Related species and phylogenetic affinities
Boloponera ikemkha is readily distinguished from its only known congener by the highly polished torular lobes and lack of standing hairs on the scapes, mesosoma, petiole and first gastral tergite. The discovery of a second species in the genus has confirmed the consistency of several characteristics, as well as enabling the recognition of new autapomorphies, that are of significance in elucidating relationships with other closely related genera, as discussed below.
It is pertinent first to correct some errors in previous discussions of the relationship of Boloponera with other members of the Plectroctena genus-group. Firstly, Schmidt and Shattuck (2014) state that Boloponera has abundant short pilosity but no pubescence, but both B. vicans and B. ikhemka do in fact possess pubescence; B. ikhemka, however, almost entirely lacks standing pilosity. Secondly, Schmidt and Shattuck (2014), as well as Fisher (2006), described the propodeum of both Boloponera and Plectroctena as having posterolateral margins expanded into lamellae. However, reference to the descriptions in Bolton (1974), and inspection of images on AntWeb as well as specimens of three species represented in the AFRC, indicate that the majority of Plectroctena species have what can at most be described as a distinctly marginate propodeum (e.g. P. strigosa, Emery, illustrated here in Figure 3C, D). Some species have propodeal ridges that could be described as laminae but only a few could be considered to have lamellae (which are thin and generally translucent) developed to a significant degree. In contrast, both B. vicans and B. ikhemka have well-developed, thin and translucent lamellae that run the entire length of the margin between the lateral propodeum and the declivity. The presence of lamellae thus cannot be considered a synapomorphy linking Plectroctena with Boloponera. Thirdly, Schmidt and Shattuck (2014) state that, as in Loboponera, the anepisternum in Plectroctena appears fused to the mesonotum and metapleuron. While this may be the case for a few of the 16 Plectroctena species illustrated on AntWeb, in the majority of these, and in all three species represented in the AFRC, the anepisternum is moderately to well-defined by impressed sutures and is more distinct from the mesonotum and metapleuron than it is from the katepisternum. This character therefore cannot be considered a synapomorphy linking Plectroctena and Loboponera. In B. vicans and B. ikemkha, the anepisternum is contiguous with the metapleuron, but remains separated from the mesonotum by a weak but distinct suture.
Fisher (2006) noted similarities between Boloponera and Plectroctena, but also highlighted several differences and concluded that Boloponera should be excluded from the Plectroctena genus-group (Loboponera + Plectroctena + Psalidomyrmex) as defined by Bolton and Brown (2002). However, based on molecular and morphological evidence, Schmidt and Shattuck (2014) subsequently placed Boloponera within their expanded Plectroctena genus-group (Boloponera + Centromyrmex + Dolioponera + Feroponera + Loboponera + Plectroctena + Psalidomyrmex), as sister to Plectroctena and/or Loboponera, and suggested that Boloponera might even be nested within Plectroctena. The discovery of B. ikhemka has however confirmed several of the previously recognised differences between Boloponera and Plectroctena including:
1. the lack in Boloponera of a highly modified mandibular articulation (incorporating differences in the structure of the mandibular articulation itself as well as of the clypeus and genae);
2. the presence of both an apical and a preapical tooth on the mandible in Boloponera, as well as an additional tooth on the masticatory margin close to the basal tooth. Plectroctena have a blunt or truncated mandible with neither an apical nor a preapical tooth and, when present, the masticatory margin tooth is widely separated from the basal angle or tooth;
3. the absence in Boloponera of a longitudinal groove on the inner half of the dorsal mandibular surface running parallel to the masticatory margin;
4. the absence of a mesofemoral gland in Boloponera, and
5. the differing form of the anterior disc of the petiole in Boloponera.
