Two new endemicspecies ofAbantiades Herrich-Schäffer (Lepidoptera: Hepialidae) from Kangaroo Island, Australia Author Moore, Michael D. 0000-0002-8796-3330 Biological and Earth Sciences, South Australian Museum, SA 5000, Australia. michael.moore@samuseum.sa.gov.au Author Beaver, Ethan P. 0000-0002-0613-7046 Biological and Earth Sciences, South Australian Museum, SA 5000, Australia. ethan.beaver@live.com.au Author Velasco-Castrillón, Alejandro 0000-0002-3516-6655 Biological and Earth Sciences, South Australian Museum, SA 5000, Australia. a.velascocastrillon@gmail.com Author Stevens, Mark I. Biological and Earth Sciences, South Australian Museum, SA 5000, Australia. & University of Adelaide, Biological Sciences, SA 5005, Australia. text Zootaxa 2021 2021-04-07 4951 3 571 597 journal article 10.11646/zootaxa.4951.3.9 eeae93f1-66d5-4ca0-a1d9-946ad810fb99 1175-5326 4668500 45D34CAC-0EA1-47AB-A919-CA157741549A Systematics of Abantiades Molecular analyses (see Fig. 1 ) The Bayesian analysis for each species ( Fig. 1 ) reveals a well supported (PP = 0.84) monophyletic Abantiades clade compared to our six outgroups taxa (PP = 1.0). Although this is only based on the single mtDNA (COI) gene, it does corroborate the morphological distinction of the two new species from the 16 other Abantiades species in our analysis, which are each also defined by unique morphological characters (see Taxonomy section below). Table 1 details all specimens used, including GenBank or BOLD accessions and locality information. For instances where we obtained multiple individuals within species, we found close intraspecific sequence divergences (Supplementary Table 1 ). For example, the greatest sequence divergence observed within any of our Abantiades species varied up to 1.7%, indicating variation that we would expect within species included here (Supplementary Table 1 ). These distances also correspond to those reported elsewhere for Lepidoptera that generally indicate below 3% intraspecific while above 3% interspecific comparisons using the same COI gene ( e.g . Grund et al. 2019 ; Beaver et al. 2020a ,b; Moore et al . 2020b ). However, sequence divergences have been found to vary outside these ranges for some Abantiades species ( Simonsen et al. 2019 ; Moore et al. 2020a ,b). We were also interested in the divergence within species that we had specimens for from Kangaroo Island and the mainland. For A. marcidus Tindale, 1932 specimens showed intraspecific variation up to 0.8%, and our outgroup taxon Aenetus tindalei Simonsen, 2018 specimens showed intraspecific variation up to 0.6%, both consistent with population level divergences and the topology from Figure 1 (see Supplementary Table 1 ). Sequence divergence comparisons among the Abantiades species compared to our outgroup taxa ranged between 6.7–16% (Supplementary Table 1 ). Examining between Abantiades species reveal interspecific sequence divergences of up to 12.7% ( e.g. between A. magnificus ( Lucas, 1898 ) and A. aurilegulus Tindale, 1932 ). Pairwise comparisons between our new Abantiades species A. rubrus sp. nov. to other Abantiades species ranged between 7.6–11.2%, and A. rubrus sp. nov. groups with the other members of the “ labyrinthicus ” clade, A. labyrinthicus ( Donovan, 1805 ) and A. mcquillani Simonsen, 2018 (PP = 1.0; Fig. 1 ). Pairwise comparisons between our second new species A. penneshawensis sp. nov. to other Abantiades species ranged between 5.8–12.2% and A. penneshawensis sp. nov. aligns with A. lineacurva Moore & Edwards, 2014 , and place both species in the “ lineacurva ” clade ( Fig. 1 ). We also observed several interspecific sequence divergences that are below 3%, for example A. pallida Simonsen, 2018 and A. tembyi Moore & Beaver, 2020 (2.4%) and A. labyrinthicus and A. cf. mcquillani (2.2%) (Supplementary Table 1 ), which has been observed in some ghost moths (e.g. Moore et al. 2020b ). Although sequence divergence can be low between a number of Abantiades species, each correspond to clear morphological differences (e.g. Moore et al. 2020b ).