“Convergent” evolution in New World and Old World leaf-cutter ants.

March 20, 2023 • 11:15 am

Convergent evolution” refers to the phenomenon of two relatively unrelated species or groups of organisms evolving similar traits independently.  The classic example involves the various species of marsupial vs. placental mammals that independently evolved very similar forms and habits. So, for example, we have marsupial moles who look and behave like placental moles, but the underground habit and morphology evolved separately, for the species aren’t closely related. There are many marsupials with traits and morphologies similar to those of placentals: see p. 93 of Why Evolution is True for several striking examples. New World cacti and Old World euphorbs, which are often virtually indistinguishable, is another.

A new paper appeared in the journal Ecology and Evolution that describes “convergent” evolution between leaf-cutter ants in the New World vs. those in the Old World. I put “convergence” in quotes in the title because the convergence goes only so far: it involves collecting bits of leaves and flowers and bringing them to the nest. But what the ants do with those collected leaves—using them to grow fungi—has a very different “purpose” in the two areas.  What’s new here is that leaf-gathering associated with fungi wasn’t known in Old World ants before this, so the “convergence” involves simply the behavior of collecting leaves and plant material in the two groups, as well as the presence of fungi that grow on the collected leaves.

Heretofore, leaf-cutting was thought to be restricted to New World ants, and it’s been pretty well studied. It is an amazing phenomenon.  New World ants in the tribe Attini spend their time climbing up trees and other plants, biting off chunks of leaves or other vegetation and bringing them back to the nest. There they masticate the plant material into a kind of green humus that’s used to grow fungus. It is the fungus itself that’s important to these attine ants, for it is their main foodstuff and the food given to larvae to nurture them.

The fungi used by New World leaf-cutters are in the fungal group basidiomycetes, and are unknown outside of ant nests. They are obligate associates. This mutualism must have started out as a wild fungus that somehow found its way into food collected by ants, and then, over evolutionary time, became an obligate mutualist with the leaf-cutters, with the leaf cutters evolving behaviors to cultivate the fungus, which is very nutritious. The fungi are adapted to live solely inside ant nests, and the ants themselves are adapted to eat the fungus, so this is a true mutualism. Both species benefit from the association: the fungus gets protection and a place to grow, while the ants get an endless supply of food. The evolution of this mutualism involves behavior, morphology, and physiology. Here are some photos of New World leaf-cutters taken from Wikipedia:

Atta columbica cutting the bits of leaves. A colony can denude an entire tropical tree in just a day or so:

A gif of an ant carrying a leaf fragment back to the nest:

A line of ants carrying the leaf bits back to the nest. They can wield huge fragments, often weighing more than the ants themselves. The carriers are all workers: sterile females produced by the queen.

Sometimes smaller ants hitchhike on the leaves being carried back. These are not “free riders”, but doing an essential job (see below).

And a National Geographic video showing what happens afterwards:

Leaf-cutter ants are endlessly fascinating, and I want to point out two unique aspects of their lifestyle before we move on to the new paper. The first involves those “hitchhiker” ants shown above. The ones riding on the leaves have an essential function: they’re there to drive off parasitic flies that can lay eggs on the heads of the carrier ants, leading to their death. (Army ants don’t suffer the same parasitism because they move mostly at night when flies aren’t active).

Second, since the fungus is essential to the flourishing of an attine colony, and isn’t found outside of the colonies, how does it get to a new colony when it’s founded? (Colonies often fragment after the ants “swarm”: the queen grows wings, mates with males in the air, and then drops to the ground, shedding her wings and starting a new colony by herself, laying the first eggs that become workers.)  It turns out that before the female takes off to start a new colony, she tucks a bit of the fungus in her buccal pocket, a special receptacle in her oral cavity.

When she lands, she digs a hole in the ground that will become the future colony, removes the fungus, and then excretes some substance (perhaps metabolites from her no-longer-needed flight muscles) on the pellet of fungus. Then, for the next couple of months, she’s the only worker in the nest, cultivating the fungus (but not collecting leaves), feeding some of her own eggs to larvae that hatch from other eggs, and thus producing the workers of the new colony. Eventually the whole system goes into operation with leaf-cutting workers, fungus, eggs, larvae, and pupae, and of course the long-lived queen.

Oh, and note that all this complex behavior—in both workers and queen—is encoded in a brain smaller than a grain of sand. Evolution is indeed a marvel!

Okay, back to the paper, describing leaf-cutting in Old World ants (in Africa), and what they do with the leaves and flowers they harvest. Click on the screenshot to read, or see the pdf here. 

What the authors show is that a completely different group of African ants in the genus Crematogaster have evolved plant-cutting behavior, and, like the New World ants, carry the plant pieces back to their nest. In this case, though, the nests aren’t underground, but are “carton nests”: nests made of dried, heavily masticated plant material that are built up in the trees and have a cardboard-like appearance. Here’s a carton nest of Crematogaster peringueyi:

The authors observed the species C. clariventris in Cameroon cutting leaves and flowers and carrying the material back to the carton nests, where it’s masticated along with other stuff (not described in the paper) and used to build the carton nests. When the authors removed bits of the carton nests, there was a noticeable increase in leaf foraging to get materials to repair the damage.  Note that the Attines and Crematogaster groups are separated by 70 million years, with related groups not showing the behavior, so in this case leaf-cutting and carrying to the nest have evolved independently.

