The sensitivity of paleocommunity sampling strategy at different spatiotemporal scales

One of the primary differences between paleoecological and modern ecological research is the added dimension of deep geological time. The temporal dimension of paleoecological research can create added difficulties when determining the appropriate spatiotemporal sampling procedures required for any particular study. Previous paleocommunity studies have found that the spatial extent of sampling can have a grave impact on the multivariate analytical results. Here, we expand spatially on the previous paleocommunity study of Forcino et al. (2010) to determine the importance lateral variation may have on the paleocommunity signal within the Pennsylvanian Finis Shale. Spatial and temporal community gradients of bulk sediment shale samples from six localities were examined using ordinations and ANOSIM. The expanded spatial examination of the Finis Shale reveals the same two distinct communities separated at the mid-point of each sampled stratigraphic section. Thus, lateral variation did not have an effect on the primary multivariate result for the Finis Shale communities. At the temporal scale of the complete stratigraphic section, a temporal factor has the greatest influence on the community result. However, based on separation of samples in ordination space with respect to location and based on a significant difference between locality-based groupings using ANOSIM, the primary influence on the paleocommunity result at a finer temporal scale, containing 19 samples from only the lower half of each stratigraphic section, is a spatial factor. This minor variation in scale leading to a major change in the primary factor affecting the multivariate paleocommunity result is evidence that the paleocommunity results are extremely sensitive to the temporal scale of sampling. Thus, this differs from previous research; there are cases in which fewer laterally-equivalent samples are required per stratigraphic horizon, bed, or unit, because the temporal signal overwhelmed the spatial signal.

► Collected 29 samples from four localities and six stratigraphic sections.
► Multivariate statistical analysis reveals two distinct communities.
► A spatial factor is the primary control on all 29 samples.
► A temporal factor is the primary control on the 19 lower stratigraphic section.
► Paleocommunity results are extremely sensitive to the temporal scale of sampling.