FW: Confidence and keys

[Joseph Laferriere]

I agree that many published keys are analytical keys meant to justify the
taxonomy rather than artificial keys meant to help people identify the taxa.
I remember seeing one key separating two plant genera on the basis of curved
vs. straight embryos. This was not very useful because the seeds were very
small. Another key published by another author had these same two genera
separated on the basis of racemose inflorescence vs. paniculate
inflorescence. The embryo character may well be more useful to a
systematist, but for someone who merely wants to identify the plants, I
would rather deal with inflorescence shape. I have also seen many keys which
mix flowering and fruiting characters into the same key, meaning that you
must have both in order to follow the key. Having both on the same specimen
is rather rare. Ideally there should be multiple choices on the key, so that
if you don't know one character you can use the other. This is of course not
always possible, but it is a goal to shoot for. Another article I read said
that the best way to tell two species of Equisetum apart was by SEM. Sorry,
but I don't usually carry a portable SEM into the field with me.  Another
separated two varieties of Toxicodendron radicans (poison ivy) by the
character

Leaves velvety to the touch ...
Leaves not velvety to the touch ...

Just reading the key made me break out in a rash.
     My pet peeve in keys is characters which are technically correct but
counterintuitive. I saw one key which went like this:

Leaves simple ......
Leaves compound ......
     Leaflets one .....
     Leaflets more than one .......

The first time I saw this I thought, "Now wait a minute ..." Technically, it
is possible to have a compound leaf with a single leaflet. If the leaflet
has stipulules at the base, it is a compound leaf. This may hapapen if the
leaf is reduced from an ancestor with multiple leaflets (e.g. in a couple
species of Desmodium). However, 99% of the people using the key will either
not know this technicality or not think to look for stipulules if they see
the first part of the above key. The person writing the key should realize
this.
    Concerning confidence, a brief anecdote: Once I met a nun taking a
course in plant identification. She said that when it was her turn to be
God, she would decree that all plants should have the names written on the
underside of the leaves. It is not until she gets her turn at the helm that
we can ever be 100% certain of the ID of anything.

Joe Laferriere
joseph at bio2.com
 ----------
From: owner-taxacom
To: Multiple recipients of list TAXACOM
Subject: Re: Confidence and keys
Date: Wednesday 14 June 1995 5:30PM

As a taxonomists who works hard on keys, I was rather "hurt" when a
fellow taxonomist observed that keys are actually written for the benefit
of the writer, not the user. A careful introspection suggested this may
be quite true. A lot of the terms I use in keys derive from my knowledge
of the group. OK, I might not be quantifying these well enough but some
are fairly resistant to such analyses. So if I use the couplet
tibia incrassate I am drawing on my understanding for the genus of what
is normal. People who have one specimen often don't have that background
and hence the keys are limited. I am NOT advocating doing away with them,
after all I couldn't quickly identify something myself but we have to
remember the wealth of knowledge was with those who wrote the keys.
Robert J Raven, Museum Scientist (Arachnology)
Queensland Museum, Brisbane, Australia