Printable Version of Topic Discussion Board
+--Forum: Cabbages and Kings
+---Topic: Ant-fungus-parasite coevolution started by niiicholas

Posted by: niiicholas on Feb. 24 2003,20:18

Online here:

Evolution and Ant Agriculture: a response to Ilíon.
< >

Some of his other posts:
Successful predictions of evolutionary theory
< >

Here's da whole thing:

In the "Still Spinning Just Fine" thread, Ilíon brought to our attention the purportedly co-evolved ant-fungus-mold-bacterium system as a case that presents difficulties for the traditional Darwinian paradigm.  As an ant evolution researcher, I am of the opinion that the ant/fungus/mold/bacterium relationship is fascinating regardless of one's perspective, and think that the topic merits its own thread.  I also happen to think that much of what we know about the workings (and even the existence) of this complex relationship is a direct consequence of the use of Darwinian theory by biologists, and I echo Marc's sentiments that the system is a fine example of evolutionary biology in practice.

This is a long post, and it has the following structure:

  • 1. Short ant/fungus background.
  • 2. The role of evolutionary theory in ant/fungus research.
  • 3. Some quibbles over technical issues.
  • 4. Overview of evidence for common descent and natural selection in the ant/fungus/mold/bacteria system.

1. Quick background.

In a nutshell, there are some 200+ species of ants in the new world tropics that are obligate fungivores. These are mostly small, inconspicuous ants, but the group also includes the famous and economically important leaf-cutter ants.  These ants are involved in a largely mutualistic relationship with a fungus.  The fungus is eaten by the ants, but the fungus also depends on the ants for its propogation: a true agricultural system.  In the past 4 years, a twist has been discovered: an Escovopsis mold that is a specialized parasite of the ant fungus, and a Streptomyces bacterium that produces an antibiotic that the ants use to combat the parasitic Escovopsis.

< The attine ants >

< The discovery of the Streptomyces >

< Ulrich Mueller's publications >

< photo gallery of Atta leaf-cutters and gardens >

< photo gallery of Acromyrmex leaf-cutters >

< photo gallery of Mycocepurus, a more "primitive" attine. >

(note, the final three links are shameless self-promotion of my brand-new web site)

2. The role of evolutionary theory in ant/fungus research.

To say that most of what we know about fungus-growing ants and their fungi stems directly or indirectly from evolutionary research would not be an exaggeration. To say that everything we know about the Escovopsis parasite is due to evolutionary research is a simple, observational truth.

I bring this up because Ilíon wrote:      

"Why are 'evolutionary biologists scratching their heads?' Aside from the obvious interesting questions, I think it's because the facts don't fit the paradigm"

I have a hard time believing that the facts cause problems for the paradigm, because "the facts" to which Ilíon refers wouldn't even be known if it weren't for the Darwinian paradigm in the first place.  Escovopsis was re-discovered, and the details of its biology were described, as the result of research into a widely-held idea that contradicted Darwinian theory.  The following passage is from the introduction to the pioneering study:


"The longstanding assumption that ant fungal gardens are free of significant pathogenic pressure is surprising because it contradicts some fundamental theories of the evolution of parasitism... To resolve the conflict between the theoretical prediction that parasites should exist in the clonal attine fungicultural systems and the widespread yet untested belief that ants maintain their gardens free of parasites, we conducted an extensive examination of fungal parasitism of gardens of attine ants."

Source: Currie, C.R., U.G.Mueller, and D. Malloch. 1999. The agricultural pathology of ant fungus gardens. Proc.Nat.Acad.Sci.USA. 96:7998-8002. < (Download pdf) >


Briefly, evolutionary theory predicted that there should be serious parasites of this system as a consequence of prolonged asexual reproduction.  So these researchers- primarily Currie, a graduate student at the time- went chasing a Darwinian notion and discovered Escovopsis.

But this discovery is just the tip of the iceberg.  Biologists have flocked to the attine/fungus system because the system raises so many interesting evolutionary issues, some of which may be easily resolved and others of which are real head-scratchers. Besides the discovery of the fungal parasite, here are a few other bits of knowledge that owe their existence to evolutionary theory. It is by no means an exhaustive list.  (Note: by ‘evolutionary theory', I don't mean as Ilíon insinuates that observations are merely reported in ‘Darwinspeak'. I mean that these discoveries were motivated by Darwinian research questions, or that their discovery would not have happened without knowledge of common descent or natural selection.)

1. Ant fungi have been transmitted horizontally between ant nests, in addition to vertical transmission from parent to daughter colony. This result emerged from phylogenetic analysis (See Mueller, U.G. 2002. American Naturalist 160(supplement): s67-s98.).

2. Attine ant species with more complex caste systems are polyandrous (See Murakami, T. et al 2000. Behav. Ecol. SocioBiol.48:276-284.)

