A Myrmicine Phylogeny Shakes Things Up

Monomorium kiliani
Monomorium kiliani, an Australian myrmicine. The narrow, two-segmented waist is characteristic of this subfamily.

We’re only halfway through the year, but already 2014 will be remembered as pivotal for studies of ant evolution and classification. Following right on the heels of Schmidt & Shattuck’s massive ponerine revision comes an important new study from the Ant Tree of Life group. Ward, Brady, Fisher, and Schultz (2014) have reconstructed the first thorough genus-level phylogeny of the great ant subfamily Myrmicinae.

How important is this study?

Roughly half of all ants are myrmicines, both in abundance and in species diversity. Their numbers include fire ants, harvester ants, leafcutter ants, big-headed ants, acrobat ants, and so on, to the tune of some 6,000+ species.

So… Boom! Suddenly, we’ve been given a detailed picture of the evolution of half the ants. This is big. It is so big I cannot cover the paper in detail. Instead, I’ll just give a few preliminary thoughts, as follows:

1. This is a well executed study, as we’ve come to expect from the Ant Tree of Life team, applying a thorough analysis to over 250 carefully selected taxa and 11 genes. It’s also a shining example of an older generation of genetic techniques, alas, and while I am confident the stronger results will mostly endure, be aware that an incoming next-gen tide of full genomes, and the 6,000 yet-unsampled myrmicine species, may yet overturn some of the findings.

2. The deep history of Myrmicinae, starting 100 million or so years ago, mostly occurred on those continents that drifted to become the Americas. Echos of these earliest divisions are heard in six clear, genetically distinct groups that Ward et al have formally set the up as a new system of tribes, replacing an earlier, messier scheme. The six groups are listed here in their order of divergence: Myrmicini (MyrmicaManica), Pogonomyrmecini (Hylomyrma & Pogonomyrmex), Stenammini (AphaenogasterMessorStenamma, and relatives), and three sprawling groups with thousands of species: Solenopsidini, Attini, and Crematogastrini. 

The myrmicine big picture. (Sharpie on office paper, 2014, limited edition print available, unless I recycle it).

3. The news is not all good. The clarity deep in Ward et al‘s tree fades for slightly younger events. Early relationships within some of the the six tribes are discouragingly ambiguous. This study has resolved some problems, myrmicine taxonomists face a difficult road ahead. Many of the world’s greatest genera do not form natural groups and will have to be redone. These include Aphaenogaster, Pheidole, Tetramorium, and especially Monomorium, which splatters almost comically across the Solenopsidines.

What, really, is Monomorium? Modified from Figure 1 of Ward et al (2014).

Distressingly, fuzzy resolution in a data set with this many markers and taxa means achieving proper resolution, if at all, will likely be expensive. Myrmicines may have speciated so explosively that we may never be able to reconstruct what happened with confidence.

4. The authors correct a few of the more obvious instances of paraphyly. Notably, the New World “Messor“, being unrelated to their old world doppelgangers, were moved to a revived Veromessor, and several social parasites like Protomognathus and Anergates have been sunk into the host genera from whence they evolved: Temnothorax and Tetramorium, respectively. There are other changes, too; they are listed in the abstract

Most of the identified problems- such as what to do with Monomorium and Aphaenogaster were left for targeted future research.

5. Remember the dispute over Pyramica vs. Strumigenys? The argument was fundamentally over how ant mandibles evolve. Apparently, high energy trap-jaws arise easier than anyone imagined. According to Ward  et al, not only is the assemblage of trap-jaw ants formerly included in dacetini a polyphyletic splatter, even within the genus Strumigenys the trap jaw has arisen at least twice.

A phylogram of Strumigenys, modified from Figure 1 in Ward et al 2014, showing strong support for the parallel evolution of trap-jaws in the genus.

6. The rare and bizarre African myrmicine genus Ankylomyrma is not a myrmicine at all! Rather, Ward et al‘s results unambiguously tie it to the equally bizarre Tatuidris of the Neotropics, sitting on a distant branch of the ant tree. Peas in a poneromorph pod…

Ultimately, Ward et al have crafted a sobering view of how little we still know about ant evolution, and how much remains to be done.

