Leptogenys Roger

Schmidt, C. A. & Shattuck, S. O., 2014, The Higher Classification of the Ant Subfamily Ponerinae (Hymenoptera: Formicidae), with a Review of Ponerine Ecology and Behavior, Zootaxa 3817 (1), pp. 1-242 : 92-98

Leptogenys Roger

Fig. 15

Leptogenys Roger, 1861: 41 (as genus). Type-species: Leptogenys falcigera Roger, 1861: 42 ; by subsequent designation of Bingham, 1903: 52.

Lobopelta Mayr, 1862: 714 , 733 (as genus in Ponerinae [Poneridae]). Type-species: Ponera diminuta Smith, F., 1857: 69 ; by subsequent designation of Bingham, 1903: 54. Forel, 1892: 520 ( Lobopelta as subgenus of Leptogenys ); Emery, 1896: 177; Bolton 1975a:240 ( Lobopelta as junior synonym of Leptogenys ).

Prionogenys Emery, 1895b: 348 (as genus). Type-species: Prionogenys podenzanai Emery, 1895b: 349 ; by monotypy. Taylor, 1988: 33 ( Prionogenys as junior synonym of Leptogenys ).

Machaerogenys Emery, 1911: 100 (as subgenus of Leptogenys ). Type-species: Leptogenys truncatirostris Forel, 1897: 195 ; by original designation. Brown, 1973: 181; Bolton, 1975a:240 ( Machaerogenys as junior synonym of Leptogenys ).

Odontopelta Emery, 1911: 101 (as subgenus of Leptogenys ). Type-species: Leptogenys turneri Forel, 1900a: 67 ; by monotypy. Brown, 1973: 183; Taylor & Brown, 1985: 32 ( Odontopelta as junior synonym of Leptogenys ).

Dorylozelus Forel, 1915b: 24 (as genus). Type-species: Dorylozelus mjobergi Forel, 1915b: 25 (junior secondary homonym in Leptogenys , replaced by Leptogenys tricosa Taylor, 1969: 132 ); by monotypy. Taylor, 1969: 132 ( Dorylozelus as junior synonym of Leptogenys ).

Microbolbos Donisthorpe, 1948 f: 170 (as genus). Type-species: Microbolbos testaceus Donisthorpe, 1948: 170 ; by original designation. Wilson, 1955b: 136 ( Microbolbos as junior synonym of Leptogenys ).

Leptogenys is the largest ponerine genus, with 211 described extant species and one described fossil species, and is widespread in the tropical and subtropical regions of the world. It is probably sister to Myopias . The genus is notable for its ergatoid queens, frequent specialization on isopods, and for containing species that exhibit an army ant-like lifestyle.

Diagnosis. Workers of Leptogenys are easily distinguished from those of other ponerine genera. Usually the presence of pectinate tarsal claws is enough to identify the genus (as no other ponerines have pectinate tarsal claws), though the tarsal claws of some Leptogenys species are not pectinate. Other diagnostic characters in combination include: slender build, linear mandibles, triangular or distinctly lobed anterior clypeal margin, round propodeal spiracles (rarely slit-shaped), and strong gastral constriction (sometimes only moderately so, and constriction absent in many Afrotropical species of the L. guineensis and L. stuhlmanni groups). Leptogenys is morphologically most similar to Myopias , but Myopias usually has a blunt rectangular projection on the anterior clypeal margin, has simple tarsal claws, and has more mandibular teeth than Leptogenys .

