Aphaenogaster rudis

AntWiki - Where Ant Biologists Share Their Knowledge
Jump to: navigation, search
Aphaenogaster rudis
Scientific classification
Kingdom: Animalia
Phylum: Arthropoda
Class: Insecta
Order: Hymenoptera
Family: Formicidae
Subfamily: Myrmicinae
Tribe: Stenammini
Genus: Aphaenogaster
Species: A. rudis
Binomial name
Aphaenogaster rudis
Enzmann, J., 1947

Aphaenogaster rudis casent0104843 profile 1.jpg

Aphaenogaster rudis casent0104843 dorsal 1.jpg

Specimen Label

Aphaenogaster rudis is a common, widespread species found from Massachusetts south to Alabama and west to Ohio, Indiana, and Missouri (Gregg 1963, Smith 1979, Wheeler and Wheeler 1988b, Umphrey 1996, Nemec et al. 2016). It occurs in deciduous woodlands and nests in soil, under stones or logs, in decaying wood, leaf litter, hollow stems of plants, or under bark at bases of trees (Smith 1979, Nemec et al. 2016). It should be noted that as currently understood, A. rudis is thought to be a complex of species (Umphrey 1996) and further studies on its taxonomy are needed (see fulva-rudis-texana complex).

Photo Gallery

  • Worker. Photo by Tom Murray.
  • Queen. Photo by Tom Murray.
  • Male. Photo by Tom Murray.


This is the most commonly found Aphaenogaster species on the east coast of the US. It is polyphyletic and ranges in color from light to dark brown. The last four antennal segments are not lighter in color. It cannot be distinguished from Aphaenogaster carolinensis without the gene CAD, which has an intron that A. carolinensis is missing. (DeMarco and Cognato, in prep.) (DeMarco, 2015)

Keys including this Species


Distribution based on Regional Taxon Lists

Nearctic Region: United States (type locality).
Aphaenogaster rudis Distribution.png

Distribution based on specimens

Loading map...

The above specimen data are provided by AntWeb. Please see Aphaenogaster rudis for further details


Lubertazzi (2014) summarized the biology of the Aphaenogaster rudis species group. Ants from the genus Aphaenogaster are abundant in the wide ranging mesic hardwood forests of eastern North American. These ants have been the subject of dozens of published studies and are often identified and reported as A. rudis. Using range limits and habitat infinities provided by Umphrey we can infer that these studies are likely to involve three species: A. rudis or N22a in Umphrey’s coding system, the undescribed form N17 and Aphaenogaster picea or N18. From this, it was inferred that these studies could be used to describe the general biology of what is referred to as a rudis-group species.

Annual Colony Cycle

Mature rudis-group colonies exhibit an annual cycle of activity and productivity that tracks seasonal changes. The description outlined here arbitrarily starts with a renewal of colony activity in the spring and ends with winter diapause. Specific seasonal activities and responses are detailed, along with supporting data and information, in an idealized version of what occurs within the colony over the course of a full year. The seasons are used in a relative sense to indicate times of the year when temperatures are warming (spring), consistently supporting colony growth (summer), cooling (autumn), and cold (winter). The dates associated with these times of the year will vary according to the geographical location of a given population.

Spring. In Connecticut rudis-group workers are among the first ants to become active on the forest floor during the spring. Once the snow cover has gone and the ground is exposed, foragers can be found above ground on warm and sunny days of early spring (∼15◦C ambient temperature). Early spring activity explains, in part, Lynch et al.’s [22] finding in a Maryland ant community that a rudis-group ant was active for a greater portion of the year than other co-occurring ant species. Umphrey [18] suggested N18/A. picea was among the most cold tolerant North American Aphaenogaster species. This tolerance helps these ants be poised to begin their annual above-ground activities as early in the spring as possible. As spring progresses entire colonies diurnally move back and forth from underground chambers to protected cavities near the surface, especially favoring locations warmed by sunlight. Daytime spring temperatures at the top of the leaf litter can reach 40◦C and may be 20◦C warmer than 15 cm below ground [23]. Exposure to these higher temperatures raises metabolic rates, restarting brood development and queen oogenesis.

Later in the spring the winter nest is abandoned for a new site on or near the ground surface. This switch in nesting location coincides with the ground surface and subsurface becoming consistently warmer than the more thermally stable, but now regularly cooler, underground nesting chambers. The spring-season nest migration of rudis-group colonies has been observed in Missouri [24] and Connecticut. Maintaining a high degree of flexibility in where and when they move their nests help rudis-group ants hasten their transition from winter diapause to summer productivity.

