HIGHER-LEVEL SYSTEMATICS OF THE MILLIPEDES (ARTHROPODA: MYRIAPODA: DIPLOPODA) EVALUATED USING TRANSCRIPTOMIC DATA, MITOCHONDRIAL GENOMICS, AND CLASSIFICATION ANALYSES

Abstract

Knowledge of deep level phylogenetic relationships within and between many animal groups is currently lacking. Past attempts to reconstruct these ancient evolutionary relationships once relied chiefly on morphological characters and consequently suffered from deficiencies inherent to these data (homoplasy, few diagnostic characters, etc.). Advances in molecular biology over the last 20 years (namely PCR and high-throughput sequencing) have provided a whole other facet of potential characters via a vast array of loci, both nuclear and mitochondrial, which represent unlinked regions spanning entire genomes. The class Diplopoda (the millipedes) is a diverse group, 12,000 described species, which has suffered a long and convoluted taxonomic and systematic history. Containing many overinflated groups (numerous monotypic taxa) while other groups are vastly understudied, the current scheme of classification does not reflect the evolutionary history of the group but instead exists primarily for identification purposes. A classification constructed in such a manner can be characterized as an "unnatural classification". This work represents an attempt to: 1) ascertain the uniformity of millipede taxa at the ordinal level by evaluating species diversity contained in higher taxa (orders, families, genera); 2) achieve a robust and empirically derived estimate of millipede global diversity; 3) reconstruct the higher level relationships among all major millipede taxa using genomic scale molecular data; 4) map the ancestral states of key millipede characteristics; and 5) date the divergences of major diplopod lineages. Full mitochondrial genomes and transcriptomic data obtained via second-generation sequencing for exemplar taxa representing the ordinal taxa were used to reconstruct the relationships between the higher millipede taxa. Our work uncovered interesting and potentially alarming trends in millipede taxonomy. Additionally, we demonstrated the utility, or lack thereof, of using mitochondrial sequence data to reconstruct deep evolutionary relationships. Transcriptomic data analyses yielded well- supported trees containing relationships that have not been suggested previously. The transcriptome-based phylogeny was used to reconstruct the evolutionary changes of key millipede characteristics and, consequently, to assess their utility in delineating millipede taxa. This information provides a basis for future work within the Diplopoda including but not limited to: character evolution, biogeography, divergence time estimations, and comparative genomics.