Generic revision of the ant subfamily Dorylinae (Hymenoptera, Formicidae) Author Borowiec, Marek L. text ZooKeys 2016 608 1 280 http://dx.doi.org/10.3897/zookeys.608.9427 journal article http://dx.doi.org/10.3897/zookeys.608.9427 1313-2970-608-1 F865473C03374FD2915A0E3DD2299E66 Taxon classification Animalia Hymenoptera Formicidae Nomamyrmex Borgmeier, 1936 Type-species. Eciton crassicornis (junior synonym of Labidus esenbeckii ), by original designation. Nomamyrmex is a relatively commonly observed genus with only two species and two additional subspecies recognized. It is the only army ant genus that has been reported to successfully attack well-defended and often enormous colonies of Atta leaf cutter ants. Diagnosis. Worker. The workers of Nomamyrmex are easily recognized by a combination of highly positioned spiracle and lack of pronounced propodeal lobes, propodeum armed with cuticular projections, two-segmented waist, armed pretarsal claws, and absence of metatibial gland. The lack of conspicuous lighter area of cuticle on the inner side of hind tibia (the metatibial gland) distinguishes this genus from all other Eciton genus-group ants except for some Neivamyrmex , but those always have simple pretarsal claws. Male. Nomamyrmex males possess traits characteristic of New World army ants; see discussion under Cheliomyrmex for characters distinguishing New World army ant males from those of the Old World. Nomamyrmex is also easily told apart from other New World army ant males by its dense tufts of very long hairs present on the gaster. Eciton setigaster is one species that could be mistaken for a Nomamyrmex , but the setae on its gaster are not as dense or as long, not approaching front femur length. Description. Worker.Head: Antennae with 12 segments. Apical antennal segment not enlarged, not broader and longer than two preceding segments combined. Clypeus with cuticular apron. Lateroclypeal teeth absent. Parafrontal ridges reduced. Torulo-posttorular complex vertical. Antennal scrobes absent. Labrum with median notch or concavity. Proximal face of stipes projecting beyond inner margin of sclerite, concealing prementum when mouthparts fully closed. Maxillary palps 2-segmented. Labial palps 3-segmented. Mandibles triangular, with teeth. Eyes present, appearing as single large and convex ommatidium, in reality composed from fused ommatidia. Ocelli absent. Head capsule with differentiated vertical posterior surface above occipital foramen. Ventrolateral margins of head without lamella or ridge extending towards mandibles and beyond carina surrounding occipital foramen. Posterior head corners dorsolaterally immarginate. Carina surrounding occipital foramen ventrally absent. Mesosoma: Pronotal flange separated from collar by distinct ridge or not. Promesonotal connection with Pronotomesopleural suture completely fused. Pronotomesopleural suture visible, unfused partway to notal surface. Mesometapleural groove not impressed. Transverse groove dividing mesopleuron absent. Pleural endophragmal pit concavity present. Mesosoma dorsolaterally immarginate. Metanotal depression or groove on mesosoma present. Propodeal spiracle situated high on sclerite. Propodeal declivity with or without distinct dorsal edge or margin and rectangular in posterior view. Metapleural gland with bulla visible through cuticle. Propodeal lobes present, short. Metasoma: Petiole anterodorsally marginate, dorsolaterally immarginate, and laterally above spiracle immarginate. Helcium in relation to tergosternal Pronotomesopleural suture placed at Pronotomesopleural suture and axial. Prora forming a simple U-shaped margin or V-shaped protrusion. Spiracle openings of abdominal segments IV-VI oval to slit-shaped. Abdominal segment III anterodorsally immarginate and dorsolaterally immarginate. Abdominal segment III about half size of succeeding segment IV, which is strongly constricted at presegmental portion (binodal waist). Girdling constriction of segment IV present, i.e. pre- and postsclerites distinct. Cinctus of abdominal segment IV gutter-like and sculptured but not cross-ribbed. Abdominal segment IV conspicuously largest segment. Abdominal tergite IV not folding over sternite, and anterior portions of sternite and tergite equally well visible in lateral view. Girdling constriction between pre- and posttergites of abdominal segments V and VI absent. Girdling constriction between pre- and poststernites of abdominal segments V and VI absent. Pygidium small, reduced to narrow strip, without impressed medial field and simple, not armed with cuticular spines or modified setae. Hypopygium unarmed. Legs: Mid tibia with single pectinate spur. Hind tibia with single pectinate spur. Hind basitarsus not widening distally, circular in cross-section. Posterior flange of hind coxa not produced as raised lamella. Metatibial gland absent. Metabasitarsal gland absent. Hind pretarsal claws each armed