Phylogeny of Termites Inferred from Mitogenomic Data with a Large Taxon Sampling

Background: Termites (epifamily Termitoidae) represent the oldest known lineage of eusocial insects, yet their higher-level phylogenetic relationships remain contentious. Methods: We sequenced ten new termite mitochondrial genomes (mitogenomes) from Australia and China and integrated them with published data to assemble a comprehensive dataset of 560 species, including 553 termite ingroup species representing all 13 extant families and seven Cryptocercus outgroups. Phylogenetic relationships were inferred using maximum likelihood analyses of mitochondrial protein-coding genes and ribosomal RNA genes under amino acid and nucleotide datasets. Results: All newly sequenced mitogenomes contained the canonical set of 37 mitochondrial genes, exhibited a pronounced A+T bias, and retained the ancestral insect gene order. Mastotermes darwiniensis (Mastotermitidae) was consistently recovered as the sister lineage to all other termites. Among the basal lineages, the relationship ((Hodotermopsidae + Stolotermitidae) + (Archotermopsidae + Hodotermitidae)) received strong support. Kalotermitidae was recovered as monophyletic and as the sister group to Neoisoptera. The traditional Rhinotermitidae was non-monophyletic, and the analyses supported the revised family-level classification that restricts Rhinotermitidae to Rhinotermitinae while recognizing Psammotermitidae, Termitogetonidae, and Heterotermitidae as separate lineages. Within Termitidae, Macrotermitinae was recovered as the sister group to the remaining termitid lineages, whereas Sphaerotermitinae formed a distinct lineage, consistent with its recognition as a separate subfamily. Conclusions: Our results substantially expand mitogenomic sampling across termites and provide further support for recent family- and subfamily-level taxonomic revisions. The inferred relationships broadly agree with recent molecular studies, while some unresolved relationships require denser taxon sampling and nuclear genomic data.