Further progress on the phylogeny of Noctuoidea (Insecta: Lepidoptera) using an expanded gene sample

Major progress has been made recently toward resolving the phylogeny of Noctuoidea, the largest superfamily of Lepidoptera. However, numerous questions and weakly supported nodes remain. In this paper we independently check and extend the main findings of multiple recent authors by performing maximum-likelihood analyses of 5-19 genes (6.7-18.6kb) in 74 noctuoids representing all the families and a majority of the subfamilies. Our results strongly support the six family system of Zahiri et al., with the former Lymantriidae and Arctiidae subsumed within the huge family Erebidae, and Noctuidae restricted largely to the subfamilies with so-called trifine hindwing venation. Our data also strongly corroborate monophyly of the set of four families with quadrifid forewing venation, to the exclusion of Notodontidae, and removal from the latter of Oenosandridae. Other among-family relationships, however, remain unsettled. Our evidence is equivocal on the position of Oenosandridae, which are sister group to either Notodontidae alone or to all other noctuoids. Like other recent nuclear gene studies, our results also provide no strong support for relationships among the four quadrifid forewing families. In contrast, within families our analyses significantly expand the list of robustly resolved relationships, while introducing no strong conflicts with previous molecular studies. Within Notodontidae, for which we present the largest molecular taxon sample to date, we find strong evidence for polyphyly for some, or all, recent definitions of the subfamilies Thaumetopoeinae, Pygaerinae, Notodontinae and Heterocampinae. Deeper divergences are incompletely resolved but there is strong support for multiple backbone' nodes subtending most of the subfamilies studied. Within Erebidae, we find much agreement and no strong conflict with a recent previous study regarding relationships among subfamilies, and somewhat stronger support. Although many questions remain, the two studies together firmly resolve positions for over half the subfamilies. Within Noctuidae, we find no strong conflict with previous molecular studies regarding relationships among subfamilies, but much stronger resolution along the backbone' of the phylogeny. Combining information from multiple studies yields strongly resolved positions for most of the subfamilies. Finally, our results strongly suggest that the tribes Pseudeustrotiini and Prodeniini, currently assigned to the largest subfamily, Noctuinae, do not belong there. In sum, our results provide additional corroboration for the main outlines of family-level phylogeny in Noctuoidea, and contribute toward resolving relationships within families.