Polyergus

Creighton (1950) - "The activities of Polyergus are so spectacular that the genus has long been a favorite for habit studies. All the species are obligatory slave-makers and they carry out their raids with great spirit and remarkable precision." Trager has recently (2013) published a complete, and much needed, revision of the genus. We are remarkably well informed about this genus, both in terms of its taxonomy and its biology.

Identification
Workers of Polyergus are similar in habitus to larger Formica species but are readily distinguished from them by their distinctive falcate (sickle shaped) mandibles.

Species Groups
Polyergus species groups

Distribution
Holartic.

Biology
Trager (2013) - Polyergus is a well-known ant genus, at least to Holarctic myrmecologists. These ants’ summer-afternoon raids against colonies of Formica species have caught the attention of naturalists and entomologists for over two centuries. The literature on Polyergus dates at least to Latreille’s (1798) description of “Formica rufescens,” followed a few years later (Latreille 1804) by his naming of the genus. The careful observations of Pierre Huber on this species in Switzerland (1810) constitute a classic ant natural history study. A century later, diversity and host specificity of North American populations of Polyergus were suggested by an array of forms described by Wheeler (summarized in Wheeler 1910, with additions in Wheeler 1915), but his nascent (sub)species concepts were lost to later synonymization. Even M. R. Smith’s (1947) careful review of the species was soon overridden by the synonymies of Creighton (1950), though Gregg (1963) made a flawed effort to resuscitate part of Smith’s system. Buschinger (2009) reviewed Polyergus biology in the context of a review of social parasitism in ants, noting (p. 4) that relatively few new parasitic ant species had been described since 1990, and that those described did not have accompanying biological information. This revision serves to break that pattern for Polyergus. Huber published his observations on the mixed nests of Polyergus rufescens and hosts Formica fusca and Formica cunicularia in 1810. He was the first to surmise correctly that mixed colony populations of Formica and Polyergus were in nests rightly considered those of the parasite, even though comprising relatively few Polyergus and far more numerous Formica workers. He demonstrated that the Formica workers in Polyergus nests were raised from brood robbed from nearby Formica colonies. A single Polyergus queen enters and usurps the queen of a host Formica, and the Polyergus-Formica colony that ensues goes on to parasitize multiple neighboring host Formica colonies through brood robbing. After successful usurpation of the Formica queen and the eclosing of the first Polyergus workers, it is no longer correct to refer to the nest as a parasitized Formica colony. Polyergus colonies always contain an abundance of host workers, usually reported to outnumber Polyergus workers by 5–10 times, and their colonies are consistently reported as more populous and inhabiting a larger nest than is typical for the host species. Their nests are otherwise typical in habitat, structure, and materials for the particular host (Talbot 1967, Marlin 1971, King and Trager 2007, Tsuneoka 2008).

All Polyergus species are obligate parasites of one or more Formica species. Many parasitize only a limited set of species within a single species-group, at least locally. The hosts are usually members of either the F. fusca group (sensu Francoeur 1973) or the F. pallidefulva group (sensu Trager, et al. 2007). Here, study of field and museum samples indicates that published reports of F. neogagates group species as hosts of a variety of Polyergus species are often, but not always, erroneous. No member of the lucidus-group utilizes F. neogagates-group members as host, and in the breviceps-complex only Polyergus mexicanus can be confirmed to do so. Talbot (1967) reported raids by Polyergus lucidus on Formica neogagates and Formica lasioides, but she clearly stated that brood from these raids never matured in the Polyergus lucidus nest. Polyergus mexicanus samples from California, Oregon, Nevada and one from Wisconsin were collected in association with various F. neogagates group hosts, but these populations also parasitize various members of the F. fusca group nearby.

