Myrmecocystus, a genus of ants found in arid and semi-arid areas of the western United States and Mexico, is one of several ant genera that include species with a highly specialized caste known as "repletes". Repletes store large quantities of liquid food in their crops, causing their abdomens to become so distended they have difficulty moving and remain permanently in the nest as living storage vessels. When food is scarce, they regurgitate and share food with their nestmates. Repletes are drawn from the largest workers (Rissing 1984) and may vary in color from clear yellowish brown to amber. The darker crop contents apparently contain more dissolved solids, largely glucose and fructose, whereas the lighter fluid may include more sucrose and water.(Hölldobler and Wilson 1990; Kronauer and Gadau 2002)

A mature Myrmecocystus colony consists of around 15,000 workers and a single queen. One researcher investigating a Myrmecocystus colony enlisted the help of professional gravediggers to follow the nest galleries through 5 meters of Arizona desert soil to recover the queen from a small chamber at the very bottom. (Hölldobler and Wilson 1990 and references therein)

A nice overview of the biology of Myrmecocystus mendax, by Randy Morgan of the Cincinnati Zoo, is hosted by the Sonoran Arthropod Studies Institute here. Morgan also provides a summary of the taxonomy of the genus Myrmecocystus and, more broadly, of all the world's honeypot ants (including those in other genera) here (also see Morgan 1991). Snelling (1976, 1982) reviewed the systematics of Myrmecocystus.

The honey ant genus Myrmecocystus is distributed throughout the northwestern US, Baja California, and northern Mexico, where the 29 (at least) species inhabit arid and semi-arid habitats (Kronauer and Gadau 2002).

These are the well-known “honeypot ants”, whose colonies include a subgroup of specialized workers (repletes) devoted to storage of liquid food in their swollen abdomens. Species of Myrmecocytus are restricted to the western United States and northern Mexico and are found primarily in desert habitats. Snelling’s (1976) landmark revision of the genus also contains much useful biological information.

Species identification: keys in Snelling (1976, 1982b). Additional references: Diniz-Filho and Fowler (1998), Duncan and Lighton (1994a, 1994b), Hölldobler (1976b, 1986), Kay and Whitford (1978), Kronauer et al. (2003, 2004), Lloyd et al. (1989), Rissing (1984), Snelling (1969b, 1971a), Snelling and George (1979), Smith (1979), Wheeler and Wheeler (1986g).

Myrmecocystus , Wesm. Bull. Acad. Roy. de Brux. v. (1838) p. 766.

Lucas, Bull. Ent. pl. 54; Ann. Soc. Ent. Fr. (1855).

" The first abdominal segment of one node; the mandibles triangular and much toothed; abdomen swollen at times like a balloon, and then perfectly transparent. In Mexico called hor- migas micleras or mochileras, i. e. honey-ants or pouched ants. their abdomen containing a sufficiently agreeable honey."

Termites are a major food for Myrmecocystus mimicus (Hölldobler 1981; Hölldobler and Wilson 1990 and references therein).

Shaffer and Whitford (1981) report that Myrmecocystus are a major prey item for the Roundtail Horned Lizard (Phrynosoma modestum).

Chew (1979) reported that nests of Myrmecocystus were at least occasionally excavated by large mammals, probably badgers (Taxidea taxus), in a desert-scrub site in southeastern Arizona (U.S.A.). He also cited references to possible predation by coyotes (Canis latrans) and perhaps skunks.

Hölldobler (1986) documented extensive food-robbing of Pogonomyrmex ants by M. mimicus, with a strong emphasis on termite prey.

Conway (1990) reported finding a Myrmecophila cricket and springtails (Collembola) deep in a Myrmecocystus mexicanus nest, as well as Cremastocheilus scarabbeetles (as previously reported by Cazier and Mortenson 1965), at least some of which are known to eat ant larvae. In a M. mendax nest, Conway (2003) reported finding a Myrmecophila cricket, a staphylinid larva, collembolans,
and mites (Gymnolaelaps).

Hölldobler (1981) found that Myrmecocystus mimicus and M. depilis were randomly distributed relative to one another, although each was overdispersed intraspecifically (i.e., within species, colonies were spaced farther away from each other than expected by chance).

