TAPHONOMIC DAMAGE AND TAXONOMIC IDENTIFIABILITY: IMPLICATIONS FOR SECULAR BIAS IN THE FOSSIL RECORD

Fossil assemblages rarely consist exclusively of pristine specimens, but are rather a mixture of well-preserved specimens and those that have undergone various degrees of taphonomic alteration (fragmentation, bioerosion, surface alteration, etc.). The post-mortem alteration of specimens from pristine to highly degraded without actually destroying the specimen has the potential to introduce bias into paleoecological data by preventing taxonomic identification of some portion of the assemblage, effectively removing them from the data set. Bivalve mollusks and brachiopods from a number of geologic time intervals and the Recent are examined here to identify the relative importance of different types of damage in degrading the taxonomic identifiability of individual specimens. Identifiability was measured by (a) the finest taxonomic resolution possible (species, genus, class, etc.) and (b) damage to morphological features critical to taxon diagnosis (e.g., shell outline, hinge features, etc.). Results indicate that for fossil material, fragmentation that obscures shell outlines (whether caused by predation, wave action, compaction, or other means) and, to a secondary degree, surface alteration that obscures surface features (e.g., caused by dissolution, micro-bioerosion, abrasion, etc.) are the dominant processes leading to a shell becoming unidentifiable. Comparisons with the few other studies that have examined taxonomic identifiability indicate that similar processes are at work for other taxonomic groups (echinoids and corals). Both fragmentation and surface alteration, and their many causes are thought to either increase (durophagus predation; bioerosion) or remain relatively constant in intensity (abrasion, wave action) over Phanerozoic time, and thus should result in an overall decrease in preservational quality over time in the fossil record. Such a trend would lead to underestimates of diversity and evolutionary durations in the younger fossil record compared with the older record, contrary to trends expected from other factors (e.g., sampling, lithification, etc.).