Polyrhachis

Rudy Kohout has spent more than 2 decades revising members of this diverse genus. His ongoing publications focus on sets of species placed in species groups. These groups are organized by subgenera. Dorow (1995) presented the first contemporary reorganization of what had previously slowly developed into a messy tangle of poorly organized groups and forms. Kohout's revisions have altered some of Dorow's groupings. This has come about as particular species and groups have been more carefully assessed. There no longer exists any published overview detailing how all the species groups are organized under their respective subgenera. Antwiki does present some of these updated group revisions and species keys (species groups are presented on their respective subgenera for many groups; see the subgenera listed in the taxobox on the right side of this page). A phylogenetic analysis, using molecular sampling, has recently examined the relationships within and between existing groups (Mezger & Moreau 2015, see below). Not all of the groupings are supported by this analysis. Most of the modern taxonomic revisions have been clear in stating that the current groupings are logical constructs largely based on similarities in morphology and are not phylogenetic hypotheses. Both the species groups and subgenera are important considering the current taxonomic state of the genus as a whole.

Identification
Bolton (1973) - Worker. Monomorphic, medium to large (4.4 to 14.1 mm) ants belonging to the formicine tribe Camponotini. Antennae 12-segmented, the scapes inserted some distance behind the posterior clypeal margin (usually a distance greater than the basal width of the scape). Palp formula 6,4; mandibles usually with five, rarely with four teeth. Eyes well developed. Pronotum armed with a pair of spines, teeth or tubercles in all species of the Ethiopian region, the propodeum usually armed with two spines, teeth or tubercles, or a pair of ridges, rarely with only a single transverse ridge or completely unarmed. Promesonotal suture usually present (absent from khepra ), the development of the metanotal groove variable. Mesoscutellum very rarely present. In the single species in which the mesoscutellum occurs on the dorsum of the alitrunk, it is not separated from the scutum by a deep impression. Margination of the alitrunk variable, often present and complete but showing all stages through to a fully immarginate condition. Petiole usually with four but occasionally with two or six spines or teeth of variable configuration. Gaster large, globose, the first tergite extensive, usually forming at least half of the dorsal surface. The anterior face of the first gastral tergite is often truncated or concave. Acidopore not borne upon a conical projection of the hypopygium, usually concealed by the pygidium when not in use. Queen. As worker but with the alitrunk massively developed and with a corresponding reduction in armament and margination. The petiolar spines tend to be reduced and are usually smaller than those of the associated worker. Ocelli are present and wing venation is of the usual camponotine form. Male. Very poorly known, in most cases indistinguishable from the males of Camponotus.

See the nomenclature section below for links to more Polyhrachis keys.

Biology
Bolton (1973) - The nesting habits of the genus as a whole have been discussed by Hung (1967) who sums up by saying that four types may be recognised, as follows:

1. Arboreal: carton and silk nests amongst leaves and twigs.

2. Lignicolous: nests in the cavities of plants.

3. Terrestrial: nests on the ground under any object.

4. Subterranean: nests in the soil, without cover.

Robson & Kohout (2007) - Based on a review of 197 of the approximately 500 described species of Polyrhachis, the majority of species within a particular subgenus demonstrate a consistent nesting type but a few species can be highly variable, demonstrating both subterranean and arboreal habits. The inclusion of larval silk in the nests is correlated almost entirely with arboreal nesting, but the presence of spider silk in the nests of at least four species suggests caution when inferring silk origin. Pupal cocoons are present in all but subgenera Cyrtomyrma and Myrmatopa.

Africa
Bolton (1973) - All four nesting types (see above) are found in the Ethiopian regional fauna. Obviously, the first on the list is more or less restricted to forest species whilst the third and fourth are most commonly found in savannah forms. The second type listed above is rather a broad category and as far as the Ethiopian region is concerned includes such diverse nest sites as those of Polyrhachis decemdentata in rotten or termite-mined tree branches and those of Polyrhachis cubaensis inside stem galls. It would probably also hold Polyrhachis otleti which nests in rot holes or crevices in tree trunks and covers the entrance with a mesh of silk and vegetable fibres.

The distribution of the species may be divided roughly into forest and savannah forms although some may penetrate the edges of one from the other. The species Polyrhachis viscosa and Polyrhachis schistacea may be considered as typical of the savannah forms, spreading throughout the continent; the former even occurs on the coastal plains of West Africa. Nests are constructed in the earth and the ants are fast-moving, ascending grasses and bushes to forage. The forest species are typified by Polyrhachis militaris and Polyrhachis decemdentata, arboreal retiring forms foraging singly upon the trees and with a marked tendency to release their grip and fall into the undergrowth if disturbed.

