A full spectrum of opinions

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

If you'd like some of the old messages to refresh your memory, just
let me know.  Let's see if we can pick up some of these things where
we left off.  This is what I was writing when I was so rudely
interrupted by the network disconnection...

Thomas Lipka writes:

: Research seems to indicate that elephants like most other large
: herbivores are color blind and see only in black and white or shades
: of gray.

If you have any specific references to the above, I'd like to see
them.  Otherwise, I'm afraid you've been hoodwinked.  Table 1 of the
Jacobs paper I mentioned previously is a series of quotes such as:

    'Most mammals are colourblind' -- The Encyclopedia of the Animal
                                          World (1991)

Statements such as yours are clearly widespread.  However, there's
virtually nothing to back them up.  From what we know, most mammals
are dichromats.

Tom Holtz adds his color commentary (incidentally, Tom, I don't horn
in on questions about Therizinosauroids... return the favor ;-):

> ...there are several researchers (whose names have all slipped my
> mind right now) who have shown that sauropsids (birds & reptiles)
> have four or five different photosensitive pigments.

For a review, see:

Goldsmith, T.H. (1990). "Optimization, Constraint and History in the
     Evolution of Eyes", _Quart. Rev. of Bio._, 65:281-322.

Perhaps I'm confused, but I thought sauropsid doesn't include anapsid.
At least some (possibly all... I don't remember) of the turtles
studied have four visual pigments.  Turtles and birds also share many
characteristics not found in mammalian retinas, so I suspect that the
dinosaurs had those features as well.

Going back to the questions Tom was addressing:

] If mammalian colour vision re-evolved after being lost during our
] nocturnal lifestyle phase, does that mean reptiles/dinosaurs/birds
] might have different retinal sensitivity modes than we mammals
] currently do?

To re-emphasize Tom's statement, Jacobs (see the ref. in my previous
message) suggests that primitive mammals were dichromatic (i.e. they
had two visual pigments).  While I also find rather compelling the
molecular evidence indicating that basal mammals were at least
dichromatic, based on shared features between turtles, birds, to some
extent fish and even a few mammals (e.g. monotremes), I suspect that
our ancestors did lose retinal features during the Mesozoic era.  

As to different sensitivities...  As per my previous message, NO two
animals' nervous systems reconstruct the world in exactly the same
way.  There is considerable variation even within humans.  Saying that
another animal is a trichromat "like us" doesn't mean that it's visual
world is colored the same way as yours.  That's one reason to shy away
from statements such as:

] Our colour vision is based on three photosensitive pigments, whose
] sensitivities peak in the red, green and blue wavelengths, 

First off, I object to the use of color names applied directly to a
physical stimulus.  Because of:

] and our eyes and brains reconstruct a subjective `colour' based on
] the differential responses from the three types of cone cells.

it's not appropriate to refer to wavelengths by color names.  Under
different conditions, light with identical wavelength compositions
will appear to have different colors.  Furthermore, the statement
about the sensitivities isn't even correct.  In J.N. Lythgoe's _The
Ecology of Vision_, there's a plate with circles dyed to match the
peak wavelengths to which our three cone classes are sensitive.  Our
"red" pigment appears yellow.  Although people frequently (even in
peer-reviewed sources) refer to visual pigments by color names, such
usage of terminology confuses much more than it clarifies (I warned
you guys I could go on forever on this stuff :-)

] It is known that many insects have the ability to discriminate among
] colours in the near-ultraviolet:

There are rodents, fish, birds, turtles and salamanders that also have
UV sensitivity.  It's actually quite widespread.

] (Insects also sense polarization characteristics of light, which
] might also be factored into their sensation of `colour').

Polarization and color are probably always kept separate.  We can't
know what the perception of polarization is like -- it may be similar
to color perception (see, for example: Bernard, G. and Wehner,
R., (1977). "Functional Similarities Between Polarization Vision and
Color Vision", _Vis. Res._, 17:1019-1028.), but bees (in which
polarization vision has been best studied) use only a small part of
their retinas for processing polarization.  Wehner and Bernard in
a PNAS paper in around 1993 (afraid I don't have that around at the
moment) argued that bees have to keep the two systems separate, or
they'd have difficulty dealing with specular reflections off of
leaves.
 
] It is possible that mammals did not entirely lose the ability to
] produce the ancestral photopigment proteins, or even if they did
] completely lose them, convergent evolution could still have
] re-created analogues of the original set (since the three mammalian
] photopigments are variant forms of a single protein, I believe).

All of the opsin genes, for the rod as well as the cones appear to
have evolved from the same primordial gene.  It appears that the rod
opsin and the cone opsins diverged from each other 800 MY ago.  The
genes for the cone pigments started diverging from each other to yield
the one coding for our short wave sensitive opsin and the gene that
would lead to those coding for our long and mid wave sensitive opsins
about 200-300 MY ago.  Our mid and short wave opsins diverged from
each other around 30 MY ago.  (These numbers are all from Jacobs.)

] My point is, is it likely that dinosaurs' perception of their own
] colours, and our perception of their colours, would be very close?

When people ask about dinosaur colors, I always interpret the question
to mean "how would they appear to us if we could see them".  How they
would appear to each other would be very different, just as modern
animals don't look the same to each other as they do to us.  We can
only get by saying things like "tigers have orange stripes" because we
can take it for granted that your retinal physiology is similar to
mine.  Dinosaurs probably not only had more visual pigments than us,
their pigments were probably very different from ours and screened by
oil droplets like those in bird, lizard and turtle retinas.  It's also
reasonable to believe that they were sensitive to polarization.  Which
reminds me... I should be writing my thesis instead of this message
;-)

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