Complex leads in keys

Mike Dallwitz miked at ENTO.CSIRO.AU
Wed Sep 6 15:42:55 CDT 1995

                                                              6 September 1995

> From: Mike Crisp <Mike.Crisp at ANU.EDU.AU>
> To: Taxacom

> [In the following couplet,] I only used 'not as above' for only 1
> character out of 2. Thus it is possible to key the specimen without
> considering 'not as above'.
> 12    Unit-inflorescence umbelliform; flowers subtended by an involucre of
>       bracts 3-20 mm long; bracts enlarging considerably to enclose pods,
>       becoming scarious or coriaceous (D. cordata group)                18
>       Unit-inflorescence not umbelliform, or if so, then bracts not as
>       above                                                             13

It is not possible to key out an arbitrary specimen on the `whether
umbelliform' character, because of the qualification `or if so'. The
annotation `(D. cordata group)' may give a clue to the convoluted leads: if
you try to make a key reflect the classification, its usefulness for
identification is usually impaired.

> The species in couplet 27 was very difficult to key out.  It has not unique
> character, but combines those in a group of closely related species.  The
> simplest way to deal with it was to pull it out first using its unique
> _combination_ of characters.  If I had not done so, I would have had to key
> it out in several different places.
> 27    Plants prostrate or procumbent; phyllode apex acuminate or cuspidate,
>       semi-pungent; inflorescence a 2-4(-5)-flowered umbel; pedicel dilated
>       upwards, with an annular thickening at the apex         D. lancifolia
>       Not the above combination                                          28

Every taxon in a completely resolved key has a unique combination of
attributes, but it is not usually necessary or desirable to combine all of
these attributes in a single lead. D. lancifolia would still key out only once
if all the attributes (i.e. the pieces of text delimited by semicolons) were
placed in separate leads with suitable contrasts:

      Plants prostrate or procumbent
      Plants erect

      Phyllode apex acuminate or cuspidate, semi-pungent
      Phyllode apex not as above


and this would tend to produce a better key, as pointed out by Joseph

> ... the simplest way to deal with it was to pull it out first. ... I was
> picking off the odd-looking species before tackling a large group.

This may be the simplest way for the _author_ of the key, but in general it's
not the best way for the users of the key. Here is an extract from my previous
posting on the subject.

    There are many algorithms for choosing characters to be used in
    identification, but as far as I know all of them look for characters that
    divide the remaining taxa into subgroups that are as nearly equal as
    possible. This tends to minimize the number of steps required in an
    identification. For example, for a group of 100 taxa, the use of (2-state)
    characters which are optimal in this sense would lead to a key requiring 6
    or 7 steps (characters) for an identification, whereas the use of
    characters which split off one taxon at a time would lead to a key
    requiring an average of 50 steps for an identification. ...

    Characters which split one taxon from all the rest are often preferred by
    taxonomists ... . However, such preferences should be examined critically.
    A distinction may seem obvious to the expert who has a mental picture of
    all of the taxa. Novices may make the distinction accurately when
    identifying a specimen of the unusual taxon, but will they do so with
    other specimens, particularly if they have never seen an example of the
    unusual taxon? How will the overall accuracy of the identification be
    affected, taking into account the much greater number of characters that
    will have to be used?

> Using only 1 character in a couplet (see Laferriere's example below) is
> risky - the specimen may not show that feature, or a simple mistake will
> more easily lead the user astray.

Combining attributes in a lead, with a contrasting lead `not as above', does
not add any redundancy/robustness to the key. A mistake in any one of the
attributes will still lead to the wrong answer.

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

More information about the Taxacom mailing list