pdf: http://www.palaeo-electronica.org/2010_3/209/209.pdf
htm: http://palaeo-electronica.org/2010_3/209/index.html

The dominant biodiversity paradigm in the Earth sciences has been that abiotic forces such as orbital variations, bolide impacts, and volcanism, among others, are the primary drivers of extinction and biodiversification. By describing several instances when increased tectonism and biodiversity hotspots coincided, a recent study of fossil foraminifera indicated that over the past 50 million years (m.y.) the predominant driver of biodiversity has been plate tectonics. We test whether this hypothesis holds true for assemblages of marine macrofossils from temperate latitudes using one of the most comprehensive data sets of its kind, the Paleobiology Database. Our multivariate statistical approach provides a base for examining the historical biogeography of marine macrofossils and for investigating ecosystem interactions on various scales. We began by examining the biogeographical patterns across 40 phylogenetic classes of marine animal macrofossils spanning the most recent 57 m.y. in both passive and convergent tectonic settings using area-standardized alpha and beta diversity components at the species and genus level. When the diversity components of average grid cells were compared between regions, the highest alpha diversities occurred in the passive tectonic setting of the eastern United States. We tested whether this could be due to biased coverage of North American fossils by rarefying and recomputing our diversities. Although the subsequent alpha diversities were more homogeneous, eastern North America was still the most diverse.

To explore the connection between this observation and ecological niche theory, we simulated paleo-habitats for each grid cell using a multivariate k-means clustering of the lithological annotations of occurrences within grid cell assemblages. Across 83,000 occurrences, the number of simulated habitat types within a grid cell was significantly correlated to the alpha diversity of that grid cell (species-level r = 0.763, genus-level r = 0.857). Further tests using rarefied subsets of occurrences did not strongly impact this correlation (species-level r = 0.679, genus-level r = 0.762). Our results, which apply to Cenozoic marine macrofossils, imply that a more robust predictor of alpha diversity across both passive and convergent tectonic settings is the diversity of paleo-habitat types. This investigation provides statistical evidence that simulated habitat types from litholgies can be used as a rudimentary analogue for ecosystem complexity, and that the abundance of habitats scales linearly with the alpha diversity of fossil assemblages.