Museomics, molecular phylogeny and systematic revision of the Eurepini crickets (Orthoptera: Gryllidae: Eneopterinae), with description of two new genera

Natural history collections worldwide house billions of specimens, representing one of the most globally important biobanks. In recent years, the advent of next‐generation sequencing has significantly reduced the challenges of obtaining considerable genetic information from historical museum specimens. Crickets in the Australian tribe Eurepini Robillard are a good example of a taxon in which such museomic data have particularly strong potential to advance systematic knowledge, because comprehensive sampling requires decades of work over a very wide area. The tribe currently comprises 64 described species in five genera. Previous studies conflict in the generic relationships inferred for this tribe, all of which are poorly resolved, being based on limited data and sampling. In addition, there has so far been no systematic research for this tribe with extensive taxon sampling, and therefore, the consequence for genus boundaries remains to be investigated. To investigate phylogenetic relationships within Eurepini, we first applied the genome skimming approach to obtain molecular data from a comprehensive sample of Eurepini museum specimens. Of the 69 specimens sampled representing 61 described species, mainly including holotype specimens, we obtained 50 complete and 11 partially complete mitogenomes. Three nuclear genes (H3, 18S, and 28S) were also partially recovered for nearly all of these specimens. Phylogenetic analyses performed with mitogenomes plus three nuclear genes using maximum likelihood and Bayesian inference generated well‐supported and highly congruent topologies. Eurepini was strongly recovered monophyletic with eight well‐defined groups. These groups are used to revise the systematics of the tribe based on a combination of molecular phylogenetics and morphology. The phylogenetic results support the current definition of three genera (Eurepa Walker, Arilpa Otte & Alexander and Eurepella Otte & Alexander), lead us to redefine three genera (Salmanites Chopard, Napieria Baehr and Piestodactylus Saussure), and define and describe two new genera: Miripella Robillard, Tan & Su gen.nov. and Arrakis Robillard, Tan & Su gen.nov. Our results reinforce the importance of natural history collections as a repository for information on biodiversity and genetics, and provide the first comprehensive and robust phylogenetic framework for future systematic and evolutionary studies of Eurepini.


INTRODUCTION
Natural history collections (NHCs) worldwide house billions of specimens including a million described species (Short et al., 2018;Wandeler et al., 2007;Wheeler et al., 2012), which represent a substantial sampling of global biodiversity across space and time.NHCs are, therefore, an invaluable resource for scientific research (Graham et al., 2004;Suarez & Tsutsui, 2004;Yeates et al., 2016).Traditionally, voucher specimens in museums are primarily used as a source of morphological characters for taxonomic studies.Increasingly, as sequencing technologies progress, the role of voucher specimens is being extended, such that they have become an important reference in genetic studies (Droege et al., 2014;Suarez & Tsutsui, 2004).Genetic data from historical specimens have been used to address questions that could only be made possible with the help of historical specimens.Examples include investigating historical population genomics (Bi et al., 2013), phylogenetic placement of cryptic lineages (Kanda et al., 2015) and delimitating cryptic species (Hind et al., 2015;Lindstrom et al., 2015;McCormack et al., 2016).
Despite being an important genetic resource, most specimens were collected long before the extraction and generation of molecular data and were preserved primarily to retain morphological features.
Hence, specimen DNA can be heavily fragmented and degraded depending on preservation conditions and age, often making specimens difficult or impossible to sequence with Sanger technology (Wandeler et al., 2007;Yeates et al., 2016).However, the advent of next-generation sequencing (NGS) and its application to museum specimens to obtain genetic data-museomics-has made molecular studies of historical specimens more efficient (Hofreiter et al., 2015;Rizzi et al., 2012).An advantage of NGS is that it capitalises on highly fragmented and short DNA and, thus, circumvents the obstacles faced with traditional Sanger sequencing of historical materials (Bi et al., 2013;Burrell et al., 2015;Prosser et al., 2016;Staats et al., 2013;Timmermans et al., 2016).One method employing this sequencing technology is 'genome skimming' (Straub et al., 2012)shallow sequencing of total genomic DNA at random.It is a relatively simple method in terms of bioinformatics and has the advantage that it is among the most affordable NGS methods that do not require prior genetic information regarding the investigated species (Knyshov et al., 2021).In addition, compared with other NGS methods (e.g., whole genome sequencing, transcriptomics, RAD or target capture), this method requires less laboratory work and the quality requirements for samples are also lower.Moreover, genome skimming is highly efficient in sequencing organelle genomes and highly repetitive genome regions such as rDNA (18S and 28S) and histones.As sequencing costs have dropped during the past decade, genome skimming has become an affordable solution, and it has been successfully applied to obtain genetic data from historical specimens in a growing number of studies on various groups of insects (e.g., >120-year-old beetles in Jin, Zwick, Slipinski, Marris, et al., 2020;Jin, Zwick, Slipinski, De Keyzer, & Pang, 2020) and plants (e.g., >140-year-old herbariums in Zedane et al., 2016 andBakker et al., 2016).