3. Escovopsis fungi likely had a single origin (See Currie et al 2003. Science 299:386-388.)

4. The fungal cultivars do sometimes depart from the ant mutualism and become free-living (See Mueller, U.G. 2002. American Naturalist 160(supplement): s67-s98.).

I don't bring up these examples as a demonstration that Darwinian evolutionary theory is ultimately the best explanation of the ant/fungus/mold/bacteria relation. Rather, I bring them up to show that Darwinian theories have been extremely fruitful in both bringing this system to light and in stimulating further research. The ability to generate questions is *extremely* important in scientific theories, and the ‘head scratching' produced by Darwinian theories is a natural consequence of their usefulness.  Head-scratching means questions, and questions mean research.  A theory that doesn't raise questions does not provide a useful substrate for science.

It is worth noting that so far answers have been forthcoming to these questions, which I take as an indication that we are on the right track.

It is also worth noting that little, if anything, in this system can be traced to creationist or Intelligent Design thought.  Simply put, advocates of those schools will need to provide quite a bit more detail about how, where, and when design was affected in this system before a design research program will be possible.  If you would like to see design adopted by scientists working in this system, you've got to provide an alternative that is more useful in the lab and in the field than Darwinism.  This will entail doing better than making non-committal statements about the "unknowability" of the design process of the sort that permeates writings.  Meanwhile, Darwinian reseach keeps suggesting and finding new players in the symbiosis.

3. Some quibbles over technical issues.

Some of my issues are with Ilíon, but most are with the NYTimes article that Ilíon cites.
Ilíon wrote:

Also, note that the story says the mold seems to be related to a mold that devastates commercial mushroom farms. The mushrooms we grow have not been prohibited from sexually reproducing for the last 50 million years, so by your argument they have been able to continue evolving to resist the mold attacking them. And yet, it doesn't sound as though they do a much better job than the ants' fungus does.

I have a minor correction to make.  The fungi have not been prohibited from sexually reproducing for 50 million years.  They have been observed to reproduce sexually in both the lab and in the field.  (See again the Mueller 2002 Am Nat paper).  This may not have been clear from the NYT article, but then, the popular media has never been known for the accuracy of its reporting. Rather, the issue is that the fungal cultivar only rarely reproduces sexually, so its rate of production of genetic novelty is lower than that of the parasite.  It still is true, however,  that within each nest the fungus garden is a clonal monoculture.


You're forgetting that the evidence indicates that all nests are using the same clonal cultivar of the mold.

An error in the reportage of the NYTimes.  Some species of Apterostigma, one of the "lower" attines,  use a distantly-related species of fungus. But, I consider your point valid, as most species use cultivars from the same clade, and it seems clear from the phylogeny that the original domestication happened once.


In your nice little story, the ant are becoming increasingly dependent on the fungus, but the mold is increasing, reducing their yields. During this time, and during the time that the ants are 'learning' to use the bacteris to fight the mold, some of the nests will even lose their crops, which will cause some of those nests to die, but others will restart their gardens with a new batch of 'wild' fungus - thus we should find multiple cultivars of the fungus.

I think you've taken the simplistic reporting of the NY Times a bit too seriously.  Not your fault. The inference of a single domestication event comes from phylogeny, and you've got to be careful with what you can and cannot conclude from a phylogeny.  By "single event", what the original researchers mean is that all current fungal cultivars, and their ants, trace their ancestry back to a single fungal and a single ant species.  Beyond this inference, there is simply no resolution to distinguish an actual one ant-one fungus domestication event, and a population that did it gradually with many original cultivars.  Specifically, here are three situations that will look identical in phylogenetic retrospect:

1. A literal single origin, where one colony picks up one fungus and the rest is history. I think this is the interpretation that you object to.  Not without  reason, IMHO.

2. Multiple origins, but all the fungal cultivars involved are closely related and share a common ancestor.  If that ancestor left no descendants that are not cultivated by ants in the present day, then the fungi will still trace back to a single node on the phylogeny as an artifact of extinction, even though there may actually have been multiple domestication events.

3. Multiple origins, but when the ants lose their fungi they re-acquire not from the wild but from another ant nest.  There is some evidence that this is what modern attine ants do when they lose their fungus.  This way the fungus spreads laterally among colonies, and given a stochastic process of fungal lineage loss throughout the population all colonies will eventually come to cultivate a single lineage.  The other original fungal lineages, having gone extinct, will not appear in the phylogenetic analysis and the phylogeny will trace back to a single node.  

The take-home message is that "single origin" claims can be a bit misleading.  Currently there is no way to test between these different possibilities, so for the time being we are stuck with a "single origin" only in the broad sense.


The three-way symbiosis of the ant (and apparently, we're talking about more than one species of ant but they're all using the same fungus clone), the fungus, and the bacterium has to be all in place from the beginning to be able to successfully fight the mold.

