Paper No. 0
Presentation Time: 10:45 AM
TRANSCENDING PATCHINESS IN THE COMPARATIVE ANALYSIS OF PALEOCOMMUNITIES: CASE STUDIES FROM THE CRETACEOUS AND THE RECENT
Patchiness in the spatial distribution of species within a fossil assemblage is an important component of sampling error confounding attempts to quantify the composition of paleocommunities. Even relatively time-averaged fossil assemblages retain vestiges of the patchy distribution of species within the original living community. To demonstrate this, we have sampled a time-averaged fossil assemblage from the Late Cretaceous Navesink Formation on the Atlantic Coastal Plain by collecting a large number of small bulk samples distributed at three spatial scales (replicates located decimeters apart, outcrops located tens of meters apart, and localities located tens of kilometers apart). Replicate samples were used to generate 95% cluster confidence intervals around mean abundances at the outcrop and locality spatial scales. For the six macroinvertebrate species preserved, percent abundances varied greatly between individual replicate samples and there were statistically significant differences in mean abundances from outcrop to outcrop, demonstrating the patchy distribution of the shelly fauna at small spatial scales. However, estimates of mean species abundance between the two localities studied were not statistically different for any of the six species. In fact, mean species abundances estimated at each locality using replicate samples dispersed among outcrops were remarkably similar, demonstrating that by replicate sampling multiple outcrops within each locality we were able to transcend the patchiness present between replicates and between outcrops. Even the very patchy Recent subfossil remains found along the north shore beaches of Long Island provide reliable and repeatable estimates of species abundances at very modest levels of sampling effort, if many small replicate samples are taken from a variety of points along the beach. The importance of dispersing sampling effort for reliably quantifying patchy fossil assemblages cannot be overstated.