Hunting Tactics in Short-Mandibulate Strumigenys

In 1984 Keiichi Masuko published an insightful paper examining predatory behavior in Strumigenys. At the time the genus was scattered among a large number of genera and Masuko's intent was to examine a range of short mandible forms across a range of these taxa. Some of the findings of this study are summarized below.

Hunting behavior
Strumigenys solifontis, Strumigenys lewisi, and Strumigenys sp.

Singly foraging workers walk slowly, swaying the head and antennae bilaterally in a periodic manner. Encounters with prey are mostly haphazard events, and foragers become aware of prey by faint antennal touching. However, I often observed that they were able to perceive prey without direct contact; probably by air movement, when moving within a range of less than 1 mm. Immediately after the prey is perceived, the ants stop walking and lower the head against the substratum, while opening their linear-shafted mandibles to the maximum angle (about 180°) When this occurs, the paired labrum, which is much modified from the generalized form in ants, is lifted, to lock between the basal projections of the mandibles. It thus serves as a stopper, preventing the mandibles from closing, despite the opposing force of the huge adductor muscles which occupy most of the head capsule. A pair of long hairs projects forward from the labrum. These are the tactile trigger for the mandibular snap, which is effected by withdrawal of the labral stop from between the mandibles. The ants, with mandibles open, then turn to face the prey, and advance cautiously. When the extremities of the trigger hairs touch the prey, the labrum drops down, allowing the mandibles to snap violently, to impale the prey body with their apical teeth. In a slow-motioned but continuous manner the ants flex the abdomen ventrally, while lifting the impaled prey into the air. It is then stung into paralysis. Brown and Wilson (1959) characterized such hunting behavior of Strumigenys as involving a swift approach to the prey, followed by a fixed action pattern of strike-lift-sting movements.

Hunting success relies to a great extent on the powerful mandibular snapping mechanism.

Strumigenys canina (Pentastruma)

The foragers, when they become aware of a collembolan nearby (mostly within a range of 1 mm), stop all movement, lower the head against the substratum, and fold the antennae completely into the antennal scrobes. Taking such a posture, the ant approaches the prey in a extremely slow and cautious manner. It then inserts the anterior part of its head beneath the prey body. This part of the head includes the clypeus, which is greatly broadened, like a visor, and overhangs the basal half of the mandibles. The first mandibular insertion is made beneath part of the prey body nearest to the ant; and therefore occurs in various places. After this first contact, the ant remains immobile until the prey begins to move. The ant now, for the first time, changes its position skillfully, aiming at the mouthparts of the prey. When the prey resumes rest, the ant again freezes. If the prey moves with relatively slow speed, the ant is able to follow it in a crouching posture, and to maintain contact. The ants loose their prey most frequently during this position-changing operation. After various repetitions of this behavior the ant may succeed in inserting its mandibles just under the head of the prey, and once this is achieved, it will never change its position on its own, but moves only when the prey moves, and this maneuver evidently reduces the risk of escape by the prey, because movements by the ant while the prey is still could cause an escape reaction. This ant prefers entomobryomorphans, collembolans of a group which have the mouthparts protruding downwards, i.e., onto the mandibular position adopted by the ant.  Pentastruma  mandibles and labrum: all the teeth are small, with some clearly larger than others. Between the mandibles the linguiform labrum extends forward. It is morphologically very different from that of Strumigenys, and lacks special mechanism serving to lock the mandibles open. It bears hairs on the dorsum of its tip, which doubtless function as trigger hairs. The position of these coincides with that of the larger teeth. In Strumigenys, the length of the trigger hairs is adjusted, so that when their extremities first touch the prey its body is well within reach of the apical forks of the mandibles (Wilson, 1953). In Strumigenys, it is probable also that the larger teeth have some particular role in prey catching.