Synoptic description. Worker. Small to large (TL 2.1–14.5 mm) slender ants with the standard characters of Ponerini . Mandibles subtriangular to curvilinear, often without a distinct basal margin, usually lacking teeth except for a single large apical tooth (mandibles with multiple teeth in the L. processionalis group), and articulating with the head at the extreme anterolateral corners of the head; mandibles sometimes with a basal groove. Anterior margin of clypeus angular, with a blunt or sharp point medially, and usually longitudinally carinate medially; sometimes with a narrow blunt lobe medially and a sharp tooth to each side. Frontal lobes small. Eyes small to large, placed at or anterior to head midline. Metanotal groove shallowly to deeply impressed. Propodeum broad to moderately narrowed dorsally. Propodeal spiracles usually small and round, though occasionally a short slit. Tarsal claws usually pectinate, though sometimes with only four or fewer teeth and very rarely simple. Metatibial spur formula (1s, 1p). Petiole usually nodiform (more squamiform in the L. processionalis group), though variable in shape. Girdling constriction between pre- and postsclerites of A4 apparent (less pronounced in the L. processionalis group, absent in many species of the L. guineensis and L. stuhlmanni groups). Stridulitrum present on pretergite of A4. Hypopygium occasionally with a row of stout setae on either side of the sting. Head and body usually shining, though often punctate, foveolate, striate, or rugoreticulate. Head and body with scattered to abundant pilosity, and usually no pubescence. Color variable, testaceous to black.

Queen. Usually ergatoid, flightless and very similar to the worker but with a broader petiole and larger gaster, and often vestigial ocelli; ergatoids also often differ from conspecific workers in various minor characters. The queens of one species ( L. ergatogyna ) are not ergatoid, but are still wingless; unlike other Leptogenys queens, they have the ocelli and thoracic modifications typical of alate ponerine queens ( Wheeler, 1922b; Bolton, 1975a). The queens of at least one species, L. langi , are fully winged ( Wheeler, 1923a). Queens are fully absent from at least a handful of species in which reproduction is performed by gamergate workers ( Davies et al., 1994; Ito, 1997; Gobin et al., 2008).

Male. See descriptions in Arnold (1915), Wheeler (1922b), Bolton (1975a), and Yoshimura & Fisher (2007).

Larva. The larvae of various Leptogenys species have been described by Wheeler & Wheeler (1952, 1964, 1971a, 1974, 1976, 1986b) and Petralia & Vinson (1980).

Geographic distribution. Leptogenys is abundant and species-rich throughout the tropical regions of the world, and to a lesser extent the subtropical regions ( Wheeler, 1922b; Bolton, 1975a). Some members of the L. maxillosa species group are widely-distributed tramp species ( Bolton, 1975a). L. falcigera , for example, was one of the first ants introduced to the Hawaiian Islands ( Kirschenbaum & Grace, 2007).

Ecology and behavior. Leptogenys dwarfs every other ponerine genus in terms of number of described species, accounting for roughly a quarter of the subfamily’s species diversity, and is considered to be a major component of many tropical faunas ( Peeters & Ito, 2001). Though there are numerous exceptions (see discussion below), most Leptogenys species follow a general pattern of having small colonies, nesting in soil or rotting wood, having monomorphic workers which forage individually and are specialist predators of isopods, and having a single ergatoid queen which performs reproduction for the colony.

Average colony sizes vary greatly among Leptogenys species. While colonies of most species have between 25 and 300 workers, some Southeast Asian species average fewer than 10 workers per colony ( Ito, 1997), and the colonies of the mass raiding members of the L. processionalis group can have over 50,000 workers, the largest colonies known in the Ponerinae ( Maschwitz et al., 1989; Witte & Maschwitz, 2000).

Most Leptogenys occur in tropical forests, and like most ponerines construct their nests in soil, leaf litter, or rotting wood ( Forel, 1893b; Arnold, 1915; Mann, 1921; Wheeler, 1922b; Wilson, 1958b; Lenko, 1966; Maschwitz & Mühlenberg, 1975; Bolton, 1975a; Peeters, 1991b; Villet et al., 1991; Duncan & Crewe, 1994b; Ito & Ohkawara, 2000). A few species are known to nest subarboreally in dead tree branches ( Bolton, 1975a) and some have been found nesting inside abandoned termitaries (Déjean et al., 1996, 1997). Due to the typically ephemeral nature of their nesting sites, many Leptogenys conduct frequent emigrations to new nest sites ( Maschwitz & Schönegge, 1983; Shivashankar, 1985; Duncan & Crewe, 1994b; Peeters & Ito, 2001). Members of the L. diminuta and L. processionalis groups are nomadic and form temporary bivouacs similar to those of the true army ants ( Wilson, 1958b; Maschwitz & Mühlenberg, 1975; Maschwitz et al., 1989).