with a tooth. Polymorphism: Polymorphic. Male.Head: Antennae with 13 segments. Clypeus without cuticular apron. Parafrontal ridges absent. Torulo-posttorular complex vertical. Maxillary palps 2-segmented. Labial palps 3- or 2-segmented. Mandibles falcate. Ventrolateral margins of head without lamella or ridge extending towards mandibles and beyond carina surrounding occipital foramen. Carina surrounding occipital foramen ventrally absent. Mesosoma: Pronotal flange not separated from collar by distinct ridge. Notauli absent. Transverse groove dividing mesopleuron absent. Propodeal declivity reduced, without distinct dorsal edge or margin. Metapleural gland opening absent. Propodeal lobes present. Metasoma: Petiole anterodorsally immarginate, dorsolaterally immarginate, and laterally above spiracle immarginate. Helcium in relation to tergosternal Pronotomesopleural suture placed at Pronotomesopleural suture and axial. Prora forming a simple U-shaped margin. Spiracle openings of abdominal segments IV-VI slit-shaped. Abdominal segment III more than half size of succeeding segment IV; latter weakly constricted at presegmental portion (uninodal waist). Girdling constriction of segment IV present, i.e. pre- and postsclerites distinct. Cinctus of abdominal segment IV gutter-like and cross-ribbed. Girdling constriction between pre- and postsclerites of abdominal segments V and VI absent. Abdominal segment IV not conspicuously largest segment. Abdominal sternite VII simple. Abdominal sternite IX distally armed with two spines, with lateral apodemes about as long as medial apodeme, directed anteriorly (towards head). Genitalia: Cupula very long, nearing or surpassing length of rest of genital capsule and of approximately equal length on both dorsal and ventral surfaces. Basimere narrowly fused to telomere, with sulcus discernable at junction, and ventrally with left and right arms abutting. Telomere expanded at apex. Volsella laterally flattened, narrowly triangular in lateral view, narrowing towards tip. Penisvalva curved ventrally at apex, with short dorsal and longer ventral process. Legs: Mid tibia with single pectinate spur. Hind tibia with single pectinate spur. Posterior flange of hind coxa not produced as raised lamella. Metatibial gland absent. Metabasitarsal glands absent. Hind pretarsal claws each armed with a tooth. Wings: Tegula present, broad, demiovate in shape. Vein C in fore wing present. Pterostigma narrow. Abscissa R·f3 present, running toward distal wing margin and enclosing cell with Rs·f5. Abscissae Rs·f2-3 present, connecting with Rs+M&M·f2. Cross-vein 2r-rs present, differentiated from Rs·f4 by presence of Rs·f2-3. Abscissae Rs·f4-5 differentiated into Rs·f4 and Rs·f5 by 2rs-m. Abscissa M·f2 in fore wing present, separated from Rs+M by Rs·f2. Abscissa M·f4 in fore wing present, reaching wing margin. Cross-vein 1m-cu in fore wing present. Cross-vein cu-a in fore wing present, arising from Cu and distal to, at or near M·f1. Vein Cu in fore wing present, with both branches Cu1 and Cu2. Vein A in fore wing with abscissae A·f1 and A·f2 present. Vein C in hind wing present. Vein R in hind wing present, reaching distal wing margin. Vein Sc+R in hind wing present. Abscissa Rs·f1 in hind wing present, shorter than 1rs-m. Abscissa Rs·f2 in hind wing present, reaching wing margin. Cross-vein 1rs-m in hind wing present, about as long as M·f1. Vein M+Cu in hind wing present. Abscissa M·f1 in hind wing present. Abscissa M·f2 in hind wing present. Cross-vein cu-a in hind wing present. Vein Cu in hind wing present. Vein A in hind wing with abscissae A·f1 and A·f2 present. Gyne. Dichthadiiform, with falcate mandibles, small eyes, and no ocelli. Known for Nomamyrmex esenbeckii ( Borgmeier 1958 ). Larva. Not described. Cocoons present. Distribution. Both Nomamyrmex species are widely distributed and the genus is found from Texas to northern Argentina. Taxonomy and phylogeny. The two species of Nomamyrmex were known since Westwood described them in 1842, but he treated them under Labidus . Borgmeier introduced Nomamyrmex as a subgenus of Eciton ( Borgmeier 1936 ). Several names have been published for these widely distributed insects but the species-level taxonomy has been in relative stability thanks to the monumental efforts of Borgmeier (1953 , 1955 ) who examined much of the type material available and recognized the extensive synonymy, reducing the number of species to the two originally described by Westwood, Nomamyrmex esenbeckii and Nomamyrmex hartigii . There is a marked variation in the morphology of Nomamyrmex esenbeckii and this led Borgmeier and subsequent authors to recognize three or four subspecies (see Watkins 1977 b). Borgmeier reported sympatry of some of those subspecies ( Borgmeier 1955 , 1958 ) but recently the view has expressed that this variation is seen in largely allopatric populations with numerous intermediates known and that the