The relative influence of niche versus neutral processes on Ediacaran communities

A fundamental question in community ecology is the relative influence of niche versus neutral processes in determining ecosystem dynamics. The extent to which these processes structured early animal communities is yet to be explored. Here we use spatial point process analyses (SPPA) to determine the influence of niche versus neutral processes on early total-group metazoan paleocommunities from the Ediacaran Period ~565 million years in age. Preservation of these sessile organisms in large in-situ populations on exposed bedding planes enables inference of the most likely underlying processes governing their spatial distributions by SPPA. We conducted comprehensive spatial mapping of six of the largest Ediacaran paleocommunities in Newfoundland, Canada and Charnwood Forest, UK using LiDAR, photogrammetry and a laser-line probe. For each paleocommunity we determined the best-fit spatial model for each univariate and bivariate species distribution, comparing four sets of spatial models (complete spatial randomness, dispersal, habitat, and combined dispersal with habitat) using goodness-of-fit tests. Random and dispersal models are considered neutral processes while habitat and combined models are considered niche processes. We find the dynamics of these paleocommunities to be dominated by neutral processes, with limited influence from niche processes. Our findings are consistent with community model predictions of when neutral dynamics dominate, but are in stark contrast to the niche-dominated communities of the modern marine realm. Thus, while the underlying processes determining metazoan community assembly appear to have been in place since the appearance of the first macroscopic, complex animals, the dynamics of these early metazoan communities were fundamentally different to those of extant communities.