Ponerinae

Schmidt, Chris, 2013, Molecular phylogenetics of ponerine ants (Hymenoptera: Formicidae: Ponerinae), Zootaxa 3647 (2), pp. 201-250 : 209-212

Monophyly of Ponerinae and its tribes and genera

The monophyly of Ponerinae was strongly supported by analyses of the ALL_DATA matrix (BPP = 1.00, BS = 0.98; Fig. 3 ), confirming the results of previous molecular phylogenetic studies with more limited sampling of ponerine taxa (e.g., Moreau et al., 2006; Brady et al., 2006) and confirming predictions from morphological evidence. In his morphological synopsis of Ponerinae , Bolton (2003) noted two synapomorphies for the subfamily: fusion of the toruli to the frontal lobes, and the shape of the outer borders of the frontal lobes, which form ”simple short semicircles or blunt triangles and in full-face view have a distinctly pinched-in appearance posteriorly”. An examination of the inferred phylogeny of Formicidae and the subfamily-level morphological synopses of Bolton (2003) suggest an additional possible synapomorphy for Ponerinae , the loss of tergosternal fusion of the petiole (C. Schmidt and S. Shattuck, in prep), though this trait could conceivably be symplesiomorphic, with repeated evolution of a fused petiole in other poneroid lineages. All current ponerine genera not sampled in this study ( Asphinctopone , Belonopelta , Boloponera , Dolioponera , Feroponera , and Promyopias ) display these traits and their placements within Ponerinae and Ponerini are not in doubt, so it is highly unlikely that their inclusion in this study would have rendered Ponerinae non-monophyletic.

Phylogenetic analysis of the ALL_DATA matrix ( Fig. 3 ) strongly supported the monophyly of Platythyreini (Platythyrea) (BPP = 1.00, BS = 1.00). The status of the genus Eubothroponera as a junior synonym of Platythyrea (Brown, 1952, 1975) is here confirmed with molecular data for the first time: P. t u r n e r i, though never formally included in Eubothroponera despite its obvious close relationship to members of that former genus, is nested within Platythyrea with strong support (monophyly of Platythyrea excluding P. t u r n e r i: BPP <0.0001; BS = 0.002). Members of Eubothroponera have several morphological traits that differ from most other Platythyrea species, including coarser sculpturing, denser standing pilosity but sparser pubescence, and a relatively low helcium. The latter character approximates the condition found in most Ponerini , but the results of this study imply that low helcia were independently derived in Platythyrea and Ponerini .

The monophyly of the diverse tribe Ponerini was not supported in either Bayesian or ML bootstrap analyses (BPP <0.0001; BS <0.001), though Ponerini monophyly could not be rejected by an SH test (p -value = 0.661). The non-monophyly of Ponerini stems from its inclusion of Thaumatomyrmecini (Thaumatomyrmex) , a result also recovered by Brady et al. (2006). The phylogenetic position of Thaumatomyrmex has been unclear due to its bizarre structure (including pitchfork-like mandibles, widely separated frontal lobes, and unusually convex eyes; Kempf, 1975). Bolton (2003) noted that Thaumatomyrmex shares several apomorphies with Ponerini , including vestigial male mandibles and articulation of the petiole low on the anterior face of the first gastral segment, but maintained its separate tribal status. The genus lacks a principal synapomorphy of Ponerini (the narrow insertion of the clypeus between the frontal lobes, and related characters; Bolton, 2003), but this character is highly modified in Thaumatomyrmex , rendering it uninformative about the phylogenetic position of the genus. Morphological data therefore imply at least a sister relationship between Thaumatomyrmex and Ponerini , and are consistent with its placement within Ponerini . The molecular results reported here suggest that Thaumatomyrmex is simply a highly derived member of Ponerini , its strange morphological traits secondarily derived as a result of extreme prey specialization (Brandão, 1991).

Among sequenced taxa, all sampled ponerine genera were strongly inferred to be monophyletic (BPP and BS = 1.00 for all genera except Myopias , for which BPP and BS both = 0.99, and Thaumatomyrmex , for which BS = 0.99) except Pachycondyla and Odontomachus . Pachycondyla was strongly inferred to be non-monophyletic (BPP of monophyletic Pachycondyla <0.0001; BS <0.001; SH p- value <0.00001), as members of this genus were broadly distributed across the phylogeny of Ponerini . The sampled Pachycondyla species represent numerous distinct lineages which largely correspond to the boundaries of Pachycondyla “subgenera,” though Bothroponera , Mesoponera and Pseudoponera are themselves non-monophyletic ( Bothroponera : BPP <0.001, BS = 0.014, SH p - value = 0.768; Mesoponera : BPP <0.0001, BS <0.001, SH p -value <0.00001; Pseudoponera : BPP <0.0001, BS <0.001, SH p -value <0.001).

Pachycondyla is the end product of heavy synonymization at the genus level, due primarily to the work of W. L. Brown, who was in the process of revising ponerine taxonomy at the time of his death. Brown’s formal revision remains unpublished, but his numerous planned synonymizations have been published elsewhere without supporting justification (Brown, 1973; Snelling, 1981; Hölldobler and Wilson, 1990; Bolton, 1994; see also Mackay and Mackay, 2010) and in practice have been generally accepted by the scientific community. The monophyly of Pachycondyla is doubtful even in the absence of molecular data because the genus displays substantial morphological, ecological and behavioral diversity (reviewed in C. Schmidt and S. Shattuck, in prep.), is defined only by morphological symplesiomorphies such as triangular mandibles and paired metatibial spurs, and contains several lineages that show clear morphological similarities to other ponerine genera. This study confirms the extensive non-monophyly of Pachycondyla and stresses the need for a comprehensive taxonomic revision of ponerine genera (provided in C. Schmidt and S. Shattuck, in prep.).

Analysis of the ALL_DATA matrix was ambiguous about Odontomachus monophyly, with three resolutions: O. coquereli Roger sister to Anochetus (BPP = 0.58, BS = 0.37), Odontomachus monophyletic (BPP = 0.29, BS = 0.41), or O. coquereli sister to Anochetus + other Odontomachus (BPP = 0.13, BS = 0.22). To further address this question, a preliminary expanded data set was produced which included sequences for six additional Anochetus taxa and 18 additional Odontomachus taxa, plus data for the mitochondrial gene cytochrome oxidase I (COI), but excluding abdA, EF1αF1, and EF1αF2 (C. Schmidt, unpublished data). Phylogenetic analyses of these data supported reciprocal monophyly of Odontomachus and Anochetus (results not shown), but some phylogenetically important Anochetus species (such as members of the A. gladiator (Mayr) and A. cato Forel species groups; Brown, 1976, 1978) were not sampled and their inclusion in future phylogenetic analyses could still reveal the nonmonophyly of either genus.

Figures

FIGURE 3. Consensus phylogeny of Ponerinae inferred by MrBayes analysis of ALL _ DATA matrix with CAD included for Harpegnathos, showing mean branch lengths. Most outgroups are not shown (the full tree including all outgroups is given in Supp. Fig. S 1). Genus groups within Ponerini are demarcated by branch color and by the labels on the right. Branches are colored for the crown groups of all genus groups except Harpegnathos, whose stem branch is also colored to ease visualization. Clade support is indicated by solid (MrBayes BPP = 1.00) and empty (0.95 <BPP <0.99) circles and solid (RAxML BS = 1.00) and empty (0.95 <BS <0.99) triangles. Pseudo.: Pseudoponera.