Wednesday, November 11, 2009

The Amazing Housefly Part 2: Coolest Eye Ever

As longtime readers know, I am fascinated by vision, and find it absolutely fabulous that the world is bathed- all day long- in electromagnetic radiation that bounces around all over the place and reflects off of practically everything.

image I’m even more amazed that we have these funny little camera-like balls stuck in our faces with refractive lenses that focus this reflected radiation onto little screens set back in our heads, and that these screens, and the neural circuitry wiring them up to our brains, can turn this jumble of a gazillion different photons of different wavelengths and intensities into a full-blown 3-dimensional model of the world around us, extending from a few inches in front of our nose to dozens of miles away.

image Think about that the next time you’re down in the dumps. You can see! Think about what an amazing, incredible, impossible thing that is, and you do it all the time without thinking about it. Sometimes we worry about work or family or the meaning of life or whatever, but if you step back and think about it- really think about it- the mere fact that you can actually see makes your whole life worth it even if you never accomplish anything else. Everything else in life is gravy.

image I’m also fascinated by how other living creatures see, and how the world appears to them. So much of our sense of the world is wrapped up in how we see it, as is- I’d argue- our sense of self. Think about how you sense yourself. If you really think about where you are, in your body, it’s not in your hand or your leg or your belly- it’s in your head, right behind your eyes.

We’ve looked at the vision of other creatures before, including birds and other mammals. But all of the eyes we’ve looked at are camera-style eyes, very much like our own. Camera-style eyes are just one of several basic eye “architectures”, and by no means the most prevalent. No, the most common eyes in the world are compound eyes, and the trillions of creatures who bear them see the world very differently from you and me.

HumanEyeDiagram In a human eye (diagram right, not mine*), light passes through the cornea, and enters the eyeball through a lens which focuses the image, via refraction**, onto the back of the eyeball, or retina, where the various wavelengths of light received aan recognized by special receptors which communicate this information to the optic nerve. There’s a lot more detail, but that’s how it basically works, and the same general design is present in the vast majority of land-based vertebrates. There are plenty of differences between animals of course. For example, the shape of our eyes’ lenses is somewhat flexible, allowing us to focus on objects near or far. But some animals, such as snakes, have lenses that are fixed, and focus by effectively telescoping the lens***.

*Because hell, I already drew enough graphics for this post

**I explained refraction in this post.

***Arguably making them much more “camera-like” than ours.

There are also invertebrates- namely spiders- that have camera-style eyes, but the overwhelmingly majority- including Flies- have compound eyes. There are several different types of compound eye, but the most common design among insects is the Apposition Compound Eye. Here’s how it works.

Ommitidium Schematic A compound eye consists of multiple (from as few as 6 to as many as 30,000) individual “eye units” called ommatidia. Each ommatidium consists of a hexagonal tube capped by a teeny little cornea. Just below the cornea is a teeny little crystalline lens-cone, and just below that is a photo-receptive organelle called the rhabdom, and which consists of a small number of individual photoreceptor cells, called rhabdomeres, fused together toward the bottom of the tube. The light-data received by the rhabdom is transmitted through a single axon, or nerve fiber, which transmits the image to a succession of nerve layers (the lamina, medulla and lobula) where it’s combined with the thousands of other images being received simultaneously from other ommatidia and combined into a composite image in the insect’s brain.

Apposition Graphic You need to remember this part, or you’ll be lost when we get to the Fly-Eye. Go back and read it again if it didn’t make sense.

Balsamroot Arrowleaf Diagram Tangent: There’s a cool parallel here with the plant world- composite flowers. Just as a daisy or a dandelion is composed of hundreds of miniaturized flowers, a compound eye is composed of hundreds (or thousands) of “mini-eyes.”

An apposition-eye image is generally not as precise as a camera-eye image. Part of this is due to the effects of diffraction* in the small opening of the tube; it’s likely that most (all?) insects cannot see stars in the night sky. So insects that need really good vision have lots and lots of ommatidia- close to 30,000 per eye in some Dragonflies. Even so, a Dragonfly’s image of the world is far less precise than ours.

