Which mitochondrial gene (if any) is best for insect phylogenetics?

Abstract

Despite the benefits of whole-genome analysis, in many instances — particularly in applied entomology — mitochondrial genes continue to offer a reliable, rapid and cheap alternative. To date, most studies using insect mitochondrial DNA have analysed single genes and none have rigorously attempted to assess which genes are best suited for studying particular insect orders; here, we address this issue and use the ability of individual genes to recover ordinal monophyly of various insect orders as a metric. Phylogenies were constructed for nine insect orders and three outgroups, using 12 protein-coding genes and two rRNA genes; 153 genomes were analysed and trees were constructed using PhyML. The importance of gene length and region within the mtDNA genome were explored using correlation and sliding window analyses. No single gene appeared to outperform all others. Accordingly, we recommend that mitochondrial-based reconstructions of insect relationships use a multi-gene approach, using as many genes and taxa as possible.