Pheidole longispinosa

A dominant ant in the Madagascar forests where it occurs. The species was collected at between 80–1680 m in elevation, in montane rainforest, montane shrubland, thicket, tropical dry forest, and urban areas. Nests were located in rotten logs, tree stumps, soil, and rock crevasses (Salata & Fisher, 2020).

Distribution based on Regional Taxon Lists
Malagasy Region: Madagascar.

Biology
Lampasona (2015) observed a colony of Pheidole longispinosa located at 21°09′21.3″ S, 047°33′04.1″ E, 1078 m (near the village of Miaranony, northern area of Ranomafana National Park, Madagascar). Primary forest, consisting of old growth trees including Rosewood and large tree ferns, covered the mountain. The nest was located in a rotting log (approx. 10 m long), which housed Pheidole longispinosa, Paraparatrechina glabra and a termite nest. A large Pheidole oswaldi nest occurred on the ground near the long.

Pheidole longispinosa exhibited a dominant behavior over local insect fauna in the area around their nest. Workers preyed upon neighboring ant species with relative impunity, only experiencing resistance when engaged in predation against P. oswaldi. In the local forest floor of the study site, P. longispinosa acted as a top-tier invertebrate predator, exhibiting Wilson’s “extirpator” category of ant behavior on other colonies in the area, by hunting and killing large arthropod prey, aggressively defending those kills, and engaging in periodical raids on rival ant colonies. They attacked and killed or drove off intruding ants of any species that wandered onto their territory, including large predators such as Odontomachus coquereli. During all conflicts with P. oswaldi or O. coquereli, (which occurred daily at least) major workers took part in the combat, either assisting minors with dismembering large prey that had been pinned, or by crushing smaller ants in their large mandibles. Majors performed both tasks when confronted with larger ants such as O. coquereli, but did not assist in killing pinned P. oswaldi, opting to simply crush them. The presence of major workers in conflicts was only observed in conflict with ants that posed a threat to minors (P. oswaldi and O. coquereli, as opposed to the largely helpless Paraparatrechina glabra). These ants were the only insects observed to attempt to attack minors during observation periods. When bringing in non-social insects, majors only assisted minors with insects that they could not physically kill or fit into nest entrances intact. Large Lepidopteran larvae and Pentatomids were the only non-social insects seen to spurmajors into action, and that was only after several minutes of approximately 15 or more minors struggling to bring the prey in with no success.

In both cases, majors were recruited by minors during periods of conflict; majors followed minors to the area of engagement, killed and/or dismembered prey items, then returned to the nest within seconds of completing their task. Majors never patrolled, except for periods directly after battles with P. oswaldi, during which time they antennated all encountered insects heavily, attacking non-nestmates immediately after recognition. Majors engaged enemies directly, crushing their head capsules or gasters, and leaving the bodies behind, which were gathered by minors, who would sweep the combat zone shortly after fighting concluded. Majors were never observed gathering food far from the main nest, only killing or dismembering prey brought closer by minors. Throughout the day, majors could be observed waiting directly inside nest entrances, checking any workers attempting to return, and barring entrance to any foreign insects. At night, no ants were visible at entrances.

Rescue Behavior Workers of P. longispinosa were observed rescuing a nestmate from predation by a Clubionid spider on their nest log. This individual spider had a small nest under the bark, and it’s web was stretched between two raised sections of bark. It had been observed on several occasions preying upon ants that wandered into this web, and its nest was saturated with P. glabra exoskeletons. An individual P. longispinosa minor worker became trapped on the surface of the log, in this same small web. In this instance, the trapped worker began vibrating and repeatedly tapping the log with its abdomen. This method of “substrate borne communication” has been observed in ants as a method of alerting nearby nestmates (Virant-Doberlet and Cokl 2004). In less than 60 s, 6 minor workers appeared from over the opposite side of the log. Two of these workers attacked the spider, chasing it from the log onto the path. The other 4 remained with the trapped worker. Over the next 18 minutes, these workers used their mandibles to cut the trapped worker down from the web and clean its eyes and antennae. Two workers constantly cleaned the trapped worker for the entire duration of this time, while the other 2 shifted about half their time between cleaning and engaging in allogrooming with the damaged ant. Ant rescue behavior has been observed in several contexts, and is quoted by Nowbahari and Hollis (2010) as “cooperative self defense,” particularly in reference to Driver Ants predating upon Megaponera analis, and the defenses that the latter ants mount in response. However, in this observed instance, the rescuers were not in any immediate danger, and purposefully moved towards the spider, putting themselves at personal risk, while it could have simply been ignored.

Conferring workers of both P. longispinosa and P. oswaldi were observed communicating in tight knots of workers. When encountering annelids or gastropods, workers would often make antennal contact with the potential prey, before retreating to a group of nestmates. This “messenger” ant would then move its head between each worker in the group, antennating each heavily. The workers then moved in a similar fashion as the original communicator, antennating frequently between each other. The workers would then either depart or attack the prey in unison.

Pheidole longispinosa workers behaved in an aggressive manner to nearby ant and termite colonies, yet never initiated any large-scale attacks on neighboring colonies. Their aggression was limited to raids on workers and occasional skirmishes along their foraging territory. They were also observed taking workers from the termite colony in the log they resided in. Excavation of sections of the log revealed that Pheidole longispinosa and Paraparatrechina glabra very likely nested in tunnels produced by these same termites.

Communication among Pheidole workers was also observed closely during study, and several behaviors emerged that merited more attention. The “conferring” behavior seen in P. longispinosa and Pheidole oswaldi was only observed when foragers were confronted with unappealing, potentially inedible prey such as leeches and land snails. The workers involved may have spread the scent of the potential prey amongst its nestmates. The group, after several seconds of such contact, either attacked or abandoned the prey in unison.

During the completion of transects near the studies colony, as well as on a main trail leading from Sakabe to Miaranony, P. longispinosa workers were observed carrying other prey ants, often still alive, along the forest floor. Two separate groups of workers, from different colonies, were also observed on separate occasions attacking and killing large predatory Odontomachus coquereli foragers by pinning them to the ground, and recruiting majors to dismember them. P. longispinosa was observed in the primary forests of the Sakabe Mountain, in disturbed forest edge adjacent to Miaranony village itself, and agricultural fields, including fallow rice paddies and active tavy fields. While hiking along study transects (in areas separate from Main Nest), I observed P. longispinosa workers capturing P. glabra, indicating that these ants may be a food source for other colonies of P. longispinosa with some regularity. Based on their consistent and concerted raids on rivals and aggressive predation of various insects, including large and aggressive rival ants, P. longispinosa acted as the locally dominant ant in this section of the forest floor.

Worker
minor

Nomenclature

 *  longispinosa. Pheidole longispinosa Forel, 1891b: 170, pl. 5, fig. 4 (s.w.q.) MADAGASCAR.

Type Material

 * Lectotype (designated by Salata & Fisher, 2020): major worker (top specimen, CASENT0101682): Madagascar, Antananarivo, Ambatomanjaka, Miarinarivo (MHNG) [examined].
 * Paralectotypes: 1 major worker (CASENT0810542, bottom specimen, the same pin as lectotype) (MHNG) [examined], 3 minor workers (1 pin, CASENT0101598, CASENT0923202, CASENT0923203) (MHNG) [examined], 1 dealate queen (CASENT0101657) (MHNG) [examined]: the same data as lectotype.