Tapinoma erraticum worker. The rectangle shows the opening location of the left metapleural gland (Scanning Electron Micrograph, Roberto Keller/AMNH)

The metapleural gland is the definitive character of ants. It is unique to the family. Nothing homologous or similar is found anywhere else in insects. Within the tree of life of Hymenoptera, myrmecologists agree that the appearance of this gland provides a good cutting point to marks-out ants as a monophyletic group¹. You have it? You are an ant. You don’t? Sorry, you don’t qualify, get the hell out of here lousy wasp². It is the ultimate ant synapomorphy.

Metapleural gland opening of a Tapinoma erraticum worker, left side. Note the turf of setae (in blue) arising from inside the storage chamber (Scanning Electron Micrograph, Roberto Keller/AMNH)

The gland is a paired organ hosted inside the ant’s mesosoma. Each side consists of a cluster of glandular cells, the content of which is drained by membranous tubes into a rigid storage chamber³. The chamber is a relatively large structure in itself and its presence is usually marked on the outside as a swelling of the posterolateral corners of the thorax. The most visible evidence of its presence is, however, the wide opening of the chamber to the exterior. This has allowed us to confirm, for example, the occurrence of this glandular structure in ancient fossil ants where soft tissues are not preserved.

Storage chamber of the metapleural gland in a Tapinoma erraticum worker, left side. Internal turf of setae in blue; inner surface of cuticular skeleton in light green (Scanning Electron Micrograph, Roberto Keller/AMNH)

Upon removing parts of the mesosoma and cuticular wall of the storage chamber, we can appreciate how the atrium of this organ is formed by an invagination of the metapleuron (i. e., the lateral skeletal wall of the third thoracic segment), hence the gland’s name. We can also see that, in this species, there is a group of setae (in blue) that arise at the deepest end of the chamber and project towards its opening. I cleared off the skeleton of any soft tissue before taking this micrograph, but if you dissect an ant previously pickled in ethanol under a light stereoscope you will see the clusters of glandular cells inside as a pair of tight cotton balls (white and everything).

Now, this mass of glandular cells (of the same size as the chamber itself in this exemplar species) sits exactly at the place where the setae arise but on the inner surface of the chamber’s wall, colored here in light green (technically the chamber’s cavity is still the outside of the body). Thus, the cells float inside the body in the nutrient-rich haemolymph, anchored to the storage chamber by way of the membranous ducts. The gland’s secretion is discharged into the chamber via tiny pores in the chamber’s wall (which can be seen at higher magnification in the image below), and carried to the opening of the chamber by surface tension through the projecting setae.

Pores in the inner cuticular wall of the storage chamber (light green). Membranous ducts in white. Myrmecia brevinoda worker (Scanning Electron Micrograph, Roberto Keller/AMNH)

What does this specialized gland secretes? The gland’s secretion seems to act primary as an antiseptic against microorganism, with some ants producing considerable amounts of phenylacetic acid. Secondarily, the secretion has been demonstrated to act as an alarm signal, triggering a defensive response by nestmates⁴. Some authors attribute the ecological success of ants to this evolutionary innovation. Other social insects (bees and wasps) rely on elaborate and costly cells produced from wax and paper to rear their young in dry, protective enclosed environments that prevent the growth of harmful bacteria and fungi. Ants on the contrary will nest in wet soil and leave their larva laying down in the mud, to the dismay of concern parents. The entire ant nest is kept fairly sterilized by the constant secretion of antiseptics by the adult females, hence allowing ants to nest almost anywhere.

The metapleural gland has been lost at least twice in some arboreal and dry-wood nesting ants (In the weaver ant Oecophylla and in Camponotus carpenter ants and related genera) and in some species that are social parasites of other ants. This is consistent with the ecological success hypothesis in that the arboreal nesting provides a much drier habitat, while in parasitic life you let the host do the sterilizing for you along with filling all your other needs.

Back to structure, I know Barry Bolton was interested in the internal anatomy of the metapleural gland for taxonomy, since some years ago he showed me some dissections he had made on a couple of Ponerinae ants while I was visiting the Natural History Museum. I did a more extensive survey of the structure throughout the family during my doctorate research, but still too patchy to be useful for phylogenetic reconstruction— a comprehensive survey would require dissecting material from many groups known by a few, precious specimens in museums. But I can tell you that the morphological complexity of the metapleural gland across ants is remarkable. Compared to its true sophisticated nature, the phylogenetic information we currently derive from it is shamefully little and rough: basically if the opening is directed laterally or backwards. Oh yeah, and there is an external flap involved. My assessment is that there is great potential for cladistic characters useful at various levels of the phylogenetic hierarchy, notably between some of the subfamilies that remain grouped by tiny branches in our phylogenies.

The spherical storage chamber illustrated here with Tapinoma (and representative of what is found in Dolichoderinae, Formicinae and Aneuretinae) is as simple as this organ gets, though this is not necessarily the plesiomorphic condition for the family. From there, the morphology of the chamber goes crazy. Most other ants (e.g., Myrmeciinae, Pseudomyrmecinae, most Myrmicinae and most Ponerinae) have an elongated, sausage-shaped chamber occupying a considerable part of the mesosoma. In most dracula ants amblyoponines the chamber is a wide ball, although genera like Myopopone have a finger-like structure. In some Pachycondyla species (e.g., Pachycondyla porcata) and related genera it coils clockwise, while in Ectatomminae the chamber coils counterclockwise (see image below). Ecitoninae ants have an elongated chamber but there is a system of tracheal-like branching tubes arising midway through it.

Ectatomma tuberculatum worker (April Nobile, antweb.org)

Even the presence of a turf of setae peeking through the chamber’s opening is complex and deceiving. For example, Myrmecia ants have a group of setae that exactly correspond to the one pictured here in Tapinoma. But since the elongated chamber is three times larger the apices of the setae never reach the opening and won’t be seen by exterior inspection. Conversely, acacia ants (Pseudomyrmecinae) have an apparent group of setae coming out of the chamber, but closer inspection reveals that these setae arise right at the chamber’s opening rather than truly from the inside.

A great morphological project for anyone interested would be a detailed comparative study of the internal morphology of the metapleural gland. The result will be a reconstruction of the complex evolutionary history of an important and interesting structure that was without a doubt key in the diversification of the group.

Further reading and notes.