LITERATURE VERSUS FIELD-DERIVED BIODIVERSITY ASSESSMENTS: REGIONAL SAMPLING EFFECTS AS DIRECT FUNCTIONS OF LITHOLOGY, TAPHONOMY AND POPULATION STRUCTURE

Biodiversity trends at any scale may be influenced by sampling effects and variation introduced by the nature of the rock record. Researchers attempt to remove the biases imparted by sample size, rock volume, lithology and population structure so that the biological signal can be detected. However, when working at a regional or global scale, these properties vary from sample to sample, and it is difficult to resolve their effects with respect to the many aggregated samples that comprise a typical database. Here, I investigate these effects directly by comparing diversity trends derived from literature and field-derived data obtained from the same geographic and stratigraphic interval.

As a precursor to the present study, I investigated variation in eleven submembers that comprise the Kope, Fairview, and Bellevue formations within the type Cincinnatian (Upper Ordovician) of Kentucky. Using a literature-derived database of some 11,000 fossil occurrences, I determined that, even after accounting for variable sampling effort, rock volume, lithology, and population structure, there was significant variation in diversity, exemplified in particular by the two stratigraphic end members of the section, the shale-rich, high diversity, low evenness Fulton submember of the Kope Formation, and the limestone-rich, low diversity, high evenness Bellevue formation. This variation may result from taphonomic, lithologic, or monographic effects or, alternatively, may reflect a true biodiversity signal.

With this in mind, the properties of these two end members are currently being investigated using new, field derived samples to determine whether the relative characteristics of the literature-derived data in the two intervals are overprinted in some way by taphonomy or other aspects of preservation. In addition, by analyzing multiple samples from each interval, it should be possible to determine whether the literature-derived database adequately captures their preserved aggregate and population-level biodiversity signals. Understanding the relationship between field and literature-derived biodiversity and the sampling effects that influence them should help to identify which methods of sampling and standardization generate biodiversity trends that most accurately characterize the preserved record.