Taxaonomic database structure

[Norman F. Johnson]

Here's another option for encoding taxonomic names (no matter what kingdom
you're dealing with).  Since we're dealing with a hierarchy, it's possible to
get by with just a single table.  It needs only have the taxon identifier
(name, id number, whatever) and the corresponding identifier for its parent.
 Of course, that parent has a parent, on down/up to the most inclusive taxon in
your database.  We're using Oracle as our RDBMS and it has a nice SQL add-on
feature that facilitates traversing such hierarchies.

With the hierarchy represented in this manner, it seemed easiest to have
species names represented as the combination of Genus + species. (Trinomials
would work the same way.)  I went in this direction because the genus in which
a species is classified may change.  This, in turn, may require modification of
the ending of the specific epithet.  Rather than try to program in grammatical
rules, I thought it would be easier to have it just typed in the way the name
was used.  This also makes it possible to use names in which the gender of the
genus and species names were incorrectly matched.

You may view our database structure (most of it anyway) at
http://iris.biosci.ohio-state.edu/projects/tpp/tables.html
This was based on the original Association of Systematics Collections
information model (http://www.keil.ukans.edu:70/11/standards/asc).  The newer
and still developing model may be seen at
http://gizmo.lbl.gov/DM_TOOLS/OPM/BPSL/LIB/ASC.html

Norm Johnson
-------------------
Norman F. Johnson        Johnson.2 at osu.edu
Associate Professor      phone: (614) 292-6595
Director, Ohio State     fax: (614) 292-7774
  Insect Collection