Answer to the Monday Night Mystery: Proatta and the Attines

Yesterday’s mystery ants were Mycocepurus and Proatta. The inclusion of this pair in the Monday Mystery series echoes a long-standing debate in the myrmecological literature about the relationship between the exclusively American fungus-growing attines (including Mycocepurus) and the tropical Asian Proatta.

Mycocepurus smithi
Proatta butteli

Proatta butteli has puzzled myrmecologists since Auguste Forel first named it in explicit reference to the resemblance. His original 1912 description leads with a note of Proatta‘s superficial similarity to the attines, especially in general habitus and abundance of spines, before listing differences in antennomere count and other characters. Although Forel discusses a potentially shared ancestry, he concludes his description with the wry comment:

Je préfère pour ma part m’abstenir d’hypothèses.

Attinologist Neal Weber dismissed a connection in a 1958 paper:

While it is true that Proatta butteli is strikingly like an attine, this is taken here to be an example of convergence in worker morphology and not necessarily an indication of phylogenetic relationships. The spinosity is especially like that of Mycocepurus, and another and unrelated ant with comparable spinosity is Orectognathus antennatus of New Zealand. Proatta butteli differs markedly from attines in having the clypeus produced as an angular lobe, which is impressed in the middle and with lateral carinae, and in having a single, large median spine on the basal surface of the epinotum anterior to the usual epinotal spines. The antenna of the worker is 12-segmented and the anterior tarsi are not dilated. There is no evidence that Proatta is a fungus-grower and it is not considered here to be a member of the Attini.

But not everyone agreed. More recently, Mark Moffett compiled a list of characters supporting shared ancestry, including some aspects of the shape of the clypeus and the presence of several spines on the head, thorax, and petiole.

Proatta queens and workers in a nest (Borneo)

Larval morphology gurus George and Jeanette Wheeler, on examining the immatures, weighed in on Moffett’s side:

We conclude that the larva of Proatta is definitely attine. We have a prejudice against attaching a small monotypic genus found locally in the Oriental Realm to a large wide-spread tribe in the Neotropical Realm; hence we had hoped that the larva would be either strongly attine or strongly non-attine. It is neither, but it is as good an attine as Myrmicocrypta. It lacks the coarse spinules on the mandibles, which is an attine character, but so does Apterostigma, which is otherwise like the higher attines. The weightiest evidence is said to be that Proatta is not known to be a fungus-grower; but is it really necessary that the ancestral attine already have that habit?

Now that the molecular revolution has hit, DNA sequence data have swung the pendulum the other way. Shultz & Brady (2008) used sequences from 4 nuclear genes to look at attine phylogeny. They did well within the attines, and although deeper relationships were poorly supported there was enough signal to reject a Proatta – Attine link. Here’s a simplified tree:So at present it seems as though Weber’s contention may be correct. Insofar as one trusts genetic data and distrusts larval morphology, the spines of Mycocepurus and Proatta are convergent.

Anyway. Time to tally the points.

6 to Scot for guessing Proatta and answering the convergence question first. 6 more for MarekB, who additionally got the Mycocepurus and based the convergence on a more thorough set of references.

Proatta butteli

6 thoughts on “Answer to the Monday Night Mystery: Proatta and the Attines”

  1. Well, if it’s not related to the attines, then what’s it related to besides Crematogaster and Meranoplus? And do we still not have a particularly close relative of the Attini? Bummer.

    If it is indeed convergent evolution, this has to be one of the most complex cases ever. Spines, larval forms, even their queens look look similar to my untrained eye. Taxonomy never ends, I guess.

    1. This seems to be the general trouble with Myrmicinae. It wasn’t clear from morphology how many of the genera were related. And the first molecular data has done little other than show how difficult the myrmicine phylogeny will be. There’s strong support for some non-intuitive groups (Myrmicaria + Monomorium, for example), but mush all along large parts of the backbone of the tree. It’s not just attines that are problematic. It’s also dacetines, solenopsidines, etc. It looks as though large parts of the subfamily arose nearly all at once, a long time ago, in a fashion that will be very difficult to reconstruct.

      Brady et al (2006), using several genes, inferred 5 general groupings:

      1. Myrmicini (Pogonomyrmex + Manica +Myrmica)
      2. Messor and associated forms (Aphaenogaster + Messor + Stenamma)
      3. Solenopsis and associated forms (Solenopsis + Monomorium + Myrmicaria)
      4. Large Odd Group 1 (Pheidole, Attines, Dacetines, Cephalotines, Wasmannia, and others)
      5. Large Odd Group 2 (Leptothorax, Temnothorax, Tetramorium, Crematogaster, some “Solenopsisdini”, and others). Proatta fits in here.

      But even these groups might be overturned with more data and different kinds of analyses. So you’re right- taxonomy never ends.

    1. It’s a rank between subfamily and genus. Like all categories other than species, it’s an arbitrary grouping that reflects the opinion of the taxonomist rather than anything about the biology of the organisms.

        1. Up to a few decades ago attini were defined, as you state, on the basis of their fungus-growing habits and some common morphological features.

          As systematics has become more evolutionary, taxa are now defined genealogically. So attini might properly be thought of as all the descendants of the common ancestor of Atta and Apterostigma. (Those two lineages are the most distantly related of the genera).

Leave a Reply