The millipede family Xystodesmidae includes 486 species distributed primarily in temperate deciduous forests in North America and East Asia. Species diversity of the family is greatest in the Appalachian Mountains of the eastern United States, with 188 species. Although the group includes notable taxa such as those that are bioluminescent and others that display Müllerian mimicry, producing up to 600 mg of cyanide, basic alpha-taxonomy of the group is woefully incomplete and more than 50 species remain undescribed in the Appalachian Mountains alone. In order to establish a robust phylogenetic foundation for addressing compelling evolutionary questions and describing species diversity, we assembled the largest species phylogeny (in terms of species sampling) to date in the Diplopoda. We sampled 49 genera (out of 57) and 247 of the species in the family Xystodesmidae, recollecting fresh material from historical type localities and discovering new species in unexplored regions. Here, we present a phylogeny of the family using six genes (four mitochondrial and two nuclear) and include pivotal taxa omitted from previous studies including Nannaria, Erdelyia, taxa from East Asia, and 10 new species. We show that 6 of the 11 tribes are monophyletic, and that the family is paraphyletic with respect to the Euryuridae and Eurymerodesmidae. Prior supraspecific classification is in part inconsistent with the phylogeny and convergent evolution has caused artificial genera to be proposed. Subspecific classification is likewise incongruent with phylogeny and subspecies are consistently not sister to conspecifics. The phylogeny is used as a basis to update the classification of the family, diagnose monophyletic groups, and to inform species hypotheses.

For the past several centuries, millipede taxonomists have used the morphology of male copulatory structures (modified legs called gonopods), which are strongly variable and suggestive of species-level differences, as a source to understand taxon relationships. Millipedes in the family Xystodesmidae are blind, dispersal-limited and have narrow habitat requirements. Therefore, geographical proximity may instead be a better predictor of evolutionary relationship than morphology, especially since gonopodal anatomy is extremely divergent and similarities may be masked by evolutionary convergence. Here we provide a phylogenetics-based test of the power of morphological versus geographical character sets for resolving phylogenetic relationships in xystodesmid millipedes. Molecular data from 90 species-group taxa in the family were included in a six-gene phylogenetic analysis to provide the basis for comparing trees generated from these alternative character sets. The molecular phylogeny was compared to topologies representing three hypotheses: (1) a prior classification formulated using morphological and geographical data, (2) hierarchical groupings derived from Euclidean geographical distance, and (3) one based solely on morphological data. Euclidean geographical distance was not found to be a better predictor of evolutionary relationship than the prior classification, the latter of which was the most similar to the molecular topology. However, all three of the alternative topologies were highly divergent (Bayes factor >10) from the molecular topology, with the tree inferred exclusively from morphology being the most divergent. The results of this analysis show that a high degree of morphological convergence from substantial gonopod shape divergence generated spurious phylogenetic relationships. These results indicate the impact that a high degree of morphological homoplasy may have had on prior treatments of the family. Using the results of our phylogenetic analysis, we make several changes to the classification of the family, including transferring the rare state-threatened species Sigmoria whiteheadi Shelley, 1986 to the genus Apheloria Chamberlin, 1921—a relationship not readily apparent based on morphology alone. We show that while gonopod differences are a premier source of taxonomic characters to diagnose species pairwise, the traits should be viewed critically as taxonomic features uniting higher levels.

BackgroundWith an estimated 80% of species remaining undescribed (but see Brewer et al. 2012), millipede taxonomy offers the opportunity to discover new species and explore biodiversity. The lack of basic alpha taxonomic information regarding millipedes belies their significant ecological role and potential as premier models in ecological and evolutionary studies. The group possesses many fascinating biological properties (e.g., bioluminescence, mimicry, and complex chemical secretions) that have been the focus of several recent studies and are emerging avenues of future investigation.New informationHere we summarize a methodology for large-bodied millipede collection, curation, and preservation for genetic analyses with the hope that sharing these techniques will stimulate interest in these charismatic detritivores.

Millipedes in the family Xystodesmidae (Polydesmida) are often referred to as “colorful, flat-backed millipedes” for their bright aposematic coloration and tendency to form Müllerian mimicry rings in the Appalachian region. However, there are many species of Xystodesmidae that do not display colorful warning patterns, and instead have more cryptic appearances. Perhaps for this reason, groups such as the genus Nannaria have remained understudied, despite containing a large number of undescribed species. Before his death in 2012, R. L. Hoffman worked on a revision of the genus Nannaria, and synthesized material and drawings since 1949. Here the work is continued, inferring a molecular phylogeny of the Nannariini (Nannaria + Oenomaea pulchella), and revealing two clades within the genus. One clade is named the minor species group, and the second is the wilsoni species group. This revision, using a molecular phylogenetic framework, is the basis for descriptions of 35 new species in the minor species group. A multi-gene molecular phylogeny is used to make taxonomic changes in the taxon. Eleven putative species of Nannaria are also illustrated and discussed. Additionally, detailed collection, natural history and habitat notes, distribution maps, and a key to species of the Nannaria minor species group are provided. These items are synthesized as a basis for a revision of the genus, which hopefully will aid conservation and evolutionary investigations of this cryptic and understudied group.

The name “millipede” translates to a thousand feet (from mille “thousand” and pes “foot”). However, no millipede has ever been described with more than 750 legs. We discovered a new record-setting species of millipede with 1,306 legs, Eumillipes persephone, from Western Australia. This diminutive animal (0.95 mm wide, 95.7 mm long) has 330 segments, a cone-shaped head with enormous antennae, and a beak for feeding. A distant relative of the previous record holder, Illacme plenipes from California, it belongs to a different order, the Polyzoniida. Discovered 60 m below ground in a drill hole created for mineral exploration, E. persephone possesses troglomorphic features; it lacks eyes and pigmentation, and it has a greatly elongated body—features that stand in stark contrast to its closest surface-dwelling relatives in Australia and all other members of its order. Using phylogenomics, we found that super-elongation (> 180 segments) evolved repeatedly in the millipede class Diplopoda. The striking morphological similarity between E. persephone and I. plenipes is a result of convergent evolution, probably for locomotion in similar soil habitats. Discovered in the resource-rich Goldfields-Esperance region and threatened by encroaching surface mining, documentation of this species and conservation of its habitat are of critical importance.