Species of Dickinsonia Sprigg from the Ediacaran of South Australia

An iconic member of the Ediacara Biota, Dickinsonia Sprigg is one of few such taxa with multiple species. Here we use Gaussian finite mixture models to assess the validity of species distinctions for this genus. Our results indicate that the five described species of Dickinsonia from the Ediacara Member, South Australia are better classified as two based on multiple approaches. Two different methods for dimension reduction both provide strong support for two groups, with overlapping but distinct mixture models. The variable selection method produces the most biologically realistic clusters, indicating that the two species can be primarily differentiated based on the greater relative size of the anterior most unit of Dickinsonia costata Sprigg compared with Dickinsonia tenuis Glaessner & Wade. Despite differences in aspect ratio and number of modules, both species regulated growth to maintain overall shape. The greater likelihood of preservation of a midline and an irregular outer margin in D. tenuis highlights differential structural integrity and flexibility. Co-occurrence in the Ediacara Member indicates that both species occupied the same environments and temporal distribution. Smaller maximum and average size for D. costata, combined with higher abundance, may suggest a comparatively shorter lifespan and increased rates of reproduction.

Automated summary

This study uses Gaussian finite mixture models to assess the validity of species distinctions for Dickinsonia Sprigg, an iconic member of the Ediacara Biota with multiple species. The results suggest that the five described species of Dickinsonia can be better classified as two distinct species based on multiple approaches. Variable selection method provides the most biologically realistic clusters, indicating that the two species can be primarily differentiated based on the relative size differences of specific units. The co-occurrence of these species in the same environments and temporal distribution highlights their ecological similarities.