Diacamma

Diacamma is a moderately sized genus which ranges from India to Australia. It is notable for its reproduction by gamergates and control of reproduction by nestmate mutilation.

Identification
Schmidt and Shattuck (2014) - Diacamma workers are highly distinctive and are easily identified by the presence of deep striate sculpturing, deep pits (gemmal pits) on the sides of the mesosoma, and a bispinose petiole. The gemmal pits are autapomorphic, but may be confused with the wing scars of dealate queens in other taxa. Diacamma workers lack the other characters of winged queens, however, such as ocelli and modified thoracic sclerites. The combination of deep striate sculpturing, prominent arolia, bispinose petiole, and laterally-opening metapleural gland orifice (with a posterior U-shaped cuticular lip) also differentiates Diacamma from the queens and workers of any other ponerine genus.

Distribution
The range of Diacamma extends from India east to Japan, and from southern China to northeast Australia (Emery, 1911; Wheeler & Chapman, 1922; Suwabe et al., 2007). Laciny et al. (2015) revised species from SE Asia, including new species descriptions.

Biology
Schmidt and Shattuck (2014) - In most respects Diacamma are fairly typical ponerines. The workers are monomorphic, forage individually on the ground and on low vegetation, and show a remarkable degree of directional fidelity when foraging (Abe & Uezu, 1977; Karpakakunjarum et al., 2003; Eguchi et al., 2004). They are apparently generalist predators of arthropods (Abe & Uezu, 1977; Karpakakunjarum et al., 2003). Ke et al. (2008) found that workers of Diacamma rugosum are effective predators of termites in artificial arenas, and Karpakakunjarum et al. (2003) observed that termites made up the majority of the diet of Diacamma ceylonense. Colonies contain on average a few hundred workers or less (e.g., D. ceylonense: 200–300 workers; Karpakakunjarum et al., 2003; Ramaswamy et al., 2004; Baratte et al., 2006a; Diacamma cyaneiventre: 214 workers; André et al., 2001; Diacamma indicum: 90 workers; Viginier et al., 2004; D. rugosum: 40 –50 workers; Wheeler & Chapman, 1922; Wilson, 1959b; Diacamma sp. (Japan): 118 workers; Abe & Uezu, 1977; Diacamma sp.: 86 workers; Sommer et al., 1993; Diacamma sp. 'nilgri': 275 workers; Bocher et al., 2008).

Nests are usually constructed in soil (often in the middle of clearings), rotting logs, or even in trees (Wheeler & Chapman, 1922; Abe & Uezu, 1977; Fukumoto & Abe, 1983; André et al., 2001, 2006; Eguchi et al., 2004; Viginier et al., 2004; Allard et al., 2007). The nests of many Diacamma species are deep and complex, allowing workers to retreat to deeper chambers in response to nest disturbance, though nests of Diacamma indicum are shallow and colonies emigrate after only minor disturbances (Viginier et al., 2004). Diacamma sp. (from Japan) also emigrates readily in response to disturbances or unfavorable environmental conditions, and utilizes both tandem running and social carrying during emigrations (Abe & Uezu, 1977; Fukumoto & Abe, 1983). Nestmate recruitment during emigration also occurs through tandem running in D. rugosum, with the hindgut fluid apparently acting as a long-term trail pheromone (Maschwitz et al., 1986). Moffett (1986) found that colonies of D. rugosum in India are polydomous, with multiple shallow nests separated by a meter or more, and also discovered that these ants surround their nest entrances with feathers and ant corpses, apparently in order to collect dew (Moffett, 1985).

The reproductive behaviors of Diacamma are highly unusual and have been heavily studied. Diacamma colonies are queenless, with reproduction instead being performed by a single mated gamergate worker (Wheeler & Chapman, 1922; Peeters & Higashi, 1989). Several other ponerine lineages have similarly lost the queen caste, but the control of reproduction in Diacamma is unique among ants. All Diacamma workers eclose with a pair of novel thoracic appendages called gemmae (Tulloch, 1934; Peeters & Billen 1991). In virtually every Diacamma species studied (see below for the one known exception), the presence of intact gemmae is essential for a worker to become sexually receptive and ultimately a gamergate (Peeters & Higashi, 1989; Cuvillier-Hot et al., 2002). The gemmae have minute pores on the outer surface that are connected to exocrine cells, thus providing olfactory information about gemma presence (Peeters & Billen 1991; Billen & Peeters 1991). Gemmae are covered with sensory hairs with a mechanoreceptive function, and mutilation results in degeneration of the neural connections to the central nervous system; (Gronenberg & Peeters 1993). Loss of the gemmae thus causes an individual to behave as a non-reproductive worker (Allard et al., 2005). The single gamergate in a colony therefore mutilates the gemmae of newly eclosed workers to maintain its reproductive dominance (Peeters, 1993). The social and reproductive conflicts in Diacamma colonies were reviewed by Monnin & Ratnieks (2001) and Baratte et al. (2006).

