During the present study, DNA sequence data, gross morphology and scanning electron microscopy (SEM) were used to examine cryptic species boundaries in the velvet worm, Peripatopsis clavigera species complex, from the southern Cape Afrotemperate forest belt in South Africa. Sequence data were generated for the mitochondrial COI and the nuclear 18S rRNA loci and phylogenetically analysed using both a Bayesian inference and a maximum-likelihood approach. Both the COI data and the combined DNA sequence topology (COI+18S) revealed the presence of five clades within the Peripatopsis clavigera species complex, and revealed that specimens from Tulbagh were distantly related and represented a sixth clade. The evolutionary distinction of the five clades was corroborated to varying degrees by the four species-delimitation methods (ABGD, PTP, GMYC and STACEY); however, both the gross morphological data and the SEM provided limited diagnostic differences between the five clades. Furthermore, the COI haplotype network and phylogeographic analyses provided evidence of genetic isolation between lineages that are currently syntopic. The distribution of genealogically exclusive and widespread maternal lineages was atypical among velvet worms and did not reflect the general trend of genetic and geographical isolation. Instead, lineages exhibited admixture among localities, a result most likely due to fluctuations in climatic conditions affecting the southern Cape Afrotemperate forest during the Pliocene–Pleistocene period as evident from our divergence time estimations. Four novel, narrow-range endemic species – P. ferox, sp. nov., P. mellaria, sp. nov., P. edenensis, sp. nov. and P. mira, sp. nov. – are described within the P. clavigera species complex, whereas the Tulbagh specimens are described as P. tulbaghensis, sp. nov. Collectively, these results demonstrate that Peripatopsis likely contains several undescribed species.

Pseudospirobolellidae is a poorly known family of spirobolidan millipedes with only two genera and five described species. Yet, the descriptive taxonomy and molecular systematics of this group have been largely neglected. Therefore, the present work presents an integrative taxonomic study of new pseudospirobolellid taxa in Thailand. To this end, two mitochondrial gene fragments (COI and 16S rRNA) combined with morphological characters were used to define the genus Coxobolellus, gen. nov. with 10 new species, viz. C. albiceps, sp. nov., C. compactogonus, sp. nov., C. fuscus, sp. nov., C. nodosus, sp. nov., C. serratus, sp. nov., C. simplex, sp. nov., C. tenebris, sp. nov., C. tigris, sp. nov., C. transversalis, sp. nov. and C. valvatus, sp. nov. The interspecific COI sequence divergences among the new species ranged from 6 to 15%. The intergeneric COI sequence divergence between species of Coxobolellus, gen. nov., Benoitolus birgitae and Pseudospirobolellus sp. ranged from 20 to 23%. Three major morphological differences separate Coxobolellus, gen. nov. from Benoitolus and Pseudospirobolellus, namely (1) the protruding process on the 3rd (and 4th) coxae on male legs, (2) the posterior gonopod telopodite divided into two parts, and (3) a conspicuous opening pore at the mesal margin at the end of the coxal part of the posterior gonopod. Thus, the new genus is well supported by both mtDNA and morphological evidence, while the delimitation of the 10 new species is supported by the congruence between mtDNA and morphological data. Yet, with respect to the relationships of Benoitolus birgitae, morphological data suggest a similarity with Coxobolellus, gen. nov. and Pseudospirobolellus, whereas mtDNA data place this species in the Pachybolidae. Further phylogenetic analyses are needed to explore this apparent incongruence and test the monophyly of Pseudospirobolellidae.

The perceived cosmopolitanism of polychaete worms could be an artefact of historical factors such as poor original species descriptions, lack of type material and the European taxonomic bias, to name a few. Thus, it is possible that several cosmopolitan species hide complexes of cryptic and pseudocryptic species. Two putative cosmopolitan species, Platynereis dumerilii and Platynereis australis, collected in South Africa were investigated here (1) to determine whether the South African taxa are conspecific with the morphologically identical taxa from France and New Zealand (the respective type localities of P. dumerilii and P. australis), (2) to compare the South African species morphometrically to determine whether their morphological characters are reliable enough to separate them, and (3) to investigate whether these species have geographically structured populations along the coast of South Africa. Molecular data (COI and ITS1) confirm that P. dumerilii and P. australis do not occur in South Africa. Instead, the South African taxon formerly thought to be Platynereis dumerilii is new and is described here as Platynereis entshonae, sp. nov.; the identity of the other South African species is currently unresolved and is treated here as Platynereis sp. Surprisingly, Platynereis massiliensis (type locality: Marseilles) nested within the South African Platynereis sp. clade but, since it is part of a cryptic species complex in the Mediterranean, the name is considered doubtful. Morphological characters traditionally used to define these South African Platynereis species are not reliable as predefined morphological groupings do not match phylogenetic clades and principal component scores revealed no separation in morphological characters that could distinguish between them. Haplotype networks and phylogenetic trees revealed that P. entshonae, sp. nov. and Platynereis sp. have geographically structured populations along the South African coast.

Triaenonychidae Sørensen in L. Koch, 1886 is a large family of Opiliones with ∼480 described species broadly distributed across temperate forests in the Southern Hemisphere. However, it remains poorly understood taxonomically, as no comprehensive phylogenetic work has ever been undertaken. In this study we capitalise on samples largely collected by us during the last two decades and use Sanger DNA-sequencing techniques to produce a large phylogenetic tree with 300 triaenonychid terminals representing nearly 50% of triaenonychid genera and including representatives from all the major geographic areas from which they are known. Phylogenetic analyses using maximum likelihood and Bayesian inference methods recover the family as diphyletic, placing Lomanella Pocock, 1903 as the sister group to the New Zealand endemic family Synthetonychiidae Forster, 1954. With the exception of the Laurasian representatives of the family, all landmasses contain non-monophyletic assemblages of taxa. To determine whether this non-monophyly was the result of Gondwanan vicariance, ancient cladogenesis due to habitat regionalisation, or more recent over-water dispersal, we inferred divergence times. We found that most divergence times between landmasses predate Gondwanan breakup, though there has been at least one instance of transoceanic dispersal – to New Caledonia. In all, we identify multiple places in the phylogeny where taxonomic revision is needed, and transfer Lomanella outside of Triaenonychidae in order to maintain monophyly of the family.

The genus Holopodostreptus Carl, 1913 is a poorly known monotypic genus endemic to Ecuador. It differs from most other Pseudonannolenidae by a set of characteristics of the gnathochilarium and gonopods. Herein, we provide descriptive notes of males of H. braueri Carl, 1913 based on recently collected material. Females of the species are described for the first time, and images of the vulvae are provided. In addition, we attempt for the first time to place Holopodostreptus in a phylogenetic context. With a total of 14 taxa and 41 morphological characters, and under implied weighting (k = 6), a single most-parsimonious tree is recovered (total fit = 35.86). The genus is recovered as sister-group of Physiostreptus Silvestri, 1903. The clade Holopodostreptus + Physiostreptus is supported by one synapomorphy (absence of setae on the prefemoral process of leg-pair 1 in males) and one homoplastic transformation (narrow telopodite). Based on the phylogenetic results, notes on the subfamilies Pseudonannoleninae, Cambalomminae and Physiostreptinae and a key for all pseudonannolenid genera are provided.