*I haven’t explained diffraction in this blog, and unfortunately this post is long enough. If you’d like an explanation right now, you can check out the Wikipedia entry for (a longer than necessary) one.

Side Note: Many apposition-eye insects, such as dragonflies and drone (male) bees*, have areas of the eye with “acute” vision characterized by larger facets which help mitigate the effects of diffraction. This “acute vision” portion of the eye is analogous to foveal vision in birds.

*Drones have better eyesight (and flight-speed) than workers or queens, since pretty much the only thing they do in life is intercept a flying queen to mate.

But the apposition-eye has a couple of advantages over our camera-style eyes.

The first has probably already occurred to you- the field of vision can be huge. A Dragonfly for example has a nearly 360 degree field of vision.

Extra Detail: And that understates it. It’s 360 degrees (or close to it) with respect to a 2-dimensional plane. But the dragonfly is seeing up and down as well in a total field of vision that is nearly spherically 360 degrees. As if that weren’t enough, Dragonflies see more colors than we do; most are tetra- or pentachromatic*. In particular Dragonflies seem to have enhanced color sensitivity in the blue-to-ultraviolet range. It’s thought that this “brightens” the sky and makes prey insects stand out more clearly.

Dragonfly Field Vision *This applies to day-flying Dragonflies. Many species are active only around dusk, and it’s not entirely clear what color-vision capabilities they have.

BTW, when I was doing this post I dug around in through some old photos for a Dragonfly* pic. HPennant2 This one’s from June, 2005, up the South Fork of Dry Creek. If you’re a Salt Lake-area trail-user, this is the major fork off to the right as you climb Dry Creek, about ½ way up. It’s a big, deep, seldom-hiked canyon, much deeper than Dry Creek proper, with a year-round stream. The day I hiked it the canyon was filled with these amazing-looking black & orange dragon flies, and I snapped a few pics (with, unfortunately, a lesser camera than I have today.) I filed it away and forgot about it.

*Dragonflies are of course Way Cool, and I intend to post about them next Spring/Summer.

Anyway I dug them up this week and think I ID’d it as a female Twelve-Spotted Skimmer, Libellula pulchella. Is she gorgeous or what?

The second benefit is that the apposition-eye forms and processes images much more quickly than a camera-style eye, which translates to a much, much higher flicker rate.

Humans have a flicker rate of roughly 50 to 60Hz, or images per second, and in fact if we see more than about 30 images per second we perceive constant motion. This is why old-style fluorescent lighting caused headaches; the refresh rate of the bulbs was only about 60Hz*. But other animals have different, and often higher, flicker rates. Most birds for example have flicker rates in excess of 100Hz, which is one reason they’re able to fly in and amongst tree limbs without colliding with branches all the time.

*This is not a problem with modern CFL bulbs, which have a refresh rate in excess of 10KHz.

Tangent: Know what other animal has a higher flicker rate than us? Dogs. In fact if you’re reading this on an old-style CRT monitor, and your dog is with you in the room, there’s a good chance that he/she sees your screen flickering right now. Most CRT displays are set to 70-90Hz. Dogs have flicker rates as high as 80Hz.

This also the case with CRT-design television sets, but not with modern flat-panel monitors or TVs, which have refresh rates of around 200Hz, and don’t really refresh the same way anyway.

But large-eyed flying insects have flicker rates in excess of 200Hz, which is incredibly useful when you’re in fast flight, chasing down another fast-moving, highly-maneuverable, flying insect.

So a Dragonfly or a Bee sees the world less precisely than we do, but they see far more of it and more real-time than we do. To a certain extent, they think faster.

Side Note: This BTW is the real reason it’s so darn hard to swat a fly. Yes, they can ride out the wind-wave pushed ahead by your hand or magazine coming down to swat them, but they have to be ready- and actually jump- to catch that wave. A Housefly can move into jump-ready/take-off position in 100 milliseconds, which is just 1/3 of the time it takes your brain to tell your hand to “swat.”