The origin of the gemmae is somewhat controversial. Based on neurological and developmental studies, Gronenberg & Peeters (1993) and Gotoh et al. (2005) suggested that the gemmae are homologous with the forewings (see also Tulloch, 1934), but Baratte et al. (2006b) disagreed, instead arguing that the gemmae are novel organs whose development simply co-opted some of the same genes and processes as wings. Whatever their origin, the gemmae are filled with secretory cells which seem to have evolved de novo (Billen & Peeters, 1991; Peeters & Billen, 1991). Though the exact mechanism of action of the gemmae have not yet been worked out, Allard et al. (2005) found that the mutilation of gemmae in young workers of Diacamma sp. (Japan) caused their bursa copulatrices and spermathecae to degenerate, leaving them incapable of mating. Tsuji et al. (1998) experimentally determined that the gemmae of gamergates are not directly involved pheromonally in the suppression of worker reproduction. Bitsch & Peeters (1991) examined the structure of gemmae in Diacamma australe.

In a classic study, Peeters & Higashi (1989) worked out many of the basic details of the reproductive and social behavior of D. australe. They found that colonies contain a single gamergate, which has intact gemmae, active ovaries, and sperm-filled spermathecae, and which monopolizes egg-laying in the colony. All other workers lacked gemmae and were unmated, and most of these had completely undeveloped ovaries. Only the gamergate actually mutilated the gemmae of newly eclosed workers, though other workers assisted. In experimentally orphaned colonies, some workers laid haploid eggs, and the oldest unmutilated callow worker (the future gamergate) became aggressive and began mutilating other callow workers with intact gemmae, began laying eggs, and became receptive to mating.

Subsequent research has confirmed many of these observations in other species of Diacamma and have filled in many additional details. In a scenario similar to that in Diacamma australe, the first worker to eclose (i.e., the oldest unmutilated worker) in an orphaned colony of Diacamma ceylonense immediately becomes aggressive toward her nestmates, and after three weeks begins to lay haploid eggs, ceases her aggression, and becomes receptive to mating (Cuvillier-Hot et al., 2002). In both D. ceylonense and D. australe, newly eclosed workers are aggressive toward other unmutilated callow workers (potential future gamergates), but do not resist their own mutilation by mature gamergates (Baratte et al., 2006a).

As with other ponerines in which alate queens are absent, colony reproduction in Diacamma occurs through fission (Fukumoto et al., 1989; André et al., 2006). When this occurs, one of the colony fragments is headed by the gamergate of the mother colony and one fragment is orphaned. A colony may also be orphaned through the death of the gamergate. Serial polygyny characterizes Diacamma colonies, as two matrilines coexist in a colony for some time after the death of the gamergate (André et al., 2001). André et al. (2006) studied the pattern of worker and gamergate turnover in Diacamma cyaneiventre and estimated that the average tenure of gamergates is about 200 days. Interestingly, gamergate turnover in this species does not significantly affect the average worker relatedness, which is very close to the expectation for a monandrous and monogynous ant colony (André et al., 2001). As expected from the limited dispersal abilities of Diacamma queens, populations of D. cyaneiventre were found to be highly genetically isolated, with most gene flow occurring via male dispersal (Doums et al., 2002).

