A mathematical (?) spider from the African deserts

May 1, 2014 • 5:37 am

A reader sent me the link to an old scientific article (from 1995) about a “mathematical spider” living in the Namibian desert.  It turns out that the adjective “mathematical”  is pretty misleading, but since the article was interesting I thought I’d give it a brief shout-out.

The paper, published by G. Costa et al. in the Journal of Arid Environments, (reference below, can join ResearchGate and get free copy, or ask me), is about a new species in the family of tube-dwelling spiders, Segrestriidae; the species is named Ariadna sp. (the “sp.” means “species not identified”, although it may well have been in the last 18 years).

24 specimens of this ground-dwelling spider were studied near Gobabeb, a well-known research station in the desert of Namibia. The spiders dig burrows in the ground from whence they venture to get prey. Here’s a picture of the spider:

Picture 1

Here’s its bleak habitat. Life is nearly everywhere on this planet:

Picture 2

Some gypsum casts of its burrows. There are about 2.5 cm to the inch, so the burrows are about five inches long. The spiders line them with silk.

Picture 1

Now for the “mathematical” part. For reasons yet unknown, the spiders pile stones around the entrance of their burrows. The stones are fairly uniform in size, and there are usually about seven, though the number ranges from five to nine. The centimeter scale is at the top, along with a Namibian five-cent coin for extra scale (useful only to Namibians!)

The interesting thing about the stones is that they are usually placed radially, with the narrowest parts near the burrow, and of fairly uniform size. This makes the burrow look a bit like a flower. Stones are clearly selected for size, as you can see by the surrounding stones. Perhaps the spiders use the biggest stones that they’re able to carry. The photo below shows a typical array of 7 stones.

Picture 3
The “mathematical” part, which the authors make a great deal of, is that the mean number of stones (and the mode) is seven, with other numbers distributed fairly symmetrically around that. Here’s the table showing the percentage of stones falling in each category:

Picture 4Well, that’s interesting, but hardly mathematical.  It doesn’t show at all that the spiders can count, and it’s not surprising in any way that the distribution is a bell-shaped curve. What may be going on here is simply that the spiders heft the largest stones they can carry, that their burrows are of a relatively fixed size because spiders are of a relatively fixed size, and the average number of spider-heft-able stones that can surround a burrow happens to be seven.  The spiders do apparently exercise a preference for quartz stones, though again this preference isn’t documented statistically.

What is more interesting to me is that the stones appear to be placed radially (though I’d like to see more photos and measurements), and, especially, that the spiders even bother to highlight their burrow this way. Why? The authors raise three possibilities:

a. Detection or attraction of prey. The authors suggest that “the stone ring could perhaps attract prey or facilitate the detection of prey by the spider waiting inside its burrow.” Well, maybe, though the attraction hypothesis seems more viable than the detection one. At any rate, this could be tested, even in the lab, by removing the stones and seeing fewer prey approach the burrow. But neither of these hypotheses seem really convincing.

b.  Strengthening the burrow and making it impervious to sand or debris. This seems more likely to me. The raised stones could keep dirt or sand blowing along the ground from entering and clogging the burrow. Again, this could be tested fairlly easily.

c. Deterrence of predators.  Here’s what the authors say:

The stone ring might be a way of reducing predatory risk. The evenly lighted circle around the burrow’s entrance could make this appear like, [sic] a black stone or a shaded area. On the other hand, spider holes may simulate the little black stones that are scattered over the gravel plain. The characteristic alternation of light and dark areas on the gravel ground complicates the detection of real burrows by predators.

That’s possible, but again requires testing.  Another possibility, which the authors don’t mention, is that the symmetrical pattern may help the spider find its burrow in a complicated patchwork of stones and ground. In other words, the stones could act as landmarks, much the same way that some ground-dwelling “digger” wasps recognize their burrows by the patterns of debris on the ground nearby. (That work, a classic study of animal behavior by Tingergen and Kruyt, showed that the wasps could be confused by simply moving the landmarks around a burrow.) This “recognition” hypothesis may not be likely, though, if the spiders’ vision is poor.

At any rate, it’s a cute behavior whose significance is not yet determined, but would seem to be tractable to easy experiments.

h/t: Hardy


Costa, G., A. Petralia, E. Conti, and C. Hanel. 1995. A mathematical spider living on gravel plains of the Namib Desert. Journal of Arid Environments 29:485-494.

37 thoughts on “A mathematical (?) spider from the African deserts

  1. Do both sexes do this? It would be a good way for one sex to advertise its quality to the other. Bigger stones = “bigger stones”. Or, “Hey, look at me, I can count!”

    1. Good question. The authors don’t mention this, so it’s implied that it’s not sexually dimorphic. Nor do they mention finding burrows without stones. But the short answer is: we don’t know.

    1. If the spider really did that, I would expect the number of stones to peak at 8. One per leg, in fact.

      Maybe there’s a high accident rate and there are a lot of seven-legged spiders around.

  2. It may be that the spider gains some advantage by making it’s burrow look like one of the unusual Lithops-like flowers (rock or stone flowers) native to the Namib, that would protect if from some predator, say birds? A little mimicry in action…?