Correction: there are multiple species of fungi, some of them specific to particular attine species.  The fungus is only clonal at the level of the colony.  They are all descended from a 50my or so old fungus, but they've had plenty of time to diversify.  Interestingly, the pattern of diversification is *very* suggestive of coevolution with the ants.  I'll get back to this point in a bit.


If the queens didn't 'know' from the beginning to take a start of the parent nest's fungus (and, BTW, don't they have a special 'chamber' in their heads in which they transport their sample?), there should be multiple cultivars.

You aren't the first to have thought of this. Ulrich Mueller also finds this idea a bit suspect, and he has proposed that what originally happened was a fungus exploiting the ant for dispersal.  Ants of many kinds commonly pick stuff up in their infrabuccal chamber (the ‘special chamber' of which you speak), mostly stuff that they've filtered from food, and dump it out when it's full.  The symbiosis may have arisen as a result of the fungus getting picked up and dispersed by ants.  This is interesting, because it is a fungal-centered hypothesis.  Most explanations are ant-centered, and IMHO have come off as a bit weird.

4. The evidence for common descent and natural selection in the ant/fungus/mold/bacteria system.

One compelling reason to look for evolutionary processes as explanations for the ant/fungus/mold/bacteria relationship is that there is a great deal of evidence for common descent and for Darwinian evolution in these organisms.  Here it is, in no particular order:

  • Geography.  Attines and their cultivars are restricted to the new world tropics.  The estimated age of the symbiosis is 50 million years, which is after the break-up of Gondwana and the isolation of South America.  Given limited across-water dispersal (Attine queens are heavy and clumsy fliers, if you've seen them fly...), the agreement of date and location are supportive of common descent.  It is also relevant that the ant groups that come out as sister to the attines are also South American.  In contrast, many ant groups whose estimated ages (from phylogenies that were calibrated with fossils) are older than break-up of Gondwana are found in both the old and the new worlds. These include army/driver ants and elongate twig ants.  Does design theory have anything to say about the geographic distribution of the ant/fungus system?
  • Large scale phylogenetic congruence between ants, cultivars, and Escovopsis. This is the subject of the Jan 2003 Science article that sparked the NY Times commentary.  I strongly recommend that you read the paper:
    < Ancient Tripartite Coevolution in the Attine Ant-Microbe Symbiosis >
    Here is the abstract:      

    Currie CR, et al.2003. Ancient tripartite coevolution in the attine ant-microbe symbiosis. Science 299:386-388.
    The symbiosis between fungus-growing ants and the fungi they cultivate for food has been shaped by 50 million years of coevolution. Phylogenetic analyses indicate that this long coevolutionary history includes a third symbiont lineage: specialized microfungal parasites of the ants' fungus gardens. At ancient levels, the phylogenies of the three symbionts are perfectly congruent, revealing that the ant-microbe symbiosis is the product of tripartite coevolution between the farming ants, their cultivars, and the garden parasites. At recent phylogenetic levels, coevolution has been punctuated by occasional host-switching by the parasite, thus intensifying continuous coadaptation between symbionts in a tripartite arms race.


    Two congruent phylogenies is good evidence for common descent, but three pretty much clinches it.  I suppose that it is true that congruence could also reflect an intelligent design process as well (probably one integrated with descent), but since no mechanisms exist for how or when or to what effect such design might take place they are of little immediate use to researchers.
  • Complexity and specialization of the attines correlates with phylogeny, as expected under co-evolution and common descent. The oldest lineages of Attines have the smallest nests, a single worker morphological caste, the least specialized foraging habits, and the least prevalence of Escovopsis in their nests. They also have the highest rate of fungal swapping.  The most recent lineages of attines, have enormous colonies, with a dozen or so morphologically different worker castes (the largest workers are orders of magnitude larger than the smallest workers), multiple fungus gardens, highly specialized foraging habits, and the highest prevalence of Escovopsis.  And the lineages that are of intermediate age, such as Trachymyrmex+Sericomyrmex, are somewhere in between. This pattern makes sense under common descent.
  • Present day attine colonies have some life-history features with the signature of natural selection.  Most notable, sex investment ratios.  A ton has been written on how the differing Darwinian "goals" of queens and workers in ant colonies play out in manipulations of population sex investment ratio, such that species with particular mating systems should produce different sex ratios.  Attines, with some exceptions, fall into the generally predicted pattern of monandrous species investing very heavily in females. It's a bit too complicated to explain here, but I find the evidence compelling and you can read more about it in this book:

    < Social Evolution in Ants >

    In the past, I've also written threads on ARN about sex ratios and evolution, here:
    < Successful predictions of evolutionary theory >


There.  Now I'm done.



Powered by Ikonboard 3.0.2a
Ikonboard © 2001 Jarvis Entertainment Group, Inc.