Erosion of phylogenetic signal in tunicate mitochondrial genomes on different levels of analysis.

The molecular phylogenetic position of Tunicata and internal interrelationship of higher tunicate taxa is controversial. High substitution rates and extreme gene order variability hamper phylogenetic analyses. We describe the sequence and organization of the mitochondrial genome of the aplousobranch ascidian Clavelina lepadiformis and use mitochondrial genomes to investigate phylogenetic information content on different molecular levels of comparison. Despite agreement in phylogenetic analyses of nucleotide and amino acid sequences, split analyses revealed little phylogenetic signal. Split analyses on molecular data sets deemed increasingly conservative, demonstrated that the lack of signal pervades all levels and that it is Tunicata the taxon of interest that introduces noise in the data sets. The strongest signal present in our molecular data sets as revealed by split analyses is not present in the optimal cladograms and supports a sister group relationship between cephalochordates and craniates. Phylogenetic analysis of gene order using common interval algorithms shows that phylogenetic signal is also eroded in respect of gene positions. Even functional constraints, such as partial gene overlap as exemplified in the case of the commonly observed adjacency between cox2 and cytb are subjected to homoplasy. However, rare phylogenetic events like this hold some promise to retain phylogenetic information even in such cases of extreme variability. We therefore caution to rely on sequence analysis alone and recommend investigation into the signal content of molecular data sets in order to assess the strength of phylogenetic signal.