When situated just under the prey mouthparts, the canina mandibles are opened slightly, to a parallel position, and the labrum is raised to the level of the dorsal surfaces of the mandibles. In such conditions, if the prey moves its head, particularly in a vertical direction, its mouthparts inevitably push down against the ant's labrum. At that instance, if the stimulus to the trigger hairs is sufficiently strong, the ant snaps at the lower part of the prey head. Stinging follows immediately. Soon after paralysis, the ant returns to the nest, bearing its victim. Release of prey from the mandibles does not occur until the ant is in the nest, as usual in dacetines.

Closer examination revealed that the region seized by the ants is not the mouthparts themselves, but the outer circumference of the cephalic cavity from 'which the mouthparts protrude. Grasping at the circumference of the mouthparts of a collembolan serves to prevent grasped regions from being torn off before stinging. The presence of such hunting maneuvers suggests that this ant preys exclusively upon collembolans.

Strumigenys hexamera (Epitritus)

S. hexamera is a completely subterranean species in Japanese evergreen broadleafed forest. Colonies almost always nest in relatively solid earth at a depth of around 10 cm. This ant is obligatory-monogynous, and workers in colonies usually number less than 50. The foragers are seldom seen in leaf litter. This suggests that Epitritus workers hunt for prey in pre-existing subterranean spaces, such as crevices, or in narrow tunnels which they themselves excavate in the soil.

To enhance hunting success in the laboratory, I imitated such presumed natural conditions by excavating grooves in all directions in a plaster nest, thus restricting pathways for both ants and prey animals. In these confines the workers catch prey in the following manner: Because of the narrow pathways, the forager comes across the prey in front. The ant immediately crouches, folding the antennae completely into the antennal scrobes. The mandibles remain closed. Even though there is little distance to the prey, the ants never advance at this point. In some cases I saw that a worker remained immobile, close by the prey, for more than 20 minutes. In this position, the ant waits for the prey to approach, and to step on its lowered head. Suddenly an upward snap of the mandibles impales the prey on the long apical teeth. Prey was usually caught by its trunk, from below. The “inverted” type of mandibular apex is characteristic of Epitritus, so that the dorsal apical teeth are larger than the ventrals (Brown, 1962). This appears to be a morphological adaptation associated with the upward snap of the mandibles in hunting. The tenacious waiting for the approach of prey is probably related to the subterranean hunting of this ant, for movement in narrow subterranean crevices or tunnels would restrict the course of potential prey, resulting in a high probability that it will move in the direction of a waiting ant. Actually, in preliminary observations using an open foraging arena Epitritus workers were poor hunters. Furthermore, their hunting success appears to depend greatly on the body shape and the locomotory behavior of prey animals. From the ant's hunting behavior it may be expected that animals with long bodies, which advancing rapidly (or carelessly), are hunted more easily. In the artificial nest, Epitritus workers are actually more successful in catching long-bodied arthropods, such as diplurans and centipedes, than in catching collembolans. When prey are such long or large animals, it usually takes more time to completely paralyze them than it does for collembolans. Even after a first bout of successful stinging, the motionless ant continues to hold the prey in its mandibles, with the gaster curled forward, for more than a few minutes. The preference for long-bodied animals was also ascertained in the field.