Leptogenys workers are typically nocturnal or crepuscular epigeic foragers (e.g., Arnold, 1915; Wheeler, 1922b; Bolton, 1975a; Shivashankar, 1985; Maschwitz et al., 1989; Kumar, 1990; Déjean & Evraerts, 1997; Witte & Maschwitz, 2000), although some species forage diurnally (e.g., L. breviceps ; Wilson, 1958b; L. intermedia: Duncan & Crewe, 1994b ) and some are cryptobiotic (e.g., L. testacea: Bolton, 1975a ). Two species, L. khammouanensis from Laos, ( Roncin & Deharveng, 2003) and an undetermined species from Texas ( Cokendolpher et al., 2009) dwell deep inside caves and are the only known troglobitic ponerine species. Leptogenys workers are generally very agile, and while some species respond aggressively to nest disturbance and have painful stings, others are more timid ( Jerdon, 1851; Wheeler, 1922b; Wilson, 1958b; pers. obs.). As is typical for ponerines, these venomous stings are also used in prey capture. For example, Maschwitz et al. (1979) found that L. chinensis invariably stings and paralyzes its termite prey. Some Leptogenys species also have additional chemical defenses. L. processionalis and other species have an abdominal gland (the Jessen’s gland) whose secretions have an unpleasant smell and apparently serve a defensive function ( Jessen, 1977; Buschinger & Maschwitz; 1984). L. processionalis also produces compounds in an unidentified cephalic gland, which are hypothesized to act as an early-warning system for predators ( Fales et al., 1992).

Déjean & Evraerts (1997) classified the predatory behavior of Leptogenys species in three categories: solitary hunting and retrieval of prey, group predation after recruitment by a scout, and swarm raiding. While this classification is useful, it is somewhat simplistic since Leptogenys species show a continuum of foraging behaviors from completely solitary foraging to mass raiding, with various degrees of nestmate recruitment between these extremes (see also Maschwitz & Steghaus-Kovac, 1991). Most species practice solitary hunting and retrieval of prey, and most of these are specialist predators of isopods, though some species are known to specialize on earwigs (L. sp. nr. kraepelini: Steghaus-Kovac & Maschwitz, 1993 ; also L. rouxi to a limited extent: Wilson, 1958b), termites (e.g., L. binghami: Maschwitz & Mühlenberg, 1975 ; L. unistimulosa: Mill, 1982a ) or ant queens ( L. neutralis: Wheeler, 1933b ).

The frequent specialization on isopods is a unique ecological characteristic of Leptogenys and is probably the plesiomorphic condition within the genus ( Déjean & Evraerts, 1997). Numerous Leptogenys species have been reported as partially or exclusively specializing on isopods, including L. attenuata and L. schwabi ( Duncan & Crewe, 1993; Davies et al., 1994), L. bohlsi ( Lenko, 1966) , L. bubastis , L. camerunensis , L. donisthorpei , L. mexicana , L. wheeleri , and three undescribed species (Déjean, 1997; Déjean & Evraerts, 1997; Déjean et al., 1999), L. conradti and an unidentified species ( Lévieux, 1982, 1983), L. elongata ( Wheeler, 1904) , L. falcigera ( Kirschenbaum & Grace, 2007) , L. manni ( Trager & Johnson, 1988) , L. “propefalcigera” ( Freitas, 1995), L. rouxi ( Wilson, 1958b) , L. stuhlmanni ( Arnold, 1915) , and L. triloba ( Wilson, 1958b, 1959a). The mandibular structure of Leptogenys seems well suited to predation on isopods: the mandibles are typically long, narrow, curved, and articulated at the extreme anterolateral corners of the head ( Bolton, 1975a, 1994; Déjean et al., 1999), making them excellent tools for clasping round objects such as rolled-up isopods. The medially pointed clypeus probably contributes to this process as well ( Trager & Johnson, 1988). Déjean & Evraerts (1997) studied the behavior of isopod-hunting Leptogenys and found that the behavioral sequence employed by a Leptogenys worker depends on the length of its mandibles and the species and size of the isopod, with different Leptogenys species better adapted to hunting different prey species and sizes. Interestingly, at least one species ( L. mexicana ) chemically attracts its isopod prey to its nest, a remarkable strategy which has not been recorded in any other ant species ( Déjean & Evraerts, 1997).

