Heteroponera species groups

Heteroponera relicta group
Taylor (2015) added nine new species to this group and sorted the species into two complexes:

H. relicta species complex

 * Heteroponera darlingtonorum
 * Heteroponera lioprocta
 * Heteroponera monteithi
 * Heteroponera relicta
 * Heteroponera rhodopygea
 * Heteroponera viviennae
 * Heteroponera wilsoni

H. ecarinata species complex

 * Heteroponera ecarinata
 * Heteroponera pendergrasti
 * Heteroponera trachypyx

Morphology and Phylogeny
Antennal scrobes: Configuration of the frontal carinae and antennal scrobes differs within the relicta group, and largely defines the two species complexes recognized here.

(1) Fully configured scrobes are present in species of the H. relicta complex. The scrobes and frontal carinae are strongly developed and the carinae extended and apically curved to enclose the scrobes dorsally and posteriorly. A second carina delineates the scrobe ventrally above the eye. The scrobes may carry traces of incised sculpture, but their surfaces are often highly reflective, sometimes smoothly rippled or microsculptured. Each is partly divided longitudinally by a fine “dividing carina” which arises anteriorly and extends back to the level of the eye. These appear to provide partial separation or location of the scape and funiculus when the antennae are protectively folded in the scrobe. The dividing carinae separate lateral- and medial scrobal areas. This construction almost certainly represents the untransformed synapomorphic condition in the H. relicta group.

(2) Vestigial scrobes are present in H. ecarinata, H. pendergrasti and H. trachypyx, which constitute the Heteroponera ecarinata species complex. The frontal carinae are absent or minutely vestigial and the scrobes differentiated from the frons as relatively lightly-sculptured, slightly depressed tracts. A trace of the dividing carina is evident, and a very small stub possibly representing the ventral carina is often present against the eye. These structures are considered here to be vestigial relative to the relicta configuration, directly evidencing phylogenetic derivation.

(3) In further apparent derivation scrobes and associated carinae are absent or lacking in the H. leae group (see figs in Taylor, 2011), and in the H. imbellis group. In these species groups the frontal carinae do not extend far beyond the frontal lobes and the scrobal areas are neither sculpturally differentiated nor depressed.

None of the known Neotropical Heteroponera species have fully configured scrobes (Feitosa, 2011, and pers comm.). Most species illustrated in the online Antweb gallery and specimens of 6 Neotropical species in the Australian National Insect Collection (ANIC), Canberra, differ in detail from the ecarinata configuration. Nonetheless some specimens exhibit apparent vestiges of formerly more completely structured ancestral scrobes (traces of a dividing carina and/or an apparent remnant of the ventral carina).

Scrobes possibly homologous with the fully-configured structure in the H. relicta group (but less elaborate in detail) are present in the neotropical genus Acanthoponera (see Antweb illustrations), which is a recognized sister taxon to Heteroponera (Brown, 1958) with recent phylogenetic accreditation (Moreau et al., 2013). Here the frontal carinae are less-well-developed than in the H. relicta group, the dividing carina is clearly present and the ventral carina unrepresented.

Humeral Epaulets: Most species of the H. relicta-group have the humeral shoulders quite strongly angulate or subdentate when viewed from above and shaped lateroventrally beneath to form a curved, smooth surface against which the adjacent posterolateral occipital corner of the cranium is nested when the head is depressed, as in the protective rolled position adopted by disturbed workers (see below under “bionomics”) and in the cephalic/ mesosomal posture of many cabinet specimens. These structures are identified here as “pronotal - ” or “humeral epaulets” and discussed under relevant species below. Similar protective juxtaposition of the head and mesosoma occurs in H. leae (pers obs.) and probably in the other leae-group species, but epaulet structures are barely developed.

Gastral II Spurs: These are a bilateral pair of minute, anterolaterally-directed lateral cuticular denticles or forwardly-hooked nodules on the exoskeleton of the second gastral segment (abdominal lV) at the posterior rim of the girdling constriction which divides its fused cylindrical structure into (pretergital) anterior and (tergital) posterior sections. They appear to act as stops limiting the depth of insertion of the pretergital section into the bowl-shaped preceding gastral I segmental exoskeleton by engaging its posterior rim. Gastral II Spurs are an apparent synapomorphy found in only two known species, H. monteithi and H. lioprocta. Their presence might relate to the exposed gastral II exoskeleton being tapered, narrowing posteriorly, and thereby perhaps at risk of excessive insertion. Such spurs are unknown in Neotropical Heteroponera (Rodrigo Feitosa, pers comm.) and other Australasian species.

Reproductive females known in the H. relicta group are permanently wingless with the mesosoma reduced in structure and worker-like and the gaster more or less physogastrically inflated and enlarged (Wheeler, 1915). Their cephalic and mesosomal sizes and proportions are close to those of the workers. Those of H. relicta were acknowledged by Peeters (1991, 1997) to be ergatoid gynes derived phylogenetically from winged–gyne ancestry, rather than gamergate workers. Gynes of three of the ten species (lioprocta, monteithi, relicta) are described below. Future records will likely discover that ergatoid gynes are typical for all species of the H. relicta group, in which collected or observed colonies appear to be monogynous. Alate gynes are present in at least one species of the related H. leae group (H. crozieri, illustrated by Taylor, 2011, figs 9–12—see also another specimen identified incorrectly as H. leae in the Antweb gallery). They are relatively very large compared to conspecific workers (see measurements for H. crozieri in Taylor, 2011).

If ergatoid gynes are found present in those H. relicta-group species which have reduced antennal scrobes it would be arguable that (1) reduction of the antennal scrobes occurred later in phylogenetic succession than the acquisition of ergatoid gynes (assuming reasonably that the latter are homologous across the H. relicta-group and not homoplasious), and that (2) the scrobes were reduced independently (to the point of loss) in the phylogeny of the H. leae group, in which the alate gyne caste persists (assuming that the presence of scrobes is an apomorphic character in Heteroponera). Ancestral scrobes in the genus at large need not have been as complex as in the H. relicta group, in which elaborated scrobal structure could be phylogenetically derived from less-complex ancestry, perhaps with scrobes more similar to those of Acanthoponera.

Distribution
Heteroponera relicta-group species seem more liable than other ants in the Queensland wet tropics to segregate at morphospecies level in relatively small areas of geographical isolation. This might relate to their apparent restriction to “islands” of rainforest habitat and their limited ability for physical dispersal and correlated interpopulational gene flow due to the possible (or likely?) universal presence of ergatoid gynes. For this reason they can be considered equivalent to flightless insects such as those discussed in the present biogeographical context by Yeates et al. (2002).

H. leae group records are not known from the collective range of the H. relicta-group, but H. crozieri is present in rainforest at Eungella (Dalrymple Heights), west of Mackay (Taylor, 2011), about 80km SSW of Brandy Creek, the type-locality of H. monteithi and most southern known locality for the H. relicta group. Species of the H. imbellis group are broadly sympatric with the H. relicta group in the Queensland Wet Tropics, as they are elsewhere with species of the H. leae group.

Biology
Feitosa (2011 & pers comm.) reports that known gynes in Acanthoponera are alate and have appropriate morphology. Their colonies are monogynous. He also reviewed alternative gyne morphologies among Neotropical Heteroponera species. Some such as H. dolo (Roger) and H. robusta Kempf have alate gynes. The gynes of Heteroponera carinifrons, H. inca and H. monticola (species, as noted above, known only from the western side of the Andes) are permanently wingless, with very inflated gasters (a condition identified as “physogastric” by Feitosa) and thus very similar to, though not necessarily directly homologous with, those of the Australian H. relicta group. Their colonies are also strictly monogynic. In most Neotropical species such as H. mayri Kempf, H. dentinodis (Mayr) and H. panamensis (Forel) the egg-laying function in colonies is divided between alate “primary” gynes, and “secondary” ergatoid gynes with morphology similar to workers except for the presence of ocelli and an enlarged mesosoma. In these species colonies are normally polygynous. There are no known gamergates in Neotropical (or Australian) Heteroponerinae. Gynes of Acanthoponera and Neotropical Heteroponera species are comprehensively illustrated by Feitosa (2011). It is arguable that ergatoid gynes first evolved as secondary reproductives (as apparently in H. mayri and H. panamensis) and later phylogenetically superseded alate gynes, which were arguably lost phylogenetically in some lineages like those of H. inca, H. monticola, H. carinifrons and the Australian endemic H. relicta group. A comparable history might be represented in the Australia-based ectatommine genus Rhytidoponera Mayr, where most species appear to depend exclusively upon gamergate (mated worker) reproductives, but several are known in addition to utilize reproductive alate gynes (Ward, 1986).

Most specimens reviewed here were collected as foraging workers taken in Berlese funnel samples, siftings of forest leafmould or moss, or stick brushings. Colonies of Heteroponera relicta, Heteroponera rhodopygea and Heteroponera lioprocta are found in and under rotting wood fragments or logs on the forest floor, or in soil under stones (pers obs.). Little is known of other species. Living H. relicta and H. rhodopygea frequently carry numbers of very small phoretic mites clustered in their antennal scrobes or elsewhere. Their remains may be present on cabinet specimens, including those of other species. Their systematics and phylogeny relative to the host ants would likely be of scientific interest. Heteroponera workers roll their bodies ventrally when disturbed, with the antennae enclosed in the scrobes, and are sometimes almost imperceptible against the substrate in this position. The structure of the head/ mesosoma interface including the pronotal epaulets, and the gastral configuration are likely adapted to this purpose.

The major rainforests of these areas have long been historically isolated from one another by dry sclerophyll forest and/or lower elevation river gorges (Williams, 1997). They have been collectively described as constituting a mesotherm archipelago in a sea of tropical lowlands (Nix, 1991). Some tracts, particularly on the Atherton Tableland, are fragments of formerly more continuous forest cover fractionated by human forest clearing, mostly since European settlement (e.g. Lake Eacham and Lake Barrine National Parks).