Insights into Ant Phylogeny from Mitogenomic Sampling Across All Extant Subfamilies

Ants (Formicidae) represent an ecologically dominant terrestrial lineage, yet resolving their deep phylogenetic relationships requires extensive taxon sampling. In this study, we sequenced four new ant mitochondrial genomes (mitogenomes) and assembled 34 others from public genomic data. Combining these with published sequences yielded a broad dataset of 215 species representing all 17 recognized subfamilies. Three of the four newly sequenced mitogenomes were complete and contained the canonical set of 37 genes, all exhibiting a strong A+T bias. Notably, three species displayed distinct gene rearrangements relative to the ancestral insect mitogenome. Phylogenetic inferences based on nucleotide and amino acid datasets supported the monophyly of Formicidae and all subfamilies represented by multiple species. Under concatenated Maximum Likelihood analyses, Martialis heureka (Martialinae) was recovered as the sister group to all other extant ants, with the remaining lineages dividing into poneroid and formicoid clades. Within these groups, key relationships were firmly validated, including Aneuretinae + Dolichoderinae, Heteroponerinae + Ectatomminae, Apomyrminae + Amblyoponinae, and an intriguing Myrmicinae + Formicinae clade that diverges from nuclear phylogenomic topologies. However, extensive sensitivity assessments (including likelihood mapping, alternative partitioning, and Bayesian inference) revealed underlying signal conflict and dataset dependence regarding the exact placements of Martialinae and Leptanillinae. These deep-node uncertainties are driven primarily by substitution saturation and compositional bias. Overall, while mitogenomic data provide powerful resolution for most intersubfamilial relationships, fully resolving the deepest ant divergences will require integrating nuclear genome-scale data under sophisticated models accounting for sequence heterogeneity.