Re: Fastovsky vs Archibald

[Phil Bigelow <bigelowp@juno.com>]

On Sat, 25 Jun 2005 11:33:39 -0700 (PDT) Tim Donovan <uwrk2@yahoo.com>
writes:
> 
> 
> --- David Marjanovic <david.marjanovic@gmx.at> wrote:
>  
> > Eh, but why? :-)
> > 
> 
>    Looks that way. Compare the Campanian record with
> unit 4 and the latter with the Scollard.


<groan>
A few caveats to consider:

It is unwise to compare two different paleoecosystems (e.g., the upland
Canadian formations  vs. the lowland Hell Creek Formation) or times (late
Maastrictian vs. Campanian), or both, and then draw conclusions about
extinction based on that comparison.  There are too many variables
involved (most of them unknown), and the discussion quickly becomes an
exercise in arm-waving.  The data collected from a particular formation
must be considered unique to that formation, and conclusions drawn from
that data should only be applied to that formation and to no others. 
Only later, when the data is placed in its proper paleoecosystem context,
can different formations and times be compared and a "big picture"
hypothesis produced.

In any scientific study, the ideal is to eliminate as many variables as
is possible.  Unfortunately, the disappearance of a particular taxon from
a particular unit can be the result of many unrelated and unknown
phenomena.  "Extinction" is a CONCLUSION that a worker makes; it is not
"raw data" nor are its effects directly observable in the strata.

On the other hand, "dinosaur diversity as a function of depositional
environment" *can* be compared between formations or between times, but
*not* extinction.  If I had only one wish for the betterment of science,
I would wish that biostratigraphers would concentrate more on
paleoecosystems of small areas and within smaller time parcels, and not
overemphasize "big picture" chronological relationships as much as they
currently do. In geology/paleontology, "time" has meaning only when it is
applied to a particular space.  Earth scientists often forget that little
detail, and instead focus only on chronostratigraphy.

<pb>
--