Abstract

The first well-resolved phylogeny of stylonurine eurypterids (30 taxa, 58 characters) is presented, prompting a taxonomic revision at the familial and superfamilial levels. The monophyletic suborder Stylonurina consists of four superfamilies: Rhenopteroidea, Stylonuroidea, Kokomopteroidea and Hibbertopteroidea. The enigmatic hibbertopterids – large sweep-feeding forms from the Carboniferous to end-Permian – are therefore demonstrated to be an in-group Stylonurina clade, within Eurypterida, in contrast to some earlier hypotheses. Furthermore, the genus Drepanopterus is shown to be polyphyletic: ‘Drepanopterus’ bembycoides is transferred to Moselopteridae fam. nov. along with Moselopterus and Vinetopterus at the base of the Eurypterina, defined by their possession of a pediform appendage VI bearing a modified podomere 7a. Evolution towards a sweep-feeding mode of life occurred independently in stylonuroids and hibbertopteroids, involving either multiple rows of fixed spines on the prosomal appendages (in stylonuroids) or paired movable flattened spines (‘blades’) on the prosomal appendages alongside a posteriorly cleft metastoma and coxal laden (in hibbertopteroids). The Stylonurina have a relatively poor fossil record (RCI 15%), when compared to more derived Eurypterina clades (e.g. Adelophthalmoidea RCI 66%; Pterygotoidea RCI 53%), but is relatively more complete than basal Eurypterina clades (RCI -21%). The fit between phylogeny and stratigraphical occurrences of stylonurid taxa is good (SCI 0.65 and GER 0.77, with only 0.3% of 1000 randomisation tests yielding greater congruence; GER* 0.995), and generic-level collector curves of the Stylonurina and Eurypterina show no major discrepancies in their sampling histories. These differences could be explained by geographic collection bias, taxa having different habitat preferences (and hence fossilisation potential), and ontogenetic factors: these results support previous suggestions that stylonurine eurypterids are oversplit.