Limited recruitment to clustered or large prey may occur in some predominantly solitary foraging species. For example, the African species L. schwabi , which preys on termites, isopods and amphipods, and L. attenuata , which feeds on isopods and amphipods, hunt individually but sometimes recruit nestmates to large or clustered prey sources ( Arnold, 1915; Duncan & Crewe, 1993; Davies et al., 1994; Déjean & Evraerts, 1997). Limited recruitment was also reported in L. peuqeuti ( Janssen et al., 1997) .

True group predation, in which workers engage in obligate collective foraging, is characteristic of several African and Asian Leptogenys species ( Déjean & Evraerts, 1997). Perhaps the simplest manifestation of this strategy occurs in the generalist Asian species L. diminuta and its relatives (e.g., L. purpurea ; Wilson, 1958a, 1958b; Maschwitz & Mühlenberg, 1975; Maschwitz & Steghaus-Kovac, 1991). Foraging in L. diminuta begins when scouts leave the nest in search of arthropod prey. Upon finding prey the successful scout returns to its nest and lays down a chemical trail. Once back at the nest, it recruits a group of from three to nearly 300 workers and leads them to the prey, which they attack and retrieve cooperatively ( Maschwitz & Mühlenberg, 1975; Kumar, 1990). Another Asian species, L. chinensis , utilizes a similar strategy but is a specialist predator of termites and hunts in groups of 10 to 50 individuals ( Maschwitz & Schönegge, 1983). Wheeler (1936; also Bingham, 1903) lists additional Asian Leptogenys species that are known to conduct organized raids on termites, including L. aspera , L. binghami , L. birmana , and L. kitteli , and provides interesting observations of termite raiding by some other species.

Several Leptogenys species have done away with the use of scouts altogether, instead employing swarm raids akin to those of the true army ants. For example, workers of the African species L. intermedia form foraging trails from which groups of workers (approximately 30 to 100) cooperatively search for and retrieve leaf litter arthropods ( Duncan & Crewe, 1993, 1994b). Members of the Southeast Asian L. processionalis species group have increased the scale of this swarm raiding, hunting in massive groups of up to 40,000 workers ( Jerdon, 1851; Maschwitz & Mühlenberg, 1975; Maschwitz & Steghaus-Kovac, 1991; Witte & Maschwitz, 2000). The foraging behaviors of three species in the L. processionalis group have been studied to date, and while there are many basic similarities among these species, there are also some interesting differences. L. processionalis itself is a crepuscular forager which preys predominantly on termites, though it also takes other arthropods and annelids ( Shivashankar, 1985). Foraging workers utilize permanent branching trails radiating from the nest, and successful foragers recruit nestmates from the trail to harvest prey, which is collectively dispatched and dismembered but individually retrieved ( Maschwitz & Mühlenberg, 1975). Unlike some other members of the L. processionalis group, L. processionalis does not emigrate frequently, but instead may persist at a nest site and utilize the same trails for up to several months; emigration is initiated by environmental factors ( Maschwitz & Mühlenberg, 1975). Ganeshaiah & Veena (1991) studied the formation and topology of L. processionalis foraging trails and found that they were constructed to maximize travel and prey retrieval efficiency. The flow of individuals in L. processionalis trails was modeled by John et al. (2008).

