Arthropod Silk

Silk production has apparently evolved several different times, if my cursory internet research is to be believed. It’s pretty amazing that such a complex set of glands, ducts, spinnerettes, etc. evolved multiple times.

But if silk is so useful that evolution has produced such complexity independently more than once, then why does it only appear among the arthropods? Why not chordates or echinoderms or whatever? Or am I wrong and are there non-arthropod silk spinners?

Potential utility does not mandate the development of a biological function. There are many, many conceivably useful features which just never happened to be exploited by any given branch of living things.

Just a guess, but there’s probably a natural diameter for a silk fiber, and strong though the stuff might be, such a thin fiber isn’t strong enough to be of use to a large animal. Only to something the size of a bug is it practical.

I’m not sure silk would be as useful for most chordates as it is for arthropods. We don’t pupate (an immensely common use for silk) or lay soft egg clutches on land (hard egg shells are probably more effective, actually). Silk is also very biologically expensive, so it’s not necessarily the best tool for the job (but spiders have a hammer so all their problems look like nails).

The closest analogues I can think of are hagfish and cave swiftlets. Hagfish release a substance that turns seawater into slime. It’s not unweblike, but obviously lacking the detail and elegance of silk. Cave Swiftlet saliva hardens, which they use to make nests. Not exactly silk, but it’s sort of a naturally-produced chordate structure and at least superficially similar to an invertebrate egg sac.

It also wouldn’t surprise me if it turned out there was some common overlap between building chitinous exoskeletons and producing silk, so arthropods share some of the basic silkish-producing structures that chordates have just never evolved.

I’ll sign on to this idea, as well.

Mammals spin silk from thousands of pores. It’s a slow process, and the utility of the silk is limited once removed, but while still being spun it’s useful for insulation, blocking the sun, sexual attraction, and rock music.

Another fair analogue to silk. Hair has other uses, as well, such as protection (porcupine quills), weaponry (rhinoceros horns), algae farms (sloths), and sensory information (whiskers). Invertebrates make limited use of hairs, but mammals dominate in this field.


I tried to make a funny that was neither funny nor correct.

Or it would just take a bigger evolutionary leap, which has just never happened and is unlikely to. The strongest spider silk woven on a large scale is strong enough to be useful for bulletproof vests, so it ought to be useful for animals. But…evolution wouldn’t “know” that; it has no foresight. If it would take a lot of evolution before proto-silk becomes a version of silk strong enough to be useful to larger animals, then it’s not going to happen; natural selection doesn’t work on what will be useful, only on what is useful.

Another guess. Arthropods perhaps all have Gene X, which with a very few mutations turns into Gene Y which produces some silk precursor that’s potentially useful enough for evolution to select for. Mammals don’t have Gene X, and it would take many more mutations before their closest gene would become something like Gene Y. So, it’s simply much less likely that the ability to produce silk at all will ever appear.

I don’t think this is right. What Chronos was saying is that a strand of spider’s silk only works on the scale of a strand of spider’s silk. There are very complex biochemical reactions going on in and around a spinneret which simply don’t scale up. A water hose with a bigger nozzle sprays more water, but a over-sized spinneret would just squirt out useless goo.

As you said, even the thin strands of existing spider silk can be useful for larger animals, but it would take thousands and thousands of spinnerets working together to make something reasonably useful (perhaps if it was twisted into a thread or something similar). Considering a typical spider can only manage a half dozen, it would be pretty amazing for a rat to evolve a few hundred seemingly useless, resource-intensive biochemical factories.

Thinking about it now, the biggest spiders tend not to use webs all that much, either. Tarantulas and similarly big-bodied spiders don’t spin ornate webs for catching prey, but may lay tripwires and such. For hunting, they pretty much pounce and bite. Those spiders are only about as big as the smallest mammals. That may speak to the impracticality of webs for larger animals and / or the efficacy of just plain pouncing and biting.

That’s partly what I meant by “bigger evolutionary leap”. If some large complex setup is necessary for silk to be useful for a mammal, then evolution isn’t at all likely to produce silk weaving mammals because it doesn’t make huge leaps like that.

Silkworms are not arthropods and they certainly make silk. Anyone who is interested further in this should listen to the segment, “Outsourcing spider silk” at Essentially they made gmo silkworms spin a silk that had a small amount (5%, IIRC) spider silk in it that resulted in substantial improvement in strength and other properties. The big advantage of silkworms is that they spin a large quantity of silk all at once, instead of a little bit at a time.

Wait, what?

Arthropods include insects, right? And moths are insects? And a silkworm is a baby moth, right? So how are they not arthropods?

I think you’re confusing arachnids with arthropods.

And that was the same mistake for which I deleted my ‘hilarious’ post above.

Evolution is, at its heart, driven by a random process - mutation. Thus, nearly always, the only real answer to the question “Why did evolution not do…” is “because it didn’t”. We can speculate, as you see here, about how something that seems useful may not be as useful as we think, or how the same function is done by something else, but as has been said, just because something’s useful doesn’t mean it’s going to pop up by a random mutation.

We could wax philosophical here about fitness landscapes and local versus global maxima, but that’s really just saying the same thing.

Interesting responses, thanks everyone.

I guess the question that I failed to articulate is, is there something special about arthropods that predisposes them to evolving silk production? Some have touched on that - small size, something about having an exoskelton, etc. But I wonder if anyone has more specific knowledge on that.

As an aside, I’m not sure I agree with the silk/hair analogy. Silk is produced on demand for various purposes, none of which has an analog in hair, afaik. Also, chemically, keratin and silk aren’t all that similar. I mean obviously they’re organic polymers but that’s about where the similarities end, no?

I think the similarities really shine in industrial uses, where fibers can be manipulated. It’s true that their direct biological uses don’t have much overlap. As I said, spiders use hairs for hair-like things and webs for web-like things.

If I were to sit down and try to approximate silk in a mammal, I might use hair as the basis. Imagine a squirrel with a long, pre-grown hairs coiled up in modified follicles with some sort of adhesive on the tip. If being chased by a particularly agile predator, a quick flick of the tail against a tree branch and the squirrel could jump away, with a drag line to hang from. I don’t know that the mechanics work out, and it’s certainly not practical (better to invest in faster feet and better eyes or ears). It seems particularly unlikely an adaptation (though so does the claw of a pistol shrimp, so there’s that).

Oh, I thought of another possible mammalian analogue to silk. Males do forcefully expel fluids that coagulate rapidly when exposed to air. Again, I can’t imagine it being used as silk is, but I never underestimate the font of human ingenuity.

They are both proteins, therefore they are chemically very similar. It’s the three-dimensional fold that is different (coiled-coil in keratin, beta-strand in silk)

Yes, I was confusing arthropods with arachnids. But that broadcast was most interesting.

I think hair is a good analogue for silk. But growing hair and spinning silk are very different. Hair doesn’t grow on demand, and it grows too slowly, and often not very long, for the kind of independent utility that a spider web has.