We explored the phylogenetic utility of mitochondrial DNA sequences in Daucus and compared the results with prior phylogenetic results using the same 36 accessions of Daucus (and two additional outgroups) with plastid DNA sequences and with other nuclear results. As in the plastid study we used Illumina HiSeq sequencer to obtain resequencing data of the same accessions of Daucus and outgroups, and analyzed the data with maximum parsimony and maximum likelihood. We obtained data from 47 of 71 total mitochondrial genes but only 17 of these 47 genes recovered major clades that were common in prior plastid and nuclear studies. Our phylogenetic trees of the concatenated data set of 47 genes were moderately resolved, with 100% bootstrap support for most of the external and many of the internal clades, except for the clade of D. carota and its most closely related species D. syrticus. There are areas of hard incongruence with phylogenies using plastid and nuclear data. In agreement with other studies, we conclude that mitochondrial sequences are generally poor phylogenetic markers, at least at the genus level, despite their utility in some other studies.

Aligned sequences from 47 mitochondrial genesAligned sequences from 47 mitochondrial genes for phylogenetic analysis. Sequences also available as GenBank accessions NC_017855 and MN260368 through MN262020aligned_sequences.nexchecksum for aligned_sequences.nexMD5 checksum for nexus filealigned_sequences.nex.md5