Megaponera analis

Known as the Matabele ant, this species is a highly specialized termite predator. Megaponera analis raids on termite nests can occur throughout the day and into the evening, and last up to one hour. Raiding columns, made up of large numbers of sizeable workers, are readily noticeable and often seen when they cross a road. Colonies will occasionally relocate and during these migrations a large column of workers carrying brood, males and myrmecophilous associates (thysanurans and staphylinids) may be seen following a trail to a new nest site.

Distribution based on Regional Taxon Lists
Afrotropical Region: Cameroun, Central African Republic, Eritrea, Ethiopia, Ghana, Guinea, Ivory Coast, Kenya, Malawi, Mali, Mozambique, Senegal, Sierra Leone, South Africa, Uganda, United Republic of Tanzania, Zambia, Zimbabwe.

Biology
Frank and Linsenmair (2017) provided the following summary of the biology of this species (please refer to the publication for citation details):

Megaponera analis has caught the attention of various researchers for its very pronounced continuous allometric size polymorphism (Crewe et al. 1984; Villet 1990), a rare phenomenon in ponerines, and its specialization on group hunting only termites of the subfamily Macrotermitinae (Longhurst et al. 1978; Yusuf et al. 2014). Worker size in Megaponera analis varies greatly, with majors being twice the size than minors (Schmidt and Shattuck 2014). Megaponera analis has a clear work division inside the nest, by partitioning their work by size, with the smaller workers taking care of the smaller larvae and eggs (Villet 1990). This work division is not only restricted to size but a clear age polyethism can also be observed, with younger workers conducting nest work and older workers going out to forage and scout (Villet 1990). This division of labor is not only confined to nest tasks but also plays a vital role in their rather unique foraging activity. The general foraging pattern of Megaponera analis starts with scout ants searching an area of approximately 50 m radius around the nest for termite foraging sites (Frank and Linsenmair 2017; Bayliss and Fielding 2002). These scouts always belong to the largest ants in the colony (Longhurst and Howse 1979). Once a scout ant has found a potential hunting site it starts to investigate it, while avoiding contact with the termites, before returning in a direct route to recruit approximately 200–500 nestmates and lead them to the termites in a column formation (Bayliss and Fielding 2002; Longhurst and Howse 1979). The recruiting scout now is the raidleader and positioned at the front of the column followed by other large ants and scouts. The number of ants recruited by the scout depends on the number of termites at the hunting ground, in-line with what optimal foraging theory predicts (Frank and Linsenmair 2017). During the raid division of labor occurs (Corbara and Dejean 2000); larger ants break open the protective soil cover created by the termites, while the smaller ants rush into these openings to kill and pull out the prey (Corbara and Dejean 2000). After the hunt, the larger ants collect the dead termites, the column forms again and the hunting party returns together to the nest. More recent studies even show the existence of rescue behavior, with ants that got injured during the raid in the form of lost extremities and clinging termites getting carried back to the nest to recover (Frank et al. 2017).

Hölldobler et al. (1994) - Colonies have a single ergatoid queen (Villet 1990). She is the only large individual that is carried during colony emigrations. The queen is often surrounded by many workers, and this attractiveness is based on chemical signals, possibly released by the thick glandular epithelium lining her entire body.

Kistner (1993) - This is one of the few ponerine species with a diverse assemblage of myrmecophiles living in their nests. The staphylinid beetles Degalliera megaponerae, D. peetersi and Ponerilla hoelldobleri were collected from an emigration column in Kenya.

Longhurst et al. (1978) - At Mokwa, in the Nigerian southern Guinea Savanna, Macrotermitinae are the only recorded prey of the obligate termite predator, Megaponera foetens (=M. analis). The main prey species is Macrotermes bellicosus (141 termites m-2 a-1) followed by Odontotermes spp. (42 m-2 a-1). Predation on other species, Ancistrotermes cavithorax, Macrotermes subhyalinus and Microtermes spp., amounts to only 10 termites m m-2 a-1. Macrotermes beIlicosus is not the most abundant termite in the primary savanna.

Longhurst and Howse (1978) - Scout ants of this obligate termite predator respond to cues in fresh soil sheeting laid down by foraging termites by returning to their nests laying a recruitment trail. Effective cues are found in the sheeting of Macrotermes and Odontotermes, but not Microtermes and Ancistrotermes. They can be extracted using organic solvent, but have disappeared from the sheeting after about 24 h. Solvent extracts of minor worker termites have the same effect on scout ants as fresh sheeting. The ability of M. analis to detect fresh soil sheeting may increase the efficiency of predation on Macrotermes and Odontotermes species.

Frank and Linsenmair (2017) studied field colonies in Comoé National Park, northern Côte d'Ivoire to examine raiding behavior and flexibility in task allocation. This included assessing the role of worker polymorphism in these processes. One conclusion of their research was the suggestion that continuous allometric size polymorphism in Megaponera analis evolved mainly to allow for greater flexibility in task allocation, and this arose in response to the unpredictability of task requirements in an irregular system such as group hunting of termites. Their findings also include the following:

Colony size for 12 excavated colonies was between 900 and 2300 ants, a result comparable to previous studies in other regions (Villet 1990; Yusuf et al. 2013 ).

