Gronenberg (1996) reported on the trap-jaw mechanics and associated morphology of two Dacetine ants.
- Gronenberg, W. 1996. The trap-jaw mechanism in the dacetine ants Daceton armigerum and Strumigenys sp. Journal of Experimental Biology. 199:2021-2033.
"Abstract: Ants of three different subfamilies, among them the tribe Dacetini, have evolved very fast snapping mandibles called trap-jaws. The two dacetine genera examined, the large Daceton and the small Strumigenys, employ the same mechanism for their mandible strike. Video analysis reveals that, in Strumigenys sp., the strike takes less than 2.5 ms. It is released within 5 ms by contact of trigger hairs on the labrum. The ants employ a catapult mechanism to generate such a fast movement. Before the strike, the mandibles are opened wide and locked in the open position by the labrum, which functions as a latch. They stay open even when the large slow closer muscles contract. Upon trigger hair stimulation, the labrum is pulled backwards by a small, fast trigger muscle. The mandibles are thus freed from the catch and close rapidly. This reflex is controlled by giant sensory and motor neurons in the labral neuromere that are probably monosynaptically coupled. The short latency of the reflex thus results from the combination of a catapult mechanism, fast trigger muscles, high neuronal conduction velocities and small synaptic delays. Comparison with the trap-jaw mechanism of the ant genus Odontomachus reveals a remarkable example of convergent evolution."
Within the paper is the account of Daceton armigerum and a Strumigenys sp. with long, linear mandibles. Some of the detailed description of the Strumigenys sp.:
"In the foraging arena, Strumigenys sp. generally walk around slowly with closed mandibles. The scapes of their antennae are held perpendicularly to the body’s long axis while the flagella point forwards. These small ants (body length without mandibles, 1.2–1.5 mm) have to approach very close to the potential prey in order to initiate prey-capture behaviour. In the present experiments, most collembolan prey was smaller than the ants, but Strumigenys sp. also attacked prey of about their own size.
Upon brief antennal contact with the prey, the mandibles are opened completely (the angle between the mandibles is approximately 220 °) and the ants move more slowly so as to avoid any disturbance of the prey. At the same time, they perform swaying search movements with the head and body in order to re-establish antennal contact with the prey (see detailed description by Dejean, 1986). Vision does not seem to be involved in this apparently chemosensory searching behaviour. In the absence of antennal contact, the springtails can literally step on the ants without releasing prey-catching behaviour. If, however, the collembolan’s position is established by another antennal contact, Strumigenys sp. stalks the springtail until the forward-pointing trigger hairs on the labrum contact the prey. The mandibles snap shut immediately after the prey touches the trigger hairs. As in other trap-jaw ants, the trigger hairs thus act as a rangefinder and trigger the strike when the prey is in the correct position between the mandibles. The mandible strike occurs within 2.5 ms. However, in several cases, prey catching was not successful: large collembolans may struggle free or the strike may be triggered by a leg or antenna of the springtail. In this case, the ant clasps a limb of the prey rather than its body and the springtail may then be able to perform an escape jump. After a successful strike, when the prey is penetrated and firmly held by the mandibles, the ant immediately bends the gaster forwards and stings the prey to prevent further struggling.
The latency of the entire trap-jaw reflex (including trigger hair contact and mandible strike) could not accurately be assessed from the high-frequency video recordings because the hairs are too small to be resolved by the system. Only two video sequences allowed an approximation of the reflex latency (judging from the distance between the ant’s head and the springtail). In both cases, the supposed trigger hair contact and the mandible strike occurred within two video frames (within 5 ms)."