[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Convergence and Coding Characters

Jerry D. Harris wrote:
>        You're a bit off base here, but only because I was taking a few
>things for granted in my explanation (such as the nature of streptostyly in
>a bird).

        A. Streptostylic cranial kineses: 0 absent, 1 present
>differentiated between having the quadrate swing forward and having a
>hypothetical fused quadrate-quadratojugal complex swing forward because in
>the former, the quadratojugal has been lost entirely; it isn't present to

        B. Quadratojugal: 0 present, 1 fused to quadrate, 2 absent UNORDERED

        Note that this effectively separates the quadratojugal from your
complex of streptostyly character states, which is as it should be. I'm sure
there are other situations in which losing the qj might be advantageous, and
we should not discount potential homology by becoming overly focused on
skull kinesis.

>For a taxon with a streptostylic quadrate (with a lost
>quadratojugal) to have given rise to a taxon with a fused
>quadrate-quadratojugal complex would require that the intermediate taxa
>(which may or may not be known) would have to reevolve the quadratojugal
>and then fuse it;
        Or re-evolved it in the (hypothetically) ancestral condition of
already being fused. It is, for example, possible that the bone might not be
lost, simply integrated into the quadrate during development (effectively
cutting the corner of having to grow the bone then fuse it), then reduced
beyond detection. "Reversal" might then consist of merely skipping the
integration, and going back to the ancestral grow-then-fuse mechanism. Of
course, lacking embryological evidence, this is extremely hypothetical. But
then, so is phylogenetic reconstruction. :)
>in this case whilst simultaneously moving the hinge joint
        Ooooohnoyoudon't! "Simulataneously" is something you cannot, will
not, and should not (practically, not ideally) be able to prove. Suggest,
perhaps, but I doubt anyone has a very tightly constrained hypothetical
ancestor-descendant lineage which shows this transition occurring during the
same speciation event or within the same species anagenetically.
"Simultaneously" is a loaded word, and is being used here in a somewhat
rhetorical (hyperbolic? I'm not all that great with the word "hyperbole",
but Dr. Champman seems to like it :) manner here. I prefer a less a priori
approach to assigning credit to evolutionary hypotheses.

>from the squamosal to the exoccipital (I think I wrote "paroccipital" in my
>original post; I meant to say "paroccipital process," but somehow
>forgot...good catch!).
        C. Streptostylic quadrate: 0) hinges on the squamosal, 1) hinges on the
                paroccipital process  UNORDERED  UNPOLARIZED
                NOTE: This character must be coded ? or NA if character A = 0
>As you surmise, the likelikhood of this is small;
        "A priori" is truly one of my favorite words in the context of
phylogenetic reconstruction. Please don't be insulted that I use it so
often. I simply am very very unimpressed by statements such as the one you
make above. Sure, it doesn't sound very likely the way you present it, but
that does have something to do with, well, how you presented it.
        My point is that, in coding characters, making this sort of
consideration is tantamount to weighting characters. If you have no problem
with this, I suppose that's ok, but you should be aware of what you are
doing and acknoledge it in any publication resulting from the study.

>more likely the two taxa evolved steptostyly (each with their own
>variation) to solve the same problems that having streptostyly alleviates. 
        I agree it doesn't seem likely, but I would rather not close my mind
(and my data) to the possibility. And I don't think you have suggested that
I do this. :)

>But a blind analyst (the cladistics program) would not be able to
>differentiate the two types if we code _only_ for the presence or absence
>of streptostyly;
        Well, that depends. You have to be pretty blind not to code for the
ansense of the quadratojugal, and (once it has been pointed out) pretty lame
not to code the differing joints.

>coding for the nuances of streptostyly are necessary for
>the program to see that they are quite different and more likely convergent
>than homologous.
        Not, however, more necessary than coding many other characters as
well, especially since, in the end, these are all just characters. If you
multistate them, bistate them, overstate them, or whatever, these characters
seem to be solid, but they won't do anything without additional charcters
and taxa in the analysis.

     Jonathan R. Wagner, Dept. of Geosciences, TTU, Lubbock, TX 79409-1053
 "Only those whose life is short can truly believe that love is forever"-Lorien