Fungi

 * Unknown species of Polyrhachis are hosts for the fungi and  (Araujo et al., 2018).

Phylogeny
(Mezger & Moreau 2015) Spiny ants (Polyrhachis Smith) are a hyper-diverse genus of ants distributed throughout the Palaeotropics and the temperate zones of Australia. To investigate the evolution and biogeographic history of the group, we reconstructed their phylogeny and biogeography using molecular data from 209 taxa and seven genes. Our molecular data support the monophyly of Polyrhachis at the generic level and several of the 13 recognized subgenera, but not all are recovered as monophyletic. We found that Campomyrma Wheeler consists of two distinct clades that follow biogeographic affinities, that the boundaries of Hagiomyrma Wheeler are unclear depending on the analysis, that Myrma Billberg might be treated as one or two clades, and that Myrmhopla Forel is not monophyletic, as previously proposed. Our biogeographic ancestral range analyses suggest that the evolution of Polyrhachis originated in South-East Asia, with an age of the modern crown-group Polyrhachis of 58 Ma. Spiny ants dispersed out of South-East Asia to Australia several times, but only once to mainland Africa around 26 Ma.

Species Uncertain

 * Polyrhachis sp.(ANIC-1):
 * Polyrhachis sp.1:
 * Polyrhachis sp.2:
 * Polyrhachis sp.3:

Nomenclature

 *  POLYRHACHIS [Formicinae: Camponotini]
 * Polyrhachis Smith, F. 1857a: 58. Type-species: Formica bihamata, by original designation.
 * [Polyrhachis Shuckard, in Swainson & Shuckard, 1840: 172. Nomen nudum.]
 * Polyrhachis subgenus of Myrma: Wheeler, W.M. 1911c: 860; Wheeler, W.M. 1911f: 170.
 * [Myrma subgenus of Polyrhachis: Forel, 1915b: 106; Forel, 1917: 251; Wheeler, W.M. 1922a: 702, 993; Emery, 1925b: 198 (see note under Myrma).]
 * Subgenera of Polyrhachis (alphabetical order with synonyms indicated): nominal plus Aulacomyrma (= Johnia), Campomyrma, Chariomyrma, Cyrtomyrma, Hagiomyrma, Hedomyrma (= Dolichorhachis, = Morleyidris), Hemioptica, Hirtomyrma, Myrma (= Anoplomyrma, = Hoplomyrmus, = Pseudocyrtomyrma), Myrmatopa (= Irenea), Myrmhopla (= Cephalomyrma, = Florencea), Myrmothrinax (= Evelyna).
 * [All subgenera were given as provisional junior synonyms of Polyrhachis by Brown, 1973b: 178-184. The list was repeated in Hölldobler & Wilson, 1990: 19 with all subgenera listed as junior synonyms. They reverted to subgeneric status in Bolton, 1994: 50; see under individual entries. The entry of Chariomyrma, Hagiomyrma and Hedomyrma under the synonymy of Polyrhachis by Taylor & Brown, D.R. 1985: 131, is not considered as confirmation as the three subgenera all range outside the Austral Region.]
 *  Polyrachis Arnold, 1924: 741, incorrect subsequent spelling.

Dorow (1995) provides the latest revision of the species into subgenera and species-groups. These hierarchical arranged names are not supported by any strong evidence that these groups between the genus and species level represent monophyletic units. Regardless, much of the framework for these names has been in place for some time and they do serve as a welcome convenience for organizing the complex diversity of species. Rudy Kohout has been revising species by geographic areas and within specific subgenera and species-groups for more than two decades. Kohout's ongoing taxonomic work has largely followed the arrangement of Dorow (1995) with some modifications becoming necessary as more species have been described.

Aulacomyrma - key to Polyrhachis (Aulacomyrma) species
 * Polyrhachis porcata species-group
 * Polyrhachis dohrni species-group

Campomyrma
 * The gravis species-group and a key to species.
 * The micans species-group and a key to species.
 * The xiphias species-group and a key to species.

Chariomyrma

Cyrtomyrma  - key to Australian Cyrtomyrma species

Hagiomyrma - key to Polyrhachis Hagiomyrma group species

Hedomyrma 

Hemioptica  - species key

Hirtomyrma

Myrma 

Myrmatopa 

Myrmhopla 

Myrmothrinax

Polyrhachis
 * Key to the subgenus Polyrhachis workers (includes lamellidens and bihamata species-groups)