Ancient spiders


Spiders are amazingly sophisticated animals, and probably the premiere complex adaptation of modern spiders is the ability to spin silk. They have multiple internal glands that can produce multiple kinds of silk — webs contain different kinds, from structural strands to adhesive strands, and other kinds are used for spinning egg cases and for wrapping prey — and they are sprayed out through small spigots mounted on swiveling spinnerets, which are modified opisthosomal (abdominal) limbs. Obviously, these detailed features did not spontaneously appear all at once, but had to have evolved progressively. A couple of fossils have recently been described that reveal a) silk spinning is ancient, from at least the Permian, but that b) these early spiders did not have the full array of modern adaptations.

Here is a pair of fossils: Permarachne novokshonovi, from the Permian in Russia, and a more recent specimen, and Palaeothele montceauensis, from the Carboniferous in France. Both are eight-legged arthropods, and if you saw one scuttling about now you wouldn’t hesitate to call them spiders. There are some differences, though: Permarachne in particular shows a little less tagmosis, or fusion and specialization of segments, than we usually see in spiders, and it also has that prominent flagellum (which is completely different from a bacterial flagellum!), a long segmented ‘tail’ covered with sensory hairs that was probably a sense organ; it has no sign of a web-spinning function.

(Click for larger image)

Paleozoic Araneae and Uraraneida. (A-C) Permarachne novokshonovi, Permian of Russia, PIN 4909/12. (A) Holotype part in rock matrix. (B) Explanatory drawing of A. (C) Close-up of flagellum showing whorls of setae. ch, chelicera; cx, coxa; fe, femur; mt, metatarsus; pa, patella; pl, ventral
plate; st, sternum; ta, tarsus; ti, tibia. (D) Palaeothele montceauensis, Carboniferous of France, In 62050a, X-ray CT scan showing appendages buried in the rock matrix; note, anal tubercle (arrowed)
is not a flagellum. (Scale bars: B, 1 mm; C and D, 0.1 mm.)

What about the production of silk and webs in these old spiders? Here’s another specimen, Attercopus fimbriunguis, a 376 million year old fossil. It’s a little less dramatic because these are fragments of cuticle that have been carefully extracted by dissolving the rocky matrix with acid; it means, unfortunately, that it is more fragmented, but the advantage is that now we can zoom in microscopically and see far more detail in the structure. What we can now see in pieces of the ventral plates of the opisthosoma are small spigots, and in a few cases, there are even strands of spider silk still extended from these pores. In F, there’s also a nice shot of a chelicera (fang) from the spider — it’s wicked sharp, but the small holes seem to be preservation artifacts, and there’s no sign that venom secretion, another important spider adaptation, has evolved yet.

(Click for larger image)

Attercopus fimbriunguis, Devonian of New York (localities: G, Gilboa; SM, South Mountain), macerated from matrix with HF and slide-mounted. (A) First-described “spinneret,” G 334.1b.34; darkness of cuticle reflects number of layers, so this fragment is folded over
twice. (B) Palpal femur, SM 1.11.12; arrow indicates patch of distinctive spinules. (C) Piece of cuticle from corner of opisthosomal ventral plate showing setae, spigots, and possible silk strand, SM 1.11.4.
(D) Close-up of E showing possible silk strand emerging from spigot shaft, SM 1.11.4. (E) Flagellar structure with 12 segments (including possible distalmost) from original Gilboa locality; segments show distal
collars and setae, G 334.1a.4. (F) Close-up of cheliceral fang showing a number of holes (arrowed), the most distal of which had been interpreted as a venom-gland
opening, G 329.22.9. (Scale bars: 0.5 mm, except F, 0.25 mm.)

One of the critical observations here is very simple: no spinnerets. These spiders did not have the modified limbs with sets of spigots that we see nowadays, but instead, had a series of spigots arrayed across the bottom of the abdomen. They almost certainly were not able to make webs: what they could have done was produce sheets of silk, of the kind that could be used to make egg cases or wrap around prey. These are another example of a transitional fossil, forms that have only some of the capabilities of a later organism.

(via Cheshire, who promises to have his own post on this paper soon.)

