TAXACOM Digest - 30 Sep 1995 to 1 Oct 1995

[Joseph Laferriere]

> >Joseph Laferriere wrote (in part)> >Sometimes a person doing an
> >ecological or a floristic survey will be forced to collect specimens that
> >are less than perfect. Writers of keys should realize this and enable the
> >users to figure out the specimen's identity anyway. They should, if at
> >all humanly possible,  avoid using obscure characters which they know are
> >difficult or impossible to observe, so long as alternatives are
> >available. I once saw a key separating two genera using the following
> >couplet:
> >
> >Embryo curved
> >Embryo straight
> >
> >The seeds of both taxa were about a millimeter in diameter. Another book
> >separated the exact same genera thus:
> >
> >Inflorescence a spike
> >Infloresence a panicle
> >
> >Which would you rather use? Embryo shape might be a more useful character
> >in determining the relationships of these genera to other groups, but
> >inflorescence shape is infinitely easier to determine.

Robin  Panza wrote:
>
> This argument is self-contradicting.  The very fact that a particular specimen
> is damaged is exactly why the key's creator cannot hope to make a key for all
> uses.  Granted, inflorescence shape is more likely to be of use on future
> identifications.  Granted, if seed and inflorescence are both available, the
> latter would still be easier to use.  However, if I had to do an environmental
> survey right now (fall, around here), I couldn't possibly identify this plant
> based on inflorescence shape.  I would be better off with the embryo character.
>

Joe responds:

Self-contradicting??? Hardly. Sorry, but you've lost me on why you think
that. I suggest you re-read my statement. The  ideal situation (as I have
said on Taxacom before) is to have multiple characters, i.e. both embryo
shape and inflorescence shape. No key can cover all possibilities, but
a good one should at least anticipate likely problems. Even the best
botanist in the world is forced to collect less-than-ideal material on
occasion, and the key-writer should anticipate what types of situations
are likely to arise in a fair number of cases.