semi-paraphyletic taxa: a new paradigm?

Doug Yanega dyanega at MONO.ICB.UFMG.BR
Wed Jan 6 15:48:22 CST 1999


Fred Schueler wrote:

>Ken Kinman wrote:
>
>>      "A truly cladistic classification of life would
>> require an enormous number of intermediate categories
>> and would therefore sacrifice stability and usefulness
>> in favor of predictive power."
>
>* I've never understood why, if there are no quantitative criteria of age
>or phenetic distance-or-gap to define categories, Linnean categories
>continue to clutter up putatively phylogenetic classifications. We deduce
>nested hierarchies of taxa, and give each a name, but I've never
>understood what we gain by imputing some sort of equivalence between the
>Parulidae and Staphylinidae by calling them both "families."

Actually, I don't think is what we do at all, at least not literally - and
even conceptually, I have my doubts. We most certainly do NOT give names to
EACH nested set of taxa, for one thing. If we *tried* to do that, yes, it
would become hopelessly complex and unwieldy, as Ken's quote suggests. That
is an unreasonable approach to classification because of impracticality, as
also pointed out by Joe Laferriere.
        Moreover, I don't think anyone familiar with phylogenetics would
ever assume that two taxa of the same Linnean rank are in fact equivalent
in *any* other sense, and attempts to recategorize taxa to enforce "taxon
equivalence" are misguided. After all, which way would we polarize this
process - e.g., do we split all arthropods into a million smaller taxa to
make them equivalent with vertebrate taxa, or do we lump all vertebrates
into a single subfamily, to make them equivalent with arthropods? What
about lineages like the coelacanth, nautilus, or horseshoe crab? We don't
*need* to have equivalence, so why even ask the question? Equivalence is a
red herring, a non-issue, and can NOT be used as an argument against
Linnean ranks. For me, a rank like "Staphylinidae" only has relevance in
the context of other families of its superfamily - that is, each is
monophyletic with respect to the others - once you get beyond that, then
you are effectively correct in that there is nothing to gain. But no one
who understands things would (or should) ever MAKE such comparisons.

>> "The semi-paraphyletic markers solve one of the
>> major problems that cladists have with eclectic
>> classification, namely the paraphyletic placement of two
>> sister groups on different taxonomic levels.
>
>* but this dodge (or something else like it) is needed only if there's
>some real benefit to retaining categories. Without categories you can
>name and discuss as many monophyletic groups as you find, and maybe talk
>about the phenetic distance & overlap between sister groups in terms of
>various character sets, but I just don't see the need for retaining
>Aristotelian categories in a phylogenetic discourse.

Maybe this can be confusing "in a phylogenetic discourse", but of course,
if all you're doing is looking at cladograms, one can conceivably HAVE such
a discourse without once referring to a formal Linnean category or even a
group name (assuming the cladogram is of species), just by pointing to
cladogram nodes. But when you come around to the discussion section of that
paper, you'd better have SOMETHING that places the results back into a
classificatory framework! At the very least you need categories so you can
communicate your results without having to show the cladograms. Otherwise
you're essentially talking about doing phylogenies without having any
formal classification system AT ALL. I'd also be interested in how you
would give names to monophyletic taxa that is not simultaneously a process
of categorization, unless it's purely ad hoc (more below).
        It's also not necessarily difficult to get people to adjust to a
more phylogenetically accurate classification scheme, depending on how
entrenched the old categories are, so keeping paraphyletic taxa just for
tradition's sake is not *necessary* - no "dodge" is required. An example of
recent vintage in my own area of interest has to do with bees and sphecid
wasps; both were historically treated as superfamilies (Apoidea and
Sphecoidea), the latter taxon often as a single large family Sphecidae -
but phylogenies show that the bee "superfamily" is sister taxon to one of
the traditional subfamilies of Sphecids. Accordingly, people have adjusted
to the idea that there is only one superfamily encompassing all bees PLUS
sphecids (Apoidea is the older name, thus has priority), and that Sphecidae
should now be split into separate monophyletic families, one of which is
now sister to all the bee families (some extremists have even lately
proposed to collapse all bees into a single family to make it more
"equivalent" to the sphecid families, which is an unnecessary and highly
artificial approach that is not likely to gain acceptance). Voila, the
paraphyly problem goes away with minimal effort, and the system continues
to reflect our (now improved) understanding. In what way would a "rankless"
scheme improve upon this?

>I'm obviously wondering about a bigger subject than the "Kinsman System"
>here, but I'd be interested to see a non-nostalgic defense of the use of
>categories.

It's easier to say "Apoidea" than to give a cladogram of all the 50,000+
included species, which is the only true alternative if one literally has
no categories (after all, if there are no "families" or any lower taxon
groupings, like genera, you can't give a simplified cladogram, now, can
you?). True, a cladogram of 50,000 species has more information on the
actual relationships than some shorthand system like Linnean ranks, but it
isn't practical until we all have Cray supercomputers wired into our brains
- and ranks convey SOME information, which would be lost if they were
abandoned.
        But you obviously realize you can't define every functional taxon
by giving a list of its included species. You HAVE to give names to groups
of taxa, at SOME level above species, which you even admit above. But we
also still need to have some hierarchy for it to make sense beyond the
context of a single paper; if we just imagine replacing all the
rank-specific endings with -idae, for example, how is one supposed to know
whether author A's "Apidae" is the same as author B's "Apidae" (consider,
also, that there would be no way for the reader to ever look at two taxon
group names and know whether one included the other, as opposed to
something like Apoidea/Apidae/Apinae/Apini/Apis) - and even something like
replacing "species" with "primary taxon", "genus" with "secondary taxon",
and so forth would still ultimately be the same thing. Sure, there are
times when two authors use the same name and mean different things (Apidae
is defined three different ways in current usage, for example), but as bad
as the problem may be, in a "rankless" scheme it would only be WORSE. It
likewise conveys a *lot* when I am told that the superfamily Pelecinoidea
is represented by only one species, for example - and if that's only
shorthand for giving a species list for its putative sister taxon of 2000
species, then fine. It's not based on nostalgia, it's based on the fact
that categories give us information content that a "rankless" system would
lack.
        The problem is not that we have or need to use categories (I
believe we do), the problem is that our understanding leads us to make
changes in them (though we've been dealing with that just fine for over 200
years), a problem for which I don't think a practical "quick fix" exists.
Cladograms aren't stable either, so there is no possible gain there no
matter what we do - any names for groups whose boundaries are redrawn will
be just as invalidated, whether they are "ranked" or not. Unlike Curtis, I
don't think that future generations are going to do anything any
differently without some compelling reason - and I haven't heard any
compelling reasons yet.

Sincerely,

[NOTE THAT I WILL BE LEAVING BRAZIL ON FEB 13]

Doug Yanega    Depto. de Biologia Geral, Instituto de Ciencias Biologicas,
Univ. Fed. de Minas Gerais, Cx.P. 486, 30.161-970 Belo Horizonte, MG   BRAZIL
phone: 31-499-2579, fax: 31-499-2567  (from U.S., prefix 011-55)
                  http://www.icb.ufmg.br/~dyanega/
  "There are some enterprises in which a careful disorderliness
        is the true method" - Herman Melville, Moby Dick, Chap. 82




More information about the Taxacom mailing list