One of the best studied members of the L. processionalis group is L. distinguenda , which is a generalist predator and will take anything from arthropods and other invertebrates to small vertebrates ( Witte & Maschwitz, 2000). L. distinguenda has taken the army ant lifestyle even further than has L. processionalis , as its massive colonies are even larger than those of L. processionalis (nearly 50,000 versus 16,000 workers, respectively: Maschwitz & Mühlenberg, 1975; Witte & Maschwitz, 2000). More importantly, L. distinguenda colonies emigrate frequently, with emigrations initiated when suitable nest sites are discovered in the course of a raid ( Witte & Maschwitz, 2000).

Finally, an undescribed species closely related to L. mutabilis was studied by Maschwitz et al. (1989) and found to have the largest colonies of any ponerine (over 50,000 workers). These generalist predators forage nocturnally in massive swarms of up to 40,000 workers and capture and retrieve their prey cooperatively. Like L. distinguenda , this species emigrates to new bivouac sites frequently, from every several hours to every 10 days. The combination of obligate collective foraging and nomadism in this and related Leptogenys species makes them quite close to embracing a full army ant lifestyle (sensu Brady, 2003), with the main differences being that they do not display dichthadiigyny or pulsed brood production ( Maschwitz et al., 1989).

Chemical communication has been studied in several Leptogenys species , especially those which exhibit mass raiding behavior. Jessen et al. (1979) characterized the abdominal glands of L. processionalis and L. chinensis . In both species, recruitment pheromones are produced in both the venom and pygidial glands ( Maschwitz & Schönegge, 1977, 1983; Duncan & Crewe, 1994b; Witte & Maschwitz, 2002). In L. processionalis , the secretions of these glands act as orientation pheromones, maintaining the cohesiveness of groups of workers, and are utilized in both foraging and emigrations. Due to differences in chemical composition, workers are able to distinguish between raiding and emigration trails ( Witte & Maschwitz, 2002). L. diminuta produces recruitment pheromones in the pygidial gland ( Attygalle et al., 1988, 1991; Steghaus-Kovac et al., 1992), and also produces pheromones in the venom and Dufour glands ( Maile et al., 2000). Kern & Bestmann (1993) studied the electrophysiological response of the antennae of L. diminuta workers to trail and recruitment pheromones. L. binghami also produces a trail pheromone, but it is primarily utilized during emigrations ( Maschwitz & Mühlenberg, 1975). Janssen et al. (1997) characterized the trail pheromone of L. peuqueti , which is produced in the venom gland and is the most complex trail pheromone known in any ant.

The diversity of foraging strategies and recruitment pheromones in Leptogenys make it an excellent system for understanding the evolutionary origins of cooperative foraging ( Maschwitz & Schönegge, 1983), as well as chemical communication and prey specialization. Most likely, the ancestral Leptogenys was a solitary forager (though perhaps with limited recruitment) which specialized at least partially on isopods. Mass raiding species probably arose from this ancestral archetype, through stages with progressively greater emphasis on collective foraging. This progression is illustrated well by extant species of Leptogenys . While Schmidt’s (2013) phylogenetic data for Leptogenys are limited, it appears that the L. processionalis group may form the sister group to the remainder of the genus. L. diminuta and L. attenuata , both species which exhibit at least limited recruitment, are closely related to each other and to species not known to engage in much recruitment. Greater taxon sampling and natural history observations of more species will be required in order to reconstruct the evolutionary history of Leptogenys foraging behavior.

