One neat thing about the Antennoseius mites is that they have two adult forms: a ‘normal’ free-living mite that lacks the peg setae and a phoretomorph with the cone setae that searches out the beetles for a ride. Thus, the homology of the setae is readily determined (in these mites, every seta has a designation).

We assume that the cone setae help the mites hold on (in the case of Fig. 2 by pressing against the underside of an elytron). More mysterious are similar setae with a long distal process (Fig. 6) usually found on the venter of a basal limb segment. If you get a chance to look at this figure, I’d be interested if you see similar things in ants.

They provide good histological evidence that these teeth-like structures are mechanoreceptive sensillae, thus strengthening the homology case that they are indeed modified setae. And, as they also point out, such type of “oral” hairs are commonly present throughout the family rather than being unique and novel in Amblyoponinae. No doubt that in all ants such sensillae tell the insect that something is right in front of them, and thus will incite a response (e.g., close the mandibles).

In groups where these sensillae are a crucial component of the prey-catching mechanism, trap-jaw ants for example, the structures are rather elongated and thin, projecting forward away from the head as deadly triggers (as you probably know and illustrated on these pictures of Anochethus). Amblyoponinae shows quite the opposite- the dentiform setae are minute in most cases and very close to the cuticular surface. An object needs to be in the ant’s face for it to excite these pegs. On the other hand, all Amblyoponinae have long and thin setae along the mandibles, clypeus and labrum that project way beyond the teeth-like pegs (with the exception of Amblyopone pluto and Apomyrma stygia which have these setae hypertrophied in the clypeus and labrum respectively). As Gronenberg et al. point out in the case of Mystrium, the dentiform setae doesn’t come in contact with the prey since the mandibles are already closed for the snap mechanism to function.

So we are left with the situation in Amblyoponinae, where having “oral” mechanoreceptive sensillae is a plesiomorphic feature, but having such structures modified into blunt, short teeth evolved later. My guess is that they do secondarily function rather like teeth, providing some traction for the mandibles that do not have wide masticatory margins that meet when closed.

http://www.ncbi.nlm.nih.gov/pubmed/12769958?dopt=Abstract

?

I guess I can take suggestions from readers.