In Diacamma sp. from Japan, workers are apparently inhibited from laying haploid eggs by a non-volatile pheromone produced by the gamergate (Tsuji et al., 1999). This was confirmed by studying Cuticular Hydrocarbons in Diacamma ceylonense: Cuvillier-Hot et al., 2001). Both gamergates and non-gamergates also police reproduction by workers through aggression and egg cannibalism (Kikuta & Tsuji, 1999, Kawabata & Tsuji, 2005). Nonetheless, non-gamergate workers do succeed in laying a small number of eggs, especially in larger colonies (Nakata & Tsuji, 1996; Kikuta & Tsuji, 1999). In a test of the extent to which Diacamma workers control reproduction in the colony, Nakata (1998) observed that workers of Diacamma sp. (from Japan) do not differentially rear male or female brood, and therefore do not control the sex ratio of the colony’s sexual brood. Peeters & Tsuji (1993) found that orphaned workers of Diacamma sp. (from Japan) aggressively competed with one another and formed a non-linear dominance hierarchy with a definite alpha and beta; only the alpha reproduced, and she ate the eggs of other workers. In an unidentified Diacamma species from Malaysia, orphaned workers also compete and form a dominance hierarchy, with only the alpha reproducing (Sommer et al., 1993).

The age and reproductive status of Diacamma workers is communicated by their cuticular hydrocarbon profile (in Diacamma ceylonense: Cuvillier-Hot et al., 2001). Gamergates have a distinct hydrocarbon profile, and this seems to play a role in controlling reproduction by nestmates (Cuvillier-Hot et al., 2002). Suwabe et al. (2007) found that workers of a Diacamma sp. can distinguish nestmates from non-nestmates, presumably by their cuticular hydrocarbons, and are hostile toward non-nestmates. Marukawa et al. (2001) and Masuda & Mori (2002) described the biochemistry of cuticular hydrocarbons in this species.

One Diacamma species is known in which the control of reproduction is not mediated through gemmae. In a species closely related to D. ceylonense (referred to as Diacamma sp. 'nilgri'; Baudry et al., 2003), reproduction is controlled by aggressive dominance interactions among workers (Peeters et al., 1992; Cournault & Peeters 2012), similar to the situation in other ponerines with gamergates such as Dinoponera. In 'nilgiri', the dominant worker begins laying eggs and eventually will mate and become gamergate (Peeters et al., 1992). Ramaswamy et al. (2004) found that the cues for mutilation originate in the gemmae of the victim, as D. ceylonense callows introduced into D. sp. 'nilgri' colonies are mutilated, but ‘nilgiri’ callows introduced into D. ceylonense colonies are not. Baratte et al. (2006a) hypothesized that the mutilation mechanism in Diacamma may maximize colony productivity relative to other queenless ponerines in which dominance interactions determine the dominant reproductive individual. In support of this hypothesis, Bocher et al. (2008) found that dominance interactions in D. sp. ‘nilgiri’ lead to reduced colony work efficiency and reduced immunocompetence in colony members. The selective conditions favoring the dominance interaction strategy over the gemmae mutilation strategy are unclear.

Diacamma appears to be monandrous, with single mating reported for several species (e.g., Diacamma cyaneiventre: André et al., 2001). In laboratory colonies of Diacamma australe, foreign males encountered by foraging workers were carried into the nest, where mating occurred (Peeters & Higashi, 1989). New gamergates of Diacamma sp. (from Japan) wait outside the nest entrance and call to males using pheromones derived from the metatibial gland (Fukumoto et al., 1989; Nakata et al., 1998; the metatibial gland was described by Hölldobler et al., 1996b). Peeters et al. (1992) also observed calling behavior by new gamergates of Diacamma sp. ‘nilgiri’. Despite very rapid sperm transfer, copulation in Diacamma is exceptionally long, with males remaining attached to females for as long as two days; males actually have to be killed and forcibly removed by the gamergate and her nestmates (Allard et al., 2002, 2007).

Relatively little work has been done on the division of labor in Diacamma colonies, other than reproductive division of labor. Nakata (1995, 1996b; also Dahbi & Jaisson, 1995) found that Diacamma sp. (from Japan) has a typical age-related polyethism, though workers remain behaviorally flexible. Nakata (1996a, 2000) found that smaller colonies of this species have lower temporal stability in colony productivity, and that the behavioral flexibility of workers does not fully compensate for drops in colony productivity due to fluctuating colony demographics.