Refraction Superposition

So apposition-eyes have some real advantages, but they’re not the only kind of compound eye. Moths, for example have a completely different type of compound eye.

Moth Eyes Moths, as you’ve probably noticed, are active at dusk and in the evening, when very little light is available. So they need eyes with an architecture that makes the most of what light they do receive. Moth eyes don’t aggregate images like Dragonfly eyes, but rather superimpose multiple images through a technique called refraction superposition.

While a moth’s eye looks superficially similar to an apposition eye, internally it’s radically different. Light is refracted through the lens element of each ommatidium in such a way that it emerges from the bottom of that ommatidium at the same angle, and then onto a teeny little retina.

Refraction Graphic So in a weird way, it’s- internally at least- a little bit more like our eye. The ommatidium achieves this by acting like a two-lens telescope, which it accomplishes by containing a cylindrical lens with a gradiated refractive index that continually bends the light as it travels through the tube. In other words, the refractive index of the lens varies depending on where it is along the tube. (This is way, way cooler than your telescope.)

The guys who figured this out BTW, an Austrian physiologist names Sigmund Exner, did so way back in the 1880’s, but it wasn’t until the 1970’s that sophisticated-enough instrumentation* was developed to prove him right.

*The interference microscope.

Reflection Superposition

Superposition is such a great idea that it’s evolved multiple times in compound eyes, using very different architectures. Lobsters for example, also live in light-poor environments where superposition helps improve image formation. But refractive lenses are trickier to effect in water, which has a much higher refractive index than air. So the lobster’s eye achieves superposition via reflection. The insides of a lobster’s ommatidia are mirrored, and each facet uses reflection internally to focus the image onto the teeny retina.

Reflection Graphic More recently (1988) a third type of superposition compound eye was discovered in a genus of crabs (Macropipus) that use both crystalline lenses and parabolic mirrors, and is called the Parabolic Superposition Eye. Superposition eye of all (optical) types BTW need to be very spherical in form to superimpose the images correctly.

Wow. That’s a lot of cool bug-eyes. But what does all this have to do with flies?

Everything. Because flies have the coolest* compound eyes of all.

*Totally my opinion, and I have absolutely no authority or credentials to back it up. But it’s true. Really.

Neural Superposition

A fly’s eye is structurally very similar to an apposition eye, like a dragonfly’s or a bee’s.

HF Facets Closeup But it’s wired differently. It’s also a superposition eye, but it achieves superposition neurally. Remember when we walked through the structure of the apposition eye, after the images were received by the 8 rhabdomeres, or photoreceptor cells, they were joined into a single rhabdom which was connect by a single axon, or nerve fiber, to the lamina, or top nerve layer below.

But in a neural superposition eye, the rhabdomeres stay separated clear down the tube, and emerge from the bottom as 7* separate nerve fibers, each of which is then joined at the lamina to 6 other fibers from rhabdomeres from the 6 adjoining facets surrounding it which are angled in the same direction**. Neural GraphicSo the image received at the lamina is 7 times brighter than the image at any given photoreceptor, which helps flies form superior images in low light conditions. And because superposition is achieved via wiring (vs. optics) the shape of the eye doesn’t need to be as strictly spherical, allowing for greater flexibility of form. Neural superposition eyes are are also thought to be even more sensitive to motion detection than apposition eyes and might experience an improved signal-to-noise ratio.

*There are 8 rhabdomeres, but #7 and #8 sit on top of each other in the center, and are effectively one.

**The angles between individual rhabdomeres in a given ommatidium are the same as the angles between adjacent ommatidia.

Frontal Facet View Now that’s a cool eye.

Halteres, pulvilli, neural superposition eyes- they’re just 3 of a whole slew of amazing features in Houseflies. As is over and over again the case in this project, each little living thing turns out to have this amazingly cool story when you stop, check it out and learn a little bit about it.