Raiding for brood, mostly pupae of the host species, is essential for maintenance of the mixed species population of a Polyergus colony over the years of its life. The importance of scouts in initiation of raids has been confirmed for at least Polyergus rufescens (Le Moli et al. 2001) and Polyergus lucidus (Talbot 1967), and is probably a generic behavioral trait. Scouts head out late morning to early afternoon to find a target host nest, then return to their nest around the time of day when raids normally begin, typically mid- to late afternoon. Scouts actively recruit to the raid by running about jerkily, and probably secreting recruitment pheromone(s), among conspecific nestmates that often are already milling about the nest entrance. Talbot (1967) was able to induce short and somewhat atypical raiding behavior by painting a trail of a whole body extract of Polyergus workers from a nest entrance to a colony fragment of the host. In warmer climates, raids are typically late afternoon, after the daily maximum temperature, but they may coincide with the daily maximum in cooler climates of higher latitudes and altitudes. Wheeler (1910) reported raids of bicolor occurring in early afternoon, in a forest near Rockford, Illinois. Talbot (1967) saw one P. lucidus colony in Michigan raid during the morning four times, and I have seen this twice by P. lucidus in Missouri and once by Polyergus topoffi in Arizona (Trager, unpublished observations). Nonetheless, morning raiding still must be described as unusual.

Based on hundreds of raids of various Polyergus species seen over the years (Trager, unpublished observations), I will summarize the events of a typical raid. Instigated by a successful scout, a dense swarm of raiders heads off, slowly at first, then organizing into a column and hastening the pace. The columns are periodically interrupted as the raid progresses. When this occurs, there is the appearance of the group’s not knowing where to go next. Then, a worker that apparently had run ahead of the main group (a scout/leader?) may be seen to return to the group and recruit the column in the right direction once again. Upon arriving at the host nest, Polyergus workers hesitate and amass outside the entrance, often clearing pebbles, twigs, and other impediments from the entrance of the host nest before they enter. Shortly after the first Polyergus enter, they begin emerging and head home bearing pupae, prepupae, or less commonly, last-instar larvae. Sometimes, especially early in the raiding season, raiders return one or more times the same afternoon to pillage additional pupae from a particularly productive host colony. Less often, a Polyergus colony raids more than one host colony in an afternoon, either simultaneously or sequentially. A commonly reported impression is local depletion or diminution of host species colonies in the neighborhood of the Polyergus colony (e.g., Talbot 1967, Hasegawa and Yamaguchi 1997). Though the year’s raiding begins roughly contemporaneously with the pupation of sexual brood of the host Formica, there are no reports of Polyergus transporting host sexual brood. Further study is needed to appreciate the impact on fitness of raided colonies.

Mating behavior is variable. In some species or populations, mating takes place on the ground near the natal nests or in raiding columns, before or during raids, and then the newly mated gynes enter into raided host nests shortly after mating (Talbot 1968, Topoff and Zimmerli 1993). In other species or populations of Polyergus, mating often takes place in early afternoon, away from the nest in unknown locations, apart from raiding activity. Which of these mating patterns happens may be a species or population-specific behavior. For example, Polyergus mexicanus and Polyergus lucidus gynes in eastern Missouri always emerge and mate mid-day to early afternoon, 2–4 hours before raiding occurs, while females of P. mexicanus in California (form “umbratus”) at least sometimes accompany raids and mate along side them. Dealate females of P. mexicanus also may seek out host colonies to invade alone, as do at least some P. lucidus dealates, but a few P. lucidus dealates, after having mated earlier, accompany a raid, skirting along the sides as it progresses, then enter the just-raided host nest near the end of the raid.

What happens next has been elucidated in laboratory studies (Topoff and Zimmerli 1993). The recently mated Polyergus gyne creeps in or is dragged in by host workers. She seeks out the host queen, and if the colony has numerous workers (i.e., the invaded colony is well established), the Polyergus gyne immediately attacks and kills the Formica queen, clamping onto her oozing corpse for several minutes, thus acquiring the host queen’s scent (cuticular hydrocarbons), and probably gaining some nutrition by consuming her hemolymph. If the host queen is young and has few or no workers, Johnson et al. (2002) reported that the parasite gynes do not kill the Formica queens until the host colony reaches about 200 days of age. In Johnson’s experiment, the Polyergus gynes were withdrawn after each of the trials over the experimental period of 29 weeks. This laboratory study extended into the winter, a highly unnatural circumstance, but the study nonetheless has interesting implications for the field. The flight season of Polyergus follows the peak flights of its host species, extending up to two months beyond it, and encounters of a Polyergus foundress and a young host gyne, with or without her nanitic first workers, may be common. If the parasite can manage to cohabit with the gyne and her young offspring into the next growing season, then kill the host queen and inherit her workers, this could result in a successful colony founding. In this study, I examined several colony samples containing nanitic workers (typical of young colonies) of both host Formica and Polyergus together. These might well have been the result of such a colony foundation scenario (See also Topoff and Mendez 1990).