Selden PA, Shear WA, Sutton MD (2008) Fossil evidence for the origin of spider spinnerets, and a proposed arachnid order. Proc Nat Acad Sci USA 105(52):20781-20785.


  1. recovering catholic says

    Let’s see, which is easier here–

    …to just forego reading all this technical stuff entirely and just make some biblical reference to the factoid that god made each species individually and immutably OR

    …to have studied biology, geology, and evolution; to carefully read the post and read up on the parts you might be a bit rusty on; to think about the post critically from a scientific point of view.

    No contest!

  2. 'Tis Himself says

    These are another example of a transitional fossil, forms that have only some of the capabilities of a later organism.

    But…but…but…all the best creationists say there are no transitional fossils. Do you mean they’re wrong?

  3. recovering catholic says

    @1 Oh yes, it also takes effort to check up on someone else’s rustiness as well! (You’re right, of course…)

  4. 'Tis Himself says

    Wasn’t Carboniferous before the Permian?


    Carboniferous from the end of the Devonian, about 359.2 ± 2.5 Ma (million years ago), to the beginning of the Permian, about 299.0 ± 0.8 Ma, which ran until to 251.0 ± 0.4 Ma.

  5. says

    It’s still micro-evolution if we don’t find the LCA between insects and spiders.

    Plug your ears and repeat after me:

    There are not transitional fossils because I don’t want there to be.
    There are not transitional fossils because I don’t want there to be.
    There are not transitional fossils because I don’t want there to be.

    What is the evolutionary advantage of fused segments of body parts? I am seriously curious about it.

  6. KnockGoats says

    The only time I ever worked in biology, it was on spiders – specifically, modelling the web-construction capabilities of Araneus diadematus as my first post-doc. Can’t say it was particularly successful, although a couple of journal papers resulted; and my successor got quite a bit further, using genetic algorithms (I was using GOFAI methods and a bloody useless implementation of an excellent programming language). However, it did help to confirm me in what still seems an important (though not original) insight: how far cognition depends on using the environment as a memory, problem-solving and action-directing resource. My boss thought the spider must have an internal “cognitive map” with something close to Euclidean properties, in order to work out what to do at each stage. From the start I was doubtful about this (from earlier work on human wayfinding), and observation of the spiders confirmed me in my view. At the start of the web-building process, the spider wanders around seemingly aimlessly, attaching its thread to convenient points it encounters, but as the process continues, its actions are increasingly determined by what it has already built – which effectively acts as its “memory” for where it has got to. It also gets less and less able to deal flexibly with interference – such as burning away strands of the web. I ended up modelling only the final stage – the “sticky spiral”, which was amenable to the kind of linear-hierarchical algorithm my methods would support – the earlier stages needed something more like a set of simple, autonomous rules. Ah, if I had the time over again…

  7. Rowen says

    But, that’s just microevolution. Show me a transitional fossil between a whale and spider!!!! HA!!! I was right! JEbus is Gawd!!!

  8. Brad D says

    My wife would still scream and beg me to squish these guys, but I would still try to scoop them up in something and toss them outside.

    What spiders ought to do is to evolve more cuteness, that could be a good survival tactic, at least while there are so many humans around.

  9. Sclerophanax says

    Attercopus, eh? Makes you wonder if the discoverer was a fan of Tolkien or just well versed in old-fashioned English.

  10. JakeS says

    Thanks PZ. Now I have a ready conversation topic the next time I’m with someone and we see a spider!

  11. clemmie says

    While at Oxford Uni 20 years ago, looking through past Finals papers to get an idea of likely or possible questions my favourite was:

    “Spiders are great. Discuss”.

    I wonder how many people attempted an answer! This would have to be in the essay now! Very interesting.

  12. Nerd of Redhead says

    Why is delusional Charlie so infatuated with us? Charlie, do you have the cojones to remove this site from your bookmarks, and stop coming here for say 3 months?

  13. says

    Have you ever watched a spider spin a web? Years ago, when The Royal Spawn was a boy, we sat in our kitchen watching a spider spin a web right next to the back door. This was a smart spider. It knew to spin its web next to our floor-to-ceiling glass door because we kept the light on in the kitchen. Bugs were attracted to the light, and they’d get caught in the web. This spider ate well that night. It was fascinating to watch it spin away, catch a bug, spin around it, and run back to the center of the web waiting for the next entree.