Here’s something kind of gross about me: I routinely swat bugs with my bare hand. I’m not proud of it, but a couple of my apartments in college were so bug-infested that if I’d had to search for a tissue or swatter every time I IMG_3167 spotted a roach, I’d never have managed to eek out a degree in 4 years. The habit’s stuck, and over the years I’ve become not-half-bad at hand-swatting flies. But this gal had been a fine subject, posing patiently on my driver’s-side window as I fumbled with focus and lighting to get my shot. I opened the car door, gently shooed her out, and walked across the lot to my office.


KanyonKris said...

Now I'm going to have a hard time killing flies, even though I find them annoying (if they just wouldn't land on me it would improve our relationship). Why? Because now when I pull back to swat one I'll be thinking of all the cool technology they have - and because I appreciate cool tech - it will be akin to smashing an iPhone with a hammer.

But I'll be happier today because I can see. Vision is indeed way cool.

Anonymous said...

Wow - I missed your earlier posts about vision so this one was an "eye opener". Ha, I crack myself up. Sometimes your post about bugs/spiders creep me out but I found this one fascinating (with extremely cool graphics).

I never realized there were so many different types of eyes, particularly in the insect world. No wonder I have such a hard time killing them - they are watching me move in slow mo. BTW, your camera is killer.

So, comparing neural superposition to apposition, what difference does this make in seeing color?

mtb w

Enel said...

Epic post. Very interesting stuff.

Watcher said...

mtb w- that’s a great question. The short answer is that color vision in flies doesn’t seem to have been nearly as well-studied as color vision in bees, dragonflies, and butterflies, and so it doesn’t seem clear how many colors most flies can actually see. (See this paper for more details.)

But I’ll hazard a conjecture; since the color-receptors of insect eyes seem to work better in brighter light (as they do in vertebrates, such as us) it would make sense that neural superposition, which effectively brightens images in dim light, would improve color vision in those conditions. So while we don’t know how good flies’ color vision is, I’d guess that neural superposition makes it better under dim lighting conditions than it would be otherwise.

cmsparks said...

Coming late to this party, but... so where in the evolutionary chain was the divergence to camera type and compound type eyes? Any idea where that can be traced back to?

Great post.

Aaah diffraction, usually it makes my days easier, somedays not so much.

Watcher said...

cmsparks- They evolved completely independently. The eye has evolved dozens of times (how many is way up for debate- I’ve read ranges from 40 to 70) along 9 basic architectures. The simplest eyes are just photosensitive patches that that determine light or dark. Any curvature of that patch can also provide directional information (where the light or shadow is coming from.) Compound eyes evolved from simple convex-surface eyes, camera-style eyes from concave ones.

BobT said...

Thank you for a very lucid discussion of the differences among apposition, superposition, and neural superposition compound eyes. Your graphics are excellent and really serve to get the details across. Are these available in the open literature so that they can be used by others with appropriate attribution and permissions? Thanks again for your post.

Watcher said...

BobT- The graphics in this post, and as a rule the graphics in this blog, are my own originals unless otherwise noted. You’re welcome to use them however you like. If you do so, I’d appreciate it if you gave attribution to this blog, but if you forget or otherwise blow off doing so I won’t sue you or send you nasty emails. This is a non-commercial project, and I’m happy if the graphics or content are of use to others.

Keep in mind that I have zero qualifications or scientific training, and this blog almost certainly contains many errors.

Lidia Badarnah said...

Nicely explained topic.
Did you make the drawings?
Is it possible to have your permission to use the schemes of the compound eye in my PhD dissertation?

I'm a PhD candidate at Delft University of Technology.

looking forward your reply

Watcher said...

Lidia- yes the graphics are mine (with the clearly marked exception of the human eye diagram, which is courtesy of

You're welcome to re-use any of them in your dissertation. If you do so, attribution to this blog is appreciated but not required. (I covered graphic re-use in this post.) Good luck with your dissertation.

Anonymous said...

do you know what a facet eye is??