Raiding Behavior

Raiding activity was highest in the morning and afternoon hours between 6:00–11:00 and 15:00–19:00 local time. Night raiding was also observed.

Raid column formation: During the outward journey, a clear over representation of large workers was present at the front and tail of the column. On the return journey, large workers tended to be more present at the front half of the column, although a disproportionate amount of large ants was again present at the tail of the column (see video: ).

Column Formation: Column formation in previous studies of Megaponera analis focused on the front of the column (Longhurst and Howse 1979). We can confirm the results from Nigeria that more former scouts and large ants were present at the front and that these ants seemed to initiate the return journey (Longhurst and Howse 1979). In addition, we now showed that the formation is even more complex, with termite carriers concentrated at the center of the column and a rearguard predominantly occupied by individuals best adapted to fighting off potential predators (raid-guards) (Breed and Harrison 1988; Dejean and Feneron 1996; Holldobler and Wilson 1990). All these positions are ecologically meaningful. The role of front ants during the return journey could be guiding the column back to the nest in case of trail disturbance. While it has been confirmed multiple times that Megaponera analis follows a pheromone trail back to the nest (Holldobler et al. 1994; Longhurst and Howse 1979), this trail could be interrupted by external influences. Former scouts, which are accustomed to moving around the environment without a pheromone trail, could thus be better suited to lead the way back to the nest in case the pheromone trail disappeared, which some qualitative observations of trail disturbance by us seemed to suggest (pers. obs.). The more vulnerable termite carriers are best protected at the center of the column, these ants would arguably have more difficulties following a pheromone trail or fending off predators when carrying up to 12 termites in their mandibles. A further benefit of the rear guard, apart from fending off predators, comes in the form of picking up lost prey or injured individuals, which are essential in colonies of Megaponera analis (Frank et al. 2017). Raid-guards are already known for instance in the driver ant genus Dorylus, with guards standing still, facing outwards, with open mandibles for long periods of time on the sides of a column (Holldobler and Wilson 1990). While this is a good strategy in a mass-raider with ant columns lasting for days, this strategy would be less efficient in a species in which the entire column passes a certain point in one to two minutes (such as Megaponera analis). The sophisticated formation in Megaponera analis shows adaptations to their rather unique foraging behavior, which most likely improve their defense capabilities. The underlying mechanisms regulating this formation (how the ants ‘‘know’’ where to position themselves) remain unclear.

Roles and Tasks

We were able to identify three new tasks during the hunting process of Megaponera analis: lingerers, runners and raid-guards. Most of these tasks seemed not to be predetermined before the raid but were filled out as need arose during the raid and were mainly conducted by the larger ants. The formation within the raid column was also much more sophisticated than previously thought (Longhurst and Howse 1979 ), with larger ants being present both at the tail and front of the column and performing special tasks, such as patrolling on the sides of the column (runners). Furthermore, the formation was more stable than expected, with the same individuals resuming similar positions in subsequent raids and front ants even returning to their position if experimentally displaced in the same raid.

Work division: We identified seven different roles during the foraging process of Megaponera analis for the larger ants and three for the smaller ants.

Scouts: Scout ants always belonged to the largest ants in the colony (N = 100 scouts), although only a fraction of large ants were scouts at a time. Only 14 ± 5 scouts were active per foraging activity period (N = 10) in colony sizes between 700 and 2000 individuals. The tasks of a scout involved leaving the nest at the beginning of the activity period, searching for food sources, investigating them and recruiting nestmates to the food source. At this point, the scout became the raid leader. During the hunt itself the raid leader did not participate in the hunting process, standing at the periphery of or moving around the hunting ground. After the hunt, when the ants started to gather again, the raid leader was observed to join the column with the last 10% of the returning ants (N = 10 raids), but then clearly moved forward to be at the front of the column. When the raid leader recruited ants to a foraging site, there were always formerly marked scouts that also joined the raid. Former scouts are ants that normally scout for food sources but decided to participate in a raid instead (before resuming their normal scouting behavior in later activity periods). Former scouts also showed a clear over representation at the front of the column during the outward journey (40% of marked scouts; N = 15 raids). Once at the hunting ground, these ants did not participate in the actual hunt but instead positioned themselves at the periphery of the hunting ground, potentially searching for nearby termite sites. Furthermore, the first ten ants initiating the return of the column back to the nest were in 20 ± 16% of the cases front ants from the outward journey, while marked ants from the center contributed 4 ± 5% and tail ants were never observed to lead a return column (N = 5 raids; definition: front: first 10%; center: 10–90%, tail: last 10% of the column).