Schmidt & Shattuck (2014: 162) downplayed the significance of the last character and, based on the rather indistinct published image of this feature in B. vicans, suggested that the shape of the median depression in the petiolar articulatory surface in Boloponera is more similar to that of Plectroctena than is its shape in Loboponera. However, this cannot be said of B. ikhemka, in which the median depression is very distinct from that of Plectroctena (compare Figure 2F with figures 28–29 of Fisher, 2006) and is more like a foreshortened version of that found in Loboponera (see figures 25, 26 in Fisher, 2006). The median depression in Boloponera is broadly rounded anteriorly and approximately equal in width to the horseshoe-shaped articulatory strip, which perfectly matches Bolton & Brown’s (2002) description of the ancestral structure. In strong contrast, in Plectroctena (and Psalidomyrmex), the median depression is reduced in width anteriorly to a very narrow groove by inward expansion of the articulatory surfaces (as described in Bolton & Brown, 2002), with an overall arrowhead-like appearance in most species. Loboponera appears intermediate, with the median depression significantly narrower than the lateral strips, but approximately uniform in width and by no means as compressed anteriorly as in Plectroctena. On current evidence I therefore disagree with Schmidt & Shattuck’s (2014) assessment of this character and consider Boloponera to display a more ancestral form, with Loboponera showing slight modification and Plectroctena + Psalidomyrmex with the most derived state.
Boloponera, Loboponera and Plectroctena share a longitudinal metafemoral gland as a synapomorphy, but Boloponera lacks the similar mesofemoral gland found in both Loboponera and Plectroctena. There is no visible groove in Boloponera, in which the gland is indicated merely by a thinning of the cuticle. However, the development of the grooves appears variable in the Plectroctena material I have inspected, and thus not much weight can be placed on this character; only the presence/absence of the mesofemoral glandular structure is of significance here and again Boloponera appears to display the less-derived state.
While Bolton and Brown (2002) considered the strongly arched A4 tergite autapomorphic in Loboponera, several Plectroctena (e.g. P. dentata Santschi, P. strigosa and P. thaui Fisher) have the A4 moderately arched (tergite approximately three times the length of the sternite) and in at least one (P. laevior Santschi), the A4 is even more strongly arched (tergite more than four times the length of the sternite) than that of some Loboponera, while most other Plectroctena species have the A4 slightly arched (tergite roughly twice the length of the sternite). The variability of this character within Plectroctena means that in addition to not being autapomorphic for Loboponera, it also cannot be considered a synapomorphy for Plectroctena + Loboponera. In Boloponera the A4 tergite (Figure 1B, C) is straight or weakly arched anteriorly and only weakly arched posteriorly, the tergite being approximately 2.5 times the length of the sternite. Although I believe that relatively little weight can be afforded to the character, Boloponera does again appear to have the least derived form, with Plectroctena and Loboponera showing increasing modification.
In contrast, the extreme protrusion of the torular lobes beyond the posterior clypeal margin, which itself is extended anterad of the anterior clypeal margin and overhanging the mouthparts, appears unique to Boloponera and hence more derived. While the anterior clypeal margin also overhangs the mouthparts in Loboponera, the torular lobes do not extend anterad of the clypeus as they do in Boloponera. In Plectroctena the torular lobes are less developed and the clypeus is approximately vertical, not overhanging the mouthparts. Movement of the antennal scapes in Boloponera (as well as Loboponera) is highly constrained by the enlarged torular lobes. Despite the complex shape of the basal portion of the scape (see Figure 2C), which allows some additional dorsal movement, the entire scape is prevented from rising above a plane at approximately the level of the top of the lobes. The adaptive significance of this is unknown.