Aphaenogaster fulva, photographed in Illinois.

source: Ward PS, Brady SG, Fisher BL, Schultz TR (2014) The evolution of myrmicine ants: phylogeny and biogeography of a hyperdiverse ant clade (Hymenoptera: Formicidae). Systematic Entomology, online early. DOI: 10.1111/syen.12090

disclosure: I received my Ph.D. from Phil Ward’s lab where much of this study was completed, and I contributed a few of the samples, but I was long gone by the time the study was initiated and have had no other involvement with the research.

12 thoughts on “A Myrmicine Phylogeny Shakes Things Up”

  1. From Facebook, my remarks:

    “I’m very interested in what happened with Tetramorium and related genera… Yes, Teleutomyrmex and Anergates… but also Strongylognathus… Had already posted some remarks on The Ant Farm and asked some questions at Prof Buschinger but want to know what they have done in this article…”

    “When I see the supplementary information and read the abstract, I wonder why they didn’t synonymize Strongylognathus and Tetramorium… Sanetra and Buschinger in 2000 didn’t want to go so far to put them all 4 in synonymy… Or they are all 4 synonyms or Tetramorium must be divided in many parts! Same situation as in 2000…”

    “I went quickly through the article and found they postponed the synonymy of Tetramorium and Strongylognathus for a ruling of the Int Comm Zool Nom to change in this case the priority-rules. Sanetra and Buschinger hesitated to call them all Strongylognathus, the oldest name, but I think they shouldn’t change the rules in this case… Yes, Tetramorium is a big genus but Strongylognathus is a good name, even longer in use than Tetramorium…”

    1. Thanks for your comments, Teleutotje. The primary argument in favor of retaining Tetramorium is the degree of distruption that the change will cause:

      1. I’m not convinced that how much a name is used is the criterion to keep a name and not the older one. Strongylognathus is a good name that was is use, not forgotten, and means also a lot for many, researchers and amateurs. I’m in favor to keep Strongylognathus!!!

  2. One unfortunate consequence is that the fungus-growing ants are left without a formal name (now just called the Atta genus group). Such a biologically important monophyletic group should have a FORMAL name! One option would be to synonymise all the genera of the group under Atta, so they become all one genus.

    1. Yeah, that thought had occurred to me, too. The “Massive Atta” solution would also resolve the polyphyly of Cyphomyrmex and Trachymyrmex.

  3. I’m afraid the taxonomy stuff is much too arcane for me as an amateur ant fan… but I still had to comment on how gorgeous that photo of Monomorium kiliani is. Amazing!

  4. All these name changes for some of my favorite social parasites are going to take some getting used to, especially that very generic-sounding Temnothorax americanus. In the meantime, I’m interested in the parasitic clade within Tetramorium arising from T. caespitum ((Teleutomyrmex+Anergates)+Strongylognathus), which seems to imply some sort of progression from non-parasitic to dulotic to inquiline. Hasn’t that hypothesis been difficult to prove before?

  5. Alfred Buschinger

    First of all, I must say that this paper will cause an excessive amount of work for museum curators as well as for private collectors. Identification keys have to be rewritten, species lists have become invalid, and all students of ants will have much trouble when reading “old” literature (until July 2014) to figure out what now are the correct names, and what are synonyms!

    Second: I disagree with the synonymisation of social parasites with their host genera. E. g. Myrmoxenus (= Epimyrma) is a monophyletic group of parasites with a fascinating evolution from active slavery to worker loss and a “derived state of slavery”, to “degenerate slavemakers”. The same holds true for Chalepoxenus. These names tell you something on all species of the group!

    Third: The revision is inconsistent: Harpagoxenus, Formicoxenus, Leptothorax muscorum (as a representative of Leptothorax, including L. acervorum and the former Doronomyrmex) have remained separate genera, though Harp. and Formic. according to our former knowledge and to the new revision clearly derive from Leptothorax. I am not sure about the consequences of a synonymization, but probably then all the group had to become genus Formicoxenus?

    Fourth: One of the intentions of the ICZN was stability of the system. This Revision destabilizes the System. Next time Polyergus will become Formica?
    I believe that ICZN has to be reworked to avoid such detrimental consequences. Is it really necessary to change nomenclature always when new records (e. g. of fossil insects) reveal “new” relationships? – I do not question the results of modern phylogenetic work. But I could well live with a phylogeny on the one hand, and a different but stable system/nomenclature on the other!