subspecies are best treated as synonyms of esenbeckii (Gordon Snelling pers. comm., Wild 2007 ). However, the formal synonymization of two of these subspecies, Nomamyrmex esenbeckii wilsoni and Nomamyrmex esenbeckii mordax has yet to be made. The species-level taxonomy of Nomamyrmex would benefit from a thorough morphometric and molecular phylogenetic study. Brady et al. (2014) and genomic data (Borowiec, in prep.) recover a well-resolved clade of Labidus sister to Nomamyrmex plus Eciton . It may be noted that Borgmeier (1955 : 137) wrote ' Nomamyrmex stands between Labidus and Eciton ' when referring to the genital morphology of Nomamyrmex as showing similarities to the latter two genera. Biology . Henry Walter Bates was perhaps the first to report on the habits of Nomamyrmex in his famous narrative ( Bates 1863 ), describing a '(...) very stout-limbed Eciton , the Eciton crassicornis , whose eyes are sunk in rather deep sockets' that '(...) goes on foraging expeditions like the rest of its tribe, and attacks even the nests of other stinging species ( Myrmica ), but it avoids the light, moving always in concealment under leaves and fallen branches' . Borgmeier (1955) and Rettenmeyer (1963) summarize what was known about Nomamyrmex to date, most observations being on Nomamyrmex esenbeckii . The summary below regarding raids and emigrations is based on these resources unless stated otherwise. Nomamyrmex presumably forms bivouacs which are always subterranean and have never been directly observed. Based on the durations of emigrations observed, Rettenmeyer (1963) estimated that the colonies must be enormous, perhaps in the excess of a million workers. The diet of these army ants consists mostly of immatures of multiple species of other ants, although they have been observed raiding nests of other social insects, including termites and bees (see also Souza and Moura 2008 ). It appears that raids are primarily subterranean, although columns of these ants are also observed above ground. The raid columns are narrow, not forming swarms. The raids have been observed both at night and during the day and often last throughout the day. Rettenmeyer reports that Nomamyrmex esenbeckii on Barro Colorado, Panama conducted raids mostly during the day but there are reports of the same species raiding at night and being strongly photophobic ( Sanchez-Pena and Mueller 2002 ). Given the large size of the colonies, raids and emigrations can take a very long time and last well over 24 hours ( Rettenmeyer 1963 , Powell and Clark 2004 ). Numerous myrmecophiles have been observed in emigration columns, including multiple limulodid beetles riding the emigrating queen. The brood is synchronized. A remarkable aspect of Nomamyrmex biology is the capability to successfully raid the huge colonies of leaf-cutting ants in the genus Atta , otherwise mostly ignored by army ants. Most published records of Nomamyrmex foraging contain observations of raids on leaf cutters ( Swartz 1998 , Sanchez-Pena and Mueller 2002 and references therein) and Powell and Clark (2004) conducted the most comprehensive study of interactions between these ants to date. They show that Nomamyrmex is capable of successfully raiding both young and mature colonies of Atta and that the latter respond in a specific manner to the presence of workers of Nomamyrmex but not Eciton . The leafcutters defend their nests by mobilizing large numbers of major workers and plugging nest entrances with cut leaves. Nomamyrmex and Atta workers that directly engage in combat are most often the largest ants in the colonies of both species and the encounters usually result in the ants becoming locked head-to-head. Furthermore, slightly smaller workers of both species also participate in combat but in a slightly different way. On the Atta side, they assist in spread-eagling the attacking army ants while the 'primary combatants' are locked with their mandibles. On the Nomamyrmex side they overrun and sting the leaf-cutter majors to death. Nomamyrmex is capable of inflicting significant damage on a raided Atta colony. A subterranean raid on a partially excavated Atta mexicana colony was observed where the army ants killed a large proportion of adult Atta , including the queen ( Rettenmeyer et al. 1983 ). Swartz (1998) reported that an army ant raid on a young Atta cephalotes colony extirpated the leaf-cutters and eventually turned into an emigration, the Nomamyrmex colony relocating into the abandoned nest. Powell and Clark (2004) estimated that during one nearly 36-hour raid the Nomamyrmex removed over 60,000 brood items from an Atta cephalotes colony, possibly over a half of all the brood present in the nest. Species of Nomamyrmex Nomamyrmex esenbeckii (Westwood, 1842): Brazil Nomamyrmex esenbeckii mordax (Santschi, 1929): Mexico Nomamyrmex esenbeckii wilsoni (Santschi, 1920a): United States Nomamyrmex hartigii (Westwood, 1842): Brazil