Summer. The abandonment of the winter nests putatively delimits a time when conditions become consistently conducive to colony growth. In Connecticut this point is reached in late May or early June. Some larval growth and egg production take place in the spring and fall but the bulk of egg laying, larval growth, pupation, and eclosion occurs during the summer. Productivity peaks in late summer and declines sharply in the fall [25, 26]. Maintaining the optimal temperature/humidity micro environment for the brood and queen is an important focus of the workers. This may involve daily movements within the nest or, if a better nesting location is found, can include a summer-season nest migration [12, 27–29]. Structures such as downed wood that contain hollow chambers or loosely attached bark, areas between exposed rocks and the soil, and other similarly sheltered warm locations can all provide suitable nesting sites during the summer.

Food demands are highest during this season. Prior to the summer individuals had to rely on internal reserves, built up during the autumn, to sustain their low metabolic and developmental needs. During the warmer months food is now needed to supply nutrients to the brood, the queen, and the workers. Foraging, worker development, and the production of new sexuals all assume greater importance during the summer.

Autumn. The onset of shorter days and cooler temperatures leads to changes in nesting location and colony productivity. There is also a general slowing of overall colony activity. Above-ground or near-ground level nests are eventually abandoned for deeper below-ground nesting sites. This autumn nesting site is likely different from the nest used during the summer even if the summer nest contained soil chambers [24]. By October in Missouri [24] and Connecticut, rudis-group ants have all moved to what will be their underground winter nests.

Egg production decreases as temperatures decline, with oogenesis eventually stopping altogether. Colony activity slows [30], metabolic and development rates fall, and individuals increasingly rely on internally stored nutrients for their decreasing metabolic needs. Foraging slows and eventually ceases as the temperatures cool.

Winter. In winter months colonies avoid freezing temperatures by maintaining their nests below ground. Talbot [24] found the average depth of 5 winter colonies in Missouri to be 25 cm. Colonies in Connecticut appear to prefer deeper nests, to a depth of at least 50 cm. Developmental processes enter a diapause and worker activity within the nest is minimal.

Nesting Biology

Downed wood provides a good nesting resource for rudis-group ants in Connecticut and in other locations [12, 31]. Limbs and boles greater than 10 cm in diameter and slightly decaying appear to be particularly favorable. In the Connecticut study sites almost all of these suitable nesting sites were occupied by a thriving rudis-group colony. Wooden nesting structures are often in short supply relative to the high density of nests. The combination of the relative scarcity and suitability of wooden nest sites [31] likely contributes to the attractiveness of artificial wooden nests.

In mature hardwood forests in eastern Connecticut soil nests were common in the summer. Below-ground nest chambers are not likely to reach temperatures as conducive to larval developmental as can be found in ground surface structures or among the leaf litter [23]. Regardless of their limitations soil nests may be the only option available for many colonies.

Headley [25] described the structure of soil nests surveyed from central Ohio. A typical ground nest had a single entrance, which was an inconspicuous circular hole ∼6mm in diameter. A few nests had multiple entrances but in every case all the entrances for a single colony were located within 10 cm of one another. A central shaft, or a few bifurcating shafts, lead down from the entrance and connected the underground nest chambers. Chamber number averaged 6.5 and ranged from 2 to 17. The chambers’ dimensions averaged 12mm high, 12 cm wide, varied in length from 18 to 50mm and were found from just below the surface to a depth of 84 cm. Similar ground nests were reported from Missouri [24] with notable differences being slightly shallower chamber depths and, on average, fewer nest chambers. In both studies some colonies were found inhabiting various co-opted cavities such as areas between rocks and in downed wood.

Ants of the rudis-group maintain a concentrated central nest chamber regardless of where their nest is situated. In Connecticut the majority of the colony’s biomass (brood, nurse workers, and idle foragers) was found within 20 cm of the queen.

The average density of nests was 0.5 nests/m2 across three Connecticut populations, 1.3 nests/m2 in Missouri [32], and 0.5 nests/m2 in Ohio [12]. Aphaenogaster species are also known to be common in other forests [33, 34].

