Elucidating the phylogenetic position of Gnathostomulida and first mitochondrial genomes of Gnathostomulida, Gastrotricha and Polycladida (Platyhelminthes).

Gnathostomulida is a taxon of small marine worms, which exclusively inhabit the interstitium. The evolution of Gnathostomulida has been discussed for decades. Originally regarded as primitive animals with affinities to flatworms, the phylogenetic position of Gnathostomulida has been debated. Given the lack of an anus a close relationship to Platyhelminthes has been maintained (i.e., Plathelminthomorpha hypothesis). Alternative hypotheses proposed Gnathostomulida as being close to Gastrotricha due to the presence of a monociliary epidermis (i.e., Monokonta/Neotrichozoa hypothesis) or to Syndermata based on the complicated jaw apparatus (i.e., Gnathifera hypothesis). Molecular analyses using only few genes were inconclusive. Recent phylogenomic studies brought some progress by placing Gnathostomulida as sister to Syndermata, but support for this relationship was low and depended on the analytical strategy. Herein we present the first data of complete or nearly complete mitochondrial genomes for two gnathostomulids (Gnathostomula paradoxa &G. armata), one gastrotrich (Lepidodermella squamata) and one polyclad flatworm (Stylochoplana maculata) to address the uncertain phylogenetic affinity of Gnathostomulida. Our analyses found Gnathostomulida as sister to Syndermata (Gnathifera hypothesis). Thorough sensitivity analyses addressing taxon instability, branch length heterogeneity (also known as long branch attraction) and base composition heterogeneity showed that the position of Gnathostomulida is consistent across the different analyses and, hence, independent of potential misleading biases. Moreover, by ameliorating these different biases nodal support values could be increased to maximum values. Thus, our data support the hypothesis that the different jaw apparatuses of Syndermata and Gnathostomulida are indeed homologous structures as proposed by the Gnathifera hypothesis.