The fungus-growing ants, or “attines” are a group of about 140 described species of ants that cultivate a fungus as their primary food source. The best known attines are the leafcutter ants (Acromyrmex and Atta), which are so proficient at cutting leaves from living plants that, despite their small size, have been described as the dominant herbivores of Central and South America. The ants use the leaves as a source of nutrients for their symbiotic fungus, which is maintained in massive underground nests that can house upwards of five million individual ants.
To better understand the evolutionary history of this group, I worked together with Dr. Mauricio Bacci and his lab to utilize nuclear and mitochondrial genes for the first phylogenetic reconstruction of the genus Atta (Bacci et al. 2009). Leafcutter ants and their symbionts are a fantastic model system for examining how the interactions between organisms influence the geographic range of each species.
Recent work with Ulrich Mueller and others showed that the Texas leafcutter ant, Atta texana, has adapted to the relatively harsh winters in northern Texas by adjusting where it houses its fungal gardens (deeper during winter) and via the evolution of cold-resistant strains of fungi. We were able to show that even colder winters currently prevent these ants and their symbionts from spreading further north (Mueller et al. 2011). Unlike the leafcutters, the majority of fungus-growing ants do not typically cut leaves. Instead they use small bits of organic matter to grow their fungus gardens and have small, simple nests with only several hundred or several thousand worker ants, all of which look nearly identical. The evolutionary transition between the “lower” fungus–growing ants and the highly derived leafcutter ants is one of the most important and fascinating events in the history of the insects. By using fresh vegetation as a substrate for fungal cultivation, the ancestor of the leafcutters gained access to an incredibly abundant and reliable food source, especially in the tropical forests of South America.
We are currently working to determine whether the switch to fresh vegetation provided the opportunity for the ants’ colonies to grow larger and become more specialized, or whether an increase in colony size and complexity evolved first, and required the use of an abundant and reliable food source like leaves in order to be maintained. The order of evolutionary events is therefore key to understanding the evolutionary history of the higher attines. In order to piece together the evolutionary history of the higher attine symbiosis, I am working with colleagues at the Smithsonian Institution and several universities in the USA and Brazil to complete a detailed examination of the closest living relatives to the leafcutters.
This critical position is occupied by Trachymyrmex, one of the most diverse genera of fungus-growing ants, with 49 currently recognized species. Trachymyrmex species share several biological and ecological traits with the lower attines, and others with the more derived leafcutters, emphasizing their position as transitional species. We are currently working on elucidating the evolutionary relationships among this important group of attine ants and their microbial symbionts using molecular systematic techniques. In addition to the fungus kept as food for larvae and worker ants, other microbes inhabit the nests of attine ants, including specialized fungal parasites in the genus Escovopsis.
Recently, thanks to our extensive sampling in Brazil, we discovered that the diversity of Escovopsis parasites found in higher attine nests is much greater than was previously thought. We documented several previously unknown lineages of Escovopsis, one of which we are describing as a new species (Meirelles et al. in review). Furthermore, we found that despite high-level congruence between the evolutionary history of the ants, their mutualistic fungi, and the parasite, higher attine nests of different genera appear to share similar strains of Escovopsis, contradicting the prevailing model of co-cladogenesis (Meirelles et al. in prep).
Indeed, the nests of fungus growing ants are susceptible to growth by a wide range of microbial species, and the ants benefit from antibiotic compounds produced by actinomycete bacteria that grow on their exoskeletons. It has been suggested that such bacteria have co-evolved together with the ants, their fungal cultivars, and the parasite Escovopsis. However, we recently found that bacteria in the genus Pseudonocardia found in association with Trachymyrmex nests have broad inhibition activity, suggesting they are not highly specialized symbionts. Indeed, one compound we isolated from a nest of Trachymyrmex ants is effective against the fungus Candida. In addition to suggesting the generalized nature of the activity of the bacteria found in attine nests, these results suggest that the compounds found in attine nests may be of medical importance.
Diversity and phylogeography of Amazonian ants
The Amazon Basin contains the greatest diversity of land-dwelling species on earth, yet there are many species that remain undiscovered. Ants are especially common in these forests. In fact, there are so many ants in the Amazon rainforest that their collective biomass exceeds that of all vertebrates combined!
My work with Dr. Ted Schultz of the Smithsonian Institution and Dr. Heraldo Vasconcelos of the Federal University of Uberlandia in Brazil, as well as other Brazilian and American collaborators seeks to further explore the diversity of ant species in the Amazon and adjacent regions. The evolutionary reasons why this region developed such a rich biota are likewise not well understood. Several hypotheses have been proposed, each attempting to explain the origins of this hyperdiversity by proposing a different mechanism for how geographic barriers arise in what appears, at least superficially, to be a relatively homogeneous area.
