Horseshoe crabs aren’t really “living fossils”

September 14, 2012 • 7:48 am

There are four species of horseshoe crabs in the world, with the most familiar to Americans being Limulus polyphemus, the Atlantic horseshoe crab. Here’s a group of them mating:

They are arthopods, but not “true” crabs, which are in the subphylum Crustacea.  Horseshoe crabs are in a different subphylum, the Chelicerata, more closely related to spiders and scorpions than to the crabs we know.

A good website on the group describes its morphology, behavior, and medical uses (yes, they’re useful in making drugs; see here).  Plus they’re really cute. Here’s a description of the legs of modern animals:

The horseshoe crab has 6 pairs of appendages on the posterior side of the prosoma. Five pairs of walking legs or pedipalps enable the horseshoe crab to easily move along benthic sediments. Each has a small claw at the tip except the last pair. The last pair of legs has a leaflike structure at the terminal end that is used for pushing and clearing away sediments as the crab burrows into marine bottom. The base of each leg is covered with inward pointing spines called gnathobases that move food towards the mouth located between the legs. As the legs are moving, food is crushed and macerated. There are also 2 small chelicera appendages that help guide food into the mouth.

And this is what they look like upside down.  Note the five pairs of walking legs, which are “uniramous,” that is, each leg comprises a series of segments attached end to end:

Horseshoe crabs are famous for having changed little in morphology since they first show up in the fossil record over 400 million years ago (!): they are thus called “living fossils”. (In fact, the title of Richard Fortey’s new book on living fossils is Horseshoe Crabs and Velvet Worms.  Well, living species are remarkably similar in general external features to their long-dead (and extinct) relatives, but there are distinct differences, and of course we know nothing about the difference in their internal features, nor in the structure of the DNA of ancient species (go here for more criticisms of the notion that horseshoe crabs are living fossils).  Nevertheless, there is surprising “stasis” of morphology over a very long period of time, and we’re not sure why that is.

A new paper in the Proceedings of the National Academy of Sciences  by Briggs et al. (reference below, the “Briggs” is Derek Briggs of Burgess Shale fossil fame) casts further doubt on the “living fossil” status of horseshoe crabs, for not even their external morphology has remained reasonably similar. In particular, their legs have changed drastically.

The authors examined a fossil horseshoe crab found in Herefordshire (UK) dated at about 425 million years old.  This is the specimen, which doesn’t look promising, but they’re able to make out telling details of the appendages:

And what they find is that the limbs, unlike the uniramous limbs of modern horseshoe crabs, are biramous: each limb branches into two parts, with each branch comprising a series of segments attached end to end.  As the paper describes,

Appendages 2 to 6 project just beyond the head shield (Fig. 1 F and H). Appendages 2 to 5 are biramous (Fig. 1 F and I). The inner rami (endopods) insert in a series posterior of the chelicera, surrounding a raised central area occupied by the mouth. The outer rami insert along the outer margin of the ventral body wall. There is no evidence that these two rami were connected by an elongate limb base like the coxa in Limulus. Nonetheless the rami clearly represent elements of the same limb rather than successive limbs alternating in morphology.

Here’s a reconstruction showing first the group of biramous limbs (each bifurcated limb has a green section and a blue section) and then two pairs of limbs, with one part attached, as noted above, to the outer margin of the body wall, and the other part attached further inside around a raised ventral area (the “inside limbs,” as you can see from the above diagram of a modern crab, are the ones that remain). Phys.Org describes how these reconstructions were made:

The name of the new fossil, Dibasterium durgae, refers to the double limbs and to Durga, the Hindu goddess with many arms. It was reconstructed in three dimensions by stacking digital images of physical surfaces exposed by grinding away layers in tiny increments.

Two pair of limbs.  Modern crabs are missing the blue member of each pair.

What has happened, evolution-wise? Obviously, an entire set of limbs, the outer member of each biramous pair, has disappeared somewhere during the evolution of modern horseshoe crabs.  We don’t know exactly why this happened, or, if it was due to selection, what form of selection would favor such loss.  But the authors do speculate about the genetic basis of the loss.

In particular, in modern crabs the areas where the outer limbs would have been express the gene Distal-less (Dll) only in a transitory fashion, while the inner limbs—the ones that are present—express more marked expression of Dll.  The gene is known to be required for limb formation in other arthropods, like the fruit fly Drosophila, so it’s possible that the loss of the outer set of limbs somehow occurred by loss of Dll expression at some time in the distant past.  The authors emphasize that this is a hypothesis that is untested, for of course there’s no way we know of to see what gene expression was like in the ancient biramous fossils.

I suppose along with that theory goes a “macromutational” corollary that perhaps a single mutation in the Dll gene changed its expression so that the outer set of limbs was eliminated in one fell swoop.  If Dll is indeed involved, I would find the single-mutation scenario unlikely.  When we see a change in animals that looks like it could be caused by a single mutation that has similar effects in modern species (like the Bithorax mutation in Drosophila, which causes the appearance of four wings instead of the usual two), it’s dangerous to say that, for example, the gain or loss of wings in insect evolution was due to single mutations at a gene like Bithorax. Such changes could well be due to the accumulation of several to many genes of smaller effect.

Regardless of its cause, the difference in limb number between ancient and modern horseshoe crabs shows that what are regarded as “living fossils” may well be at one extreme of “changeability” among long-lived groups, but they have by no means remained completely unchanged.


Briggs, D. E. G., D. J. Siveter, D. J. Siveter, M. D. Sutton, R. J. Garwood, and D. Legg. 2012. Silurian horseshoe crab illuminates the evolution of arthropod limbs. Proceedings of the National Academy of Sciences, online publication, 10.1073/pnas.1205875109.

28 thoughts on “Horseshoe crabs aren’t really “living fossils”

        1. Didn’t we all eat the odd mud pie when we were kids? Isn’t the taste of mud, however experienced in the first place, familiar to most adults?

          Certainly, it is to this adult: some years ago I bought some Thai black rice. On cooking, it tasted of MUD. Yet at that date I was decades past the mud pie stage of life.


  1. Is there any indication of variability of legs amongst other then-contemporary crabs?

    That is, do we have any indication of whether extra legs were universal, common, rare, etc., and how those proportions changed over time?


  2. Very interesting! I’ve long been suspicious of the label ‘living fossil’. Have any biologists or ecologists argued that some animals occupy niches that are so stable that their evolution has reached an equilibrium state and in some sense ‘stopped’? It seems like genetic drift alone makes that unlikely, before we even get to sexual selection, etc, suggesting that the ‘living fossil’ idea exists mostly in the popular press.

    Still, 400 million years of fairly stable morphology is it pretty neat by itself!

    1. While limb morphology was quite different 400 million years ago in the Paleozoic, one could still argue that these arthropods are ‘living fossils’ by looking at Jurassic (Mesozoic) forms that I believe ARE quite similar to modern species. I believe that horseshoe crabs from a little more than 150 mya are similar enough that taxonomists place them in the same genus, Limulus, as the modern species. That is pretty remarkable – all the mammals from the Jurassic are QUITE distinct from any modern mammal.

      1. Sure. It’s a relative thing. You’ll always find differences if you go far enough back into any animal’s ancestry.
        What’s interesting about horseshoe crabs is how far back you have to go to find significant differences.

    2. In his book, Fortey makes it clear he doesn’t much like the term “living fossil”: there is no such thing as no change, since there’s always the DNA changes that we don’t see. If he uses the term at all, it’s more for animals that were first discovered in the fossil record, after which live ones were found.

      I just happened to finish his book last night. It was a great read. And, I am thinking of going down to Delaware once, to try and see the horseshoe crabs come ashore, in spring.

    3. Still, 400 million years of fairly stable morphology is it pretty neat by itself!

      Neat? Most assuredly. But hardly unusual!

      Cladoselache, the Devonian-era ancestor of modern sharks, is still easily recognizable as a shark.

      Trigonotarbida, also from the Devonian era, are the spinneret-less ancestors of modern spiders — and, again, unquestionably spiders. And we have amber-encased spiders from the early Cretaceous, complete with spinnerets.

      And Lituites, from way back in the Middle Ordovician, are definitely quite similar to their modern descendants, the nautiloids.

