RE: "Common ancestor" in cladistics

["Thomas R. Holtz, Jr." <tholtz@geol.umd.edu>]

So much to say, so little time...

> On Thursday, July 29, 2004, at 07:59 AM, Mike Milbocker wrote:
>
> > Michael Habib wrote:
> >
> > "This means that the generally
> > assumption is that all nodes should have two branches; more (a
> > polytomy) are said to imply a lack of agreement/resolution."
> >
> > I find the statement that a node should have only two branches or else
> > a
> > lack of agreement/resolution is implied, to be troubling. (I'll set
> > aside
> > for later debate the implication here that a character should have
> > just 2
> > states.)I'll start by positing that each node represents at least a
> > species,
> > maybe multiple very similar species, and not a particular mating pair.

Urrr, actually, I would argue that both ideas are incorrect!

The nodes ultimate represent *populations*, which are the entities that
actually do the evolving.  Branching at a node represents the historical
event whereby members of one subpopulation wound up not breeding back into
the rest of the population.  As such, it is closer to the "at least a
species" idea than the "particular mating pair" idea.

> > The hypothesis that only 2 branches should occur suggests
> > to me
> > an assumption that extreme selection pressure is in effect. That as
> > soon as
> > a superior variation occurs, the "basal" species becomes extinct.
> > While this
> > may be valid for densely populated niches, it certainly should not
> > hold for
> > periods just after mass extinction, where niche filling is less
> > constrained
> > by competition. Trees that cross multiple extinction events should not
> > be
> > expected to preserve a bifurcating structure.

Part of this argument here is a misconception about the reality of
phylogenetic analysis. In fact, we can only ever sample a small, small
subset of the actual sets of populations that lived at any given time in
Earth history.  Thus, we resort to the method of phylogenetic systematics to
reconstruct the most likely pattern of divergences among that small, small
subsample of the total connections in life history.

Thus, the likelihood of recovering representatives of multiple lineages that
were produced at the same exact speciational moment is extraordinarily
small, if theoretically possible. (see below, after my comments on Michael's
response, for different type of polytomy that certainly does crop up).

Access to total knowledge of all populations over time would render
phylogenetic analysis moot, and something other than cladistics would be
used.

> -----Original Message-----
> From: owner-dinosaur@usc.edu [mailto:owner-dinosaur@usc.edu]On Behalf Of
> Michael Habib
>
> You make a strong argument, and I agree: there are definitely scenarios
> that can create real polytomies.  These are called 'hard' polytomies.
> However, almost all authors assume that polytomies in their trees are
> 'soft', and trees with branches that have polytomies are not considered
> completely resolved.  My point was that this is a general assumption
> made in most cladistic analyses, not that it was biologically true in
> all cases.  After all, the treatment of common ancestors is not exactly
> the truth either, since it is not impossible to find the precise
> ancestor, just incredibly unlikely.  In the same way, the chances of
> having several character changes all arise from the exact same ancestor
> at exactly the same time is very unlikely (remember that time component
> must be there as well).  Even Darwin's finches probably diverged
> somewhat sequentially (though the time intervals in between would have
> been extremely small).

Incidentally, the Grant's and colleagues have recently developed some quite
reasonable divergence patterns for the Darwin's finches, so these do not
seem to have been a single massive polytomy.

>  However, hard polytomies are probably real, and
> it is simply very difficult to determine when this is case.

Here's another type of polytomy:
A geographically widespread species (being a perfectly good species under
nearly all species concepts) has multiple cases at different points in time
where subpopulations become entirely isolated, each developing its own set
of unique features. The main body of populations of the ancestral species
keep on keeping on, even long after each of the subpopulations have diverged
(perhaps multiple times).

>From a morphological view, this is a hard polytomy: each isolate clade
shares no more morphological features in common than any other isolate
clade. (From a molecular standpoint, there might be some additional hope).

A real case of this is _Branta canadensis_ (Canadian geese), which have had
multiple insular isolate populations which have diverged from the ancestral
form (some of the Hawaiian forms being the most interesting and most
divergent), even while the good old-fashioned ones are doing just fine.

                Thomas R. Holtz, Jr.
                Vertebrate Paleontologist
Department of Geology           Director, Earth, Life & Time Program
University of Maryland          College Park Scholars
                College Park, MD  20742
http://www.geol.umd.edu/~tholtz/tholtz.htm
http://www.geol.umd.edu/~jmerck/eltsite
Phone:  301-405-4084    Email:  tholtz@geol.umd.edu
Fax (Geol):  301-314-9661       Fax (CPS-ELT): 301-405-0796