Chew (1987) found that M. mexicanus, M. depilis, and Aphaenogaster cockerelli were all overdispersed relative to one another. Chew interpreted this as the outcome of interspecific competition among these species.

Cole et al. (2001) found that although intraspecific dispersion was highly uniform at their study site (i.e., nests of M. mexicanus were more uniformly dispersed than expected by chance), colonies were significantly positively associated in space (i.e., clumped) with at least larger nests of the harvester ant Pogonomyrmex occidentalis. Colonies of M. mexicanus were more likely to be found within 3 meters of P. occidentalis and less likely to be found as far as 10 meters away. The protein component of the diet of M. mexicanus at the study site consisted almost exclusively of dead or dying workers of P. occidentalis. Myrmecocystus mexicanus thus appear to build their nests close to one of their main food sources. An earlier study in a different area with very different densities of the two species found a negative association between the two species, a difference Cole et al. attribute to differences between the two studies in the importance of Pogonomyrmex as a food source for Myrmecocystus. (Cole et al. 2001 and references therein)

J.R. Conway published a series of small papers on various aspects of Myrmecocystus biology, including nest architecture (Conway 1983, 2003).

Sanders and Gordon (2003) investigated resource competition among M. depilis, M. mimicus, and Aphaenogaster cockerelli.

Chew (1987) reported on a 23 year study of demography and temporal and spatial dynamics of colonies of the three largest ant species in a desert scrub community: the diurnal Myrmecocystus depilis, the nocturnal M. mexicanus, and Novomessor cockerelli.

A number of ant species have long been known to capture brood from colonies of particular other species and raise the foreign ants as workers in their own colonies. This is often referred to as "slave-making" (or, in more technical jargon, "dulosis"). The phenomenon of intraspecific slavery (which is more in line with the usual use of the term "slavery" among non-entomologists) was not described until 1976, when Hölldobler documented it for Myrmecocystus mimicus. More recently, Kronauer et al. (2003) used mitochondrial and nuclear DNA markers to document and study the effect of both intraspecific (within-species) and interspecific (between-species) brood raiding on colony structure in populations of M. mimicus and M. depilis at a study site in Arizona (U.S.A.) and discussed implications for the evolution of slave-making in ants.

In contrast to the many species of ants that engage in (often deadly) physical fighting when territorial borders are challenged, Hölldobler (1976) found that colonies of M. mimicus conduct ritualized tournaments in which hundreds of ants perform highly stereotyped display fights and few individuals are injured. These fights may last for several days (breaking for the night, during which these ants are generally inactive). Opposing colonies attract their workers to the tournament area by means of an alarm-recruitment system. When one colony is considerably stronger than the other, the tournament quickly ends, the weaker colony is raided, and the losers are "enslaved". Some tournaments may be initiated as a way of protecting a rich food source from other colonies (Hölldobler and Lumsden 1980; Hölldobler 1981).

New M. mimicus colonies are often founded by groups of foundresses ranging in size from 2 to 9 females, with groups of 2 to 4 females being most common. Founding nests are aggregated together in patches which are distant from existing M. mimicus colonies (Bartz and Hölldobler 1982).

The maximum age recorded for Myrmecocystus is more than 11 years for a M. mimicus queen in a laboratory colony (Hölldobler and Wilson 1990).

J.R. Conway published a series of small papers on various aspects of Myrmecocystus biology, including reproductive biology (e.g., Conway 1980, 1981, 1983).

Schooley and Bestelmeyer (2000) studied the behavior of M. mimicus queens as they searched for sites on the ground to initiate nests.

Kronauer et al. (2004) published a phylogenetic analysis of Myrmecocystus based on mitochondrial DNA. Their data set included 15 of the 29 described species (including all recognized subgenera and seven of the eight recognized species groups), as well as several undescribed species.

The bodies of some honey ants are used for liquid storage via their expandable abdomen.

Kay and Whitford (1978) investigated critical thermal limits of five species of Myrmecocystus (M. romainei, M. depilis, M. mimicus, M. mexicanus, and M. navajo) and discussed possible ecological implications of their findings.

Kronauer and Gadau (2002) developed 5 polymorphic microsatellite markers to study the population and colony kin structure of Myrmecocystus mimicus (at least three of these markers appear to be useful for studying related species as well).