Pyramica (or is it Strumigenys?) rostrata, Illinois

I’ve been thinking today about the Wikipedia edits to the Pyramica page, and my curiosity about the controversy prodded me to attempt a quick phylogenetic analysis.  Before I get to the analysis, though, here is some background.

The Ants.  Forests in warmer regions around the world hold a great number of tiny, sluggish ants covered with bizarre hairs of unknown function.  These oddly ornate little insects are predators of other arthropods.  Mites, springtails, and the like.  Because of their size, their preference for below-ground prey, and their habit of freezing when disturbed, these are not easy ants to find even where they are abundant.  Most belong to the genera Pyramica and Strumigenys.

While all species are stealth predators, approaching prey by edging almost imperceptibly slowly towards it, species show either of two distinct ways to subdue prey once they get within striking range.  Pyramica use their mandibles like pliers, holding fast while they swing their abdomen around to sting.  They are slow but grippy.  Strumigenys are trap-jaw ants, stunning their prey with a sudden blow when their mandibular trap slams shut.  They are fast but lack a powerful grip.

The photos below show the two mandible types.  However different their jaws may be, the ants are otherwise quite similar.

Pyramica reflexa

Pyramica reflexa

Strumigenys louisianae

Strumigenys louisianae

I’ll also point out that while mandible length alone often indicates the hunting strategy a particular ant uses, the trait isn’t absolute.  Some Pyramica slow-grip with long mandibles, rather more like needle-nose pliers, while a few Strumigenys sport surprisingly stubby trap-jaws.  If you wish to peruse the diversity, Antweb hosts extensive specimen galleries of Pyramica here and Strumigenys here.

The Taxonomy. The taxonomic issue is more substantive than the usual nomenclatural squabbling that systematists frequently engage in.  It’s not just about what to call the ants, but a deeper disagreement about how they evolved.  Because the ants have different mandible shapes and hunting strategies, the phylogenetic arrangement of the various species will have profound implications for how we infer traits to evolve.

For instance, if Pyramica turns out to be an ancestral radiation from which Strumigenys later emerged, then the trap-jaw must have arisen as a special case of the slow-grip mandibles, and the evolutionary trend over time in some lineages is towards longer mandibles and prey capture by shock-and-awe.  In the converse situation, Strumigenys birthing the slow-grip Pyramica, then the trend is instead towards shorter mandibles, and with the slow-grip being a subsequent modification of a trap-jaw.  Finally, it could also be that both groups are independent, each having created its specialized form by modifying the generalized mouthparts of a common ancestor.

At the moment, myrmecologists hold conflicting views on which scenario they prefer.  E. O. Wilson and Bill Brown famously raised what I call the Incredible Shrinking Mandible hypothesis, preferring trap-jaws to be ancestral to the slow-grippers.  This view is shared by Barry Bolton, who revised the larger tribe Dacetini in 2003.  Cesare Baroni-Urbani prefers the opposite, postulating that trap-jaws emerged second, and that the groups are so blurred in places that they ought to all be placed in a single genus, Strumigenys.  Apparently, adherents of both views have taken to alternating edits of Wikipedia.

My Half-Baked Analysis. This morning, my interest piqued by controversy, I visited Genbank to see if I could find species with enough overlapping gene fragments to generate a phylogeny.  Turns out that 5 species from each genus had sequences posted from the same 2 genes: the 28S ribosomal gene and the COI mitochondrial gene.  Considering the hundreds of species distributed across the two genera, this is certainly a paltry representation.  But it’s enough for a preliminary look.

I took about an hour to align the sequences and prepare a file for analysis, and another hour for my computer to run it.  (If you’d like the details, email me).  Here’s the result:


A phylogeny of Pyramica (red) and Strumigenys (blue) based on DNA sequence data from the genes 28S & COI. Generated using the freeware program MrBayes; numbers are the Bayesian posterior probabilities supporting the indicated relationships.

What does this tell us?

To be honest, not as much as I’d hoped.  Support values are low throughout the tree, undercutting the confidence we may have in the depicted relationships.  But it does show that, for these species and for these genes, there is no immediately clear-cut distinction between Strumigenys and Pyramica. Contra Baroni-Urbani, we do get a hint that the trap-jaw came first.  But contra Bolton, the two groups may be intermingled to an uncomfortable degree. It might be that everyone is wrong!

A proper resolution to the Pyramica problem will require sampling hundreds more species than I did here, using more genes and a more exhaustive set of analyses.  More that just a couple hours’ work on a Saturday.  But this is exactly the kind of approach we need.  In an ideal world, the genetic research would be accompanied by behavioral assays to determine what each species eats and how each uses its mandibles, as well as a careful biomechanical assessment of the mouthparts themselves.  Only then will a clear picture emerge about how these ants came to acquire such strange jaws.

footnote:  The DNA data I used is publicly available, and the software programs I used (Mesquite and MrBayes) are both freeware.  A surprising amount of evolutionary biology can be done without spending a dime.