Reproduction in nearly all Leptogenys species is performed by ergatoid queens. Exceptions to this general rule include one species with fully winged queens ( L. langi: Wheeler, 1923a ) and one species with queens that are flightless but have other characteristics typical of normal alate queens ( L. ergatogyna: Wheeler, 1922b ; Bolton, 1975a). Finally, a few species have completely lost the queen caste and instead reproduce via gamergate workers ( L. schwabi: Davies et al., 1994 ; Peeters, 1991b; L. peuqueti and three undescribed species: Ito, 1997; Gobin et al., 2008). The near universality of flightlessness in Leptogenys queens is probably a major cause of the immense species diversity of the genus, as their poor dispersal ability undoubtedly contributes to reproductive isolation and subsequent allopatric speciation.

Most Leptogenys species which have been examined are monogynous (e.g., L. arnoldi: Arnold, 1915 ; L. attenuata and L. castanea: Villet et al., 1991 ; L. diminuta , L. kraepelini , L. myops , and five undescribed species: Ito, 1997; L. intermedia: Duncan & Crewe, 1994b ; L. sp. nr. kraepelini: Steghaus-Kovac & Maschwitz, 1993 ; L. sp. nr. mutabilis: Maschwitz et al., 1989 ; Leptogenys in general: Wheeler, 1922b). Among species with ergatoid queens, polygyny has only been observed in a single Neotropical species (K. Okhawara & S. Higashi, pers. comm. cited in Ito, 1997). L. diminuta and its relatives are unusual in that their colonies contain a single mated ergatoid queen but also additional unmated ergatoids which perform the same activities as workers; interestingly, the morphological distinction between queens and workers is even less pronounced in this species group than in other Leptogenys ( Ito & Ohkawara, 2000) . All known Leptogenys species with gamergate workers are polygynous ( Davies et al., 1994; Ito, 1997; Monnin & Peeters, 2008). Davies et al. (1994) found that mated gamergates in L. schwabi pheromonally inhibit reproduction by unmated workers; this study also includes one of the only examinations of Leptogenys division of labor. Asexual reproduction by Leptogenys workers presumably occurs, as it does in many other ponerines, but to our knowledge this has not been documented. While workers of those species with gamergates are obviously sexually reproductive, the workers of several Leptogenys species lack ovaries and are completely sterile (e.g., L. intermedia , L. castanea , and L. attenuata ; Villet et al., 1991). Probably not coincidentally, queens in these species have an unusually large number of ovarioles (Villet et al., 1991).

As with foraging behavior, the diversity of reproductive systems in Leptogenys makes the genus an excellent model system for understanding the evolution of flightlessness in queens, the loss of reproductive ability in workers, and the loss of the queen caste (Villet et al., 1991). While Leptogenys is almost certainly descended from an ancestor with winged queens, the almost complete absence of winged queens in extant Leptogenys implies that ergatoid queens are plesiomorphic within the genus. The phylogenetic placement of L. langi (which has alate queens) and L. ergatogyna (with queens intermediate between alates and ergatoids) are unknown but are of particular interest as these species may represent a sister group (or groups) to the rest of the genus. Alternatively, they may represent reversals from ancestors with ergatoid queens. Also of great interest are the phylogenetic placements of species with gamergates or sterile workers.

Virtually nothing is known about Leptogenys mating behavior, but since queens of most Leptogenys species are flightless, mating may occur in the natal nest of the unmated queen, with the males finding the nest through chemical means. Alternatively, males may locate emigrating colonies and mate en route. Hölldobler & Engel- Siegel (1982) examined the abdominal glands of L. diminuta males and discovered that they have a huge sternal gland. They speculated that the secretions of this gland might mimic queen pheromones, enabling the males to enter foreign nests and thus mate with virgin queens. In L. processionalis and L. chinensis , males fly from their natal nest and then land and search for recruitment trails laid by conspecific workers. Once located, the males follow these trails, and could mate with ergatoid queens on the trails ( Maschwitz & Mühlenberg, 1975; Maschwitz & Schonegge, 1983; Peeters, 1991a). Maschwitz et al. (1989) reported that males of an undescribed species closely related to L. mutabilis are carried by workers during emigrations.

