Ian Musgrave has just posted an excellent article on the poor design of the vertebrate eye compared to the cephalopod eye; it’s very thorough, and explains how the clumsy organization of the eye clearly indicates that it is the product of an evolutionary process rather than of any kind of intelligent design. A while back, Russ Fernald of Stanford University published a fine review of eye evolution that summarizes another part of the evolution argument: it’s not just that the eye has awkward ‘design’ features that are best explained by contingent and developmental processes, but that the diversity of eyes found in the animal kingdom share deep elements that link them together as the product of common descent. If all we had to go on was suboptimal design, one could argue for an Incompetent Designer who slapped together various eyes in different ways as an exercise in whimsy (strangely enough, though, this is not the kind of designer IDists want to propose)…but the diversity we do see reveals a notable historical pattern of constraint.
How different are animal eyes? In the metazoans, about a third of all phyla don’t have any eyes at all, although light sensitive molecules are ubiquitous (for example, in sea urchin tube feet), and can be found in bacteria and, obviously, plants.
Another third of the metazoan phyla have light sensitive organs, specialized epithelial patches that respond to changes in light levels. They may have some morphological specializations, but they don’t focus an image in any way, and can’t resolve patterns of light.
The rest have distinct, image-forming eyes that focus light on an array of light-sensitive cells, and can detect patterns of light and shadow that are used to perceive a picture of the world around them. For us motile animals, eyes seem to be part of the recipe for success—the six animal phyla (cnidaria, mollusca, annelida, onychophora, arthropoda, and chordata) that contain 96% of all animal species also primitively evolved eyes.
These animal eyes fall into two major categories, with further subdivisions. Chambered eyes have a single optical element—a slit, a lens, a mirror—and focus an image on a 2-dimensional array of photoreceptors, the retina. Compound eyes use multiple optical elements.
Eyes can be further categorized as rhabdomeric or ciliary by the nature of the cellular elements that make up the photoreceptors, by the kind of opsin molecule used to transduce the light signal, and by the signaling pathway used to convert a conformation change of the opsin molecule into a change in the electrical potential across the cell membrane.
Whoa…the differences are all over the place. Eyes look different, function differently, develop differently, and use different molecules, so where are the signs of common descent? The differences tell us that eyes have arisen in evolutionary history multiple times, but there are still deep homologies; in particular, look at those opsins, specifically the Type 2 opsins.
It’s one big happy gene family, with members all related to one another. The major family members are the r-opsins, used in the rhabdomeric eyes of invertebrates, and the c-opsins, used in the ciliary eyes of vertebrates, but note that there is considerable overlap. We vertebrates also have an r-opsin: melanopsin is a visual pigment molecule expressed in ganglion cells (not classically considered photoreceptors) in our eyes, and are involved in detecting general light levels to reset our circadian clocks. Some invertebrates have both rhabdomeric and ciliary eyes and use both r-opsin and c-opsin in vision. Note also that some of these opsins have unknown functions—neuropsin, for instance, is expressed in human testes, a curious pattern that makes me wonder if there’s some analogy that could be made with the tube feet of sea urchins.
You might argue that the relationships are all spurious—maybe there is only one chemistry possible for transducing photons into a chemical signal (which is absurd on the face of it, but let’s be thorough and make the suggestion anyway). That’s easily countered: those are Type 2 opsins, what about Type 1? Type 1 opsins are found in the Archaea and in eukaryotic microbes, and while both Type 1 and Type 2 interact with retinal, Type 1 opsins have a different molecular size, a different structure, and a different function—they couple photoreception to transmembrane ion pumping, rather than to activation of a G protein signal transduction cascade. The similarities of these various phototransduction molecules are not necessary outcomes of their function, but instead reflect a contingent historical connection between them. We can put together a good explanation for these relationships with evolutionary theory.
Taken together, these data show that at least two kinds of photoreception existed in the Urbilateria, before the split into three Bilateria branches at the Cambrian. Moreover, each branch of the family tree still carries versions of both of these photoreceptor types, along with other opsin-dependent photodetection systems yet to be fully described. In the course of evolution, vertebrate vision favored ciliary photodetection for the pathway that delivers images, whereas invertebrates favored rhabdomeric photodetection for their main eyes, although why this might be remains unknown. Along both evolutionary paths, secondary photodetection systems remained to give additional information about light, possibly to instruct circadian rhythms, phototaxis, or other light-dependent behaviors. But, if vertebrates are an example, these two photodetection systems functioned together, rather than remaining separate. Although the remaining five families of opsins have not been fully characterized, it seems probable that they also respond to light, and organisms use the information they provide.
One other thing that I would note from these examples is illustrated by the c-opsin and r-opsin pathways. These are “molecular machines” or “biochemical pathways” of the sort that Intelligent Design creationists like Behe talk about, but they don’t dig into the specifics of these because they undercut their point. I look at pathways like the c-opsin → phosphodiesterase (PDE) vs. r-opsin → phosphatidylinositol (PIP) pathway, and what I see are two common signal transduction pathways (PDE and PIP show up in lots of other places, too) that have been coupled to slightly different sensors. What we find in molecular biology is flexibility and modularity, attributes that lend themselves well to combinatorial changes that can easily increase complexity—a complexity that is a hallmark of unguided evolutionary change, not design.
Fernald RD (2006) Casting a genetic light on the evolution of eyes. Science 313:1914-1918.
99 bottles says
Good review, thanks for posting it. Reminds me of the (IIRC, Onion) parody that scientists discovered animals were intelligently designed, but by a committee.