Colony Demographics

Two lines of evidence, genetic and observational, suggest that new colonies are begun claustrally by single queens. Genetic studies show that rudis-group nests contain workers produced from a single queen that has mated with one male. A few incipient colonies found in Connecticut contained a single queen and 25–35 minim workers. These colonies were discovered during mid-summer and were presumably founded the previous fall.

Productivity can be highly variable among nests within and among populations. Worker number in mature colonies has been surveyed in a number of populations and ranges from a mean of 266 to 613 workers per nest. Colonies with less than a hundred to more than a thousand workers can produce new sexuals but large colonies are more likely than small colonies to allocate energy towards reproduction (Mann-Whitney U-test = 460, P < 0.04). The number of reproductives produced is also highly variable among nests ([14, 24, 25]). Some large colonies produce no sexuals, suggesting that mature colonies do not produce new sexuals every year.


Foraging distances for three eastern Connecticut populations were similar among populations (F2,61 = 0.04, P = 0.9) and collectively averaged 57 cm (SD = 31). Despite this short foraging range the high density of nests provides for the abundant presence of rudis-group foragers across the forest floor. The running speed of individual workers returning to the nest with food has been found to vary with the number of workers in a colony [35]. Laden foragers returned to their nest faster in colonies with 140–150 workers than in colonies with 30–40 workers.

Talbot [32] estimated that total worker density of a rudis-group ant, above and below ground, averaged ∼425 workers per square metere in a Missouri woodland. In a Maryland forest rudis-group workers occupied 27% of ground baits [22]. In Connecticut more than half of the ground baits placed on the forest floor were typically found by these ants within 30 minutes.

Solid food is primarily brought back to the nest by individual foragers. Some recruitment of nestmates does occur at concentrated food finds (Lubertazzi, personal observation). A trail pheromone used for recruitment to food has been isolated fromthe poison gland of a rudis-group ant species [36]. The pheromone is amixture of N isopentyl-2-phenylethylamine, anabasine, and anabasiene, and 2,3' bipyridyl. Even with recruitment rudis-group ants do not maintain more than eight nestmates at a food item at any given time [22, 36].

Ants of the rudis-group are timid when encountering workers of other species and do not defend foraging territories. These ants are readily displaced at large food items by a number of co-occurring ant species [22, 37]. Their propensity to avoid confrontations is also evident in their intraspecific interactions. It was not unusual to find individuals from two or three colonies of a rudis-group ant foraging on the same bait in Connecticut forests. Here and in Maryland paired individuals could occasionally be found engaging in paired battles that “involve long, seemingly inconclusive “wrestling” bouts that result in few if any casualties” [22]. In Connecticut, workers were observed indifferently walking around any intertwined pair of fighting ants that they encountered.


Ants of the rudis-group are general scavengers. In Connecticut the majority of food items observed being carried by foragers were small invertebrates or parts of insects. Workers have been observed preying on termites (Reticulitermes flavipes) in the field in Connecticut and Indiana [38] and in a laboratory study [39]. Small invertebrates are likely the staple of their diet. Other food resources are also exploited and are clearly important, but are either temporally limited or spatially uneven in their availability. For example a mushroom species is known to be foraged upon by a rudis-group ant [40]. This opportunistically encountered resource may provide nutrients that are not readily found in their typical diet but is unlikely to even be encountered within the foraging range of many colonies.

These ants are keystone seed dispersers in the mesic forests of eastern North America [41, 42]. Ants of the rudis-group move a majority of the diverse myrmecochorous seeds that are produced in these habitats. The collection of eliasome-bearing seeds by rudis-group ants is well documented, as is the floral diversity of myrmecochorous plants throughout these ants’ ranges (e.g., [13, 43, 44]). Sexual production within colonies can be altered by elaiosome consumption [12, 45] despite the fact that colonies become satiated quickly when provided with myrmecochorous seeds. Seed foraging ceases within hours of a colony being presented with elaiosome-bearing seeds and this response can persist for many days [13].

Foragers opportunistically imbibe liquid food resources and behaviorally overcome morphological limitations in how much liquid can be held in their crops [46]. Foragers recruit nestmates to particularly rich finds and can also use absorbent objects to collect liquids [37]. Saturated materials are brought back to the colony and the liquids they hold are consumed within the nest. Workers can store an average of 0.13 mg of liquid in their crop but can transport up to 10 times this amount of liquid using an absorbent tool [37].

