Molecular Phylogenetics and Evolution this morning has the first detailed molecular phylogeny of the leafcutting ant genus Atta. Maurício Bacci et al sequenced several mitochondrial genes and the nuclear gene EF-1a from 13 of the 15 described species-level taxa, using them to infer the evolutionary history of the genus.
This is an important paper. Atta is the classic leafcutter ant of the new world tropics and a model system for co-evolution among the ants, the fungi they cultivate, and a suite of microbes that either parasitize or protect the ant-fungus mutualism. These ants are also one of the most important agricultural pests in the Neotropics, capable of defoliating entire trees overnight. An understanding of the evolutionary history of this group will be very helpful in uncovering details about the nature of the ant-fungus relationship and the evolutionary factors that have shaped it.
I’ve reproduced a simplified version of the Bacci et al tree in the figure at left, stripping out the species-level detail to show only the subgenera, each of which comprises several species. The geography of the genus is represented by the color of the branches.
Two patterns jump out at me. First is the geography. The older lineages are northern and the younger ones southern, suggesting that Atta arose in Central America and speciated as it dispersed southward. Oddly, the highest diversity of Atta occurs at the southern part of its range, near Paraguay, as if the farther afield the ants travelled the more species they generated.
Second, the molecules offer unambiguous support to classical taxonomy. Each of the subgenera named by various morphologists- Archeatta, Atta s. str., Epiatta, and Neoatta- emerge with strong support, with Archeatta as the oldest.
I can’t leave any paper without a bit of criticism, and mine is this: for all the different genes representing this phylogeny, the total number of base pairs is just a hair above 1,000. That’s not very many, and the results could be prone to statistical artifacts stemming from a small sample size. Nonetheless, support for the classical arrangement is high, suggesting that this tree is a good one and should hold up as more data are added.