The arrangement of clypeal setae described here for B. ikemkha appears identical in B. vicans and, at least within the Plectroctena genus-group, seems to be unique to the genus. Inspection of AntWeb images of all other Afrotropical members of the Plectroctena genus-group sensuSchmidt and Shattuck (2014), including 10 Centromyrmex, one Dolioponera, one Feroponera, nine Loboponera, 16 Plectroctena and six Psalidomyrmex species, did not reveal the pattern repeated. Although Feroponera ferox Bolton & Fisher, Plectroctena anops Bolton, Loboponera nobiliae Fisher, Centromyrmex bequaerti (Forel), C. praedator Bolton & Fisher and C. sellaris Mayr each have a moderately convergent pair of clypeal setae, which in two cases meet at the tips, none have setae which cross substantially proximal of the tips as in Boloponera and in no case are there two divergent pairs in addition to the convergent pair. This also appears to be a character for which Boloponera displays the most modified state.
In both Loboponera and Plectroctena, as in all other members (except Feroponera) of Schmidt & Shattuck’s (2014) expanded Plectroctena genus-group, the metapleural gland opens laterally and is clearly visible in profile view, while both Boloponera species have the metapleural gland bulla strongly laterally expanded ventrally (visible in ventral view in Figure 2F and in dorsal view in Figure 3B), resulting in a dorsally oriented orifice (Figure 3B). This is obscured in profile view (Figure 3A), especially posteriorly, by the dorsad extension of the ventral flap on the metapleural gland opening; in this character Boloponera again shows an apparently more derived state than either Loboponera or Plectroctena.
The prora in Boloponera is fairly inconspicuous, while in Plectroctena the prora is evident as a much more prominent projection on the anterior face of the first gastral sternite; in Loboponera the prora is often exceptionally well developed and in several species forms a very prominent antero-ventral projection.
In summary, Boloponera and Loboponera are similar with respect to petiole articulation, the expansion of the torular lobes, orientation of the clypeal surface, mandibular articulation and absence of a dorsal groove on the mandible. Loboponera and Plectroctena are similar in having the torular lobes not overhanging the mouthparts, configuration of clypeal setae, presence of a mesofemoral gland, position of the metapleural gland opening, development of the prora and degree of arching of A4. Plectroctena and Boloponera are similar in the absence of posterolateral head flanges and shape of the mandibles, although the latter may not be equivalent in view of the extreme development of other mandibular characters in Plectroctena.
Thus there appear to be fewer characters suggesting a close relationship between Boloponera and Plectroctena than there are linking Boloponera with Loboponera or Loboponera with Plectroctena; this assessment implies that Boloponera and Plectroctena are less similar to each other than either is to Loboponera, suggesting an intermediate position for the latter.
The characters discussed above strongly support the retention of Boloponera as distinct from Plectroctena and suggest that Boloponera is sister to Loboponera and Plectroctena rather than being closer to or even nested within Plectroctena. Nesting of Boloponera within Plectroctena would imply either that, in addition to the appearance of several new unique characteristics in Boloponera, there had been secondary loss (with reversion to the ancestral state) of three Plectroctena autapomorphies relating to the mandibles and head capsule (mandible dentition, dorsal groove and articulation), or alternatively that Plectroctena is paraphyletic and that these same adaptations had arisen independently in two separate lineages. The latter is highly improbable, but even the former seems unlikely and the hypothesis that Boloponera split from an ancestral line, before the unique mandibular configuration of Plectroctena had arisen, is a far more parsimonious explanation.
Finally, were Boloponera to be included within Plectroctena, which ranges in total length from 7 mm to more than 18 mm, they would be by far the smallest representatives of the genus, both species being barely more than half the length of the smallest Plectroctena currently known, and hence resulting in a very disjunct size range. Loboponera includes species intermediate in size between Boloponera and Plectroctena, with lengths ranging from about 3–7 mm, which may provide further (if weak) support for Loboponera being intermediate between Plectroctena and Boloponera.
The question remains, which of the three genera is closest to the ancestral form and which is the most derived? The characters discussed here are equivocal, with each genus displaying some that appear most derived and others that appear least derived, suggesting that Boloponera, Loboponera and Plectroctena have all undergone substantial modification since separation of their respective lineages; more comprehensive DNA phylogenetic analysis will most likely be required to resolve this question.
