Myrmelachista

Myrmelachista is an exclusively Neotropical ant genus. The ants from this genus nest and forage in live or dead trees, but data on its life cycle are relatively scarce. The taxonomy of this genus is considerably complex. (Nakano et al., 2015.)

Identification
Species in this genus can be difficult to identify. Workers within species are often quite variable and there are not good characters to clearly delimit many species from others by morphological traits. Nakano et al (2015) found that a multivariate analysis of worker morphological characters could not reliably distinguish species found in the Atlantic forests of Brazil. Queens are often deemed more reliable for separating species, so much so that many species are described from a queen holotype.

Nakano et al (2015) - Characterized by a visible antennal club with one or two segments, a character that is absent in the remaining Neotropical Formicinae (Fernández, 2003). Myrmelachista species have antennae with nine or ten segments; most species have nine-segmented antennae and are found mainly in Central America (only two have been described in South America), whereas ten-segmented species are mainly found in South America (only three species have been described in Central America and Mexico) (Longino, 2006).

Biology
Longino (2006) - Among these stem-nesting ants is the formicine genus Myrmelachista. This group of ants is confined to the Neotropics and is exclusively arboreal. They are inconspicuous and have been little studied, but there are hints from the literature that they are far richer and with more complex plant associations than previously suspected. Early literature contained scattered reports of South American Myrmelachista inhabiting ant-plants (reviewed in Bequaert 1922, Wheeler 1942). Myrmelachista nigella and Myrmelachista schumanni were reported in internodes of Duroia hirsuta (Rubiaceae) (Ule 1907–1908, Schumann 1889). More recent reports have described “devil’s gardens” in South America, in which large polygynous colonies of Myrmelachista occupy monospecific patches of understory Duroia and Tococa (Melastomataceae) (Morawetz et al. 1992, Svoma & Morawetz 1992, Renner & Ricklefs 1998, Frederickson et al. 2005). These patches stand out in the typically dense and diverse understory vegetation as peculiar zones of a single plant species with bare earth beneath them and a ring of bare soil around the perimeter. These patches are created and maintained by the ants, which attack and kill foreign vegetation by spraying it with formic acid, which acts as an herbicide (Morawetz et al. 1992, Renner & Ricklefs 1998, Frederickson et al. 2005).

The first indication of something interesting in Central American Myrmelachista was Stout’s (1979) observation of a tight association between an unidentified Myrmelachista and an understory tree species in the genus Ocotea (Lauraceae). She examined 50 plants of Ocotea “pedalifolia” (probably a mix of O. atirrensis and O. dendrodaphne [Hammel 1986]) at the La Selva Biological Station, and found the stems of 49 of them inhabited by Myrmelachista. Since then little has been written about the association, although floristic works on the Lauraceae recognize that various species are routinely occupied by ants (Hammel 1986, Burger & van der Werff 1990). Ibarra-Manríquez and Dirzo (1990) reported the same phenomenon at the Los Tuxtlas Biological Station in Veracruz, Mexico. My studies of Myrmelachista in Costa Rica reveal that these observations are just the beginning and that there is a largely overlooked community of Central American Myrmelachista and their associated plants.

Throughout Costa Rica the understory of mature wet forests have high densities of plants that are associated with multiple species of Myrmelachista. Cloud forests likewise have high densities of Myrmelachista living in live stems, but they are more abundant in the canopy than in the understory. Many of these species nest entirely inside of live stems and rarely venture out onto the surface, and the plants they inhabit show no external signs of specialization for ant occupation. There are no preformed domatia, no food bodies, and no extrafloral nectaries. The result is a lineage of ants that in spite of being species-rich and very abundant is almost never collected. Many of the species reported here (Longino 2006) had never been collected prior to my work in Costa Rica, much less receiving any formal taxonomic treatment.

Nakano et al (2015) - Workers have been observed in association with 20 families and 53 species of angiosperms (see review by Nakano et al., 2013) and in leaf litter (Suguituru et al., 2011, 2013). Nests are located in the trunks of live trees (Longino, 2006). Colonies with immature and reproductive ants also occupy twigs dispersed on the leaf litter (Nakano et al., 2012, 2013). The genus Myrmelachista is characterized by forming mutualistic associations with some myrmecophytes (Renner & Ricklefs, 1998; Edwards et al., 2009), where workers use formic acid to protect the host plant (Frederickson, 2005), as well as with Coccidae and Pseudococcidae (Kusnezov, 1951; Stout, 1979; Ketterl et al., 2003; Longino, 2006). Although some studies on this genus have been published recently (Longino, 2006; Nakano et al., 2012, 2013), information on the biology of Myrmelachista species is still scarce. Ants from this genus feed on extrafloral nectaries (Frederickson, 2005; McNett et al., 2010), but workers may take plant, fungal and animal fragments (eggs, feces and larvae) into the nests (Torres, 1984).

Caribbean Taxa
Longino (2006)


 * Myrmelachista kraatzii
 * Myrmelachista rogeri
 * Myrmelachista rogeri manni
 * Myrmelachista rogeri rubriceps
 * Myrmelachista ambigua
 * Myrmelachista ramulorum
 * Myrmelachista ramulorum fortior
 * Myrmelachista gagates

All of the species from the Caribbean have 9 antennal segments.

