The unusual hook-shaped mandible closer apodeme (left one in the pair) of species in the Odontomachini genus group (Anochetus emarginatus pictured here). Piece dissected out and cleared from all muscles (Scanning Electron Micrograph, Roberto Keller/AMNH)

Last August, before taking a break from blogging, I posted an impossible-to-answer trivia. It consisted of the image above depicting an unidentified mysterious skeletal piece (sclerite) in the shape of a hook, together with two key pieces of information: a) it is entirely internal; b) it comes in pairs.

A regular visitor to this blog, Marc “Teleutotje” Van der Stappen quickly asked if it was part of the sting apparatus. This was a perfectly good guess, since it satisfies both a) and b), but the mysterious sclerite occurs on the opposite end of the ant. A subsequent comment by C.M. Wilson guessed the tentorium. It was also a good guess. We are now correctly assuming it is something inside the head, but I will argue that, since the tentorial arms are invaginations of the outer cuticle, strictly speaking they are not entirely internal.

The sclerite in question is nothing but the structure that serves as a link between the insect mandible and some of the muscles moving it. It is a specific type of apodeme: a term used to describe any internal piece of the arthropod skeleton that gives support to muscles¹. The tricky part of the trivia was that, while these apodemes exist in all ants (in all insects with mandibles in fact), they are extremely modified into strong hooks only in a small clade consisting of the ponerine trap-jaw genera Anochetus and Odontomachus (a group sometimes referred as the subtribe Odontomachini, but not currently recognized).

Ants have the basic type of mandible articulation found in most insects and known as dicondylic: each mandible interacts with the head capsule through a couple of hinges (called condyles in 1337 anatomical speak), and has two sets of muscles connected to it that pull on opposite sides– one for opening the mandible (abductor) and one for closing it (adductor). Now, the muscles don’t attach directly to the mandible but do so by way of a membranous ligament that, in the case of the mandible closer, connects in turn to an apodeme that receives all the muscle packs, hence the name mandible closer apodeme (in German, of course, all of the above information is summarized into a single, long word).

The left column shows the shape and location of the mandible closer apodemes (solid black) inside the head in Myrmecia sp. and Odontomachus chelifer. Right column: mandible closer apodemes painted (in orange) as they would appear internally on the head of an Odontomachus bauri worker ant. b, apodeme base; c, apodeme collateral branches; l, ligament; SOG, suboesophageal ganglion (Drawings from Paul and Gronenberg, 1999; SEM image by Roberto Keller/American Museum of Natural History)

In most ants the mandible closer apodemes consist of a highly sclerotized (that is, hardened) basal body (b) that branches into three long extensions (c) running towards the back of the head, as exemplified with Myrmecia sp. in the illustration above². Numerous muscle packages run from the inside of the head capsule to these collateral branches, so when the muscles contract all the force generated concentrates on the massive sclerotized base that pulls the mandible shut via the flexible ligament (l).

The mandible closer apodeme in members of the ponerine trap-jaw ant clade also has a stout base and three collateral branches. Here, however, the lateral branch is modified into a massive (really massive) and highly sclerotized hook. This is what you see in the image opening this post and in the illustration immediately above. Also above, to the right is an SEM of the head of a Odontomachus species where I painted in orange how these apodemes would look internally in place, so you can appreciate the size of these structures. They are also highly pigmented due to sclerotization: when you prepare workers of these ants for regular skeletal observation by clearing the muscles and other soft tissues with a strong base (KOH), you can see the couple of large pigmented spirals through the semitransparent cuticle of the head.

The hook-shaped branches provide extra attachment surface and support to the powerful adductor muscles, that in the case of these ants fill about two thirds of the entire head volume³. The apodeme in ants of this group also differs from the basic type in that it has a ventral projection that receives a specialized muscle called “trigger muscle”⁴. This muscle, one on each side of the head, is responsible for causing the subtle deformation of the frontal part of the ant’s head that releases the mandibles that were locked in the catapult mechanism to either strike prey or propel the ant into the air.

You can read all the details of this mechanism in the papers cited below.

Notes and references

Frontal part of the head in an Anochetus emarginatus worker, profile view (Scanning Electron Micrograph, Roberto Keller/AMNH)

This image shows the mouthparts of a trap-jaw ant in resting position. The only structures really visible are the prominent elongated mandibles (in yellow) that project forward. The rest of the pieces, laying immediately below, are retracted inside the preoral cavity.

Fully extended mouthparts in an Anochetus emarginatus worker, profile view (Scanning Electron Micrograph, Roberto Keller/AMNH)

And this is how the mouthparts look fully extended, when the ant is sticking its tongue out. There are four different sets of structures here: the labrum (in green); the mandibles (in yellow); the maxillae (in orange); and the labium (in red). Each set corresponds originally, both phylogenetically and ontogenetically, to a pair of structures, although only the last three are modified limbs properly (each correspond to a pair of head appendages).

These are very complex structures, each part deserving its own separate discussion. But I wanted to thrown these images here now to serve as reference for future posts, and mention a few important generalities.

Ants display the unmodified general architecture of a biting insect. The mouthparts of adult ants are typical for what is found when comparing different insect groups, and one can readily homologize each part with a corresponding structure in a grasshopper or a beetle for example. Even the most derived mouthpart morphologies found within ants, like that of the trap-jaw ant pictured here, preserve this general pattern.

However, ants do have some uniquely derived features. They are truly prognathous insects, something uncommon within Hymenoptera (but not exclusive). While in bees and in most parasitic and stinging wasps the mouthparts hang down below the head pointing to the ground, in ants they are directed forward, always pointing to the front.

Ant prognathy, however, is not only a function of the fact that the whole head is tilted forward. Examine the image above and you will notice that the labrum, maxillae and labium are fully extended while the mandibles remain closed. That is, unlike other Hymenoptera, prognathous or not, in ants the mandibles do not fold right on top of the rest of the mouthparts at rest. Instead the main body of the mandible “steps-up” immediately after the mandible’s articulation (the rounded yellow piece at the far right), thus laying out of the way from the remaining structures.

The much derived Anochetus pictured here provides an extreme example illustrating this, but the exceptional modification is universally shared within the family. It is another unique ant synapomorphy. Obvious as it may seem once explained, I have to confess it took me a while (a few years actually) to figure out what was happening structurally in ants that was different from the non-formicid outgroups. But since then, after the explanation clicked, I cannot look at an ant without seeing it.