Myrmicine Mutualists and Formicine Foragers:
A Tale of Ant/Fungal Coevolution

© Peter Werner
Original publication: Mycena News, November 2009

One of the many fascinating stories of fungal symbiosis is that of fungus-farming ants found in the New World tropics and subtropics. This relationship is agricultural in nature and is analogous in many ways to the human domestication of crop plants and the resulting changes in human lifestyles. The primary players in this complex interaction are ants of the attine group (Tribe Attini) and “attine fungi”, mainly lepiotaceous fungi of the Leucoprinus/Leucoagaricus group, or in some cases coral fungi allied to Pterula. It has also been found to involve at least one other partner, an actinomycete bacterium that lives on the surface of the ants’ bodies.


In all forms of this symbiosis, the ants collect a rich organic substrate which they deposit in underground chambers. The attine fungus, which grows throughout these chambers after initial inoculation by the founding queen, then spreads into on the new substrate. Unlike termiteassociated fungi, attine fungi typically do not produce macroscopic fruiting bodies, and instead are restricted to a wholly mycelial or even yeast morphology. In some attine fungal species, the mycelia produce structures called gongylidia, small sterile structures that serve as food for the ants.

The ants benefit from this system in that the fungi transform organic matter that is largely inedible to animals into an edible mass of mycelium or yeast. The fungus benefits in not only getting a rich food source from the ants, but in also being physically cultivated and protected from predators by the ants, as well as chemically protected by compounds manufactured by the actinomycete found on the ants’ bodies. The latter defense is particularly important, as this mutualism also universally involves a parasitic fungus Escovopus, an anamorphic ascomycete related to Cordyceps and Hypomyces. Typically, this parasite is largely kept in check in the attine fungal gardens, but in some cases can take over with devastating results to the colony. Escovopus and the ant/fungus/actinomycete system appear to be locked in a long-term “evolutionary arms race”, being selected for defenses and counterdefenses against each other over evolutionary time.

There is a great deal of species diversity among the fungus-cultivating attine ants, and a variety of specialized cultivation types have arisen. The most general type is engaged in by the “lower” attine ants, which live in small colonies with relatively unspecialized castes and provide a mixed substrate of plant material and detritus for their fungi. The “higher” attine ants have larger colonies and often numerous castes that are highly specialized to various tasks of food gathering and fungal cultivation. The most specialized group are the leafcutter ants, which provide substrate solely in the form of fresh leaves, which are then broken down on an assembly line-like system, until the final substrate is produced. Leafcutter colonies are often very large (with a volume on the scale of a city bus) and consume vegetation on the scale of a large mammalian herbivore.


Although it has been established that the attine/fungus symbiosis sprang from a single evolutionary event, the details of the origin and course of evolution of this system have been less well-understood. Recently, several discoveries have shed some light on these questions. A 2008 paper by Ted Schultz and Seán Brady of the Smithsonian Institution produced a broad phylogeny of 65 species of attine ants, which was compared against the cultivation specialties of this species. It was found that lower attine agriculture was indeed the most widespread and “primitive” condition among the attines. The shift from lower to higher attine agriculture was found to be restricted to a single specialized line representing a single evolutionary event, as was the shift from general higher attine agriculture to leafcutting. Similarly, fungal cultivar switching, a change from mycelial leucocoprinoid fungi to yeast forms of these species, or to pterulaceous coral fungi, were also found to be singular evolutionary events.

The details of how ants first domesticated fungi also remains an open question. There are two broad hypotheses on how this took place. One is the “transmission first” hypothesis, which posits that the ancestors of attine fungi were dispersed by ants, perhaps living on their colonies in a parasitic or commensal relationship, until eventually becoming utilized as a food source. Another is the “consumption first” model, in which free-living leucocoprinoid fungi were consumed by the ants as a food source, then brought into cultivation.

While this question remains unanswered, several recent discoveries provide some evidence for the “consumption first” hypothesis. The first is a recent paper by Tracey Vo and others of the Ulrich Mueller lab at University of Texas, who carried out a phylogenetic study of freeliving leucocoprinoid species and attine fungal strains. It was found that some strains of attine fungi are more closely related to free-living, mushroom-producing strains of Leucocoprinus and Leucoagaricus than they are to each other. These species are typically found unassociated with ant colonies and are capable of wind-dispersal of spores like other mushrooms. This lends support to the idea that attines “captured” free-living species and that they have possibly gone in and out of attine cultivation a number of times.

The second discovery is that Euprenolepis procera, a Southeast Asian species of formicine ant (a group not directly related to attine ants), subsists largely on wild fungi. This research was carried out by a pair of German entomologists, Volker Witte and Ulrich Maschwitz of University of Munich. As an initial test of Euprenolepis food preferences, Witte gathered some 80 species of wild mushrooms and placed fruiting bodies in the middle of ant trails. While the ants ignored many types of mushrooms, they swarmed and consumed some 30 species. Another experiment demonstrated a line of ants could consume an entire 40 gram oyster mushroom over a course of 3 hours. Although Euprenolepis was found to be capable of an omnivorous diet, sampling of food particles carried by non-baited ants revealed that over 99% of their diet comes from fungal sporocarps.

While Euprenolepis is not a close relative of the attines, and hence, does not directly answer the question of the ancestral diet of that group, the study does provide an example of ant consumption of wild mushrooms that had not previously been observed. It is entirely possible that many other examples of ant consumption of wild fungi exist, but have simply been overlooked. How widespread fungivory is among the ants and what role it played in ant domestication of fungi looks to be a promising area of study.

For further reading on fungus-farming ants:

Fungus-foraging ants: