Cerapachyini

Brown, W. L., 1975, Contributions toward a reclassification of the Formicidae. V. Ponerinae, tribes Platythyreini, Cerapachyini, Cylindromyrmecini, Acanthostichini, and Aenictogitini., Search: Agriculture; Cornell University Agricultural Experiment Station 15, pp. 1-115 : 14-18

Tribe Cerapachyini

Worker: Essentially monomorphic, though often with a considerable range of sizes in a single species and even in a single nest series. Smaller workers tend to have allometrically narrower heads and usually lack ocelli; larger workers apparently grade into ergatoid queens in many species and have 1 or all 3 ocelli; in some species, all workers have 3 ocelli.

Integument hard and thick, smooth, or variously sculptured, in most species at least partly punctate. Color of full adults ranging from yellow through red or brown to black, dark species sometimes with a blue opalescent overlay. Pilosity of simple, fine, tapered hairs, sparse and short to fairly abundant and long; sometimes specialized into a shorter pubescence that may be dense on limited parts of the body.

Head unremarkable in form, usually slightly to considerably longer than broad, with subparallel, moderately convex sides and transverse posterior border; cervical border concave. Undersides of posterior corners of head on each side with a carina that extends forward for a greater or lesser distance in most species (fig. 93). Frontal lobes sometimes lying horizontal or only obliquely raised, and separated, but more commonly sharply raised or vertical and contiguous, with the frontal carinae fused behind.

Compound eyes varying from very large (and placed in anterior, middle, or posterior position on sides of head) to small or even absent.

Antennae thick, with short clavate scapes and clavate funiculi, 9 - 12 segments total, inserted near front of head. Apical segment largest. From each antennal insertion a more or less flat or concave area extends laterad and is bounded near the side of the head by a curved or angled carina that runs forward across the cheek and ends at the lateral wing of the clypeus (figs. 46, 67). Clypeus narrow, crowded by the anteriorly displaced frontal lobes and antennal insertions, and extending back between these lobes only when they are separated, a condition found in a minority of species; free margin of clypeus depressed, usually with a small, rounded median lobe (often translucent). The flat area around the antennal insertion sometimes extends forward as a flat angulate or rounded process from the anterior clypeal margin on each side of the middle.

Mandibles short, thick, downcurved, subtriangular, with convex outer borders, distinct basal and masticatory borders meeting at an angle; masticatory border edentate, crenulate, or obscurely dentate; when closed, held close to clypeus.

Under mouthparts (see Gotwald, 1969: 43 - 48, plates 29, 30, 32) compact; stipites heavily sclerotized, each with a transverse groove; labrum broader than long, bilobed, with an anteromedian notch.

Palpi ranging from maxillary 6 - to 2 - merous (1 - merous in Leptanilloides) and labial 4 - to 2 - merous; it is possible that some of the many subterranean species not yet dissected have fewer segments in one or both pairs of palpi. The most common formulae seem to be maxillary 2 or 3, labial 2 segments.

Trunk rigid and box-like, the dorsal sutures obsolete or nearly so, except in the 2 fossil species and very rarely in extant species, which have a deeply cut and possibly flexible promesonotal suture. Dorsal surface convex to nearly flat, in some groups of Cerapachys marginate dorsolaterally and constricted in the middle as seen from above; propodeum with a distinct dorsum and declivity, the latter often marginate, but unarmed. Propodeal spiracle usually round or elliptical, rarely a slit, placed low, near or below level of trunk axis. Bulla of metapleural gland distinct. Posterior coxa often with genual plates well developed, and sometimes with the mesal plate extended as a prominent, rounded, translucent dorsal lobe (fig. 3).

Posterior tibiae each with a single pectinate apical spur; middle tibiae with or without a smaller apical spur, usually pectinate when present. A second, smaller tibial spur is present in Cerapachys crawleyi on both middle and hind legs. Tarsal claws in most species slender and simple, but in a minority of species, the claws each have a submedian tooth, or may even be subpectinate, that is, with a basal lobe or tooth in addition to the submedian tooth.

Petiolar node sessile, various in shape with the species, with or without anterodorsal and dorsolateral margins. Postpetiole a clearly defined segment separated front and rear by constrictions from the adjacent segments, and varying in relative bulk by comparison with these segments so as to appear either as a virtual part of the gaster ( Sphinctomyrmex ) or as a second node in the waist ( Cerapachys typhlus group), as in Myrmicinae or some army ants (e. g. Eciton , Neivamyrmex , Aenictus ). All intermediate conditions exist in various cerapachyine species, so that the tribe as a whole forms a morphocline from the 1 - to the 2 - segmented waist often used to distinguish subfamilies among the Formicidae . (See figs. 91, 95, 99, 100). The postpetiole usually, however, has a well-developed, convex sternum.

More or less correlated with the variation in postpetiolar size and proportions is the wide variation in form of the gaster among cerapachyine genera and species; this also forms a rough morphocline. In most Sphinctomyrmex , the first 3 gastric segments after the postpetiole are similar in size and distinctly set off by intersegmental constrictions.

In Cerapachys , the first gastric segment (true abdominal segment IV) dominates the succeeding segments, and this tendency is carried to an extreme in the C. typhlus group in such species as biroi (fig. 95) and edentata , where the first segment covers nearly the whole of the gaster, and the succeeding segments are reduced and crowded into an apical position, as in most Myrmicinae .

Pygidium usually impressed discad, the laterapical borders with raised edges furnished with a row or a field of minute acute denticles; these may be reduced to only 4 or 6 denticles in some presumably arboreal ( Simopone ) species, but they are usually more numerous. In Leptanilloides, in which the pygidium is reduced and shifted to a ventral position, the denticles are absent. Sting present and functional, usually well-developed, but no sign of an exserted sting in the 3 Leptanilloides specimens available.

Queen: Normal (i. e., winged when virgin, with large eyes, ocelli, and developed pterothorax), or ergatoid to various degrees in different species, or even in a single colony. The boundary between worker and queen is especially vague and problematical in this tribe. In Sphinctomyrmex , ergatoidy in some species has begun to converge toward the dichthadiiform condition of army ant queens, which are characteristically without eyes and ocelli, have broadened head and petiole, reduced wingless trunk, and relatively large gaster. Often more than one ergatoid is found in a colony, but it is not known whether these are all functional reproductives. In most species, queens resemble large workers in size, sculpture, and color, as well as in details of the mouthparts, etc.

Wings, when present, basically of the ponerine pattern, but with thick, broad, heavily pigmented pterostigma, and often with veins weak or shortened in the apical half. Reduction of venation is notable in some smaller species.

Male: (Unknown for Leptanilloides.) Similar to the conspecific queen in size and robustness, or a little smaller; head shorter, more nearly globular, with very large, semiglobose eyes. Frontal carinae a reduced copy of those of conspecific workers, usually forming half-rings around inner front sides of antennal sockets; clypeus broad and usually with translucent, rounded median apron on the free margin. Mandibles well developed, triangular, fig. 96; (falcate in some Sphinctomyrmex , figs. 97 - 98), edentate, meeting or crossing when closed. Antennae geniculate, 13 - merous, occasionally 12 - merous, with scape short, not reaching beyond ocellar triangle, but still decidedly longer than funicular segments I, II, and III, and usually longer even than apical segment; the last sometimes slightly incrassate, but in general the antenna is slender and without the strong club of the worker and queen; pedicel very short, much shorter than succeeding funicular segment (II). The strong curved or angulate cheek carinae of the workers are not developed in the males. Palpi in the few species examined like those of worker.

Trunk moderately developed; notauli present in some species of Sphinctomyrmex (e. g., asper ), but weakly developed or absent in most other known species. Wings with rather weak venation, ranging from complete ponerine pattern to lacking most distal elements back to midlength (fig. 92); pterostigma usually large, heavily sclerotized and dark in color; hind wing without anal lobe. Legs robust; mid-tibiae with 2, 1, or 0 apical spurs, and hind tibiae with 1, rarely 2, pectinate spurs. Tarsal claws with submedian tooth in some species, rarely (some Simopone ) with a second lobe or tooth, but usually the claws are simple.

Petiole usually simple, with a subcuboidal, dome-shaped or elongate-rectangular node. Postpetiole simple, a little smaller than following (first gastric) segment, and separated from it by a weak to strong constriction; gaster usually more or less cylindrical, slightly downcurved, with or without constrictions between segments ( Sphinctomyrmex ), or narrowed segmental bases amounting to constrictions (e. g., in Cerapachys augustae ). Subgenital plate very diverse in different species-groups, but usually with bilateral points or spines, often forming a " fork " (figs. 114 - 116, 118, 120 - 122). In some Australian species of the " Phyracaces groups, " the bidentate condition is lost or nearly lost (fig. 116). Genital capsule with all main parts (parameres, volsellae, aedeagus) present, but very diverse in form with genus and species (figs. 123 - 129). The sample available is still too small and unrepresentative to allow the analysis of phylogenetic trends and group characters, and it sometimes seems that intrageneric variation is more extreme than it is among genera. The figures at least serve to show the variety of forms among the genitalia; further analysis will, of course, require mesial views of the separated parts. Note that even within single genera, parameres may be separated into basal and distal pieces (gonocoxites and gonostyli?), as in figs. 123 and 127; or not, as seen in figs. 124 and 129. In a few species ( Sphinctomyrmex ) the pygidium is margined and truncate; in some others, denticulate margins occur.

Sculpture and pilosity varies with the species, but most have some smooth and coarsely punctate areas, with unremarkable, fine pilosity. Color variable, often black or brown, but sometimes bicolorous.

Larvae: G. C. Wheeler (1950) and G. C. and J. Wheeler (1964, 1973, 1974) have published descriptions, mostly with figures, of 11 species of Cerapachyini (4 Cerapachys , 4 " Phyracaces ", 1 " Lioponera ", 1 " Eusphinctus ", and 1 Simopone ). Their key to genera in the 1964 paper (p. 67) is based entirely on hair form. Clark (1923) gives data and figures on larval hairs for 2 species of " Eusphinctus . " These hairs are bifid, trifid, or 6 - branched, but very young larvae may have simple hairs. In the same paper, Clark gives hair shape for 2 species of " Phyracaces " as bifurcate, with head hairs simple in at least one species.

The 1964 and 1973 papers by the Wheelers show that Cerapachys edentata (= australis ) has both bifid and simple hairs on both head and body; body hairs may show secondary branching. Nearly the full range of Cerapachyini hair types (including hairs absent) is shown by the genus Cerapachys in the old sense (i. e., excluding Phyracaces and Lioponera ), and it is evident that even though only a small, poorly representative fraction (at most 10 % of the species) of cerapachyine larvae has been looked at, this sample fails to show a convincing division into generic lines on larval characters, at least on the characters that have thus far been emphasized. Even the few long, hooked hairs ringing each somite of the Lioponera luzuriagae larvae look like only a species or species-group character, adaptive in twig-inhabiting forms, and even some of these larvae had bifurcate body hairs in some samples examined (G. C. Wheeler 1950), though they were thought to be dried-out specimens.

Pupae: So far as known, normally enclosed in cocoons.

BIONOMICS of CERAPACHYINI Two of the 4 genera of Cerapachyini ( Cerapachys , Sphinctomyrmex ) include a number of species that have been observed raiding the nests of other ants (see especially Wheeler 1918; Clark 1923, 1924, 1941; Wilson, 1958). I can add observations on Cerapachys mayri and C. lividus in the rain forest of Madagascar, and on C. indicus in the Western Ghats of southern India. All 3 Cerapachys species were found attacking, or returning from attacks on, nests of Pheidole species, and the booty being transported consisted of larvae, pupae, pharate adults, and even in some cases dead but fully pigmented Pheidole soldiers and workers. These raids were all observed in progress near midday in shaded situations, and the raiding workers returned one by one over logs and roots along an obvious odor trail with their prey.

In Australia, a raid of C clarki was observed in 1951 in open eucalypt woodland at Berrimah, near Darwin, Northern Territory, over bare soil against a nest of a small Iridomyrmex in bright morning sunlight. This raid was very loose and involved only 3 or 4 Cerapachys workers while I watched it, so it may have been the final stages of an action begun much earlier.

A raid of Sphinctomyrmex steinheili observed by me in Victoria, Australia (Wilson, 1958: 136) was waged against a small Stigmacros ( Formicinae ) species during the afternoon; the Sphinctomyrmex ran over bare soil, but took advantage of cracks in the earth where they could. The Stigmacros workers were seen' scattering and hiding on pieces of eucalypt bark and dead leaves lying on the ground, often carrying their own larvae. Their behavior recalled that of Formica fusca- or pallidefulva group species raided by F. sanguinea-group slavemakers in the Northern Hemisphere, and may imply the use of " propaganda " allomones such as those sprayed by the slavemakers to spread panic in the slave Formica colonies they are attacking (Wilson and Regnier, 1971). The behavior of army ant prey ant species is also often of the " panic flight " kind, so army ants should also be investigated for propaganda allomones. It seems that actual combat has rarely or never been observed in raids of Cerapachys on prey ants in nature, but the damaged and blackened antennae and legs of Sphinctomyrmex imbecilis series taken by Clark in southwestern Australia [71] testifies to possible struggles between the raiders and potential prey.

An S. caledonicus colony fragment observed over several weeks in an artificial (plaster-bottomed) nest was placed in contact with a colony fragment of Acanthomyops claviger , a North American formicine that Sphinctomyrmex would of course never normally meet in nature. (The Acanthomyops defends itself with formic acid spray and terpenoids such as citrinellal.) The Sphinctomyrmex raided the adjacent Acanthomyops colony, were duly sprayed and smeared with the noxious allomones, but nevertheless returned to their chambers with Acanthomyops adults and brood as prey. The Sphinctomyrmex workers spent an inordinate amount of time grooming themselves and their nestmates, especially just after raids — evidently a behavior pattern aimed at removing the toxic substances they receive from prey species.

The Sphinctomyrmex workers and ergatoids also spend a great deal of time in the nest with their long, flexible bodies wrapped around their own brood, thus forming round, dense clusters that may fill large chambers.

