Molecular sequence analysis has become the method of choice to address phylogenetic questions. The applied techniques improve continually and the rapidly growing amount of available data helps to broaden our knowledge of phylogenetic relationships within the animal kingdom. Nevertheless, different molecular datasets often show conflicting phylogenetic signals, so that results relying on just one dataset may be interpreted with caution . Unlike nuclear DNA, the mt-genome of animals is normally rather small and simply structured: haploid, without or only few non-coding segments, repetitive regions and transposable elements. Derived from endosymbiotic bacteria only a few genes are retained in the mitochondrial genomes of Bilateria: 13 protein subunits (nad1-6, nad4L, cox1-3, cob, atp6/8), 2 ribosomal RNAs (rrnL, rrnS) and 22 tRNAs are found encoded on a circular doublestranded DNA molecule sized about 15 kb [2, 3]. As such sequencing and annotation of mt-genomes is much easier and faster than analysing nuclear genomes, making mt-genomes one of the commonly used sources of sequence data for phylogenetic analyses. Apart from sequence data other features of the genome may contain phylogenetic information, too. Taxon-specific gene order often remains identical over long periods of time [4–6]. Simultaneously, the intra-taxonomic variances of these characteristic orders are quite distinctive and convergent changes in the positioning of single genes are rather unlikely, due to the vast number of possible combinations . Thus changes in the mitochondrial gene order have proved to be valuable tools in phylogenetic analyses [8–10]. Less often secondary structures of tRNAs or rRNAs show distinct differences between taxa (e.g. loss of a stem/loop region) and hence may also contribute to a phylogenetic analysis .
The taxon Sipuncula (peanut worms) comprises about 150 species, being found in all water depths of different marine habitats. The hemisessile organisms dwell in mud and sand, but settle also in empty mollusc shells or coral reef clefts for instance. Their body shows no segmentation, but a subdivision into a posterior trunk and an anterior introvert that can be fully retracted into the trunk is observeable . Fossils that date back into the later cambrian  suggest that sipunculans have undergone little morphologically change over the past 520 Myr. The monophyly of this morphologically uniform taxon is well founded by morphological  and molecular data . However, the phylogenetic position within Bilateria was highly disputed. Based on morphological characters, very different phylogenetic positions of Sipuncula were discussed. Early in history an affinity to Echinodermata, especially holothurians was mentioned and later again propagated by Nichols , but with little acceptance from other authors. Scheltema  proposed a close relationship to molluscs based on the presence of the so calles "molluscan cross" organization of micromeres during spiral cleavage. The usefulness of this character for phylogenetic inference was neglected by Malaskova . Other analyses found Sipuncula to be sister group of Mollusca, Annelida and Arthropoda , Articulata (Annelida and Arthropoda) , Echiura , Mollusca , Annelida  or Annelida+Echiura . More details about the different hypotheses of sipunculid relationships are reviewed in .
In contrast to all these studies, molecular analyses of large datasets from 18S/28S data , ESTs [26, 27] or mitochondrial genome data [28, 29] favour an inclusion of Sipuncula into annelids. An implication of this hypothesis is that we have to assume that segmentation has been reduced within Sipuncula . A derivation from segmented ancestors of Sipuncula was recently also supported by a segmental mode of neural patterning in ontogeny .
Relationhips within Sipuncula are well investigated [15, 24, 32, 33, 34]. An analysis using combined molecular and morphological data recovered five major clades and supports that Sipunculus is the sister group to all other sipunculids .
Up to now mt-genome data from Sipuncula was restricted to a partial mtDNA sequence from Phascolosoma gouldii , comprising only about half of the complete genome. Here we describe the first complete mitochondrial genome for another representative of the Sipuncula, Sipunculus nudus. We analyse sequence data in comparison with mitochondrial genomes of various Bilateria to evaluate the phylogenetic position of Sipuncula. In addition we compare gene order among Lophotrochozoa and evaluate the most parsimonious explanation for gene order changes.