At least three cases are known of social parasitism or commensalism between Diacamma and other ants. Maschwitz et al. (2000, 2004) discovered that an undescribed Diacamma species acts as host to the formicine Polyrhachis lama, feeding and protecting the Polyrhachis adults and brood. A P. lama colony may simultaneously parasitize multiple Diacamma colonies (Maschwitz et al., 2004). This symbiosis may have originated through mimicry of the Diacamma host by the Polyrhachis parasite. Maschwitz et al. (2001) observed Diacamma leading their Polyrhachis guests to new nest sites during emigrations via tandem running. Kaufmann et al. (2003) found two instances of compound nesting involving Diacamma and either Strumigenys or Pheidole, with the latter genera nesting in small chambers adjacent to the Diacamma nest and feeding on mites and collembolans in the Diacamma nest (Strumigenys) or on the Diacamma refuse piles (Pheidole). Eguchi et al. (2005) discovered myrmecophilic gastropods living with Diacamma sp. nr. sculpturatum. The gastropods probably feed on the Diacamma refuse piles and apparently have adaptations to ensure their spread during fissions of the host colony.

Biochemical examinations of Diacamma include studies of cuticular hydrocarbons and the glandular properties of the gemmae (both discussed previously), as well as the contents of the Dufour’s and venom glands (Morgan et al., 2003) and of the mandibular glands (Morgan et al., 1999). Doums (1999) and Gopinath et al. (2001) identified microsatellite loci in D. cyaneiventre and D. ceylonense, respectively.

Peeters, Heraty & Wiwatwitaya (2015) recently reported the first known case of a Diacamma species being parasitised by a eucharitid wasp. Different immature stages and adults of Schizaspidia diacammae (Chalcidoidea: Eucharitidae) were found inside cocoons of Diacamma scalpratum. Wasp larvae were feeding on ant pupae, while other host cocoons yielded five wasp pupae and both male and female adults. Parasitized cocoons are cut in a distinct manner by the wasps when they exit, and this feature can be used to assess the prevalence of parasitism. Dissection of the ovaries of one recently emerged physogastric female revealed thousands of eggs ready to be laid.

Castes
The genus Diacamma is characterized by the loss of the morphological queen caste. All female individuals look like ordinary workers in other ants, except that when eclosing from cocoons, they bear two thoracic appendages known as “gemmae” that are derived from wings. The gemmae of young ants are systematically removed by other workers except for one which retains the gemmae and will be able to copulate with a foreign male (e.g. Cuvillier-Hot et al. 2002). This mated individual lays eggs and is the gamergate.

In a southern Indian population of Diacamma, referred to as ‘nilgiri’, the gamergate does not mutilate her nestmates and yet monopolises reproduction (Peeters et al., 1992). This is a derived condition, because ‘nilgiri’ is highly related to Diacamma ceylonense (Baudry et al., 2003), in which mutilations occur routinely. In 'nilgiri', dominance interactions regulate the ability to become gamergate (Cournault & Peeters 2012).

Species Uncertain

 * Diacamma sp.1:
 * Diacamma sp.2:
 * Diacamma:
 * Diacamma:
 * Diacamma:
 * Diacamma:

Nomenclature

 *  DIACAMMA [Ponerinae: Ponerini]
 * Diacamma Mayr, 1862: 718. Type-species: Ponera rugosa, by subsequent designation of Bingham, 1903: 75.

Description
Schmidt and Shattuck (2014):

Worker
Medium to large (TL 8–16 mm) ants with the standard characters of Ponerini. Mandibles triangular and usually without a basal groove. Anterior margin of clypeus convexly triangular. Frontal lobes of moderate size. Eyes large and convex, located at or just anterior of the head midline. Mesonotum very short. Large gemmal pits present laterally at the base of the mesonotum. Metanotal groove reduced to a simple suture. Propodeum moderately narrowed dorsally. Propodeal spiracles slit-shaped. Metapleural gland orifice large, opening laterally, with a posterior U-shaped cuticular lip and at most a shallow lateral depression. Metatibia with a conspicuous, depressed, usually pale glandular area on the posterior surface. Metatibial spur formula (1s, 1s) or (1s, 1p). Arolia prominent. Petiole nodiform, usually roughly cuboidal, with a pair of short spines on the posterodorsal margin. Gaster with a moderate girdling constriction between pre- and postsclerites of A4. Head and body heavily striate, with scattered short pilosity and usually dense pubescence. Color variable, generally gray or black but often metallic. A study of thoracic structure in Diacamma australe was given by Bitsch & Peeters (1991), and a detailed morphological study of workers in another Diacamma species was given by Okada et al. (2006).

Male
See description in Ogata (1987) and Okada et al. (2006).

Larva
Described by Wheeler & Wheeler (1952) and Baratte et al. (2005).