Kenelm Philip fnkwp at
Tue Feb 8 02:28:50 EST 2000

	Since my (odd) ideas re subspecies were referred to earlier in
this discussion, I might present them again here.

	First a little background. When I moved to Fairbanks in 1965, I
had my nose rubbed in a phenomenon that had previously been summed up
by someone (not sure who) as a sort of taxonomic law: the boundaries
between described arctic subspecies always fall in uncollected areas.
When you make widely separated forays into the range of an extensive
cline, you end up with a set of 'subspecies' that are totally arbitrary
and based entirely on historical factors relating to human intrusions
into arctic regions. So I became somewhat disenchanted with subspecies.

	I decided to ignore subspecies entirely in the Alaska Lepidoptera
Survey collection, and lay out the material in strictly geographical
sequence--on the assumption that any patterns would thereby be revealed.
Ideally, one would pin each species onto a large map of Alaska at its
collection site--not a practical method of specimen storage. So I broke
the state down into several biologically relevant regions, and then filed
specimens within each region in order of the 1:250,000 USGS quadrangles
within that region. Not exactly a perfect method, but _any_ attempt to
map a two-dimensional distribution into a one-dimensional display is
going to be imperfect somewhere...

	After doing things that way for a while, I suddenly realized that
the _name_ of a subspecies was the least interesting thing about it. If
someone walked in with a new Alaskan subspecies, and gave me its name--
I would have _one_ question to ask: where is it from? Since I also had
some doubts as to the reality of some northern subspecies, it didn't take
long to decide that one could shortcut the whole process by simply re-
ferring to each subspecies as '_species_ [locality]', which would answer
that first question instantly. After all, subspecies are defined as
_geographic_ races (in most cases, at any rate)--so why hide the most
important factor?

	Ideally, now that we all have access to computer mapping programs,
the information in the square brackets could even be lat/long coordinates.
That would let me take all the new subspecies in 'The Systematics of
Western North American Butterflies', bring up a map of the western U.S.
in RangeMapper, and instantly plot the localities of all those taxa and
then be able to click the cursor on each dot to find out what it was.
For those who don't think in coordinates, a place name in the brackets
might do the job.

	I have no particular quarrel with Cris' definition of a subspecies:
"...subspecies names are merely labels for geographically defined groups of
populations that show consistent relatively minor phenotypic differences."
However, I always worry about ecotypes. Subspecies, as I understand the
concept, have (small) genetic differences. If two populations look somewhat
different because of environmental effects on a constant genotype, then
those are not subspecies. Unfortunately, without a lot of work there is
no way to tell what's going on when you first encounter a new and
slightly different population. Using a locality indicator at least lets
you refrain from jumping to conclusions too early.

	This is, of course, rank heresy. But having a few heretics around
keeps life interesting...

							Ken Philip
fnkwp at

More information about the Leps-l mailing list