Polyergus mexicanus

Identification
Trager (2013): This species, described from an undetermined mountain locality in Mexico, now turns out to be the most widely distributed member of the species group, all called breviceps by Creighton (1950). During my early sorting of specimens (and for a long time thereafter), I tried to associate samples with the types of fusciventris, laeviceps, and umbratus, as well as a fluctuating number of “new species,” but as the study progressed, there was an accumulating residue of samples that seemed transitional, or in which some members of a colony sample appeared to be one “species” and others a different “species,” rendering them non-differentiable. The morphology of this widely distributed species does have some notable geographic trends:


 * Samples from Mexico, Arizona (typical mexicanus) and the US and Canada west of the Rocky Mountains (mexicanus sensu stricto, “laeviceps” and “umbratus”) most often have at least moderately shiny heads. Samples with subpolita and neogagates group hosts from southern and central coastal California, described as “laeviceps” by Wheeler, average the smallest and have on the average, proportionally the smallest, roundest, shiniest heads. It may be noted here that these and all other samples associated with F. neogagates-group hosts average smaller than the rest of the species’ populations.


 * Samples from the Santa Cruz Mts. and Sierra Nevada of Central California, to British Columbia, Idaho and western Alberta (“umbratus“) often have convex (“bulging”) mesonotal profiles. These samples match Wheeler’s (1915) “umbratus”, but curiously, just a year after describing this taxon, even Wheeler (1916) failed to recognize it when studying rather typical “umbratus” near Lake Tahoe (vouchers examined). An unpublished study (U. C. Berkeley Ph.D. dissertation, Candice W. Torres, personal communication) shows the California populations are distinct genetically from those east of the Sierra Nevada, but I have been unable to find consistent biological or morphological characters to distinguish them.


 * Samples from north of southern Arizona and from eastern California and the Rocky Mountains eastward most often have less shiny heads. This is especially true of the populations of higher elevations in the Rockies, and those of the northern Great Plains and the Canadian prairie provinces.


 * Samples from the Dakotas and Canadian prairie provinces are often a bit smaller than average, and bicolored, with orangey foreparts and partially dingy brown gasters, superficially resembling P. bicolor, but always more pilose. Even in this region, many specimens (often nest mates of bicolored individuals) are robust with heads a little narrower than long and nearly all red, like those from farther south and southwest. The bicolored samples from North Dakota led the Wheelers (1963), without the benefit of measurements and pilosity patterns observed in this study, to conclude that bicolor was yet another synonym of the catch-all taxon breviceps, sensu lato, but bicolor from the Dakotas can easily be recognized by pilosity and habitat characteristics (no to sparse pilosity, moist and forested habitat).


 * Finally, samples from the New Mexico and Colorado Front Range and foothills (“silvestrii” Santschi[?], also called “umbratus” by Gregg 1963, but differing in their less convex mesonota from Wheeler’s West Coast form) trend a bit larger than samples from elsewhere in the west, and this trend continues eastward, such that those from KS, IA, SD, IL, MO, AR, and western IN are conspicuously larger than western mexicanus or any other breviceps-complex species. These latter are associated with the large host species, F. subsericea.

In the upper Mississippi drainage south of MN, P. mexicanus is easily distinguished from P. breviceps by its larger size, silvery gray pubescence, and association with F. subsericea, contrasting with P. breviceps’ smaller size, greater hairiness, yellowish pubescence and its association with F. montana. The greatest difficulties in recognizing mexicanus arise in the Dakotas, where it is sympatric with both P. bicolor and P. breviceps, and in southern Arizona and Mexico, where it is parapatric (possibly very narrowly sympatric) with P. topoffi. Where P. mexicanus occurs in parapatry with P. topoffi, the two are separated by elevation and host species. Polyergus topoffi has longer appendages and parasitizes F. gnava and other warm-climate Formica species, while P. mexicanus occurs at higher elevation, using montane conifer forest Formica hosts. Though distinct in Arizona and Mexico (character divergence), in the Midwest the proportions of P. mexicanus come very close to those of P. topoffi (character convergence), but P. mexicanus has on average shorter appendages throughout its range, pointing to the value of measuring several specimens from good samples. Polyergus breviceps averages smaller and consistently more pilose, usually most easily seen by examining the vertex pilosity. Where there is broad distributional overlap of the latter two species in the Rockies and the upper Mississippi Valley, careful study of morphology and metrics of 5 or more individuals per colony will give greater assurance of proper identification.