Termite Carrier: The second already known task was that of the termite carriers. After the hunt, the killed termites were piled up on the foraging ground and the ants started to collect them in their mandibles, forming balls of up to 12 termite workers. In an average raid, this task was carried out by the larger ants in the colony, which generally only needed 30% of the ants to carry all killed termites (median termite carriers per raid 29 ± 19%; N = 134 raids). However, if the raid was overly successful, more and more smaller ants carried termites, resulting in a positive exponential relation between smaller and larger termite carriers (Pearson’s test of ln small carriers against large carriers: t12 = 7.2; R2 = 0.81; P < 0.001). If more than 63% of the larger ants had to carry termites, a tipping point was reached. After this point, the increase of smaller ants performing this task was greater than the increase for larger ants. Termite carriers on the return journey positioned themselves towards the front of the column, although not in the first 10% block of the column, which was occupied by non-carrying large ants. This formation could only clearly be recognized in raids that were not oversaturated with prey (< 70% termite carriers). In oversaturated raids (> 70% termite carriers), no clear formation was recognizable for the termite carriers.

Helper: Injured ants were always carried by larger ants (N = 154 helper ants). The injured ants were mostly picked up during the phase when the termites were collected at the hunting ground, but were also picked up during the return journey. A median of 3 ± 2.9 ants conducted this task per raid. Helpers showed no clear position fidelity in the column on the outward and return journey.

Lingerer: The first newly described role in this study was the lingerers. These ants actively returned to the hunting ground for another 1–2 min (N = 10 raids) after the returning raid column started to form and leave for the nest. A mean of 5 ± 1 ants per raid conducted this task (N = 11 raids). These ants were always large ants that did not carry anything (N = 51 lingerers). They ran over the foraging site and showed a clear intention to pick up left over live ants or in rare cases termites. After picking up a termite or an ant, they immediately tried to return to the column. While lingerers started at the tail of the returning raid column, they showed no clear position fidelity within the column afterwards.

Runner: Runners were ants that during the returning raid column moved in the opposite direction of the moving raid at the sides of the column from the front to the tail and overshot the column by up to 1 m. These ants were always large and did not carry anything (N = 17 runners), with generally having a mean of 1.4 ± 0.5 ants conducting this task per raid (N = 12 raids). These ants picked up fallen termites or injured ants if encountered, although these incidences were rare (17% picked up a termite; 23% picked up an ant; 59% remained empty; N = 17 runners).

Raid-guard: In the returning raid column, all large ants that did not perform one of the previously described roles were classified as raid-guards, i.e., non-carrying large ants in the column. When an alarm pheromone of Megaponera analis was elicited in the vicinity of the column (20 cm distance) the raid guards were always the first ones to respond and attacked the source of the distress, unlike small ants with empty mandibles or termite carriers, which did not seem to react to it. The raid-guards showed, such as on the outward journey, a clear preference for a position at the front or tail of the column, irrespective of the saturation of termite carriers. In addition to defense, raid-guards at the tail of the column seemed to act as a ‘‘safety net’’, picking up fallen termites or injured ants when offered (92% of offered termites picked up; 85% of injured nestmates; N = 13).

Position Fidelity: Ants were marked during the outward journey of a raid for their respective positions. Front ants showed a clear fidelity to be at the front of the column on the outward journey. On the return journey, these ants were at the front but over time started to fall back, showing no clear position fidelity by the time the column was quantified. Center ants were always observed to be in the center of the column, both on the outward and return journey. Tail ants tended to stay at the tail but were also found in the center of the column in subsequent outward journeys. On the return journey, tail ants were either at the front or center.

Fungi
This taxon is a host for the fungi and (Shrestha et al., 2017).

Nomenclature

 * foetens. Formica foetens Fabricius, 1793: 354 (w.) GUINEA. [Junior primary homonym of Formica foetens Olivier, 1792: 503.] Replacement name: Formica analis Latreille, 1802c: 282. Emery, 1897e: 597 (m.); Arnold, 1915: 48 (q.); Wheeler, W.M. 1918c: 298 (l.); Wheeler, G.C. & Wheeler, J. 1952c: 612 (l.). Combination in Megaponera: Mayr, 1862: 753. Senior synonym of abyssinica, laeviuscula: Roger, 1860: 310; of dohrni: Forel, 1909b: 64. See also: Wheeler, W.M. 1922a: 64; Longhurst & Howse, 1978: 1213.
 *  analis. Formica analis Latreille, 1802c: 282. Replacement name for Formica foetens Fabricius, 1793: 354. [Junior primary homonym of Formica foetens Olivier, 1792: 503.] Combination in Pachycondyla: Brown, in Bolton, 1995b: 302; in Megaponera: Schmidt & Shattuck, 2014: 107. [For earlier references and synonymy see under foetens Fabricius.] Current subspecies: nominal plus amazon, crassicornis, rapax, subpilosa, termitivora.
 * abyssinica. Ponera abyssinica Guérin-Méneville, 1849: 352, pl. 7, fig. 6 (w.) ETHIOPIA. Junior synonym of foetens: Roger, 1860: 310.
 * laeviuscula. Ponera laeviuscula Gerstäcker, 1859: 262 (w.) MOZAMBIQUE. [Also described as new by Gerstäcker, 1862: 505.] Junior synonym of foetens: Roger, 1860: 310.
 * dohrni. Megaponera dohrni Emery, 1902c: 30 (w.) GHANA. Junior synonym of foetens: Forel, 1909b: 64.