Food Types/Preferences
Sugary solutions

All species except  Strumigenys hexamera licked sugary solutions eagerly. Transfer of liquids by regurgitation between adults 'was observed in three species of Strumigenys. In addition, at least for Strumigenys solifontis, regurgitation of liquids to the larvae was easily observed, in which, about the time that trophallaxis between adults is completed, workers begin to regurgitate a droplet on the ventral surface of each larva, for its mouthparts to be immersed in it. The larvae so served, moving the mandibles, imbibe the droplet by themselves. Acceptance of sugary solutions in the laboratory is a 'well known phenomenon in many dacetines (Brown and Wilson, 1959); it suggests utilization of homopteran honeydew in nature. Only Daceton armigerum and Strumigenys yaleogyna have been observed, however, to attend coccids (Bodkin and Crawley, 1916; Wilson and Brown, 1956). In my collecting experience of Japanese dacetines, coccids have been collected together with Strumigenys on two occasions: 1) On 19 August 1979, during excavation of the earthen slope by a forestry road at Mt. Kiyosumi, I found a coccid-infested root in the soil, with several S. solifontis workers nearby. It is probable that my excavation disturbed the ants while they were attending the coccids. The ant nest was located in crevices of a split rock buried in the soil about 50 cm from the root. 2) On 6 April 1980, at Shin-myo-ike, Miyake-jima Island, a nest of Strumigenys sp. was found excavated in the soil under a stone, the chambers were constructed around rootlets. Later, in the laboratory, coccids were noticed among the nest contents collected indiscriminately with the ants. In both of these cases I did not actually observe coccid-attending by the ants, however the circumstances strongly suggest that both ant species collect honeydew from coccids.

Collembola

All the six species studied here easily hunted intact and active entomobryomorphans (chiefly members of Entomobryidae, Isotomidae, and Tomoceridae), and rejected neanurids and onychiurids, both of which are known to secrete repugnant substances. Sminthuridae were tested only with two species of Strumigenys, and were hunted by both. Such preferences are well-known in dacetines (Brown, 1953; Wilson, 1950, 1953, etc.). It is of interest that upward-snapping Strumigenys hexamera workers were able to capture isotomids much more successfully than tomocerids ; since, the former walk while dragging a rather long body, and are less agile than the latter.

Diplura and Symphyla

Campodeids, japygids, and symphylans were all accepted whenever tested. Strumigenys canina could not catch active symphylans, but accept injured ones.

Chilopoda

Three species of Strumigenys hunted small intact Lithobiomorpha and Geophilomorpha whenever tested.  Strumigenys canina rejected both, even when immobilized. Strumigenys hexamera captured both groups of centipedes successfully, as well as Diplura and Symphyla.

Chelicerata

Predation by Strumigenys of gamasid mites (chiefly members of Parasitidae) and small wandering spiders is well contrasted with rejection of these animals by the other genera, Hunting of intact, agile pseudoscorpions was unsucccssful at first, in most cases, but ultimately they were accepted by all species studied, except Strumigenys canina.

Other animals

Although the tests remain incomplete, it has been observed that small staphylinid beetle larvae were hunted by Strumigenys, but their adults, were not. Small larvae of terrestrial isopods (Ligiidae) and amphiopods (Talitridae) were accepted by Strumigenys, but rigorously rejected by Strumigenys hexamera.

Field data
It is rare to find captured prey (intact but paralyzed or being fed upon by ant larvae) inside nests located in the field. For Strumigenys and Strumigenys, however, a small amount of possible prey material has been accumulated from nests in evergreen broadleafed forest at Mt. Kiyosumi and Cape Manazuru. This includes: 1) an undetermined orthopteran nymph from a nest of Strumigenys solifontis; 2) three tomocerids and one isotomid, a lithobiomorph centipede, a parasitid mite, and an undetermined insect antenna from 5 nests of S. lewisi; 3) six tomocerids and 3 isotomids from 7 nests of Strumigenys sp.; 4) five campodeids, a japygid; a cryptopid centipede, an isotomid and an entomobryid from 8 nests of S. hexamera.

In addition to this material, a number of separated chitinous exoskeletal fragments were found in peripheral regions of nests. Most of this in Strumigenys colonies were parts of wandering spiders or of other ant species; while foreceps-like cerci of Japyx were found in Strumigenys hexamera nests.

From the results of laboratory observations and collections in the field, it seems that Japanese species of Strumigenys are rather polyphagous, but that they prey mostly upon entomobryomorphans, and that Strumigenys hexamera feeds chiefly on long-bodied soil arthropods, particularly diplurans, which are known to be subterranean crevice-living organism (Manton, 1977).

Related Pages
Strumigenys behavior