Distribution based on Regional Taxon Lists
Neotropical Region: Mexico.

Biology
Trager (2013): In more open habitats (fields, prairie reconstructions) in eastern Missouri, I have often found colonies of ''P. mexicanus-F. subsericea interspersed with, and within raiding distance of colonies of P. lucidus-F. incerta, and it is perhaps more often found among F. rubicunda-F. subsericea and/or F. subintegra-subsericea'' colonies in open woodland habitats. I have directly observed and seen indirect evidence of F. subintegra attacking and even bringing home pupae of P. mexicanus, as follows: One ''F. subintegra-F. subsericea colony observed in Missouri contained a contingent of a few dozen P. mexicanus workers that raided separately and later in the afternoon than the dulotic Formica'' whose nest they inhabited.

On two occasions in Missouri, I have observed mexicanus raid colonies of F. pallidefulva-group species. In contrast to their typically non-lethal interactions with their normal hosts, many workers and the queen of these nonhost species were killed, and both killed adults and some live brood were carried home. The brood of these non-host Formica never developed to adulthood in the mexicanus nests. In effect, the usually strictly dulotic Polyergus became predators.

Forel (1899) described this species from specimens collected in “Mexico”. The specimens were collected by Brinkmann, who made other collections in the mountains of Durango that were passed on to Forel (P. S. Ward, pers. comm.), so it seems very likely the series was collected from a montane conifer forest there, similar to the podzolic soil conifer forests it inhabits in western USA. A sample collected by Creighton (at LACM) near el Salto, Durango, Mexico, has workers very similar to the MHNG types, and with Formica cf. occulta host workers, as with southern Arizona samples. Southern samples are from relatively high altitude, with Chiricahua Mts. samples all from above 2200 m, a Vergel, Chihuahua, MEX sample came from about 2800 m (in an open conifer forest), and one from Nevada at a surprising 3200 m. To the north, this protean species is found at lower elevations and in various, mesic, open woodland or grassland-woodland mosaic habitats, but not in moist, closed-canopy forest (though it may occur in natural gaps and in clear-cuts within these), and only occasionally in open prairie. This is characteristically a species of woodlands with little or no shrub layer, usually of oak, oak-pine, or pine in the US Midwest, Ozark Hills and West Coast, and of airy conifer woodlands in the western mountains. It is also found in grassland-woodland ecotones, and prairie groves. Some Illinois and Wisconsin samples came from sandy prairie openings among sandy-soil black oak savannas. It occasionally nests in windbreak plantings of trees, and in less tended areas of parks, cemeteries, gardens and tree-studded lawns, where these are not heavily treated with pesticides. Forel reported that the host was unknown, but the samples I have seen reveal that this species has a wide variety of hosts in the Formica fusca and neogagates groups. In the southwestern USA and northern Mexico, the host is usually Formica cf. occulta, a conifer woodland inhabitant that is larger and darker-colored than typical Formica occulta. From elsewhere, I have studied samples with Formica argentea, Formica podzolica, Formica subsericea, Formica fusca (marcida & subaenescens), Formica accreta, Formica microphthalma, true Formica occulta, Formica neoclara, Formica pacifica, Formica neorufibarbis, Formica hewitti, Formica neogagates, Formica manni, and Formica vinculans. Moffett (2010) vividly describes and illustrates raids of this species (as P. breviceps) on ''[[Formica argentea in eastern California.