Reproductive Biology

Intra Colonial Social Structure Colonies of rudis-group ants have a simple reproductive and social structure [47, 48]. There is one singly mated queen that is the sole reproductive in her nest. Young workers may have functional ovaries but worker eggs are either not produced or eliminated in queenright colonies [49].

Although Talbot [24] and Headley [25] both found some rudis-group ant colonies with more than one dealate queen, Crozier [48] suggested these did not represent polygynous colonies. This was inferred from his genetic findings and observations that unmated rudis-group queens may spontaneously remove their wings in their natal nest. This dealation behavior was also noted by Haskins and Enzmann [50]. A few mature field colonies collected in Connecticut were found to contain numerous dealate queens after they had settled into artificial nests in the laboratory. In every case there were other winged queens present and unattached wings of queens were found in the sorting bin or in the artificial nest. Dealation was clearly occurring after the nests were collected from the field. This same behavior was observed in a few laboratory nests, originally collected in Connecticut, that had produced female reproductive. In a few colonies the unmated dealates were left in the laboratory maintained nest. Behavioral differences between a colony’s reproductive queen and her unmated dealate daughters were evident. When a reproductive queen moved within the nest and antennated a worker, the worker would typically lower her head and/or flee. Workers that initiated antennating a reproductive queen’s body typically continued to investigate the queen with their antenna. Such attention lead to the queen attracting a retinue of workers when she remained in one part of the nest. Worker-to-worker interactions within the nest produced no alarm and always quickly lead to disinterest. By contrast, dealates spend considerable time outside of the nest and appeared to be subordinate to the queen in their interactions. Antenna-to-antenna contact between workers and dealates always led to the dealates fleeing. Dealates overall were more worker-like in their actions. They were seen outside the nest, moved brood and never attracted a retinue. The one dealate behavior that was queen-like was the propensity to seek cover and remain motionless under an object when the nest was disturbed. Workers varied in their response to disturbance but were as likely to run around excitedly, pick up brood, or move outside of the nest into their foraging arena.

Worker-like behaviors in the dealates become more pronounced weeks after dealation, and, when left in the nest until the following spring, dealates no longer ran as quickly or as persistently from disturbances. In worker-dealate interactions the dealate still tended to move away from worker contacts. Workers reacted with greater alarm when they come in contact with a dealate than when contacting another worker but these differences were subdued when compared to similar interactions that occurred in the fall. The reason(s) for the spontaneous dealation of gynes and any potential adaptive value of this behavior are unknown. It has been observed in species from other genera, for example, Pogonomyrmex (Bob Johnson, personal communication) and Myrmica, [51]. The aggressive reactions of rudis-group workers to dealates, the propensity of dealates to be found outside the nest in laboratory colonies, the usual absence of supernumerary queens in field colonies in Connecticut and Crozier’s genetic data suggests that if dealation occurs in a naturally occurring colony these unmated gynes are likely driven out of their natal nest. Regardless, it is not possible to rule out that spontaneous dealation is the first part of a secondary reproductive strategy. Inbreeding within the nest, dealates walking out of their natal nest and mating on the forest floor or colonies allowing unrelatedmales into the nest to mate with a newly dealate queen are all possibilities.

Sex Ratios Sex ratios have been calculated for a number of rudis-group populations. The first such estimate, calculated to test hymenopteran sex ratio theory [5], was improperly derived. The 14-colony data set combined colony data from two states collected in two different years hence does not represent a population sex ratio. The first rigorous study of rudis-group sex ratios was a study examining the potential benefits of elaiosome food resources in an Ohio population [12]. Naturally occurring colonies presented with elaiosome-bearing seeds increased their reproductive allocation and had a more female biased sex ratio than control colonies. A follow-up study suggested the treatment response was a quantitative effect of adding more food rather than a result of specific nutrients contained in the elaiosomes [45].

A field study of a Connecticut Mohegan population also involved food supplementation [14]. No differences were found between control colonies and colonies provided with extra protein. Sex ratios were split (most colonies specialize in mostly male or mostly female investment) and the population sex ratio was estimated to be 0.86 (95% CI: 0.81 to 0.91). Colonies that produced larger broods invested slightly more in males. The sex ratio of queenless field colonies was similar to those of queenright colonies (Figure 4, Table 3). Naturally occurring queenless colonies overall invested much more in female than male production. Workers in these colonies appear to be raising their recently lost mother’s brood rather than raising their sons and nephews.