What’s odd is how the watchmaker argument falls apart so rapidly. You’re left either with incompetent design, some kind of low-level engineer hamstrung by a bogus design spec with competing requirements, or else some form of Deism as a fallback position. None of these options appeal to the Intelligent Design people, but they have no choice. To the extent they want a real scientific program based on real evidence, there are no options.
But if ID is really a kind of Neo-Deism, it can be used to undermine Evangelical Christianity. Yes, let’s talk about bacterial motors, and what they demonstrate about the nature of God….
The Lost World Museum, a Creationist museum slated to open in Phoenix, New York, USA in 2007, acquired Cy the one-eyed kitten for their collection.
They have also acquired a Chupacabra.
quork, PZ already had a post about that a while back.
Very nice article there, but I am a bit confused at your inclusion of the Onychophora amongst “the six animal phyla … that contain 96% of all animal species.” Onychophora is a minor (albeit very interesting) phylum of only slightly more than a hundred species; I would be surprised if most biologists had heard of it, and downright flabbergasted to encounter a lay person familiar with it. More than that, though, I am concerned at your omission of the nematodes, which might well be the most diverse group of animals after the arthropods. Not being a helminthologist, I cannot say offhand whether there are in fact any sighted nematodes, but I am not familiar with any eyes in Caenorhabditis (not that the one highly derived model organism is at all representative of the group in general).
This whole topic brings to mind a course that I took in sensory physiology, during which the professor had a toss-off comment to the effect that chambered eyes arose independently some forty-odd times in the animal kingdom. He was never able to provide me with a list of them, and I have never been able to compile more than a few instances myself. The only example that you have not given that I know of is in jumping spiders.
Intriguingly, I have also heard of the development of chambered eyes outside of the animal kingdom: there are some dinoflagellates with perfectly functional camera-type eyes, with lenses, even! Nobody knows what these organisms are using their eyes for, or how they are processing the information — keep in mind that these are entirely subcellular structures, in spite of their complexity. Like many subcellular photoreceptors, they appear to be descended from plastids, although in this case the eye is a compound of plastid-derived and host-derived components. Fascinating stuff, for which I wish I had more references.
john c. halasz says
So the take-away here is that both “flexibility”,- (i.e. multiplicity of divergent pathways?),- and “modularity”,- (i.e. functional integration of chemical sets?),- are both “necessary” for emergent complexity? But what would that imply about claims that evolutionary processes and their “results” are reducible to modular recombinations? Doesn’t the accumulated “feedback” of multiple, divergent pathways effect the possible combinations of modular sets? This is just a question about the “nature” of evolutionary processes.
Eyes may be diverse but you’d hardly know it from the way they are generally portrayed on toys:
as well as in cartoons etc of course.
As long as we’re talking real science, this is kinda cool too: they’ve found a (relatively) uncontaminated shard of Neandertal bone which has already yielded a million base pairs of NeanDNA, and allowed a little more precision of the divergence time with the mutual ancestor of the Saps and the Neandys–starting ~700kya and done by ~400kya.
The hope is that this one bone could yield most of the entire Neandy genome…! Yowza!
Anton Mates says
Really? I knew about velvet worms by the time I was seven or so, and thousands or millions of people have now seen one hunting on Life in the Undergrowth. They’re pretty flashy beasts.
It probably depends on whether we’re going with projected species numbers per phylum or described species numbers. From what I can see poking around the Interweb, nematodes may comprise up to a million species, but all we’ve described so far is 20-30,000, considerably less than chordates or arthropods.
Well, I stand corrected, then. (I have never personally encountered one, not that that means anything.) My point remains, though, that they are not numerically significant; echinoderms have many more species than onychophorans. As for the numbers of nematode species, even at the comparatively low number of known species, they are still a speciose group, easily outnumbering the onychophorans and those aforementioned echinoderms. Even given the tentativeness of the larger numbers, it still bears mentioning that the nematodes probably do comprise many more species than are currently described; they would definitely make my list of significantly speciose animal phyla (for whatever value that amounts to).
I hasten to point out that I mention all of this primarily because I normally find pretty much nothing to disagree with on this blog, and anything at all even remotely inaccurate is therefore quite glaring.
Anton Mates says
Nor have I, as I grew up in California. I just saw them occasionally in nature magazines and that sort of thing.
Quite true, and I don’t think PZ was claiming that all eyed phyla do better than all eyeless ones. Really, it’s mostly that arthopods skew the ratio so badly.
OTOH, “phylum” is a pretty hazy division anyway, and maybe we should just be lumping Onychophora and Arthropoda together if (as would seem likely in my utterly non-expert opinion) eyedness is a primitive trait in the Panarthropoda superphylum which contains them. In which case velvet worms wouldn’t be a particularly unsuccessful group, merely a minor branch of a particularly successful one.
Man! I am the descendant of utterly failed bacteria!
The only example that you have not given that I know of is in jumping spiders.
Q. S. says
Nice poster. But why label a beautiful worm with ‘sea fan’ ? (as it actually refers to some kind of cnidarians)
Anton Mates says
Crikey, you’re right. Even worse, the legend implies that it belongs to the “bivalve molluscs!”
That poster’s straight out of the Science article, too.
Nice article. I just wanted to gratuitously add that Russ Fernald is an old friend, and a really good guy.
PZ Myers says
Oh, yes — Russ was at the U of O when I was a grad student there.
The original publication has been made freely downloadable from the author’s website:
He has lots more interesting publications, mostly on fish:
There’s an amusing quote on his lab webpage: