Re: Re - Feathers and Flight

[rowe@lepomis.psych.upenn.edu (Mickey Rowe)]

In responding to Kevin Padian's comments:

     Let's take the diapsids, which include dinos and birds and
  lizards and crocodiles etc. Looking at the distribution of
  characters within that clade, one can make the suggestion that they
  were all (originally = primitively) visual, displaying creatures
  (Jacques Gauthier said it, not me). Lizards do it, birds do it, they
  all have color vision, and they are daylight, terrestrial animals.

Ronald Orenstein asks:

> Actually, cannot some "hard" evidence for this be derived from casts
> of the size and proportion of the optic lobes?

I think that phylogenetic analyses of extant forms are much more
informative on this particular issue.  In the first place, we don't
all do the same things with the same visual/brain structures.  Most of
our (read mammal) visual information processing occurs in areas of
cerebral cortex.  This stucture is largely lacking in sauropsids and
amphibia.  The eyes and behavior of frogs will testify that vision is
important to them anyways.  Much of their processing occurs in their
retinas, and the remainder occurs mostly in a structure called the
optic tectum on the top of the midbrain.  In us the homologous
structure is called the superior colliculus.  It's still a very
important visual structure, but it's relatively less important given
that we have the added cortical power.  Similarly birds have developed
a structure called the wulst which handles a lot of visual
information, and which doesn't really exist in any other outgroup.  In
short, you could be mislead by the prominence (or lack thereof) of the
"optic lobes" even if you had good endocranial casts.

In the other direction, analysis of modern animals can tell us a *lot*
about what their ancestors had to have been like.  Germane to this
particular topic is a lot of current research into the sequencing of
the proteins and genes necessary for the creation of the visual
pigments that grant us the first stage in a color vision system.  By
sequencing these genes in a lot of animals, we can trace lineages for
the individual visual pigments, and determine with high confidence how
many pigments any given animal probably had (although you do lose some
confidence when it comes to monophyletic groups that evolved
independently for 100 million years or so and then had the poor grace
not to leave us any living examples (e.g. the ornithischia)).  Given
that the groups of animals that have done the most losing of the color
vision structures in question are animals whose ancestors probably
de-emphasized vision (e.g. our Cretaceous ancestors were probably
nocturnal), we can reasonably safely infer that these color vision
features were conserved in the groups that didn't leave us genes to
clone.  If there's interest I'll try to find the time to elaborate on
this, so you'll know just what the heck I'm talking about.

In any case, given that much of the proteins known to be involved in
color vision in birds have other diapsid or even anapsid homologues,
it's pretty safe to say that good color vision is the primitive state
as Kevin says.

> In fact most tetrapods that are not actually blind have SOME visual
> displays, including nocturnal species.

Yes, and the point I've just tried to make is that in many cases we
can determine that the structures underlying the ability to recognize
components of these displays are homologs in different tetrapod
groups.  That means that the dinosaurs' ancestors also must have had
the same structures, and hence, most likely so did the dinosaurs
(including Archie).

I think that's pretty much what Kevin and Jacques had in mind.

-- 
Mickey Rowe     (rowe@lepomis.psych.upenn.edu)