The prey of several species of Cerapachys living in forested areas of the Old World is usually recorded as small or medium-sized species of Pheidole , but Wilson has also recorded Strumigenys and Lordomyrma ( Myrmicinae ) species serving as prey in Melanesian forests, while species of Iridomyrmex ( Dolichoderinae ) and Melophorus ( Formicinae ) are raided in more arid, open parts of Australia.

Species of the longitarsus group (= Lioponera ) tend to nest in hollow twigs, beetle burrows, or other channels in wood, branches, or bark, and it is possible that they are all arboreal or subarboreal. C. singaporensis was taken originally nesting in a hollow mango twig. C. foreli has been taken by Raignier and van Boven in hollow twigs in Zaire (personal communication), and I have found this species climbing a tree trunk in sparse single file in a copse in the Dabou Savanna, Ivory Coast.

Simopone species also appear to be dwellers in hollow twigs. Raignier and van Boven wrote me that they found S. schoutedeni and S. conciliatrix in such microhabitats in Zaire; Bolton took S. conciliatrix in hollow cacao twigs in Ghana [77], and Arnold (1915) reported S. marleyi from hollow stalks of castor bean plant in Natal, South Africa. Bequaert (label data) took an S. grandis [76] from a palm trunk in a central African swamp forest. No one has reported the food habits of Simopone , but this genus is so closely related to Cerapachys that it too may be ant predatory.

Most species of Cerapachys and Sphinctomyrmex appear to nest in the ground or in rotten wood. Nests I have seen in moist forested areas usually have one or a few relatively capacious chambers connected to the surface by a short passage, but in arid areas, there may be several chambers 10 - 50 cm or more deep in the soil, with longer tunnels. Sometimes the nest is under a rock. The entrance is usually an inconspicuous hole without a crater, but in a few cases, a small crater or turret occurs at the surface.

Nest populations can contain as few as 20 adult workers to more than a thousand; the higher counts hold for some species of Sphinctomyrmex and the Cerapachys typhlus and antennatus groups, but most Cerapachys nests probably contain less than 200 workers. Pupae are normally enclosed in cocoons. The winged males are very active when mature and may attempt to take flight when the nest is opened.

The means of nest-founding of cerapachyines is unknown. In southwestern Australia, Clark (1923, 1924) found solitary individuals of C angustatus and C. constrictus under a stone and a log respectively, and thought that the latter, at least, was founding a nest. However, it is not even certain that these specimens were reproductives, let alone founding queens, since no brood was found with either one, and both are " ergatoid " (i. e., possibly worker) in form.

The more advanced and probably more army-ant-like species of Sphinctomyrmex ( imbecilis , froggatti , perstictus , trux , mjobergi , etc.) apparently have a single, more or less dichthadiiform queen, although the picture is confused by the appearance in some of these species of additional " ergatoid " forms that could just possibly have some reproductive function. It is such species that probably have the males with falcate mandibles (see p. 31).

In other species with dealate or less completely modified reproductives (e. g., steinheili , duchaussoyi , asper , occidentalis ) each nest may contain several, up to 20 or more, of these ergatoid queens. Of course, we do not know how many such individuals in a given nest may function as true reproductives, or to what extent, if they do so function.

Whether cerapachyines are nomadic is an open question. To me, the nests I have seen look impermanent, and the broods show a strong tendency to be synchronized, like those of army ants and nomadic Ponerinae ( Onychomyrmex , Simopelta ). Nomadism in an ant-preying ant could well be adaptive, to avoid depleting the food supply in a given neighborhood, just as it is in the army ants. In the cerapachyines we probably see the early stages of developing army ant lifeways, and it is probably no accident that cerapachyines are relatively so abundant in Australia, where army ants are rare and local, and apparently are late arrivals on the continent (2 species of Aenictus ).

The feeding habits of genus Leptanilloides are unknown. This genus is known from a single collection made by Mann at Tumupasa in northern Bolivia, " from beneath a deeply embedded stone near a stream. " The collection consisted of " a small series of workers, taken with callows, but without sexual or immature phases " — a circumstance that could apply to a raiding column.

distribution: The Cerapachyini are primarily Old World inhabitants and are widespread there in tropical and warm-temperate areas, in Africa, Madagascar, and the Oriental and Australian areas (including Melanesia as far as Fiji). The tribe is unknown in Europe and the U. S. S. R., but it would not be surprising to find that hypogaeic species exist in some of the warmer parts of central Asia. Two or more species reach central China and Japan in eastern Asia.

In the New World, cerapachyines are nowhere very common, but a few species are known from the warm deserts of southwestern United States and northern Mexico, from forest country in Mexico, and very sporadically in South America south to Bolivia and southeastern Brasil. The details of distribution are cited under each genus below.