes, which have been rarely used regarding Annelida and Sipuncula, and combined these with data for six previously used genes to further elucidate the phylogenetic position of Sipuncula. We also compiled a data set for 78 ribosomal proteins from publicly available genomic data sets. These are the two largest data sets for annelids with more than 10 taxa to date. All analyses placed Sipuncula within Annelida. For the first time, topology tests significantly rejected the possibility that Sipuncula is sister to Annelida. Thus, our analyses revealed that Sipuncula had secondarily lost segmentation. Given that unsegmented Echiura is also an annelid subtaxon, segmentation, a key character of Annelida, is much more variable than previously thought. Yet, this conclusion does not support the hypothesis that the last common ancestor of Annelida, Arthropoda and Chordata was segmented, assuming several losses along the branches leading to them. As yet no traces of segmentation could be shown in taxa exhibiting serially organized organ systems such as certain Mollusca, while in Sipuncula and Echiura such traces could be demonstrated. An independent origin of segmentation in Annelida, Arthropoda and Chordata thus appears to be more plausible and parsimonious.