Neuropterida comprise the holometabolan orders Neuroptera (lacewings, antlions and relatives), Megaloptera (alderflies, dobsonflies) and Raphidioptera (snakeflies) as a monophyletic group sister to Coleoptera (beetles). The higher‐level phylogenetic relationships among these groups, as well as the family‐level hierarchy of Neuroptera, have to date proved difficult to reconstruct. We used morphological data and multi‐locus DNA sequence data to infer Neuropterida relationships. Nucleotide sequences were obtained for fragments of two nuclear genes (CAD, 18S rDNA) and two mitochondrial genes (COI, 16S rDNA) for 69 exemplars representing all recently recognized families of Neuropterida as well as outgroup exemplars from Coleoptera. The joint posterior probability of phylogeny and divergence times was estimated using a Bayesian relaxed‐clock inference method to establish a temporal sequence of cladogenesis for the group over geological time. Megaloptera were found to be paraphyletic with respect to the rest of Neuropterida, calling into question the validity of the ordinal status for Megaloptera as presently defined. Ordinal relationships were weakly supported, and monophyly of Megaloptera was not recovered in any total‐evidence analysis; Corydalidae were frequently recovered as sister to Raphidioptera. Only in relaxed‐clock inferences were Raphidioptera and a paraphyletic Megaloptera recovered with strong support as a monophyletic group sister to Neuroptera. A monophyletic Neuroptera diverged from a common Raphidioptera + ‘Megaloptera’ ancestor during the Late Carboniferous. Contrary to some previous hypotheses, Coniopterygidae, not Nevrorthidae, were recovered as sister to the rest of Neuroptera, with Nevrorthidae recovered with Osmylidae and Sisyridae. The monophyly of the universally recognized Myrmeleontiformia was confirmed, with an origin in the mid‐Triassic, but a monophyletic Hemerobiiformia was not recovered in any analysis. Dilaridae were not closely related to the clade comprising Mantispidae and Berothidae, and diverged earlier than proposed previously. The phylogenetic status and taxonomic composition of Polystoechotidae and Ithonidae are in need of re‐evaluation, as Oliarces Carpenter (presently Ithonidae) was placed well within the present circumscription of Polystoechotidae.

Background: Evolutionary relationships among the 11 extant orders of insects that undergo complete metamorphosis, called Holometabola, remain either unresolved or contentious, but are extremely important as a context for accurate comparative biology of insect model organisms. The most phylogenetically enigmatic holometabolan insects are Strepsiptera or twisted wing parasites, whose evolutionary relationship to any other insect order is unconfirmed. They have been controversially proposed as the closest relatives of the flies, based on rDNA, and a possible homeotic transformation in the common ancestor of both groups that would make the reduced forewings of Strepsiptera homologous to the reduced hindwings of Diptera. Here we present evidence from nucleotide sequences of six singlecopy nuclear protein coding genes used to reconstruct phylogenetic relationships and estimate evolutionary divergence times for all holometabolan orders.

Neuroptera (lacewings) and allied orders Megaloptera (dobsonflies, alderflies) and Raphidioptera (snakeflies) are predatory insects and together make up the clade Neuropterida. The higher‐level relationships within Neuropterida have historically been widely disputed with multiple competing hypotheses. Moreover, the evolution of important biological innovations among various Neuropterida families, such as the origin, timing and direction of transitions between aquatic and terrestrial habitats of larvae, remains poorly understood. To investigate the origin and diversification of lacewings and their allies, we undertook phylogenetic analyses of mitochondrial genomes of all families of Neuropterida using Bayesian inference, maximum likelihood and maximum parsimony methods. We present a robust, fully resolved phylogeny and divergence time estimation for Neuropterida with strong statistical support for almost all nodes. Mitochondrial sequence data are typified by significant compositional heterogeneity across lineages, and parsimony and models assuming homogeneous rates did not recover Neuroptera as monophyletic. Only a model accounting for compositional heterogeneity (i.e. CAT‐GTR) recovered all orders of Neuropterida as monophyletic. Significant findings of the mitogenomic phylogeny include recovering Raphidioptera as sister to Megaloptera plus Neuroptera. The sister family of all other lacewings are the dusty‐wings (Coniopterygidae), rather than Nevrorthidae. Nevrorthidae are instead returned to their traditional position as the sister group of the spongilla‐flies (Sisyridae) and closely related to Osmylidae. Our divergence time analysis indicates that the Mesozoic was indeed a ‘golden age’ for lacewings, with most families of Neuropterida diverging during the Triassic and Jurassic and all extant families present by the Early Cretaceous. Based on ancestral character state reconstructions of larval habitat we evaluate competing hypotheses regarding the life style of early neuropteridan larvae as either aquatic or terrestrial.

Analysis of anchored hybrid enrichment (AHE) data under a variety of analytical parameters for a broadly representative sample of taxa (136 species representing all extant families) recovered a well-resolved and strongly supported tree for the higher phylogeny of Neuropterida that is highly concordant with previous estimates based on DNA sequence data. Important conclusions include: Megaloptera is sister to Neuroptera; Coniopterygidae is sister to all other lacewings; Osmylidae, Nevrorthidae and Sisyridae are recovered as a monophyletic Osmyloidea, and Rhachiberothidae and Berothidae were recovered within a paraphyletic Mantispidae. Contrary to previous studies, Chrysopidae and Hemerobiidae were not recovered as sister families and morphological similarities between larvae of both families supporting this assumption are reinterpreted as symplesiomorphies. Relationships among myrmeleontoid families are similar to recent studies except Ithonidae are placed as sister to Nymphidae. Notably, Ascalaphidae render Myrmeleontidae paraphyletic, again calling into question the status of Ascalaphidae as a separate family. Using statistical binning of partitioned loci based on a branch-length proxy, we found that the diversity of phylogenetic signal across partitions was minimal from the slowest to the fastest evolving loci and varied little over time. Ancestral character-state reconstruction of the sclerotization of the gular region in the larval head found that although it is present in Coleoptera, Raphidioptera and Megaloptera, it is lost early in lacewing evolution and then regained twice as a nonhomologous gula-like sclerite in distantly related clades. Reconstruction of the ancestral larval habitat also indicates that the ancestral neuropteridan larva was aquatic, regardless of the assumed condition (i.e., aquatic or terrestrial) of the outgroup (Coleopterida).

The first phylogenomic analysis of the antlions is presented, based on 325 genes captured using anchored hybrid enrichment. A concatenated matrix including 207 species of Myrmeleontoidea (170 Myrmeleontidae) was analysed under maximum likelihood and Bayesian inference. Both Myrmeleontidae (antlions) and Ascalaphidae (owlflies) were recovered as paraphyletic with respect to each other. The majority of the subfamilies traditionally assigned to both Myrmeleontidae and Ascalaphidae were also recovered as paraphyletic. By contrast, all traditional antlion tribes were recovered as monophyletic (except Brachynemurini), but most subtribes were found to be paraphyletic. When compared with the traditional classification of Myrmeleontidae, our results do not support the current taxonomy. Therefore, based on our phylogenomic results, we propose a new classification for the antlions, which synonymizes Ascalaphidae with Myrmeleontidae and divides the family into four subfamilies (Ascalaphinae, Myrmeleontinae, Dendroleontinae and Nemoleontinae) and 17 tribes. We also highlight the most pressing issues in antlion systematics and indicate taxa that need further taxonomic and phylogenetic attention. Finally, we present a comprehensive table placing all extant genera of antlions and owlflies in our new proposed classification, including details on the number of species, distribution and notes on the likely monophyly of each genus.