Zombie ants controlled by fungi and worms

July 18, 2012 • 11:33 am

Many of you have probably heard of one of the most bizarre and sinister adaptations in nature: the manipulation of host behavior by parasites in a way that not only kills the hosts, but makes them behave in ways that help spread the fungus’s genes.

This can occur through either fungi making ants climb up high on trees or grass before expiring, facilitating the spread of spores that grow from the ant’s corpse, or, in the case of worms, making an ant resemble a berry before it dies. This attracts birds, which, eating the berrylike ant, become the next host in the parasite’s complex life cycle.

The ability of fungi and worms to control the behavior of insects to the parasites’ benefit is, to me, one of the most remarkable results of natural selection—which acts, as always, to promote the passing on of genes.  So, though a process of blind, mindless, and unguided selection (for what kind of a loving god would promote such a horrible scenario?), genes in the fungus that make the ant more liable to spread spores, and genes in the worm that make the ant more susceptible to the parasite’s next host, become the genes that replicate more often.

Rather than just rewrite someone else’s story, here’s the tale, from Wikipedia, of the fungus that is probably the one in the photo below, Ophiocordyceps unilateralis. Read carefully, for there are many marvels of evolution here:

The fungus’s spores enter the body of the insect likely through the cuticle by enzymatic activity, where they begin to consume the non-vital soft tissues. Yeast stages of the fungus spread in the ant’s body and presumably produce compounds that affect the ant’s brain and change its behaviour by unknown mechanisms, causing the insect to climb up the stem of a plant and use its mandibles to secure itself to the plant. Infected ants bite the leaf veins with abnormal force, leaving telltale dumbbell-shaped marks. A search through plant fossil databases revealed similar marks on a fossil leaf from the Messel pit which is 48 million years old.

The fungus then kills the ant, and continues to grow as its mycelia invade more soft tissues and structurally fortify the ant’s exoskeleton. More mycelia then sprout out of the ant, and securely anchor it to the plant substrate while secreting antimicrobials to ward off competition. When the fungus is ready to reproduce, its fruiting bodies grow from the ant’s head and rupture, releasing the spores. This process takes 4 to 10 days.

The changes in the behavior of the infected ants are very specific, giving rise to the term zombie ants, and tuned for the benefit of the fungus. The ants generally clamp to a leaf’s vein about 25 cm above the ground, on the northern side of the plant, in an environment with 94-95% humidity and temperatures between 20 and 30 °C. According to David Hughes, “You can find whole graveyards with 20 or 30 ants in a square metre. Each time, they are on leaves that are a particular height off the ground and they have bitten into the main vein [of a leaf] before dying”. When the dead ants are repositioned in various other situations, further vegetative growth and sporulation either fails to occur or results in undersized and abnormal reproductive structures.

The photo below is from Alex Wild Photography (used with permission; Alex also runs the great insect-photography website Myrmecos). His caption is below:

Ophiocordyceps fungus growing from the carcass of a carpenter ant. Note how the ant’s mandibles have gripped the leaf edge, anchoring it in place. This behavior is induced when the fungus takes partial control over the ant’s brain. Jatun Sacha reserve, Napo, Ecuador

This is an ex-ant. Bereft of life, it rests in peace

Now ponder what this fungus does: it produces some chemical that affects the ant’s brain so that it

  1. Climbs up to a specific height and location on a tree, a location that has a specific direction (north) and humidity
  2. Makes the ant clamp its mandibles onto a leaf before dying

I find that astounding. Truly, natural selection is cleverer than we are.  And here’s some project for an enterprising myrmecologist: find out exactly what chemical the fungus secretes that can control an ant’s brain in such a specific way. The results will surely be surprising.

If anything makes me “spiritual,” it is scenarios like this, gruesome as they are.

Here’s another zombie ant, this one controlled by a nematode worm that not only affects the ant’s behavior, but also its morphology, making it resemble a tasty berry and weakening the junction between the nematode-containing abdomen and the thorax, so that the abdomen is easily plucked off and ingested by the bird predator, its next host. The worm also affects the ant’s foraging behavior, making it search for food outside the nest more often, and thus be more likely to be found and eaten. Text and photo from Mental Floss:

A parasitic nematode named Myrmeconema neotropicum targets the gliding ant Cephalotes atratus in the Central and South America rain forests on its way to infecting birds. The nematodes travel to the ant’s abdomen, and as they mature, cause the abdomen to grow round and bright red! The ants will hold their abdomens, or gasters, high as if to draw attention—which they do. The red gasters look like berries and are eaten by birds who normally don’t eat insects. The nematodes then live in the bird’s digestive tract.

