Interactive keys as aids to revisions

[Mike Dallwitz]

Dmitry Dmitriev wrote (Re: [TAXACOM] interactive keys):

> Yes you are right, only few are come to the web site to use the key. ...
> [But] I myself use the key almost daily. I am now working on huge genus of
> leafhoppers Erythroneura with several hundred species. Most of the species
> have never been revised or included into any key. They can be identified
> only using original description or by comparison with identified material.
> Many of the species are to be synonymized. I cannot keep such amount of data
> in my head. I also cannot produce a dichotomic key before looking through
> all the species. What I do, before placing a new species to the datamatrix,
> I try to identify the species using the key for already scored species.
> After using 3-5 “best” characters I reduce search to less than 10 taxa,
> which I can compare with a new species and decide either the species is a
> synonym to another one already in the database, or the species is not in the
> key yet and I have to place it to the database. So the program itself is
> powerful tool for anybody doing revisionary study. It also can be very
> helpful for creating dichotomic key for paper publication.

The late Ev Britton, who started my interest in computerized keys, used
interactive keys in this way, for revisions of large genera of beetles. The
specimens in the drawers would already have been visually sorted into
species (described or undescribed). Ev used Intkey to try to obtain
identifications of the specimens in each drawer, in terms of the species
already examined and recorded. If the identification failed, the specimen(s)
would represent a new species, which would be added to the database. If the
identification apparently succeeded, he compared the new specimen(s) with
the specimen(s), from previous drawers, corresponding with the tentative
identification. Sometimes they would be the same, simply because the
specimens were separated in the collection (which can easily happen in large
groups if good identification aids aren't available when the specimens are
filed). In other cases, the specimens would belong to a different species,
in which case a character would be added to separate them.

Ev didn't publish his interactive keys - he was content to use them for his
own research. However, in one of his revisions he did publish complete
descriptions in DELTA format, as well as an automatically generated
conventional key:

Britton, E.B. 1986. A revision of the Australian chafers (Coleoptera:
Scarabaeidea: Melolonthinae). Vol. 4. Tribe Liparetrini: genus Colpochila.
Aust. J. Zool., Suppl. ser. 118: 1–135.

He commented:

"A second important virtue of coded descriptions is the fact that in a
revision, the state of all characters selected is recorded for all species,
so that the descriptions are comparable in respect of every character. This
avoids the all too frequent failure of plain language descriptions to be
properly compared and contrasted."