After the above reassessment of characters linking and separating Boloponera, Loboponera and Plectroctena, only the presence of a longitudinal metafemoral gland remains as a synapomorphy linking all three genera, other synapomorphies each being shared by only two of the three genera.
Conservation and threats
Boloponera is one of the most rarely encountered ant genera in Africa, with each of the two known species having been recorded only once to date. Whether this is an indication of true rarity, or an artefact of an extremely cryptic lifestyle, may be difficult to determine with certainty. However, Hypoponera, a related genus of cryptic, largely subterranean ants, with workers similar in size to or, more commonly, smaller than Boloponera, has 51 described species endemic to the Afrotropical region (Bolton and Fisher 2011). While some of these are also known from only a single collection, many are widespread and are frequently encountered in soil at depths similar to that at which B. ikemkha was found. During a survey of leaf litter ants in Ghana, Belshaw and Bolton (1994) collected 2410 specimens of Hypoponera representing six species, with a range from six to 1828 individuals per species. Unless Boloponera are very significantly more cryptic and deeply subterranean, this suggests that they are far rarer than Hypoponera, probably by at least 2–3 orders of magnitude.
The type locality of B. ikemkha, ca. 5 km east of the confluence of the Groot and Klein Dwars Rivers, lies approximately 400 m from the nearest boundary of the proposed new tailings storage facility for TRPM. Very little similar habitat lies within the proposed TSF site. However, approximately 2 km of this riverine fringe vegetation lies within 350–900 m down-slope of the proposed TSF. A further 1 km (within which the type locality lies) falls within 350–700 m of the proposed TSF but is separated from it by a low (10–30 m high) ridge. Dust clouds emanating from existing tailings dams in the region are frequently observed rising well over 500 m into the air and dispersing over distances of at least 8 km. It is thus clear that significant dust pollution can be expected close to the proposed new TSF unless far more effective dust control is implemented in the new facility. In addition, while possible seepage of contaminated water could not directly affect the B. ikemkha type locality due to the intervening ridge, such pollution could impact on the 2 km section of likely habitat down-slope of the proposed TSF, which thus poses a potential threat to the survival at least of part of the local population of the species; it is thus imperative that effective water management and dust control measures be implemented.
The Groot and Klein Dwarsrivier valleys and surrounding areas are subject to intense mining and prospecting pressure, with at least eight operational mines within an eight km radius of the B. ikemkha type locality and more mines further afield. Many more mines are planned for this region, which forms part of the Bushveld Igneous Complex, recognised to contain the richest deposits of platinum-group elements (PGE) in the world, representing 80% of known PGE reserves (Cawthorn 2010). There is also considerable habitat transformation due to increasing human populations (in part caused by an influx of job-seekers attracted by the mines) with concomitant increases in subsistence agriculture and livestock grazing. The cumulative effect of anthropogenic impacts and mining operations on natural habitats in the region may become highly significant and could pose a serious threat to the survival of B. ikemkha, as well as that of other species endemic to Sekhukhuneland. Habitat transformation and fragmentation would be especially significant for B. ikemkha if the species does, as seems most likely, have ergatoid queens and hence very limited dispersal abilities.
Application of the IUCN Red List Criteria (IUCN 2012) to the currently available data would indicate that B. ikemkha should be classified as Critically Endangered (CR B1ab(iii)+2ab(iii)), since it has been recorded from only a single locality which is currently under threat and other potential habitat in the region is also under threat. However, the species could alternatively be considered as Data Deficient and in urgent need of investigation. Additional surveys could either confirm a CR status or result in a lower threatened category such as Endangered (EN) or Vulnerable (VU) being assigned, but it is highly unlikely that a category lower than VU will result. Formal Red List assessments of the conservation status of species such as B. ikemkha may enable action to be taken to protect sufficient areas of natural habitat to ensure their continued survival and such assessments should be undertaken as a matter of urgency.