Here’s a figure from the paper (and its caption) showing the African ants cutting leaves:

(from paper) Crematogaster clariventris cutting (a) or retrieving (b) pieces of young nitrogen-rich leaves. The characteristic yellow gaster [abdomen] of the workers is not always as striking as in this figure (Photos Piotr Naskrecki).
When the author dissected the nests and looked at them under regular and electron microscopy, they found the walls of the carton nest were pemeated with fungal hyphae: the tubular bits of fungi that they use to digest and absorb food.  The hyphal filaments, many of which were simply dead bits of fungi, were interwoven among the nest and, according to the authors, strengthened the nest walls, making the whole nest harder to destroy and more resistant to rain. (They didn’t show that, though.)

Here’s a photo of the whole mishigass: ants cutting flowers, a carton nest, and the inside of the carton nest, showing its porous structure.

(from figure: Crematogaster clariventris in the process of cutting pieces of a nitrogen-rich Hibiscus rosa-sinensis (Malvaceae) flower (a). They also defoliate Albizia spp. (Fabaceae), Alchornea cordifolia (Euphorbiaceae), Cassia spp. (Fabaceae), Mangifera indica (Anacardiaceae) and Theobroma cacao (Malvaceae). A Crematogaster clariventris carton nest (b) and details of the carton inside the nest (c) (photos Alain Dejean).

Here’s a figure showing the hyphae forming what looks like a scaffold for the nest:

(from paper): Scanning electron microscope images of the carton of the walls of Crematogaster clariventris nests showing the fungal hyphae/mycelium that penetrate the nest carton through pores (arrows).

Using DNA analysis, the authors also identified the group to which this structural fungus belongs: it’s a Capnodiales fungus, very distantly related to the Agaracaceae basidiomycete fungus used by New World ants: they’re separated by about 640 million years! What the authors could NOT determine was whether the carton-nest fungus came into the next with the queen (as in New World leaf-cutters), or whether it was found in the outside world rather than being obligately confined to the ant nests. Perhaps it just landed on the plant material before it was carried back to the nest.

So, as I said, we have a convergent behavior: collecting leaves and bringing them back to the nest to serve as substrate for fungal growth.  But there are many, many questions unresolved in this system, beyond the origin and nature of the fungus described above.

1.) Do the Old World ants eat the fungus?

2.) Does the fungus really strengthen the nest? It appears to, but no experiments were done to determine this. (Granted, this would be hard to determine.)

3.) Are nests not strengthened by the fungus more susceptible to destruction? (This would be hard to determine, too.)

4.) Do the ants have any behaviors specifically designed to put fungus into the nest, or do the fungi simply come in as hitchhikers with the masticated “soft material” used to build the nest? Remember, carton nests are not limited to the Old World; they’re found, especially in warmer regions, in the New World, too, and fungi are also found in the walls of the New World carton nests. Dejean et al. claim that the Old World ants somehow encourage one particular fungus to grow in the nest walls, but they haven’t shown that. Perhaps one fungus just grows better in the nest material.

The behavior of leaf collecting is certainly convergent here, but it would be a lot more interesting if the convergence also included deliberate collection of a specific species fungus to help the colony. (We can’t say that the Old World Crematogaster ants are “cultivating” the fungus, as they don’t appear to do anything to facilitate the fungus’s growth.)

As always, one can end this paper with “There is more work to be done.” But the nature of that work is very clear!

h/t: Phil Ward for educating me about leafcutters.


Dejean, A., P. Naskrecki, C. Faucher, F. Azémar, M. Tindo, S. Manzi, and H. Gryta. 2023. An Old World leaf-cutting, fungus-growing ant: A case of convergent evolution. Ecology and Evolution. https://doi.org/10.1002/ece3.9904open_in_new

14 thoughts on ““Convergent” evolution in New World and Old World leaf-cutter ants.

  1. “Fascinating”.
    [Spock raises eyebrow]
    These ants are mindbogglingly “clever”. And a brain the size of a grain of sand!
    Why aren’t we humans millions of times smarter than we are? Our skull could hold millions of grains of sand. 😉

  2. Very cool! Those are definitely Crematogaster type ants, with their peculiar upside-down-looking abdomens.
    It’s possible that conventional underground ant galleries are reinforced with fungal hyphae. That would be especially valuable if they are made in sandy soil.

  3. Fascinating! Farming originated way before people started doing it. Natural selection can do amazing things.

  4. These types of ants have long fascinated me. Thanks for writing about this new paper. I’ve bookmarked it for future reference.

  5. Very interesting. A couple questions out to anyone….Does the new world fungus provide structural support to their host ant’s nests too, like these old world ones? I haven’t read the paper (yet) and wonder if the old word ants use the fungus for food too? Is the fungus adapted to life solely in the ant colony like with the new world ants?

    It may be super nitpicky; if a behavior or trait is similar between two species but it is used for different purposes, is it still considered convergent? Maybe there is a different term for such relationships?


  6. Jerry, thanks for covering this paper. I didn’t know about it. Very interesting indeed. Crematogaster is a large, widely distributed genus; many are arboreal, but I didn’t know about their nests.
    One of the authors is Piotr Naskrecki, who is associated with the MCZ. I strongly suggest you look him up. He is an authority on the large katydid family (Tettigoniidae), but more relevant to WEIT, he is a spectacularly good nature photographer, especially of insects and other small creatures. And he knows the taxonomy and biology of his subjects. I think you can find some of his photos online. He was a student in an OTS course, in Costa Rica, that I taught in the late 60s or early 70s, and he was astonishingly knowledgeable then, as well as being an extremely proficient field biologist.

    1. Doug,

      I have met Peter and we’re Facebook friends (I found him through his photos and met him at the MCZ). I found out about this paper because he announced it on Facebook! I always look at his photos there.


Leave a Reply