id keys-a plea

[Lawrence Kirkendall]

I am glad to see the interest aroused by the topic of identification keys.
Anyone doing field biology or general determinations for a museum,
herbarium or whatever is a lifelong user of keys, and has a vested interest
in how keys are developed!

I would like to raise the following point: keys (or any id. tool) have to
be tailored to the intended "audience".  Most frequently, keys are
constructed as a distillation of the knowledge acquired by an expert after
a geat deal of time and effort spent in getting to learn and understand a
group:  "Now that I can finally tell these apart, here is how I do it."
This can lead to phrases like "generally smaller", "more widely spaced",
"flowers ...,fruits...", etc., which (as has been pointed out already)
renders the key of doubtful utility in many situations.  It is such keys
that seem to be publication of the author's own comments to himself.  (In
all fairness, I realize sometimes one can't do any better than this.)

But who is the audience for id. tools?
Usually, the usergroup is assumed to be expert taxonomists who need to be
as close to 100% correct as possible.  But what about users whose mistakes
will not be life-threatening, and who need to determine large numbers of
specimens in a wide variety of groups?  Keys for them should be constructed
on rather different principals, it seems to me.  Here are two examples of
what I would like to see; they would be easiest to achieve in regional as
opposed to world-wide treatments, I would think.

(1)  For the (nonexpert-in-this-group) user trying to identify large
numbers of specimens, whether it is for ecological research or to put some
order into a museum collection, an enormous amount of time would be saved
if THE MOST COMMON SPECIES CAME OUT FIRST!  How many of you have taken that
into consideration?  Am I wrong in believing that, for most groups, 5-10
species will make up 50-75% or more of all _individuals_ in a museum
collection?  This is at least true for a variety of tropical insect groups.
Should we have to wait until couplet 52 to come to the most common species
in a large genus?  Mybe one could even start with a short key to the 10
most common species, which would each be well-illustrated; "if it is not
one of these, go to the next key".  Has this ever been attempted?  (Unless
there are some good reasons not to, I am going to try this myself!)

(2)  In any given level, there are taxa that are very distinctive to the
naive user.  Much time would be saved if there were some sort of overview
of these at the beginning of the identification process, perhaps before the
key even begins.  A table would be handy, I should think; obviously,
drawings (even crude ones, which anyone could do) of the distinctive
characters should be included.  One could even use this to "point" to
unique groups of species  ("bicolored: start at couplet 31").  This
approach should be used judiciously, though:  the utility might be diluted
if such a table were cluttered up with too many rare species.

Users need to be educated too, in a way.  We aren't always really aware
(and seldom told flat out) that some species groups can only be identified
by an expert with authoritatively id'd material at hand, and that often
single specimens cannot ever be identified with confidence.  If a key
exists, there is an implicit assumption that it is possible to identify all
the taxa it covers.

Just some thoughts from a user but soon-to-be producer of keys...


FOR FUTURE REFERENCE, NOTE:

                                           __________________________
from 22 June-22 Sept., I will be at   --->  klawrenc at cariari.ucr.ac.cr
                                           __________________________

Lawrence R. Kirkendall                        FAX:   +47 55 31 44 64
Univ. Bergen Zoological Inst.                 VOICE: +47 55 21 23 42
Allegaten 41, N-5007 BERGEN Norway
EMAIL:  Lawrence.Kirkendall at zoo.uib.no