Mating Mating flights have been described for a congener [52] but have never been described for any rudis-group ant. It is not known how far queens fly from their natal nests or if mating occurs on the ground, or elsewhere. In Connecticut it appears that populations have synchronous mating. Collecting whole nests from 3 different forests over numerous years revealed that winged reproductives disappeared from all the nests within a given population at the same time. New sexuals were typically gone from the Connecticut colonies, depending on the population and the year, sometime between late July and mid-August.

Caste Development

Larvae G. C. Wheeler and J. Wheeler [53] studied and described the morphological and developmental characteristics of the egg and larval stages of rudis-group ants. There are four larval instars. First instar larvae were found to subsist on worker-provided liquid foods. Subsequent instars were also able to ingest solid foods [53, 54].

Workers Fielde [54] found the developmental period for workers, from egg to eclosion, averaged 64 days (time for eggs to hatch: median = 19.5 days, N = 22; larval stage: median = 28.5 days, N = 26; pupal stage: median = 16 days, N = 68). Southerland [55] examined the influence of temperature on development time. She compared the productivity of artificially created rudis-group nest fragments (a queen and 50 workers) maintained at 15◦C and 25◦C. During 150 days at the cooler temperature no workers were produced. Brood were present and survived but no pupation occurred. Nests maintained at the warmer temperature were able to produce new workers.

The dry weight of Connecticut workers averaged 0.8 mg (SD = 0.16, N = 200; Figure 5(a)) and head width averaged slightly less than 1mm (mean = 0.91, SD = 0.06, N = 975; Figure 6). Connecticut and Vermont [30] head width data showed workers form a single monomorphic caste. Worker head width in Connecticut did vary significantly among colonies within populations (Nipmuck population F18,456 = 13.4, P < 0.01; Pachaug population F19,480 = 13.1, P < 0.01) and between populations (t973 = 5.96, P < 0.01). These differences are presumed to be caused by environmental variation in food availability and temperature differences experienced among colonies. A regression of average colony worker head width on colony worker number was not significant (N = 27, P > 0.3).

Southerland [55] found that worker mortality was higher in laboratory nests maintained at 30◦C than in nests maintained at 15◦C. Field colonies collected in Connecticut and maintained in the laboratory contained workers that survived for more than a year. The average life span of a worker in a natural setting, where there are many risks, is undoubtedly less.

Workers possess functional ovaries but do not lay eggs when their colony has a healthy, fertilized queen [49]. Worker-produced males are also presumed to be uncommon in naturally occurring queenless nests (as discussed in the reproduction section).

Workers of the rudis-group exhibit little division of labor and can perform a total of 41 different behavioral acts [30]. In laboratory nests it was found that 75% percent of a workers’ time is spent in nonsocial behaviors and most individuals are inactive most of the time. Worker activity levels and brood tending rates were higher in the summer relative to autumn [30].

Gynes Low temperature and a sustained drop in metabolic rate are presumably necessary to induce gyne development, as has been found for other temperate ants [56]. New queens are thus produced from a subset of overwintered female brood. Workers are likely to play a role in determining which females develop into queens by altering the diet and/or temperature environments of select larvae [57].

Early-instar gynes resume development in early spring and in Connecticut eclose in mid-June. Fielde reported a single developing queen she observed spent 17 days in a pupal state [54]. Once eclosed gynes are presumably fed by their nestmates to increase their fat stores, which is typical for ant species in which queens found nests independently.

Gynes collected from the Pachuag and Nipmuck forests averaged a dry weight of 6.5mg (SD = 0.5, N = 25; Figure 5(b)).Haskins [58] observed the survivorship schedule of 11 laboratory-housed queens, finding a median lifespan of 8 years and a maximum of 13. A number of eastern Connecticut colonies, mature when found and therefore at least a few years old when collected, were maintained in the laboratory from 2001 through 2005.

Males Males can be produced from unfertilized eggs of queens or workers but in queenright nests there is no worker reproduction [49]. Males collected from the Pachuag and Nipmuck populations averaged 0.6mg dry weight (SD = 0.1, N = 25; Figure 5(c)). Fielde [54] reported that the median duration of the pupal stage for 3 males was 19 days. Males in queenright nests are thought to be produced from overwintered brood, passing through conditions similar to those experienced by gynes [24, 25]. Adult males leave the nest within a month of eclosing and their adult lifespan is likely brief. Once leaving the nest to mate, even if they avoid being killed by a predator, they will eventually succumb to starvation.

