GEOS 26400/36400/BIOS 23255/EVOL 32300: Principles of Paleontology

Nature of the Fossil Record. III. Archiving the fossil record and growth of our knowledge of the fossil record

I. First generation compilations
      --Treatise on Invertebrate Paleontology
	--morphology, classification, stratigraphic & geographic distribution
      --Sepkoski's Compendium of Marine Animal Families (1982, 1992)
      --Benton's The Fossil Record 2 (1994)
      --Sepkoski's Compendium of Marine Animal Genera (2002)
      **Focus on first and last appearances
	--little contextual information (for example, geography, lithology)
        --can yield history of diversity, origination, extinction, faunal composition
      --In addition, numerous "private" databases developed for particular animal groups

II. The Paleobiology Database (started in 1998)
      --collections tied to specified published references
      --collections contain abundant contextual information
      --taxonomic list associated with each collection
      --occurrences throughout history of taxa, not just first and last appearance
      --enables much richer range of analyses, many of which we'll see later in course

III. Sampling
      --Importance of understanding statistical properties of samples 
        --generally not possible to observe every organism or species
      --Bias versus error in sampling (more on this later!)
      --Nature of sampling curves: diminishing returns
            --number of species as function of number of individuals
            --number of higher taxa as function of number of species
            --range of trait as function of #individuals or species
            **"Level" curve does not imply everything has been found,
                  only that modest effort is unlikely to yield much more.

IV. When is sampling adequate?
      --"Idealistic" interpretation: when do we know the truth?
      --"Pragmatic" interpretation: when is knowledge stable?
            --relation to sampling curves and principle of diminishing returns

V. Examples of growth of knowledge
      --Williams (1957) study of first and last appearances of brachiopod genera
            --Geologic periods of peak evolutionary activity vary over history
of paleontology, depending on interests and practices of leading workers.
      --Sepkoski (1993) comparison of 1982 and 1992 editions of Compendium of
Fossil Marine Animal Families
            --Diversity patterns in single groups may change substantially as
we accumulate more knowledge (e.g., foraminferans).
            --Many patterns rather stable when we lump all animals together:
                  --total diversity
                  --origination and extinction
      --Stability may depend on taxonomic scale (families vs. genera).
      --Foote (1997) study of morphological diversity
            --Large-scale sampling appears roughly random.
            --Some features of morphological diversity curves stable as we
accumulate knowledge, others more volatile.

VI. Taxonomic Standardization
      --Data compilations like Sepkoski's cover broad range of taxa; expertise
on all groups not possible.
      --Does accepting data uncritically lead to error and/or bias?
      --Example of test: Adrain and Westrop on Ordovician-Silurian trilobites.
            --A&W are trilobite experts; applied uniform taxonomic standards.
            --Found many disagreements between their data and Sepkoski's.
            --Disagreements seem to represent noise, not bias.
                  --Temporal diversity patterns unchanged by standardization.
      --Further example: Ausich and Peters on Ordovician-Silurian crinoids
	--explicitly test idea that errors are random (Fig. 5)
	--attempt to trace causes of error in unstandardized data (Table 2)
      --Further example: Wagner et al. on mollusks
        --Diversity and rates of origination and extinction affected by taxonomic vetting
        --But temporal patterns largely conserved

      **Taxonomic Standardization does not imply that True Pattern has
            been found:
            1. Competent systematists can disagree;
            2. Additional material will later be found:
                  *Temporal patterns can change simply from additional data...
                  *...and also from further revision that may be needed
                        in light of new data