    @ JasonC: If you don’t know as yet, my paper on the evolution of social parasitism might be helpful: Buschinger, A. (2009): Social parasitism among ants: a review. (Hymenoptera: Formicidae).- Myrmecol. News 12: 219-235. http://www.myrmecologicalnews.org/cms/images/pdf/volume12/mn12_219-235_non-printable.pdf

    @ Teleutotje: I have posted a PN for you in “Ameisenforum”. – I disagree with respect to “Strongylognathus”, meaning “saber-shaped mandibles”, quite misleading with respect to ordinary Tetramorium.

    1. Prof. Buschinger, the other option is, like suggested before, to divide Tetramorium in smaller genera and keep Strongylognathus as a separate genus… and maybe, at the same time, keep Anergates and Teleutomyrmex! I will prefer this above the synonymy of them all!!!

    2. What is the ideal way to curate a collection—does one want to represent the evolutionary relationships of the organisms or opt for nearly pure stability by organizing specimens alphabetically by genus? There are many pros and cons to the evolutionary organization and the alphabetical organization. Although curators will have to move specimens around in their collection if the Myrmicinae were organized by tribe, ultimately the new organization will make far more sense in light of the new relationships recovered by Ward et al. Moving species from genus to genus may be at the discretion of the curator who may wait until a new system has congealed.
      Regarding the second point, it is a matter of evolutionary vs. phylogenetic classification. Do we want supraspecific names to apply to clades or to groups with visibly distinct evolutionary histories? Biologically, what is more interesting: Placing an inquiline in a distinct genus because it looks different, or placing an inquiline in its parent genus because it is derived from within that lineage? Personally, I opt for the latter as a generality because this enriches the natural history of the lineage in question, reflecting better the complex reality of evolution and speciation.
      I would note that this is not strictly a revision. This is a phylogeny which conservatively takes steps to address taxonomy based on its discoveries. Sure, Leptothorax, Formicoxenus, and Harpagoxenus are left valid, but note the authors’ taxon sampling: One species from each genus. Within the bounds of their findings, the authors act upon their findings (they did describe the Pogonomyrmecini and split certain members of Monomorium).
      While stability is an extremely important value in taxonomy, should stability negate corrections of institutionalized errors (with respect to phylogenetic topology)?
      If Polyergus were found to be nested within Formica, either Formica would need to be divided (as may happen with Pogonomyrmex soon) or Polyergus may be subsumed. On one hand, Polyergus could be preserved, but on the other two genera would be made out of one pleisiomorphic genus.
      The Myrmicinae Phylogeny is a major stepping stone on our collective path toward understanding the evolution of the Formicidae. Would we be having the same reactions if, instead of “The evolution of myrmicine ants …” we had a “Genera Insectorum”? Do we willingly submit to tradition over evidence?
      Ultimately higher taxonomy is arbitrary. It is up to the author to weigh his/her values and determine whether reflecting evolutionary trajectories or evolutionary relationships is more important. The two aren’t necessarily incongruous; if the morphology and topology agree, why not split Tetramorium?

  6. James C. Trager

    I once heard one of the authors say something like ‘All divergent genera make their ancestral groups paraphyletic — for a while.’ In the case of most of these parasites, it has to be a long while, or the parasites would cease to exist, if you follow my pondering on this statement.

    While I find the results of this paper rather interesting, I have to agree with Prof. Buschinger that its conclusions are inconsistent and could be quite disruptive. Or maybe not — Ant Web, antwiki, DiscoverLife and such don’t even keep up with some of the recent single or partial genus revisions, so it’s a good bet this new classification won’t take hold at those sites any time soon. It will be interesting to see what the world ant catalog (“AntCat”) will do with all this.

    Aside from formalizing some parts of the classification, many of the results of this study were presaged by the recent big ant phylogeny papers. Whenever I have read one of these papers, I have come away with the feeling that I’d rather have seen them do a phylogeny of a greater number of species from a single one of the tribes or genera (or clades of the older papers, like the mess around attines and dacetines plus Pheidole), such as Aphaenogaster, or Tetramorium, or Myrmica, or what-have-you (or outside the myrmicines, something like the huge Camponotus group). Eventually, even if not done completely consistently, these could be knitted together to understand the greater phylogeny.

    I welcome the comments of anyone wishing to find the holes in this argument…

    1. The disruption aspect is nothing new. As for creating work for curators, yes; but the usual response seems to be let the dust settle for a decade or two and see how it all turns out, unless it is your revision ;P

Leave a Reply