Ectatomminae

Camacho, G. P., Franco, W., Branstetter, M. G., Pie, M. R. & Longino, J. T., 2022, UCE Phylogenomics Resolves Major Relationships Among Ectaheteromorph Ants (Hymenoptera: Formicidae: Ectatomminae, Heteroponerinae): A New Classification, Insect Systematics and Diversity 6 (1), pp. 1-20 : 7-8

Ectatomminae Genera

Our concatenated and species-tree analyses recovered a well resolved and highly congruent phylogeny for Ectatomminae , while identifying possible incongruences that need to be further investigated (Figs. 2, Supp Figs. S1–S 5 [online only]). These results, based on our 2,520 UCE loci dataset, are congruent with prior research that suggests that having a greater number of loci is beneficial ( Borowiec et al. 2015, Branstetter et al. 2017), although it remains unclear how many loci are necessary to resolve phylogenetic relationships. However, it has long been recognized that simply increasing the amount of data can exacerbate systematic bias in phylogenetic estimation ( Phillips et al. 2004, Philippe et al. 2011, Borowiec et al. 2015) and that to improve phylogenetic inference data quality is key ( Borowiec et al. 2015). We showed that, despite the incongruencies found among different datasets for some nodes (see red dots in Fig. 2), all alternative topologies are supported by good-quality data with strong phylogenetic signal. However, recoding nucleotides to RY characters suggests that composition bias may be contributing to support for nodes where gene-tree incongruence is pervasive (Supp Fig. S6 [online only]). RY-coding reduces such biases and increases the signal on internal branches relative to external, increasing phylogenetic signal in mitochondrial genome data ( Phillips and Penny 2003). Nevertheless, using RY-coding reduces the dataset size and, as shown in Supp Fig. S6 (online only), nodes that are incongruent between the nucleotide and RY-character data are supported by relatively few loci, which may suggest that dataset size may be important for resolving phylogenetic relationships in Ectatomminae . If loci are discordant, it is expected that numerous additional markers are necessary to generate a robust tree, allowing for an amplification of phylogenetic signal with the increase of the amount of data ( Camacho et al 2019).

Previous research has shown that taxonomic balance within a data set has a large impact on phylogenetic results ( Branstetter et al. 2017), emphasizing the importance of both broad taxonomic sampling (i.e., covering taxonomic disparity and geographic coverage) and taxonomic evenness across samples (i.e., having comparable samples sizes among the groups, according to their diversity).The fact that we recover alternative hypotheses for some nodes may suggest that a larger sampling of those groups might shed light on their relationships in the future. Despite the fact that our phylogeny includes a broad representation of Heteroponera , the addition of H. inca to the phylogeny could help elucidate the position of H. monticola , since both species seem to be morphologically similar and possibly closely related. Regarding the relationship among Rhytidoponera and Ectatomma , even though the 26 species of Rhytidoponera included

(CASENT0281223), Stictoponera biroi (CASENT0281519), Rythidoponera metallica (CASENT0172345), Ectatomma lugens (USNMENT00445341), Heteroponera brounii (CASENT0172105), Acanthoponera mucronata (CASENT0173540), and Boltonia microps (CASENT0173544). Images by April Nobile, Jeffrey Sosa-Calvo, Zach Lieberman,Will Ericson, Michael Branstetter, and Estella Ortega; available from www.antweb.org (Antweb 2021).

in our phylogeny represents a broad sampling for the genus, there are 104 species currently described and a more complete phylogeny for the genus could provide increased support. Similarly, a larger sampling of Holcoponera , Alfaria , and Poneracantha species could elucidate the relationships among those genera. Nevertheless, our results recover a robust and fully resolved topology that we discuss below through an in-depth discussion of the morphological hypotheses available for the group.