Regional subsets of taxa in INTKEY

[Mike Dallwitz]

                                                                19 January 1995

> From: rambold at botanik.biologie.uni-muenchen.de (Gerhard Rambold 304)
>
> Could you please tell me how I can restrict INTKEY identification
> to taxa of certain regions.

There are two ways of doing this. The best is to add a character or characters
with which you code the occurrence of the taxa in the regions. Here is an
example from Leslie Watson's Angiosperm-families package.

    472. <geographical distribution>
        1. Western and Northern Eurasia, U.S.S.R.
        2. Middle East <Turkey to North Africa, and eastwards to Iran>
        3. Mediterranean
        4. Eastern and Southern Asia <including Japan, China, India>
        5. North Africa
        6. Central, East tropical and West tropical Africa
        7. South Africa
        8. Madagascar
        9. Malay Archipelago <including New Guinea>
        10. Australasia <Australia, New Zealand>
        11. Pacific Islands
        12. North America
        13. Central America
        14. West Indies
        15. South America
        16. Arctic
        17. Antarctic
    Comprehensively encoded, and intended only for use in identification
    and for generating useful geographic subsets of the data. Assignments to
    these pragmatically defined world regions are intended to reflect
    likelihood of the family being encountered `in the field', regardless of
    floristic status. Their reliability has yet to be assessed.

The identification of a specimen from the Antarctic could be started as
follows:

    472. <geographical distribution>
        17. Antarctic
    4 taxa remain.
    156. Cyperaceae
    179. Donatiaceae
    232. Gramineae
    246. Hectorellaceae

The advantage of using this method is that INTKEY's error-tolerance mechanism
will apply to this information, so that a specimen from a taxon not previously
recorded from the region can still be correctly identified.

The second way of restricting operations (including identification) to subsets
of taxa is to use the INCLUDE TAXA command. For example, in the
Angiosperm-families package, you can restrict the taxa to those occurring in
the British Isles by the command:

    INCLUDE TAXA BRITAIN

When used in connection with identification, this method of obtaining regional
keys has the disadvantage that the error-tolerance mechanism does not apply -
taxa that have been excluded are completely out of contention as possible
answers. The advantage is that it is practical to specify regions of narrow
interest (such as the campus of a university), which would be impractical to
incorporate in the data as a character. Also, any user can define these sets,
whereas characters can only be added by the compiler of the data.

I have posted a more detailed discussion on DELTA-L. (To subscribe to this
list, send the message `subscribe delta-l your-first-name your-last-name' to
listserv at uvvm.uvic.ca.)

Mike Dallwitz                                  Internet md at ento.csiro.au
CSIRO Division of Entomology                   Fax +61 6 246 4000
GPO Box 1700, Canberra ACT 2601, Australia     Phone +61 6 246 4075