During my doctoral work, I used three species of leafcutter ants (Atta cephalotes, Atta sexdens, and Atta laevigata) to test these hypotheses, employing a combination of methods from population genetics, phylogenetics, and paleodistribution modeling. The results suggested that rivers, commonly suggested as a potential barrier for terrestrial organisms, do not play a role in diversification of Amazonian leafcutter ants. We could not rule out habitat refugia during the Pleistocene or marine incursions into the Amazonian lowlands, two of the other leading hypotheses (Solomon et al. 2008).
To determine whether these conclusions apply more generally, former Rice undergraduate student George Romar and I compiled published molecular phylogenetic studies of Amazonian species. To determine whether recent speciation events have taken place entirely within Amazonia (as assumed by the hypotheses described above) or whether areas adjacent to the Amazon Basin have played a role in such speciation events, we compared the number of sister species pairs in which both species are exclusively Amazonian and those for which the sister species occurs outside of Amazonia.
Our results suggest that speciation has occurred outside of Amazonia or at its periphery nearly as often as it has occurred within Amazonia. This suggests that the prevailing hypotheses on how Amazonian diversity originated are incomplete and that neighboring regions, such as the Andes, have been important for recent speciation. As new molecular phylogenetic analyses emerge, we will continue to test our model for the importance of adjacent regions for the origins of Amazonian diversity.
Ant species diversity and community dynamics
Biological inventories are a basic step for ecological and biogeographical studies and are critical for conservation efforts. Ants have become an important taxonomic group for biological inventories, as they are diverse, ecologically important, easily collected, and because ant diversity is correlated with diversity of other organisms (Agosti et al. 2000). I have been involved in ant inventories in Argentina, Brazil, Costa Rica, Fiji, Panama, Peru, and the United States. Two projects in particular are ongoing: Cocos Island (Costa Rica) and The Big Thicket National Preserve (Texas, USA).
Cocos Island, Costa Rica. Cocos Island is a small, volcanic island 500 kilometers off the coast of Costa Rica, and is the only island in the tropical eastern Pacific Ocean that supports a rainforest environment. Its isolation has given rise to several endemic species of plants, insects, birds, and lizards. In collaboration with Dr. Alexander (Sasha) Mikheyev of the Okinawa Institute of Science and Technology, I have been studying the ant fauna of Cocos Island in order to understand the balance between native and introduced species of ants.
We conducted a systematic survey of the ants of Cocos Island in July 2003, during the wet season, and collected 19 species of ants. Of these, the majority were widespread tramp species. However, one endemic species, Camponotus cocosensis, was found in abundance. Another endemic, Camponotus biolleyi, first collected by expeditions in the early 1900s, was not found; however, we collected several specimens of a new species of Adelomyrmex, later described by Jack Longino as Adelomyrmex coco (Longino 2012). Wasmannia auropunctata, a highly invasive species of fire ant, was found in extremely high densities (up to 1000/sq. m) near disturbed areas, but in low densities or absent from more pristine habitat.
We returned to Cocos Island in 2006 during the dry season to re-survey the island’s ants. Alexander Wu, a former undergraduate student at Rice, compared the samples taken in both surveys and found that many non-native species had increased in abundance and/or geographic distribution on the island, suggesting that non-native species may be displacing the native ants. Sasha Mikheyev and I are currently using DNA barcoding from samples collected in 2003 and 2006 to determine whether molecular diversity calculated via barcoding correlates well with species diversity as assessed through systematic surveying.
The Big Thicket National Preserve. A series of invasions by non-native ants has had major impacts on the arthropod fauna of the southeastern United States (McGown et al. 2013). In particular, the red imported fire ant (Solenopsis invicta) is well known for its ability to dominate landscapes and displace native ants and other arthropods. In the last decade, the tawny crazy ant (Nylanderia fulva) has been spreading through the southeast United States and has been shown to displace S. invicta in some ecosystems (LeBrun et al. 2013, Horn et al., 2013). N. fulva is also negatively associated with arthropod richness and abundance (LeBrun et al. 2013).
In collaboration with Dr. Tom E. X. Miller and several undergraduate students, we are currently working on: (1) characterizing the ant communities of the various units of the Big Thicket National Preserve, (2) estimating the relative proportion of native ants in each unit, (3) documenting the spread of N. fulva as it invades the Big Thicket region, (4) determining the impact of N. fulva presence on the ant community and the arthropod community in general, and (5) determining whether the size and shape of the units influences the ability of N. fulva to invade.
Two undergrduate students, Meghan Hager and Gabriela Zambrano, recently completed the first analyses of samples that were collected between May 2014 and October 2015 and plan to publish their results soon.