Here we have a simple worm that secrets a chemical or combination of chemicals that changes the behavior, color, and structural integrity of its host—a worm making chemicals that turn an ant into a mimic of a berry. What can be more wonderful than that?

66 thoughts on “Zombie ants controlled by fungi and worms

    1. You ant, nothin’ but a hound god.
      Pryin’ all the time.
      Well, you ant, never caught a rabbi.
      And, you ant, no friend of mine.

    1. one of the most remarkable results of natural selection one of the most remarkable results of Divine design Truly, natural selection is cleverer than we are. Truly, the Divine is cleverer than we are. If anything makes me “spiritual,” it is scenarios like this, gruesome as they are. No difference for the faithful here.What can be more wonderful than that? Here neither…though would give credit to the Divine vs to a process of blind, mindless, and unguided selection.

  1. So a couple of millennium ago some parasitic fungus brain(s) conceived of promising eternal life achieved by blind faith obedience to invisible spirit creatures. When the parasite(s) originated and began broadcasting this fungus meme, it slowly but surely spread ever more wildly, manipulating the behavior of increasing numbers of hosts who behaved in ways that spread the fungus’s genes even though the fungus does not stop them from dying.

    I don’t want to take anything away from the subjects of today’s WEIT post, which is far more interesting than my peevish little rant. I fear that many if not most of the people I know would not be particularly interested in these creatures or their behavior, and likely happy to swiftly glide on past with at best a casual “that god does a lot of stuff I don’t get, but has his reasons.”

    1. There’s a striking contrast between your remark “god does a lot of stuff I don’t get” and “god hates fags” (and all variations on both of these).

      Well, which is it, o ye faithful: do you, or do you not, know the mind of god?

      Henceforth, when one of the faithful starts yapping about god’s likes and dislikes, remember that it’s pure projection. When one says “god hates fags”, he really means “I hate fags.”

  2. And here’s some project for an enterprising myrmecologist: find out exactly what chemical the fungus secretes that can control an ant’s brain in such a specific way.

    How sure are we that it’s actually a chemical causing this behavior, and not, say, the selective destruction by the fungus of specific brain structures, or the short-circuiting of neural pathways by invasive mycelial filaments?

    1. To me, it’s the selective aspect that points to some biochemical cause vs. the general effect you’d expect from the physical forces exerted by a mass of mycelia. And by analogy, there are the well-known different behavioral effects elicited by different but clearly evolutionarily related cone-snail toxins on their prey (for this, look at publications from Baldomero Olivera’s lab).

      1. On the other hand, when people suffer brain tumors, strokes, and similar disorders, the precise behavioral symptoms are highly sensitive to the location of the pathology in the brain, and not just to some generalized biochemical action.

        Now it may be that the fungus produces some toxin that’s finely tuned to target specific brain structures. But I’d still argue that it’s the location of the damage, rather than the toxicity per se, that accounts for the odd behavior.

        1. Yep, but the various neuro sypmtoms of human patients depend on where the bleed is, and those occur essentially at random locations. (And some of those effects are very interesting – loss of nouns but not verbs or vice versa, loss of native language but not second language, etc.)

          For the specific behaviors in the ant to come from a generalized fungal invasion would seem to require targeting a receptor, inactivating some neurologically-important enzyme, or something of that sort. I wouldn’t expect a fungal mycelia to grow to a specific place and then stop. (By reference, the chestnut blight fungus, Cryphonectria parasitica, that has virtually wiped out the American chestnut, kills through uncontrolled growth of its mycelia, clogging the tree’s vascular system.)

          1. Mycelia growing at random could preferentially destroy particular cell types in particular locations. The mechanism of such destruction might be chemical, or it might not.

            And the growth might not be random. It might respond to particular neurochemical cues. Axons, after all, find their way to specific sites in the brain and then stop.

            I’m not saying I’m right and you’re wrong. I’m just saying I think the question remains open, and that the Wikipedia article perhaps goes too far in presuming a chemical mechanism.

            One way to settle it would be to try to isolate a fungal compound that causes zombie behavior in the absence of an actual mycelial infection.

            Another would be to image both normal and infected ant brains to see if the infected brains display any particular pattern of damage. I’m guessing current imaging tech isn’t up to this task, though.

            1. Re. your last para, not my area but I suspect that sectioning/imaging would be feasible and might show something.

              Also not my area, but Googling fungal neurotoxin turned up 3M hits, and the first few turned up Lolitrem B and Ochratoxin A, chemical structures known for each. Just from a very superficial skim, it looks like each targets elements of different neurotransmitter pathways, with Lolitrem going for an ion channel like cone snail toxins do. This is only to say that fungi are on record as producing neurotoxins with seemingly specific targets.

              Unrelatedly except as to toxins, diphtheria toxin is a fairly famous and fascinating one – it targets a specific protein in the host ribosomal machinery, ADP-ribosylating a specific histidine residue of a specific elongation factor. I’m sure there must be hypotheses (at the least) of how that translates to the manifestations of diphtheria, but that’s another story. It’s just amazing what goes on at the level of molecular warfare, though.

      1. Something in the water supply, perhaps?

        Okay, that’s a joke, but I’m not so sure that there might not be something to it.

        Quite a long while ago, I read the famous book “The Jewish Wars” by the Roman-Jewish historian Josephus (readily available in translation in a Penguin Classics edition). It was astonishing how closely the infighting between Jewish factions in the run up to the destruction of Jerusalem by the Romans in 70 CE resembled the infighting characteristic of that part of the world today.

        Something in the Jerusalem water supply, perhaps? Odder things have been known to happen.

  3. And here’s some project for an enterprising myrmecologist: find out exactly what chemical the fungus secretes that can control an ant’s brain in such a specific way.

    IMO it’d be really good to study the members of an infected colony to see how many outliers there are – i.e. to determine if the behavior is really that specific, or if we are just victims of some sort of detection bias.

    I’m also fascinated about why they get north-facing behavior in Ecuador. The first thing that sprang to my mind to explain that directionality was ‘sensitivity to sunlight,’ but that’s probably not a good hypothesis given that its in the tropics. Still, something that could be checked out by looking at seasonal variation or variation in plant populations further north and south.

    1. Detection bias?

      I would clamp my lifeless jaws onto a leaf if The Extended Phenotype (published in 1982) wasn’t already thoroughly researched and criticized.

      Dawkins has stated it is his favorite book.

  4. I ran across, many years ago, of a report that exposure to low levels of cadmium changed fish behavior in such a way as to make them more vulnerable to predators. I also saw a report of similar behavioral modification from exposure to low levels of DDT. Sometime later I read a report of a parasite causing very similar predator vulnerability behavior in fish. It may be faulty memory, as I read the reports some years apart, but I was struck that the described behavior changes in the exposed or infested fish was very similar, if not identical. I don’t recall it as a complex change, maybe just swimming slowly near the surface, or some such.

  5. Someone noted (ERV?) that since viruses like HIV are able to trick our immune system it would make them more learned about our bodies than our own cells are.

    Bad things from small minds. Reminds me of religions.

  6. Jesus loves zombies. That’s why he’s going to guide the evolution of these wonderful creations so that they can infect humans. George Romero was a prophet. We’re doomed, I tell you!


  7. As astonishing as all this is, it somehow seems equally amazing that someone looked for, and found, indications of the same thing in 48My/o fossil leaves.

  8. Sorry, I just had to mention it:
    “Here we have a simple worm that secrets a chemical…”

    This is amazing though. Is there a term for this “zombification”? I’d like to search for other cases as well (more complex animals).


  9. Reading that makes me want to change the catbox for some strange reason.

    Goddammit, I love that cat. I love him so much I could cry sometimes.

      1. That’s precisely the though that came to my mind when I read this article! Although it must be noted that at least one study of Toxoplasma Gondii contains the the following bizarre sounding assertion (emphasis mine):

        The results of
        permutation tests showed that the Toxoplasma-positive women had higher factor B (intelligence), factor O (guilt proneness) and possibly also Q4 (high ergic tension) and Q1 (radicalism).

        Perhaps someone better versed in psychometry can opine on how much salt these findings need o be taken with.

  10. Now I know why my coffee tastes like formaldehyde.

    Identifying the chemical signaling will be quite a job – the amounts are so minute and mixed in with so many other things. Perhaps pureed ants put through a solid chromatograph could help isolate the chemicals in question.

    1. Comparison of normal and Ophiocordyceps-infected ant heads by mass spectrometry, as a first approach.

  11. Coming back to this now that the dust has settled down…and I can’t believe that nobody has yet observed that all of us, each and every one of us, has personally experienced this exact type of phenomenon.

