AmericanChemicalSociety.com

Introduction.- Ecological studies suggest that rare taxa are more likely to go extinct than abundant ones, but the relationship between rarity and extinction in the fossil record has received surprisingly little attention (McKinney and Frederick 1992; Lockwood 2003). The purpose of this research was to examine the link between rarity (defined here to include mean abundance, spatial and temporal variability in abundance, geographic range, and habitat specificity) and extinction during background intervals in the molluscan fossil record of the mid-Atlantic Coastal Plain. I focused on three questions:

(1) Are rare taxa more likely to go extinct?

(2) Does the relationship between rarity and extinction differ in mass versus background extinction intervals? And

(3) Does selectivity differ according to metric of rarity?

Geologic setting.- This project focused on the molluscan fossil record of the mid-Atlantic Coastal Plain. Sampling was limited to Miocene and Pliocene units exposed in Maryland, Virginia, and North Carolina, focusing almost exclusively on the Eastover, and Yorktown Formations. These deposits accumulated in the predominantly shallow shelf, open marine setting of the Salisbury and Albemarle Embayments and contain a wealth of well-preserved molluscan material (Ward and Blackwelder 1980, 1987; among many others).

Fieldwork.- The fieldwork component of this project involved collecting and processing 45 bulk samples from 10 localities in Virginia, Maryland, and North Carolina. At each locality, the exposed units were measured, sedimentology and sedimentological features described, and replicate bulk samples collected in an attempt to assess spatial and temporal variability in molluscan abundance.

Data compilation.- Bulk sample data from field-collected bulk samples was supplemented with two large datasets donated by collaborators.

Laboratory work.- Samples were standardized (according to volume), sieved, and molluscan material greater than 2mm (following Kidwell 2001) was sorted and identified to species level whenever possible. Both right and left valves of bivalves were tallied and the larger value of the two was used to represent the number of individuals for a particular species in a particular sample. Two different metrics were used to measure abundance: rank and percent abundance. Geological durations (i.e., stratigraphic ranges) of species were compiled from the literature and the Paleobiology Database (paleodb.org) based on faunal occurrences and used as a proxy for survivorship. Spatial and temporal variability in abundance was quantified using the coefficient of variation for each abundance metric. Occupancy was quantified as the percentage of samples in which a taxon occurred. Geographic range was assessed in two ways—by calculating the great circle distance represented by geographic occurrences of each species and by calculated the number of 2° x 2° latitude/longitude grid cells occupied. We also quantified habitat breadth, using the first axis of a detrended correspondence analysis on species-level abundance data.  The link between rarity and extinction was assessed through the correlation of fossil abundance data (including mean abundance and variability in abundance) with species duration, calculated at a variety of spatial and temporal scales.

Results.- Despite the great increase in sample size (from 30 to 273 samples in the past two years), our results still showed little or no evidence of a link between abundance and survivorship in mollusks in the mid-Atlantic Coastal Plain. The relationship is extremely weak and differs according to both rarity metric (abundance, variability, or occupancy) and extinction proxy (survivorship or duration) used.

Of the metrics assessed, geographic yielded by far the strongest correlation with survivorship, while habitat breadth yielded the weakest. It is interesting to note that, although the relationships are not statistically significant, most of the trends follow predicted directions. For example, species with large geographic ranges, high abundance, low variability in abundance, and high occupancy tend to have longer durations and are more likely to survive major extinction events such as the end-Cretaceous extinction. This was not the case with habitat breadth, although it is unclear to what extent this is because habitat breadth is not tied to extinction or whether the particular proxy for breadth is not applicable for this geologic setting. The results for the end-Eocene and mid-Miocene extinctions were much more variable than the results for the end-Cretaceous extinction. Patterns varied drastically according to the extinction proxy used, and were by and large non-significant. These results are detailed in multiple GSA abstracts, as well as Lockwood (2008; in prep).

Significance.- This research is significant in two major respects: (1) it is one of the first attempts to explore the link between rarity and extinction during background intervals in the fossil record and (2) it will contribute to our understanding of how ecological, environmental, and taphonomic factors affect fossil abundance data. In total, the project has supported eight undergraduate students at the College of William and Mary and one Ph.D student at Virginia Polytechnic Institute (Barbour Wood). This project has formed the foundation for a long-term research project focusing on the macroinvertebrate paleoecology of the mid-Atlantic Coastal Plain (in collaboration with Dr. Michal Kowalewski from Virginia Polytechnic Institute, Dr. Susan Barbour Wood from Western Carolina University, Dr. Gwen Daley from Winthrop University, and Dr. Lauck Ward from the Virginia Museum of Natural History).

References

Aller, J. Y. 1995. Palaeogeography, Palaeoclimatology, Palaeoecology 118:181-212.

Kidwell, S. M. 2001. Science 294:1091-1094.

Lockwood, R. 2003. Proceedings of the National Academy of Sciences 100: 2478-2482.

Lockwood, R. 2008. Beyond the Big Five: Extinctions as experiments in the history of life. In From Evolution to Geobiology: Research Questions Driving Paleontology at the Start of a New Century. P. H. Kelley and R. K. Bambach, eds. Paleontological Society Papers 14: 207-228.

Lockwood, R. and S. Barbour Wood. In prep. Exploring the link between rarity and molluscan extinction in the Cenozoic record of the U. S. Coastal Plain.

McKinney, M. L. and D. Frederick. 1992. Geology 20: 343-346.

Ward, L. W. and B. W. Blackwelder. 1980. USGS Bull. 1482-D, p. 1-61.

Ward, L. W., and B. W. Blackwelder. 1987. Smithsonian Contributions to Paleobiology 61: 113-283.