The life span of colonies is not well documented, but it is not unusual for a student of these ants to return to a particular nest for several years of observations. Talbot (1967) reports a colony of P. lucidus at least 10 years old. There are no reports of the presence of more than a single queen in a Polyergus colony, though there are sometimes multiple ergatoids. Many or perhaps all species sometimes have these large, worker-like forms with swollen gasters containing apparently functional ovaries and occasionally, some hints of alary articulation. It is unclear whether ergatoids ever coexist in a colony with the dealate foundress. No one has documented oviposition or even the certain presence of a full spermatheca in such individuals, but if ergatoids are able to produce female offspring, they may well succeed the alate gyne after her death, and the colony may become essentially immortal.

There is a long tradition of referring to Polyergus, and other ants that build up a worker population through brood theft from a related species, as “slave-maker” ants, and to the host Formica species (or other genera) as their “slaves”. The analogy of these ants’ behavioral ecology to human slavery is imperfect, and potentially, not a little offensive (Herbers 2006). One can legitimately state that words derived from “slave” will cause some readers or students entering biology to cringe, especially as the study of dulotic ants continues to evolve from a field dominated by men of European and Japanese origin to one studied by people of many origins, including descendants of humans subjected to slavery. So it seems well advised to avoid these terms. The preferred terms for these behaviors are dulosis and dulotic, derived from the Greek word for slave. Herbers (2008) proposed alternative terminology that is to my mind equally inaccurate and possibly off-putting: piracy, pirate ants, and Hellenic “leistic”. Google Scholar searches for these terms reveal that they have almost completely failed to catch on outside of Herbers’ original suggestion, except for a publication or two by students from her lab. Of course, the behavior of these ants is not exactly analogous to any human institution or behavior, but the theft of young from their parents, against their will, to create a working caste for colony maintenance, feeding and brood rearing is not far from “enslavement.” As has been shown by the fate of Herbers’ terms (and other recent examples in biology), new terminology is often not accepted, when well-established terminology is in place and continues in frequent use. I suggest that these behaviors continue be referred to by the terms dulosis and dulotic, while doing our best to avoid the use of “slave” and its derivatives. Dulosis arguably describes the behavior of all Polyergus, as well as that of the species of the Formica sanguinea group, Harpagoxenus, Protomognathus, Temnothorax and Strongylognathus (and perhaps others with this behavioral syndrome yet to be discovered). The species of Formica and respective other genera that they parasitize may simply be called the host, or as already published in some European literature, auxiliaries or helper ants.

Castes
Trager (2013) noted: Gynes and males of the species are not described in this revision. Males exhibit subtle morphological differences, if any, among the species, as well as apparently considerable variation within species (Wheeler 1968). Males of the breviceps and lucidus complexes do differ in density of gastral pubescence, like the females. Gynes, with further study, may prove to have useful morphometric characters, but available specimens were too few to investigate differences among species.

A small number of series in the breviceps complex of western North America could not be placed with confidence in any of the species discriminated here. These may be the result of hybridization among breviceps and mexicanus, or perhaps these represent rare or cryptic species or unaccounted-for variation of the known ones. A thorough sampling of fresh material and a careful analysis of additional data not available for this study should help resolve these samples.

Nomenclature

 *  POLYERGUS [Formicinae: Formicini]
 * Polyergus Latreille, 1804: 179. Type-species: Formica rufescens, by monotypy.