    1. I was wondering whether there’s any chance the flower-like arrangement would attract pollinators, who could be prey, but it would depend on how those pollinators see these stones: not the same way we would.

      1. Any feature that would serve more than one purpose in helping to make a living in nature sould certainly be useful, and selected for. The pollinator idea is a good one, and testable, as is the camo idea…

  3. Far out, but hey: A radially symmetric flower-like structure could also lure in flying pollinators.
    Perhaps not that far out, since e.g. long-tongued flies in Aouth Africa can be fooled by artificial flowers.

      1. No worries – great minds & all that. Re: your comment below, one of my students found a (potentially new) species of trapdoor spider on Aldabra – it lives in loose coral sand, and its door looks 100% like the surrounding sand. Don’t ask me how he spotted it! (well, ok, the answer would be that the spider had a wee bit of its black, hairy leg sticking out (a bait to lure in smaller predators?)).

  4. I remember reading this paper years ago–very interesting (but show me a spider paper that isn’t!). I don’t have any blinding insights, but, in the scheme of spider evolution, there’s nothing terribly unusual about this behavior. I suspect–but don’t know–that Jerry is right that the “mathematical” nature of the stone collection depends more on the strength of the spider than on discrimination by size. (In fact, to me, when viewed from a spider’s point of view, the stones don’t seem so uniform in size.) It’s not unusual at all for burrow dwelling spiders to incorporate nearby debris into their burrow entrances–see photos of collar- and turret-web spiders, among others. In those cases, there is evidence that prey detection is at work. Radial extension of silk lines from the burrow entrance is seen in many species–I can’t remember whether these stones are connected to the inner lining of the burrow by silk, but that would make them similar to radial lines incorporating surrounding debris. I don’t think they would have anything to do with finding one’s way home, as I think these spiders have poor eyesight and in any case probably leave dragline trails of some sort when they venture forth–and they probably don’t go far at all anyway, being burrow-based predators.

    1. Further to your point, I picked up on Jerry’s :

      What may be going on here is simply that the spiders heft the largest stones they can carry, that their burrows are of a relatively fixed size because spiders are of a relatively fixed size, and the average number of spider-heft-able stones that can surround a burrow happens to be seven.

      I would bet that larger spiders have larger diameter burrows, and are stronger, so can move larger stones. Which would correlate stone size and pebble size, and also act to reduce variation in pebble count.
      So … if you’ve got acess to the full paper, do they document variations in burrow diameter?

      1. I accessed the paper. There is a relationship between diameter of hole and stone size. The mean ratio of size of hole and size of stone is M= 0.87(SD=0.1480)with the regression coefficient alpha= 0.0879 and beta=0.6754 (p<<0.001). The sample was not big enough to do a proper analysis of spider weight and hole size but spider weight and hole size were highly correlated, r=0.88(p=<0.05.

        Message me with your email if you'd like a copy.

    2. “Very interesting – but show me a spider paper that isn’t!”

      Best comment I have seen on the web for awhile 🙂
      I back the burrow protection theory

  5. What about the possibility of using the stones as a ‘heat shield’ to keep the burrow relatively cool? The symmetry and preference for a type of stone could reflect (pun intended) a selection pressure for formations that better reflect or redsitribute heat away from the burrow mouth.

    Just a thought. The ‘dust shield’ idea seems very credible too.

    1. I thought of that too.

      Also, could the stones keep the opening intact for longer periods of time, by shedding rain?

      I also wonder if this particular adroitness for radial symmetry is related to that displayed by web-spinning spiders.

  6. That is a South African 5c coin, featuring the Blue Crane.

    Stunning bird that is the republic’s national bird.

    Must know how pretty they are to be selected above the King Fisher’s, Fish Eagles etc to be the National Bird.

  7. The radial placement seems fairly easy to explain if the spider simply grabs a stone by the small end and drags it to the lip of the burrow.

    And whatever advantage is conferred by arranging stones in a circle around the burrow entrance, it seems reasonable to expect that natural selection will discover the most efficient packing (wedge-shaped stones oriented small end in).

  8. Great post; interesting comments. It would be fun to experiment and find out which of the option(s) for the explanation of the radially-arranged stones pan out.

    I’ve plugged it before, but I’ll do it again (since Leslie comments here); if you’re at all interested in spiders or evolution (who, us?), treat yourself to Leslie Brunetta and Catherine L. Craig’s book Spider Silk, subtitled “Evolution and 400 Million Years of Spinning, Waiting, Snagging, and Mating”. A superb read!

  9. This genus (Ariadna) is well known for building tubes of this general sort and spends essentially all of their time in the silk-lined tube. The exception is adult males wandering in search of mates. I’m guessing that there are silk lines radiating from the entrance attached to these stones. The radiating lines are a typical feature of Ariadna tubes and facilitate the detection of prey near the tube. When prey walk on or touch the “trip lines” the spider rushes out and captures the prey. I’m thinking that the most likely explanation is that the stones somehow provide a good substrate for attachment of these lines in a moving sand area. A simple but elegant solution. I have some close-up photos of the entrances to tubes from our local species (Ariadna bicolor) in the US I could upload if that is possible for this blog.

    1. Thanks for this information, Rich. That makes a lot of sense to me, since that “solution” is common to so very, very many species in many different families.

Leave a Reply