Molecular systematics and biogeography of an Australian soil‐burrowing cockroach with polymorphic males, Geoscapheus dilatatus (Blattodea: Blaberidae)

Abstract

An iconic group of arid‐adapted insects is the Australian soil‐burrowing cockroaches (Blaberidae: Geoscapheinae), large, wingless insects that evolved burrowing behaviour and associated forms in parallel from wood feeding ancestors in the subfamily Panesthiinae. A particularly problematic taxon within the Geoscapheinae is Geoscapheus dilatatus (Saussure, 1864), which might represent a species complex and whose delimitation has been complicated for decades by the species harbouring polymorphic males. Males can be divided into two main morphs: individuals possessing horn‐like protrusions on the anterior margin of the pronotum (‘tuberculate’) and those without these characters (‘non‐tuberculate’). A less common, third form consists of individuals that possess tubercles but are far larger than other tuberculate males and occur solely to the north of the species' distribution (‘atypical’ tuberculates). Here, we make use of whole mitochondrial genomes and nuclear ribosomal RNA data from individuals across the range of G. dilatatus to conduct the first phylogenetic analysis of this species to date. We recover all tuberculate males (including atypical forms) as monophyletic and the derived form of G. dilatatus, having evolved only once in this species, whereas non‐tuberculate forms are paraphyletic. Fossil‐calibrated molecular clock analysis revealed that the divergence between these two forms occurred during the late Miocene approximately 6.7 Mya, concurrent with an expansion of the continent's drier biomes. Environmental niche modelling suggests that tuberculate male forms are more climatically tolerant than their more restricted non‐tuberculate counterparts, and both forms' predicted fundamental niches are strongly limited by rainfall. Three species delimitation analyses implemented here failed to consistently delimit G. dilatatus beyond a single species. Ultimately, population genetic approaches paired with additional sampling will be necessary to determine these findings more concretely, but at present, we do not consider the results presented here sufficient to delimit G. dilatatus based on morphological differences found in the species' polymorphic males alone.