In article , P van Rijckevorsel
writes
Sorry to come over so strong, but there is too much of this going around. It
is as you say done by a select group of zoologists and they are very
agressive about it. Botanists with same scientific basis (actually a better
one) are much more balanced in what they say and more careful with historic
facts.

The introduction to this topic for botanists is the
_Plant_Systematics._A_phylogenetic_approach_
which was mentioned here a few times. Now in its second edition (2002).
Although not entirely free of error this is regarded as pretty good.

There is a difference between taking DNA sequence data (or chromosome
counts, or base frequencies) into account when classifying groups of
species, as has been done by APG, Judd et al, etc, as you appear to be
referencing, and making the inclusion of a DNA sequence a *mandatory*
part of the formal description of a species, a proposal for which is
described in the material written by the original poster.

Naively one would expect this sort of proposal from bacteriologists, or
even botanists, rather than zoologists; zoologists have lots of
morphology to work with, and it's not riddled with convergences the way
plant morphology is. I believe that several bacteriological taxa
(whether formally described or not) are only known from PCR amplified
DNA sequences.

I'm tempted to make a deletion in the ITS sequence the defining
characteristic of the Malva alliance (unranked infratribal taxon in
Malveae).

I can see a number of objections to make a DNA sequence a mandatory part
of a species description.

1) which locus? if there's a free choice of sequence, then it's quite
likely that the same species will be described more than once, on the
basis of sequences from different loci.

2) which locus? is there a locus which has a fine enough resolution to
resolve species, while at the same time having a coarse enough
resolution that the sequence is essentially fixed in a single species.
Is there any locus which could be guaranteed to distinguish
allotetraploid _Spartina anglica_ from its parent species, or from the
sterile diploid hybrid?

3) expense. DNA sequencing technology is improving, but it remains a
time consuming, and imperfectly reliable, process to sequence, align and
parsimonize a set of sequences. (50 species at 3 or 4 loci is a Ph.D
thesis; if it was cheap and easy to sequence DNA people would be working
with bigger sets of species, and long stretches of DNA, e.g. the whole
chloroplast.

4) typological thinking (related to 2). Defining (rather than using as
an aid to identification) a species by a DNA sequence assumes
essentialist, typological species concept. (I'm a fan of Mayr's BSC.)
--
Stewart Robert Hinsley