    The common cold is transmitted through airborne droplets of virus-contaminated body fluids, and the very first thing the virus does is induce sneezing. Bugs that spread via fecal contamination give you the runs. And I’d be surprised if there isn’t some form of STD that makes its hosts horny and / or more sexually attractive.

    Something to think about the next time you’re chewing down on a thermometer….



    1. “This exact type of phenomenon” may be overstating it. Rhinoviruses don’t alter brain function to induce sneezing; they merely irritate your nasal tract. Similarly, GI bugs affect intestinal function, not neurology.

      If there is an STD that affects sexual behavior (rather then merely hitching a ride on behaviors that would have happened anyway), that would be closer to the mark. But that’s unsupported speculation on your part.

    2. I am thinking of the religious meme. It tricks the carrier to lure its family to the source of indoctrination spread, as well as facilitate engaging in special pleading and cognitive dissonance. Clearly the brain rot is manipulating the behavior of the carrier.

      1. And it manipulates the carriers to give money weekly for the maintenance of churches and allowance for priests.

    3. I’ve never understood if sneezing was the result of viruses manipulating us for their own purposes or if it was our attempt to get rid of the viruses. Similarly, do our bodies try to rid themselves of those nasty gut parasites by a rapid-release strategy or are the parasites trying to get themselves released into the environment? Sort of a chicken-and-egg problem, I suppose.

      1. It doesn’t have to be either-or. Clearly it’s to our advantage to have a sneezing reflex to get rid of nose irritants, and a gut-cleansing mechanism to flush out poisons. Once we have those, it becomes advantageous to pathogens to exploit them by exaggerating the triggering stimulus.

  12. Yet more proof that God (the one that made all creatures great and small yadda yadda) is a sick, obsessive, sadistic, psychopathic bastard. 🙁

  13. “And here’s some project for an enterprising myrmecologist: find out exactly what chemical the fungus secretes that can control an ant’s brain in such a specific way.”
    And then synthesize it in quantity and go into the religion business.

  14. @evolutionistrue

    Moral judgments are tricky:

    “So, though a process of blind, mindless, and unguided selection (for what kind of a loving god would promote such a horrible scenario?), genes in the fungus that make the ant more liable to spread spores, and genes in the worm that make the ant more susceptible to the parasite’s next host, become the genes that replicate more often.”


    “So I’m looking for dinner in Ottawa last night and find a pub that offers food.  Famished, I order a pint and a burger, and ask for the burger to be cooked rare. (I like my beef rare, ordering it ‘saignant,’ or ‘bloody,’ when I’m in France.) …The only good burger—or steak for that matter—is one that’s cooked rare, though medium rare will suffice in a pinch.”

    In rough order-of-magnitude terms ant has c 10^5 neurons to handle its experience, a bovine c 10^9 (fairly average for mammals: cats have c 10^8 and humans c 10^10). What kind of a caring moral agent would parasitize an entire highly sentient species with such gusto?

    (The question is a serious one, but mischievous to the extent that I enjoy a good steak myself – with some qualms. My Christian acquaintances have no such worries, since their textbook advises them that only humans – plus God, angels, demons etc – have a “soul”, while animals were given to Adam for his use. So I can see that Christians have no problem here, while atheists have.)

  15. This fungal behavioral control by fungi has long fascinated me. Note that this exact kind of behavioral control of insects has arisen not once but twice among the fungi – in Cordyceps and Ophiocordyceps (which between them actually parasitize a wide range of arthropods, not just “crazy ants”), both ascomycetes related to ergot, and Entomophthora, a genus of “zygomycetes” (I use quotes, because “zygomycete” is the fungal equivalent of “invertebrate”, describing a paraphyletic assemblage) that affects the behavior of insects that they parasitize in remarkably similar ways to Ophiocordyceps, including driving the insect to take to a high place before it dies.

    A passage in my introductory mycology textbook concerning Entomophthora has always fascinated me: “Death characteristically occurs in the afternoon between 3 and 7 P.M.” The fungus certainly has its timing down – dinner by 7!

    What I find particularly interesting is that these two groups of fungi cause behaviors that are not just similar, but quite complex on the part of the insect. It is relatively easy for a plant or fungus to produce any number of compounds that cause fear reactions in insects that cause them to stop feeding and flee. (Which is what cocaine does to insects, and hallucinogenic compounds likely do as well.) But make an insect climb and die at just the right time to optimize the fungus’ reproduction? Amazing to see something like that evolve once, much less twice.

Leave a Reply