      Here’s hoping others might choose to chime in with their favorite living fossils….



      1. On the other hand, I was down at the NHM ( a few years ago to talk to their resident coelacanth expert. When going through the drawers to examine a variety of relatively recent (Jurassic) coelacanths, I was struck by their dissimilarity to the modern coelacanth in it’s formalin bath in the hall.
        We didn’t discuss it, but I suspect that Dr Forey wouldn’t have been a fan of the term “living fossil” either.

        1. I suspect that Dr Forey wouldn’t have been a fan of the term “living fossil” either.

          I can understand the reluctance…but we sure could use a term for species with notable morphological stability. “Living fossil” isn’t all that awful, and I’ve not heard any other alternates offered up.

          I’d put this one in the same bin as “Big Bang”
          and “God(damn) Particle.” Lemons, to be sure, but ones that lend themselves well to lemonade.


          1. s/species/genera/
            or some other taxonomic level ; but not “species”.
            “Taxon of unusual morphological stability” isn’t particularly pretty, but does do the job.
            In a related question, excluding single-celled organisms, what is the longest-enduring morphological taxon you can think of? My pint-voucher is sitting here on Linugula shaped brachiopods, from the mid-Cambrian to the present.

  3. So, the questions that come up include: Are the modern variants considered the same species as the ancient? How far back can the “modern” variant be traced? Is it “old” but not “ancient”?

    More generally, when do you consider a descendent of an ancient lineage to have speciated? Is there such a thing as “ancient species” of anything? Heck, are humans the same species as our homo sapiens sapiens ancestors from 50,000 – 100,000 years ago?

    Seems to me that the whole notion of “species” is kinda laid bare here.

    1. Sorry that I can’t give you the reference, but somewhere, within the lsst year, I read some remarks on this. The gist was that the concept of species is very dubious when categorizing organisms over geological time periods.

      There are living species that challenge the notion, in fact. Across Canada, there are two spruces, the Engelmann spruce (Picea engelmannii) in the west, the white spruce (Picea glauca) further east. These intergrade, but (according to a forester I used to know) no one is certain if these are merely two extremes of one species, or two good species with a hybrid swarm in the middle. (It’s possible that DNA analysis has since clarified the situation.)

    2. Well, of course the biological species concept (reproductive isolation) is impossible to apply to extinct organisms. But it’s not the only species concept. Paleontologists are forced to use a morphological species concept, but ti generally seems to work out OK.
      Of course, if every animal that ever lived was still alive, it would be more or less impossible to draw firm lines. We’re helped, in terms of classification, by the fact that most of them are gone forever.

  4. The authors emphasize that this is a hypothesis that is untested, for of course there’s no way we know of to see what gene expression was like in the ancient biramous fossils.

    Agreed. However, there’s nothing preventing us from mutating or knocking-out a suite of candidate genes in living horseshoe crabs and arriving at a plausible scenario as to what happened.

  5. …in modern crabs the areas where the outer limbs would have been…

    I take it that you’re referring to horseshoe crabs here, rather than, well, crabs?

    Sorry to nitpick, but if developmental genetic work has been done on horseshoe crabs, I wasn’t aware of it, and I’d be interested in reading about it.

  6. To me, they bear an irresistible resemblance to World War 2 Wehrmacht helmets. I think even the color is about right.

    With those eyes, it looks like they can see sideways, but not straight ahead, which leads to mental images of them bumbling around bumping into things, backing up and trying again. Not that I’ve ever seen one locomote, though.

    1. There are eyespots in addition to the big compound lateral eyes… but the horseshoe crabs I’ve seen do indeed spend a lot of time bumping into things.

  7. Modern crustaceans–including ‘modern crabs’ do have biramous appendages on the head and abdomen. Most decapods (inc. crabs) have uniramous walking legs only.
    Trilobites also had biramous appendages. Now that there’s an ancestral chelicerate with biramous appendages it’s looking more like it’s the ancestral trait for arthropods as a group. That means that uniramous appendages have evolved multiple times (chelicerates, myriapods, hexapods at a minimum).

    The reconstruction labeled “two pairs of limbs” is actually better considered a single pair of appendages each with two separate branches (rami).

Leave a Reply