Like most members of Aphaenogaster, this species is not characteristic of prairies (Trager 1998), but may be found in prairie remnants or restorations that are adjacent to deciduous trees (Kittelson et al. 2008, Nemec et al. 2016).

Ant Community Interactions

Warren et al. (2015) - In the deciduous forests of the north Georgia Piedmont, A. rudis was being displaced by an increasing presence of the invasive forest ant Brachyponera chinensis. The latter was diminishing the abundance of A. rudis and was also preying upon, perhaps more effectively than the native ant, on the termite Reticulitermes flavipes. In contrast to A. rudis B. chinensis was not serving as an effective seed disperser of myrmecochorous seeds in this forest habitat.



The following information is derived from Barry Bolton's New General Catalogue, a catalogue of the world's ants.

  • rudis. Aphaenogaster fulva var. rudis Enzmann, J. 1947b: 150 (in key) (w.q.) U.S.A.
    • [First available use of Stenamma (Aphaenogaster) fulvum subsp. aquia var. rude Emery, 1895c: 305 (w.q.) U.S.A; unavailable (infrasubspecific) name.]
    • Wheeler, G.C. & Wheeler, J. 1953b: 56 (l.); Crozier, 1970: 125 (k.).
    • Combination in Aphaenogaster (Attomyrma): Emery, 1921f: 57.
    • Subspecies of fulva: Smith, M.R. 1951a: 796.
    • Subspecies of picea: Creighton, 1950a: 147; Bolton, 1995b: 72; Mackay & Mackay, 2002: 77.
    • [Note: picea was made available earlier than rudis; hence picea has priority (Bolton, 1995b: 72).]
    • Status as species: Smith, D.R. 1979: 1362; Wheeler, G.C. & Wheeler, J. 1986g: 36 (in key); Umphrey, 1996: 558 (in key); Coovert, 2005: 49; Ellison, et al. 2012: 231.

Type Material

  • Syntype, worker(s), queen(s), District of Columbia, Virginia, North Carolina, United States, location of material unknown.



  • Creighton, W. S. 1950a. The ants of North America. Bull. Mus. Comp. Zool. 104: 1-585 (page 147, Raised to species, subspecies of picea as this name available first (picea entered as subspecies of rudis), and hence its senior name.)
  • Crozier, R. H. 1970a. Karyotypes of twenty-one ant species (Hymenoptera: Formicidae), with reviews of the known ant karyotypes. Can. J. Genet. Cytol. 12: 109-128 (page 125, karyotype described)
  • DeMarco, B.B. 2015. Phylogeny of North American Aphaenogaster species (Hymenoptera: Formicidae) reconstructed with morphological and DNA data. Ph.D. thesis, Michigan State University.
  • Emery, C. 1895d. Beiträge zur Kenntniss der nordamerikanischen Ameisenfauna. (Schluss). Zool. Jahrb. Abt. Syst. Geogr. Biol. Tiere 8: 257-360 (page 305, First available use of Stenamma (Aphaenogaster) fulvum subsp. aquia var. rude; unavailable name)
  • Enzmann, J. 1947b. New forms of Aphaenogaster and Novomessor. J. N. Y. Entomol. Soc. 55: 147-152 (page 150, worker, queen described)
  • Lubertazzi, D. 2012. The Biology and Natural History of Aphaenogaster rudis. Psyche. 2012:1-11. doi:10.1155/2012/752815
  • Nemec, K.T., Trager, J.C. & Allen, C.R. 2012. Five new records of ants (Hymenoptera: Formicidae) for Nebraska. The Prairie Naturalist 44, 63–65.
  • Umphrey, G. J. 1996. Morphometric discrimination among sibling species in the fulva-rudis-texana complex of the ant genus Aphaenogaster (Hymenoptera: Formicidae). Can. J. Zool. 74: 528-559 (page 558, (in key) Raised to species)
  • Warren, R. J., A. McMillan, J. R. King, L. Chick, and M. A. Bradford. 2015. Forest invader replaces predation but not dispersal services by a keystone species. Biological Invasions. 17:3153-3162. doi:10.1007/s10530-015-0942-z
  • Wheeler, G. C.; Wheeler, J. 1953b. The ant larvae of the myrmicine tribe Pheidolini (Hymenoptera, Formicidae). Proc. Entomol. Soc. Wash. 55: 49-84 (page 56, larva described)