The above are all the species and subspecies known from Caribbean islands. My knowledge of the Caribbean fauna is limited, but I have examined 20 collections of ramulorum from Puerto Rico, St. Croix, USA (Florida, possibly introduced and then extirpated, see Deyrup 2003), Santo Domingo, St. Thomas, and the Dominican Republic; three collections of rogeri from Cuba; syntypes of rogeri manni from Cuba, and syntypes of rogeri rubriceps from Cuba. All appear to be similar to plebecula. All are bicolored or various shades of red brown. Unlike plebecula, all have long erect setae projecting from the sides of the head (workers and queens). Eight queens of ramulorum are very small with very narrow, rectangular heads. The largest of these have the narrowest heads, with HW around 0.70mm and CI around 74, a combination not found in any Costa Rican species except for the one small longiceps-like queen described under longiceps. Unlike ramulorum, the longiceps-like queen lacks erect setae on the sides of the head. The smallest ramulorum queens and the queens of rogeri are in the same size range as plebecula queens, but with relatively narrow heads. All measured queens of plebecula have CI 85 or greater. The highest CI among the ramulorum and rogeri queens is 82. Myrmelachista kraatzii, ambigua, rogeri, and ramulorum are all older names than plebecula, and if plebecula proves to be an allopatric variant of a widespread polytypic Caribbean species it will no doubt be a synonym of one of these older names.

It is not clear that rogeri and ramulorum are distinct. When two Cuban rogeri queens I have measured are compared to eight ramulorum queens from other islands, they are at the small end of a continuum of measurements.

Myrmelachista ambigua was described from a single worker from St. Vincent. Given the relative uniformity of workers, the published description and even examination of the type will be of little use. Queen and male-associated collections of Myrmelachista from St. Vincent will be needed to compare with material from other parts of the Caribbean. Wheeler (1908) considered the worker of ramulorum close to ambigua.

Myrmelachista gagates, from Haiti, was described as being close to rogeri but solid black.

It will be important to examine multiple collections of Myrmelachista from Cuba, to ascertain whether there are multiple sympatric species there. It is unknown whether kraatzii and the forms of rogeri are distinct or represent one variable species. Myrmelachista kraatzii from Cuba and M. nigella from Venezuela are the two oldest names in the genus, kraatzii being a 9-segmented form and nigella a 10-segmented form. Thus kraatzii would have priority among all the 9-segmented forms.

One collection from El Yunque, Puerto Rico, is indistinguishable from M. longiceps. It is a collection of workers and alate queens, collected by Juan Torres. I am reluctant to identify it as longiceps until more Puerto Rican material is obtained, but there is a large size gap between the queen of this El Yunque collection and the various queens of ramulorum from elsewhere in Puerto Rico.

From these observations it is clear that more collections are needed from the Caribbean to better understand species boundaries in this group.

Southern Mexico
Longino (2006)

The following are all species from this region that have 9 antennal segments.


 * Myrmelachista skwarrae
 * Myrmelachista skwarrae picea
 * Myrmelachista skwarrae laeta
 * Myrmelachista amicta

The above are a set of species described together by Wheeler from material collected by Elisabeth Skwarra in southern Mexico. They are all very similar to plebecula and joycei. I cannot distinguish workers of amicta, skwarrae (and its two subspecies picea and laeta), and plebecula. The males of skwarrae and plebecula are distinct. The male of skwarrae differs from plebecula in having a basiparamere lobe and distinct (though minute) setose pygostyles. Myrmelachista skwarrae males are more like joycei. They differ from joycei in the relatively thinner and more delicate basiparamere lobe and relatively larger ocelli.

If the Mexican forms prove to be geographic variants of plebecula, the latter has priority. On the other hand, M. joycei could prove to be a geographic variant of one of the previously named Mexican forms.

South America
Longino (2006)

There are two species in South America with 9 antennal segments.


 * Myrmelachista brevicornis
 * Myrmelachista guyanensis

From the descriptions of both species and my examination of workers of guyanensis, these two species are similar to and could be synonymous with plebecula. But further collections of South American Myrmelachista will be necessary to further understand their status.

It is tempting to speculate that the 10-segmented forms, widespread and diverse in South America, have spawned one main radiation of 9-segmented forms in Central America and the Caribbean. In this scenario only plebecula, the “weediest” of 9-segmented forms, has spread into South America, and there is no diverse community of 9-segmented forms there. But the late discovery of a trove of 9-segmented Myrmelachista species in Costa Rica shows that this group may remain hidden from the general collector, even in the face of prolonged collecting effort. Thus we remain ignorant of extent of such forms in the South American hylea.

Nomenclature

 *  MYRMELACHISTA [Formicinae: Plagiolepidini]
 * Myrmelachista Roger, 1863a: 162. Type-species: Myrmelachista kraatzii, by monotypy.
 * Myrmelachista senior synonym of Decamera Roger (junior homonym), Hincksidris, Neaphomus: Snelling, R.R. & Hunt, 1976: 110.
 * HINCKSIDRIS [junior synonym of Myrmelachista]
 * Hincksidris Donisthorpe, 1944b: 59 [as subgenus of Myrmelachista]. Replacement name for Decamera Roger, 1863a: 166. [Junior homonym of Decamera Mulsant, 1842: 503 (Coleoptera).]
 * Hincksidris junior synonym of Myrmelachista: Snelling, R.R. & Hunt, 1976: 110.
 * NEAPHOMUS [junior synonym of Myrmelachista]
 * Neaphomus Menozzi, 1935c: 324 [as subgenus of Aphomomyrmex]. Type-species: Aphomomyrmex (Neaphomus) goetschi, by monotypy.
 * Neaphomus raised to genus: Kempf, 1972a: 152; Dlussky & Fedoseeva, 1988: 77.
 * Neaphomus junior synonym of Myrmelachista: Snelling, R.R. & Hunt, 1976: 110; Bolton, 1994: 50.