Argentine ants tending scale insects

Three years after finishing my Ph.D., I have finally published the last bit of work from my dissertation.  It’s a multi-locus molecular phylogeny of the ant genus Linepithema, a group of mostly obscure Neotropical ants that would be overlooked if they didn’t happen to contain the infamous Argentine Ant.  In less jargony language, what I’ve done is reconstruct the evolution of an ant genus using genetic data.  Here’s the citation:

Wild, A. L. 2008. Evolution of the Neotropical Ant Genus Linepithema. Systematic Entomology, online early, doi: 10.1111/j.1365-3113. 2008.00435.x

The paper is behind a subscription barrier, but I have placed a nexus file containing the aligned sequence data with the Bayesian consensus tree here: www.myrmecos.net/wild/Linepithema4gene.nex. It’s best viewed in the freeware program Mesquite I’ve also posted a quick tree figure below (click to enlarge).

The title aside, in my mind this paper is not so much about the evolution of Linepithema as it is about something deeper: the collision between old and new ways of describing biodiversity. I took the morphology-based taxonomy from the first part of my dissertation and scrutinized the results using molecular genetics.

Testing my own ideas was a delicate matter. By the end of the taxonomic project I had examined thousands of individual ants and measured countless body parts. My emotional investment in the work was correspondingly large. The monograph was 126 pages long, and on completion it felt weighty, thorough, and authoritative.

As it turns out, the test results weren’t all that pretty. The molecules suggested a tally of 35+ species in the genus, about twice the 19 I had delineated with morphology. Linepithema has finer evolutionary divisions than I had imagined. I had nearly identical species with very different genes, and I had some very different-looking species with very similar genes. I had genes that told different stories from each other. I had species giving birth to other species, and species hybridizing with each other.

The messiness gave me a lot to chew on.  Are species even real? Is such complexity in the genetic data normal? Are most genes poor indicators of species? Am I just really bad at morphological taxonomy?

On the positive side, the messes were largely confined to recent evolution, as if time smoothes out the birthing wrinkles of new species. The primary divisions I had made to sort Linepithema into four species groups were confirmed almost unanimously in the different sets of molecular data, and the genus as a whole held together in all of the analyses.

In any case, I’m enjoying the sense of freedom that comes with setting loose the last remnants of my thesis.