An answer to paraphyly?

Tom DiBenedetto tdib at UMICH.EDU
Tue Jun 17 11:48:25 CDT 1997

On Tue, 17 Jun 1997 06:01:46 -0500, Thomas G. Lammers wrote:

>  Where I differ is that I look at the continuity between the
>species-it-was-in-the-past and the contemporary mainland species and say,
>"Not much difference.  This is all one species."

Right, your criterion is the characters *per se* rather than what
they are meant to illuminate.

>(I think of species as four-dimensional: they have DURATION.)

Well, life certainly endures. At one point in time there was a single
mammalian species, has it endured? It was a fully functional
biological and taxonomic species; where is it today? It didnt go
extinct,,,probably most of its genes are still with us,,but where is
the species? Well,,species is a rank applied to a taxon. The taxon is
still here obviously. It IS Mammalia. But the taxon/species diverged,
and once it diverged it no longer is ranked as a species, but becomes
a higher taxon. That is how we name and rank taxa, by their place in
the evolving historical hierarchy. The ancestral species/population
that we have been discussing diverges when an isolate goes off to
found a new taxon. If the isolate doesnt persist, then the hell with
it, I will ignore it. But if it persists to form a part of the
diversity we wish to classify, then I dont see how you can deny that
the ancestral taxon has diverged,,it no longer is a species but is a
higher taxon. But the more widespread unchanged population IS a
species, and it takes that ranking identity from its position as a
terminal branch in the historical hierarchy. The members of the
ancestral population are no longer parts of a terminal branch. They
were part of a species when they were alive, but that species had a
different position in the historical hierarchy than do the critters
alive today. And if place in the hierarchy is your criterion, then
their different identity indicates different names.

>    Strict cladistic thinking would
>insist on two species, which seems unwise if we could not distinguish them.

Of course species, or taxa in general can only be recognized to the
extent we can  distinguish them, but the bottom line is that we are
trying to distinguish coherent lineage branches which are
historically real whether or not they have left evidence of the
branching event in their characters. No cladist would define a group
without evidence, but this example indicates to us that there is a
history which seems poorly represented by the evidence, such that the
ancestral pop and the large contemporary pop actually have different
historical identities within the branching system even though we
cannot distinguish them. And we try to indicate this in our
classifications with various conventions.
There will always be a disconnect between the actual history of life
and our attempts to represent it. Life is real. Given adquate
evidence we can have confidence that our taxa represent real
branches. Lacking such evidence we are led to less-valid
approximations, such as paraphyletic assemblages, and we should
acknowledge our suspicions as to their historical artificiality, even
if they are useful for other purposes.

>I guess the problem is, as Harvey Ballard so eloquently pointed out, that it
>is not species which are the units of evolution, but populations.  If we
>modify the statement to say that no contemporary POPULATION is the ancestor
>of another species, we have a statement that sounds even better to me.  The
>ancestor of my insular example was a POPULATION that existed long ago.  In
>my taxonomic judgment, that population is considered conspecific with
>certain currently existing populations.  If that upsets you, I'm sorry.

No need to apologize,,,it would take a lot more than that to upset
me. It sounds to me that you are merely recognizing a lower taxonomic
rank than species,,,,populations. Fine. Then for the sake of clarity
and hierarchical accuracy, why not simply include the isolate in the
same species as the larger population and distinguish it at the level
it actually diverged at,,,identify it as a separate population, one
with a unique phenotype.

>I will never accept  that the evolution of a new species from an existing
>one somehow magically transforms what is left into a new species itself, any
>more than the simple death of individuals or extirpation of populations can
>transform what remains  into a new species.  We can put forth all sorts of
>wonderful theoretical arguments and philosophical convolutions, and redefine
>terms until we're happy with the results, but in the end, it is just plain
>silly.  The plants on the mainland are the exact same species, by any
>reasonable standard, on the day AFTER the dispersal event occured as they
>were on the day BEFORE it occurred.

But, once again,,,if you are interested in a naming system which
reflects the history of lineage branching (and perhaps you are not),
then if the isolate has any persistence at all, the remaining
population has a new identity *within the lineage system*,,,it is a
descendant of an older population which has diverged. That it "looks
the same" or cannot be distinguished from the ancestral population is
a wonderful justification for calling it the same as the ancestor if
"looking the same" or having an identifiably identical gene-pool were
the naming criterion. But if its place in the historical branching
system is the criterion, then it has a new identity. The question is
whether you choose to recognize that.

> However we wish to
>deal with it nomenclaturally and taxonomically, it is a fact that groups of
>populations that meet various criteria of "species-hood" routinely emerge
>out of the middle of other groups of populations that meet various criteria
>of "species-hood", thus creating a pattern we dub paraphyly.  We cannot
>legislate it away or define it away.  We must deal with it in a rational and
>sensible fashion.

And my notion of dealing with it rationally entails agreeing to
coherent criterea of "species-hood",  and discovering the precise
relationships that you subsume under "emerging out of the middle
of...". If there is sufficient information available in the record of
character distributions to accomplsh that, and you choose to apply
ranks according to the relationships so discovered, then the
paraphyly melts away.

More information about the Taxacom mailing list