Scientists from McGill University have used data from the Paleobiology Database to look at the diversity of dinosaur species in North America, specifically from the Maastrichtian formations from 71 to 65 Ma, just before the major dinosaur extinction event. They were interested in alpha diversity (diversity within a particular area or ecosystem) versus beta diversity (diversity between ecosystems or along environmental gradients).
People tend to think that dinosaur fossils are abundant and wide-ranging. This isn't the case; fossils are actually relatively rare and patchy in their distribution. In community ecology, particularly diversity measures, sampling can make a big difference in the results, the more you sample and the more individuals you collect the better picture of the community you get. See where I'm going with this? However, there are ecological techniques to correct for this sampling bias (rarefaction, etc). The scientists in this study corrected for sampling bias by only considering four formations in the northwestern interior that each included more than 100 dinosaur specimens. They looked at the number of unique species at each site (unique meaning found in only 1 place), compensating for the unevenness in samples between sites, and used that to estimate the number of species missing from the fossil record.
They found that the entire Western Interior of North America may have once been populated by a single dinosaur community. These data suggest that differences in species in this region were relatively low, so low as to consider it a single homogeneous fauna that ranged over an entire continent. In terms of the diversity measures, beta diversity was low, there were not multiple dinosaur faunal regions. Low beta diversity is not an unknown pattern. It is something that we can observe in various communities right now, but it is on a much smaller geographic scales. Modern environments can be very patchy allowing different species to exploit various habitat types, and so low beta diversity is seen in smaller, homogeneous habitats. It is likely that the dinosaur communities discussed in this paper experienced large, homogeneous habitats exposed to stable climates. Many were not isolated to small regions and constrained by small, isolated environments. Rather, they were able to colonate and dominate vast areas. This conclusion naturally leads to more questions involving dinosaur migration, gene flow, habitat type, etc.
Its a good start to piecing together the ecology of this era and something which can be updated as more data is collected.
Here's the PNAS article:
Vavrek, Matthew J. and Hans C.E. Larsson (2010) Low beta diversity of Maastrichtian dinosaurs of North America. PNAS: published online. (DOI: 10.1073/pnas.0913645107)
(second image from www.cmnh.org)