    Not long ago, The Count told me he was bitten by a spider the size of a dinner plate. I’ve seen big spiders before, so I believed him. He had me going until he told me the spider growled at him. Spiders don’t growl! At least not outside really bad monster movies…

  14. amphiox says

    #4, #8:
    Has there been some new reworking of arthropod phylogeny? I didn’t think that insects and arachnids were on the same branch.

    Or is this part of the running joke? (My irony meter has long since expired from caffeine overdose.)

  15. says

    Wait for the wackos at Answers in Genesis and the creation “museum” to get a hold of this. I fully expect they’ll be claiming that this specimen is a perfect example of a vegetarian spider from “before the fall.”

    The Discovery Institute is still trying to pretend that they aren’t a religious organization with a desire to teach creationist BS in the classroom, so they’re likely to ignore this altogether.

    Creationists seem to be using the binary search pretty frequently to make their determinations:

    1. Find a complex biological process or organ look like something they can call “irreducibly complex.”
    2. Divide the process or organ into two parts. One part will be the one they cannot call “irredicibly complex” and the other they will.
    3. Mine just enough scientific literature to make sure they aren’t caught right away.
    4. Declare victory for ID!
    5. Claim any scientific researcher that shows contrary findings is part of the evolution conspiracy and misapprehended them.
    6. Attribute any detail they can’t explain to “the fall.”
    7. When the case is really sealed up and they look truly foolish for continuing to support it, go back to step 1.

    They also tend to do the same thing with fossils:

    1. Mine the scientific literature for new fossil discoveries.
    2. Examine the fossil depicted. Divide all fossils into only two sides. Example: If it’s a hominid and does not “look” human enough to them, claim it’s an ape. If it “looks” human (to them), claim it’s a human. Case closed.
    3. Deny that any transition is shown, or if the transition is clear to even an elementary school student claim that the scientist that found it is part of the evolution conspiracy and that the fossil is a fraud.
    4. Attribute any detail they can’t explain to “the fall.”
    5. When the case is really sealed up and they look truly foolish for continuing to support it, go back to step 1.

    Note that neither include a step “perform original research” or “get the research published in a peer-reviewed journal of science.”

  16. Jhon says

    Sorry myers,
    I dont think this is a good time for talking about Spider.
    there is a war crime out there.israel, and even the UN can’t stop them.
    it’s time for us, educated people to stop this barbarism act.
    I believe those who have brain will never set up a WAR.

  17. Lynnai says

    The rock matrix was desolved in acid? Neat. But surely with the sort of sedimentary rock formation prone to fossil making the chemical compounds of the matrix and the fossil must be pretty close…. what I’m saying is, is that technique as a big a pain in the ass as I think it is???

  18. Jhon says

    But I thinks it very interesting though…
    Spider, I love the way they make their web.
    absolutely fantastic, the greatest Architect ever.

    But Still, We have to pay attentions to they current situation on the middle east.
    this is involving our security matter as an affect!
    we Have to help the innocent.
    Open the youtube video and type this keyword ” israel attack palestine”

  19. Insightful Ape says

    Didn’t the creationist dentist from Texas say recently that given all the time in the world, he couldn’t make a spider out of a rock? I somehow doubt this will interest him, though.
    Jhon-no time is a bad time to discuss science. Speaking of times, where were you all these years when Hamas was firing rockets into southern Israel, destroying their economy completely?

  20. Nerd of Redhead says

    All PZ need do is ask.
    He refuses.
    I persist…

    Check the Dungeon on the masthead. You have been asked. No excuses. Go and stay away.

  21. Insightful Ape says

    Charlie-case for Intelligent Rubbish noted. Of course as recovering catholic said-reading the article and discussing the findings is hard, so I don’t expect you to. It’s a lot easier to say, I don’t see how it happened, so god must have done it. Could it have been Brahma that did it, though?

  22. Insightful Ape says

    Charlie the troll-you and leeches must have had the same creator. You won’t just go away, need to be removed physically.
    Speaking of creators, you didn’t answer my question. You are maintaining that the spiders needed a designer, fair enough-I just don’t think humans and spiders have the same designer, but different designers. Do you have any problem with that?

  23. clinteas says

    and it also has that prominent flagellum (which is completely different from a bacterial flagellum!

    The fact that you felt you had to point that out made me LOL……

  24. Bacopa says

    I have a few huge wolf spiders that patrol the tops of the Bacopa plants in one of my tanks. I think they come down and hunt roaches at night. I have not seen a roach in many months. One bit me once and left a little wound that festered for two weeks.

    Biggest spider around these parts is the banana spider. They are huge and their webs are huger:

  25. Fernando Magyar says

    One bit me once and left a little wound that festered for two weeks.

    If I recall, there is a nice collection of preserved amputated limbs that had suffered severe necrosis due to spider bites from certain Brazilian Lycosidae at the Butanta Institute in Sao Paulo. These spiders can cause some very nasty looking wounds leading to gangrene. I wouldn’t leave such a bite untreated and festering, as much as I like the spiders…

  26. MIke says

    I love how clear the relationships between spiders, scorpions, and horseshoe crabs are made by the distinct flagellum and other features in the Permian spiders

  27. KnockGoats says

    Actually, you get plenty of spiders with 7 legs, at least in Araneus – not to mention 6, 5, 4… They fall off rather easily, and down to about 5, the spider does pretty well, though IIRC, if they lose the legs with which they usually spin the web, the latter becomes somewhat less neat.

  28. KnockGoats says

    Sorry, #46 was a bit vague – they use one pair of legs (can’t recall which) as measuring instruments during web-construction.

  29. says

    The impression I had was that spinnerets were supposed to have evolved at a point when the predecessors of spiders had more than eight legs and then the hindmost legs shrank and in the process got co-opted as a more efficient way of getting the silk where it was supposed to go, but here it seems the number of legs is already reduced to eight before spinnerets proper have evolved. Does this mean that at some point new legs were “added” to form spinnerets?

  30. Edgar says

    #48 is a good question, perhaps the leg-making genes on the abdomen were dormant and then activated?

  31. says

    My post is up at my weblog.

    It’s kind of interesting because there’s an extant group of spiders which retains abdominal segmentation as we see in the ancestors of spiders, so I guess I have two transitional species in the post for the price of one. I guess that makes me the Wal*Mart of evolutionary biology.

    I’ve seen creationist spider enthusiasts try to argue up and down that these aren’t transitional animals-and it never ends pretty. AIG’s probably going to ignore this one.

    As for question #48, the spinnerettes (as well as many other features found in spiders) are thought to be homologous to the abdominal legs found in the horseshoe crabs which are the most basal chelicerates. It turns out that genes for complex structures can be suppressed for quite a long time and then re-evolve when needed. Function is restored through mutations which either offset the original mutation, or a reversal of the mutation which disabled the structure. There’s a paper about this on phasmids that’s been begging me to blog it for awhile…so I’ll probably get on that sometime soonish.

    The genes for silk production in spiders tend to be associated with appendages. Some species of tarantula have silk spigots in their tarsal pads (the pads at the end of the end of the tarantulas legs which often look multi-colored in pictures) and this helps them stick better to whatever they happen to be clinging to at the moment.

    Ed Yong had a post on it awhile back:
    This paper is about Aphonopelma seemani. It’s a mygalamorph, and pretty distantly related to the subjects of my latest post. I’m not sure if they’ve persued research to see how many other species do this but I wouldn’t be surprised if this was common…especially in arboreal spiders.

    If you’re wondering about silk in other animals, a common adaptation in insects is for the silk glands to be evolutionarily fashioned out of salivary glands.

  32. says

    Cheshire, thank you for your explanation.

    Amazing that such features can lay dormant for so long, but I guess the segmented structure of arthropods helps with that. You also mention that it is “common” for silk glands in insects to be made out of salivary glands. AIUI insect mouth parts are heavily modified legs, so the connection makes sense, but I was under the impression that silk glands were introduced in the basal arthropod, but they have in fact evolved multiple times in different lineages?

  33. frog says

    Cheshire #50:

    Interesting. I think we may have some evidence of non-traditional evolutionary mechanisms, then. Somehow the gene ensembles must be keeping their stability without sliding into pseudo-gene status.