As is true for most ponerines, the myrmecophile fauna associated with Leptogenys is virtually unknown, perhaps reflecting a true scarcity of such associations. In this regard L. distinguenda is exceptional, as it has a rich assemblage of myrmecophiles, including mites, isopods, bristletails, silverfish, phorid flies, springtails, spiders, various beetles, and even the only known myrmecophilous gastropod ( Ferrara et al., 1987; Witte et al., 1999, 2002; Kistner et al., 2003). It is probably not a coincidence that the other ponerine species with a rich myrmecophile fauna ( Megaponera analis ) also has relatively large colonies.

Phylogenetic and taxonomic considerations. Leptogenys was erected by Roger (1861) to house three newly described species as well as Ponera maxillosa F. Smith. Bingham (1903) later designated the first of these new species, L. falcigera , as the type species of the genus. Subsequent authors placed Leptogenys variously in Ponerini (e.g., Emery, 1895d) or in its own tribe, Leptogenyini (starting with Forel, 1893a; also called Leptogenysii by some authors). Emery (1911) retained Leptogenys in Leptogenyini based on the shape of the mandibles, the gaster, and the pectination of the tarsal claws. Brown (1963) argued that the pectination of the tarsal claws (which isn’t even universal in Leptogenys ) was a weak character to base a tribe on, and moved Leptogenys into Ponerini . Schmidt's (2013) molecular phylogeny of the Ponerinae confirms that Leptogenys is indeed nested within Ponerini .

The genus-level taxonomy of Leptogenys has also been complex, due to its numerous junior synonyms. Mayr (1862) erected the genus Lobopelta and noted its close similarity to Leptogenys , basing his generic distinction mainly on differences in the mandibular and clypeal structure. Most subsequent authors considered Lobopelta a valid genus, though some treated it as a subgenus or junior synonym of Leptogenys (e.g., Forel, 1892, Emery, 1896). Lobopelta eventually came to hold the majority of the species now included in Leptogenys . Schmidt's (2013) molecular phylogeny of the Ponerinae includes several species formerly placed in Lobopelta (including its type species, L. diminuta ), as well as the type species of Leptogenys , L. falcigera , which clearly emerges from within a non-monophyletic Lobopelta . We are therefore retaining Lobopelta as a junior synonym of Leptogenys .

Emery (1895a) erected the genus Prionogenys based on its unusually long mandibles. Taylor (1988) convincingly argued for the junior synonymy of Prionogenys under Leptogenys , pointing out that derived mandibular structure (and other cephalic characters) is a common occurrence in Leptogenys , presumably related to adaptation to different types of prey. He also suggested that the species formerly placed in Prionogenys may not be sisters, rather that their unusually long mandibles may be convergently derived. Though Schmidt (2013) did not sample either species in his molecular phylogeny, we are maintaining Prionogenys as a junior synonym of Leptogenys given the lack of any compelling morphological evidence that they are distinct.

Emery (1911) erected Machaerogenys as a subgenus of Leptogenys , again based on supposed differences in the mandibles and clypeus. Brown (1973) and Bolton (1975a) provisionally synonymized Machaerogenys under Leptogenys , without any explanation. Though Schmidt (2013) did not sample any former Machaerogenys species in his molecular phylogeny, we do not see any argument for resurrecting it as a valid subgeneric name, and in fact it appears to us to be very closely related to L. falcigera , the type species of Leptogenys . Emery (1911) also created the subgenus Odontopelta , which he differentiated from typical Leptogenys based on the mandibles, clypeus, and the petiole. As with Machaerogenys , we have seen no evidence to suggest that it is nothing more than a derived Leptogenys , and we follow Brown (1973) in treating it as a junior synonym of Leptogenys .