Nomenclature

 *  mexicanus. Polyergus rufescens r. mexicanus Forel, 1899c: 129 (w.) MEXICO. Raised to species: Trager, 2013: 515.
 * laeviceps. Polyergus rufescens subsp. laeviceps Wheeler, W.M. 1915b: 420 (w.) U.S.A. Junior synonym of mexicanus: Trager, 2013: 515. See also: Creighton, 1950a: 559.
 * umbratus. Polyergus rufescens subsp. umbratus Creighton, 1950a: 560 (w.) U.S.A. [First available use of Polyergus rufescens subsp. breviceps var. umbratus Wheeler, W.M. 1915b: 419; unavailable name.] Junior synonym of breviceps: Wheeler, J. 1968: 163; of mexicanus: Trager, 2013: 515.

Description
Syntypes (N=6 on 3 pins) [MHNG] HL 1.52–1.62 (1.57), HW 1.45–1.60 (1.54), SL 1.08–1.13 (1.09), ½ VeM 0–1 (0.33), ½ PnM 5–6 (5.60), WL 2.20–2.40 (2.29), GL 2.04–2.40 (2.26), HFL 1.58–1.66 (1.62), CI 95–99 (98), SI 68–74 (71), HFI 104–112 (106), FSI 143–154 (148), LI 3.70–4.00 (3.86), TL 5.76–6.34 (6.12).

Measurements (N=122) HL 1.28–1.96 (1.57), HW 1.24–1.92 (1.53), SL 0.94–1.36 (1.11), ½ VeM 0–3 (rarely, 5+) (1.20), ½ PnM 3–9 (6.77), WL 1.92–2.80 (2.31), GL 1.34–2.68 (2.17), HFL 1.32–1.96 (1.64), CI 91–103 (99), SI 65–81 (73), HFI 95–121 (107), FSI 134–161 (146), LI 3.24–4.76 (3.85), TL 5.01–7.36 (6.02).

Worker description. This is the most widely distributed and most variable North American Polyergus, and accounts for most literature records of “breviceps,” other than the cited works of Howard Topoff and his students (regarding Polyergus topoffi). Head variable in shape by region, but locally less so, subquadrate with nearly straight sides curving-convergent toward the vertex, to round-sided and convergent toward the mandibles (pomoid) or occasionally nearly suborbicular in outline, HL usually slightly greater than HW, to HW very slightly broader than HL, the latter corresponding with more rounded sides; vertex pilosity of 0–10 macrosetae (> 5 is uncommon); scapes not reaching vertex corners by about 2X their maximum width, curved, clavate in the apical third; HFL roughly equal to HL to slightly longer (rarely up to 1.2X, especially on West Coast); vertex weakly concave in full face view, corners often without erect setae, or each vertex corner may have 1–3 (up to 5) erect setae; pronotum with 4–10 (rarely up to 18) dorsal erect setae; mesonotum with profile flat or at most weakly convex for most of its length, but often convex and bulging in samples from along the West Coast and southwestern Canada (“umbratus” form), and occasionally inland samples; propodeum evenly rounded; petiole with sub-parallel, straight to slightly rounded sides; petiole about as broad as propodeum (above metapleura) in postero-dorsal view; petiolar dorsal margin nearly flat, or faintly convex and medially flattened, occasionally shallowly emarginate; first tergite densely pubescent; first tergite pilosity in 3 or 4 transverse arrays but concentrated in the anterior third; first tergite pilosity flexuous, suberect to subdecumbent.

Head glossy in many specimens from California, Arizona, and Mexico, decreasingly so eastward and northward, thus over most of the population weakly shining, even becoming matte in Canada and the Dakotas; mesonotum matte dorsally and somewhat to notably shining laterally, rarely entirely matte; gaster somewhat shining beneath pubescence and shinier laterally, where pubescence is dilute.

Color mostly red with infuscation of posterior portions of tergites (sometimes entire tergites, especially Canadian provinces and the Dakotas), and with slightly darker legs; pilosity browner than prevailing body color; pubescence gray (never yellowish as in P. breviceps).

Forel erred when he characterized the types of this species as lacking pilosity. The five specimens of the type series, though their pronota and gastral tergites appear hairless, bear the darkly pigmented impressions of macrosetal bases typical in this genus. These impressions come in an array and in numbers within the range of macrosetal counts of other specimens belonging to this species that are in full possession of their macrosetae.