Trager (2013):

Description
Holarctic distribution. Outstretched specimens, clypeal margin to gastral tip length: alate gynes 8–10 mm, ergatoids 7–8.5 mm, monomorphic workers 4.5–7.5 mm, males 4–7 mm. Workers with the characters of the tribe Formicini: a double row of stout setae (bristles) on the flexor (ventral) surface of the hind tibiae; anterior articulation of the petiole obscured by the metacoxae; upper surfaces of wings of both sexes with numerous suberect small setae, but lacking stout pilosity (Agosti 1994). Head shape subquadrangular or subhexagonal or with angles rounded to yield subovate or suborbicular head capsule in full face view, anteriorly truncate due to straight or weakly concave clypeal margin, and with straight to either weakly convex or weakly concave vertex margin, and with weakly to notably rounded sides, at least behind the eyes, sometimes with contrastingly straight or even feebly concave genae; vertex “corners” always rounded, HW across the outer margins of the compound eyes roughly equaling (usually slightly less, far less often a bit greater than) the head length; in broad-headed species, eye margins often not protruding beyond head margins in full face view, slightly protruding in narrower-headed species. Mandibles falcate, the inner border denticulate, and with two rows of long, straight setae arising submarginally near inner margin, dorsal line of hairs sparse, longer and oblique anteriad, ventral hairs arising about 1 per every 3 denticles and perpendicular to inner margin. Palps short, segmentation reduced, 4,2 (reportedly, rarely 4,3). Clypeus foreshortened (compare to Formica), in full-face view it is a shallow trapezoidal, triangular, or lens-shaped sclerite positioned mostly between the mandibular bases, the posterior median portion weakly umbonate, with narrow, tapering, lateral lobes behind the mandibular bases, and a nearly straight or weakly concave inter-mandibular, apical margin (but somewhat convex in samurai). Frontal carinae reduced, extending about or slightly less than half their full length beyond the antennal fossae. Compound eyes ovoid, EL/HL 0.23–0.28, in full-face view about 0.3–0.35 of EL lies (on prognathous head) anterior to the midpoint of HL, and 0.65–0.70 of EL lies posterior to the midpoint of HL. Metapleural gland opening indistinct on the weakly bullose lower rear portion of the metapleuron, a transverse narrowly ovoid slit, rimmed by short, straight setae. Propodeal spiracles are narrow crescentic slits, placed on the propodeal lateral face just below the rounded transition of the dorsal to the posterior declivitous propodeal surfaces. Both pubescence and pilosity setae longitudinally costulate (Fig. 1, but some intersegmental proprioreceptors apparently are smooth), tapering; pubescence typically curved or hooked apically, and sometimes globular at the tip; base of pilosity in a shallow pit about 3–4X the width of the seta (and the pits reliably can be used in pilosity counts when the setae themselves have been lost), pilosity tapering to a rounded, truncate, or sharp tip.

Gynes similar to conspecific workers but thorax bearing full sclerite representation, wings (before dealation) and alary articulations and musculature; larger than workers, with stouter heads, scapes and legs; petioles relatively and absolutely stouter on all axes; and the body, especially the head and legs, generally shinier than workers of the same species. Ergatoid females resemble large conspecific workers, with enlarged petioles and gasters, sometimes with weakly developed vestiges of alary articulation.

Males currently recognizable only to species group within the genus, usually a bit smaller than workers (but the few known specimens significantly larger in vinosus), superficially similar to large Lasius males or those of Myrmecocystus, but with palp formula 4,2; with longitudinally costulate dorsal macrosetae (bristles), paired rows of suberect macrosetae on ventral surface of hind tibiae (though few and fine relative to those of females and other Formicini); and as in females of the tribe, with the anterior articulation of the petiole obscured by the metacoxae (visible between metacoxae in Lasius and Myrmecocystus; Agosti and Bolton 1990). Male petiolar profile tapering dorsad, cuneate or acutely rounded; petiolar dorsal margin concave or broadly and angularly notched, thus biumbonate.

Karyotype much like that of closely related Formica, N=27 (Imai 1966). Polyergus workers have a pygidial gland, unique among formicines, consisting of a group of subcutaneous exocrine cells, the product of which exits through the porous anterior edge of the seventh tergite (Hölldobler 1984).

Etymology
The generic name Polyergus is derived from Greek poly- plus ergos, meaning much work or hard-working. This would seem an ironic name in regard to the amount of nest work performed by these obligate social-parasitic ants. However, the extent, vigor and organization of their work during brood raids is quite unmatched among other ants outside the doryloid section.