The most unusual junior synonym of Leptogenys is Dorylozelus ( Forel, 1915b) . The single species originally in this genus, D. mjobergi (now Leptogenys tricosa ), is morphologically interesting in that it superficially resembles a cross between a ponerine and the army ant genus Dorylus (hence the genus name; Forel, 1915b; Brown, 1960). It is apparently adapted to a hypogeic lifestyle, with greatly reduced eyes, flattened scapes, a very smooth cuticle, and short robust legs ( Taylor, 1969). The taxonomic placement of this species was unclear for decades after its description, with authors variously placing it in Ponerini ( Forel, 1917) , its own tribe Dorylozelini (also spelled Dorylozeli; Wheeler, 1922b; Donisthorpe, 1943b), or even incertae sedis in Amblyoponini ( Brown, 1960) . Taylor (1969) re-examined the type specimen (the only specimen of the species ever collected) and deduced that it was in fact a Leptogenys , given its pectinate tarsal claws. He also suggested that L. tricosa (as he renamed it) is likely a member of the L. processionalis species group, a hypothesis that we find plausible, as the relatively broad head and wide mandibles of L. tricosa are reminiscent of that group. If true, this would suggest the possibility of a hypogeic army ant lifestyle similar to many Dorylus , perhaps explaining the superficial morphological convergence between these taxa. An alternative hypothesis is that Dorylozelus is closely related to the bequaerti species group of Centromyrmex , with which it bears a remarkable superficial resemblance. The large number of significant morphological differences between them makes it more likely that this is a simple case of convergence, but without corroborating molecular evidence the possibility remains that Dorylozelus is in fact close to Centromyrmex .

The final junior synonym of Leptogenys is Microbolbos , which Donisthorpe (1948) described to house a single species, M. testaceus (now Leptogenys testacea ). Wilson (1955b) re-examined the holotype and concluded that it was in fact a member of Leptogenys . The species is morphologically unusual in that its tarsal claws are simple and it has multiple mandibular teeth, but Wilson (1955b) points out that other Leptogenys species have these characters. We see no reason to disagree with his assessment that Microbolbos is a junior synonym of Leptogenys .

Leptogenys is a member of the Odontomachus group, and along with its probable sister group Myopias (and possibly Mesoponera ) it apparently forms the sister group to the rest of the Odontomachus group ( Schmidt, 2013). Leptogenys is morphologically most similar to Myopias , with both genera typically having linear or at least subtriangular mandibles, round propodeal spiracles, a nodiform petiole, a strong constriction between A3 and A4, a prominent sting, and relatively smooth and shiny cuticle. Of these, the only character that is likely to be truly synapomorphic for these genera is the non-triangular mandible shape, and possibly the prominent sting, with most of the other characters likely to be plesiomorphies. Round propodeal spiracles are lacking in most other members of the Odontomachus group except Brachyponera and many Mesoponera , but the plesiomorphic condition in the group is uncertain.

Except for the small sampling of species in Schmidt’s (2013) molecular phylogeny of the Ponerinae , nothing is known about the internal phylogeny of Leptogenys . Schmidt's phylogeny suggests an early split between the L. processionalis species group and the remainder of the genus. Given this fact, as well as their divergent morphological structures and behavior, the L. processionalis group could arguably be split off into a separate genus, though we are not going to take that step. The L. falcigera group ( Leptogenys sensu stricto), with its very thin and curved mandibles, has clearly emerged from within a non-monophyletic “ Lobopelta ”. The historical biogeography of Leptogenys is likely to be complex, with much migration between Africa and Asia/ Australia ( Bolton, 1975a) and probably at least two separate invasions of the New World (including once by relatives of L. falcigera ). Given the potential utility of Leptogenys as a model system for studying the evolution of foraging and reproductive behavior, as well as historical biogeography, a detailed species-level phylogeny of Leptogenys would be extremely valuable. In fact, there is probably no better target for such work within the Ponerinae , and it should be made a priority for near-term study.

Figures

FIGURE 15. Worker caste of Leptogenys. Upper specimen: Leptogenys darlingtoni: lateral and dorsal view of body and fullface view of head (ANIC32-059614). Lower specimen: Leptogenys maxillosa: lateral view of body and full-face view of head (Philippines, G.Alpert, MCZC); world distribution of Leptogenys.