Ensiferan insects, especially crickets (Gryllidea Laicharting), are particularly well known for their calling songs.They have been used as a model system to study the ecology and evolution of acoustic communication (Bailey, 1991;Gerhardt & Huber, 2002;Song et al., 2020).However, to investigate how communication traits evolved and diversified, a clear phylogenetic context is needed, which requires clear taxonomic information for as many species as possible and robust phylogenetic hypotheses.This can be challenging, especially for cricket species from hyper-diverse regions such as Southeast Asia or Australia, where many species remain undescribed or are known from only a small number of collection specimens including the type specimen.
Among crickets, the subfamily Eneopterinae Saussure, one of the three extant subfamilies in Gryllidae, has become a model group in evolutionary biology, owing to comprehensive advances in taxonomy, phylogeny and bioacoustic studies in the field and laboratory (e.g., Robillard, 2021).Eneopterinae is a diverse group of crickets with approximately 280 described species in six tribes mainly distributed in Southeast Asia, the Western Pacific archipelagos, Africa, South America and Australia (Cigliano et al., 2023).The main phylogenetic relationships within the monophyletic Eneopterinae have been more or less congruent in previous phylogenetic studies, and the tribe Eurepini Robillard has been consistently recovered as sister to all the other clades and forms an independent lineage at the crown (Chintauan-Marquier et al., 2016;Nattier et al., 2011;Robillard & Desutter-Grandcolas, 2004b;Robillard & Desutter-Grandcolas, 2006;Tan et al., 2021;Vicente et al., 2017).
Eurepini originated in Australia in the late Palaeocene ca.58 million years ago (Vicente et al., 2017).Eurepini currently comprises 64 described species in five genera, which are all endemic to Australia: Eurepa Walker, Salmanites Chopard, Arilpa Otte & Alexander, Myara Otte & Alexander and Eurepella Otte & Alexander (Cigliano et al., 2023;Otte & Alexander, 1983;Rentz & Su, 2019).Eurepini are widely distributed in most regions of the Australian continent, including the central and western desert areas.They occupy a wide range of habitats, mainly including leaf litter, tree trunks and bark, shrubs and grasses in forests or open areas.Eurepini is also diverse in body size (ranging from 1 to 5 cm), ovipositor length (ranging from ca. 1 to 4 times hind femur length), male genitalia, forewing length (from aptery to full wings covering the entire abdomen, including different intermediate development stages) and sound producing structures (from functionally complete to absent) (Otte & Alexander, 1983;Rentz & Su, 2019).Despite a wide variety of temporal patterns in their calling songs, the calling frequencies of Eurepini crickets are relatively low (4-8 kHz) compared with some crickets in other tribes in the subfamily Eneopterinae (Robillard & Desutter-Grandcolas, 2008;Tan et al., 2021).
Although Eurepini represents a diverse species group with 64 described species and a number of undescribed species, it has been understudied when compared with other tribes in the subfamily, especially concerning historical biogeography and evolution of calling songs (Dong et al., 2018;Tan et al., 2021;Vicente et al., 2017).Phylogenetic and systematic studies with respect to Eurepini have been very scarce and confined to the inclusion of one to three representatives in each genus nested in larger taxonomic scale phylogenetic studies of Eneopterinae (Nattier et al., 2011;Robillard & Desutter-Grandcolas, 2004b;Robillard & Desutter-Grandcolas, 2006;Tan et al., 2021;Vicente et al., 2017).Over the course of previous phylogenetic studies based on morphological and/or molecular data (with the exception of Otte & Alexander, 1983), each one has recovered a subtly or strikingly different topology for major Eurepini clades (see Figure 1 for details).Overall, while Eurepini has been consistently corroborated as a monophyletic group, the generic relationships within the tribe remain unclear or untested.Due to the instability of phylogenetic placements at a generic level and significant incongruence between morphological and/or molecular evidence, a robust phylogeny to resolve generic relationships within the tribe is much needed.In addition, since the seminal taxonomic work of Otte and Alexander (1983), the systematics and classification of the tribe and component genera have not been revised or updated.
In this study, our goal was to infer the first comprehensive phylogenetic framework of Eurepini based on molecular data and to provide novel insights into genus-level phylogenetic relationships of the tribe.Specifically, we first applied genome skimming to moderately old and degraded museum specimens to obtain mitogenomes and nuclear genes (H3, 18S, and 28S), which confer phylogenetic resolution for relationships at the species level and higher levels, respectively.Second, we inferred a robust phylogeny of Eurepini with comprehensive sampling covering nearly all described species, by sequencing a large proportion of type specimens.We used the phylogeny to test generic definitions and to determine relationships within the tribe.Moreover, we thoroughly revise the taxonomy of the tribe at the generic level based on combined information on molecular phylogenetic relationships and reassessment of morphological characters, leading to the definition and description of two new genera, the redefinition of the three existing genera, and an interactive key to genera.Direct observations and dissections of types and non-type materials were performed under a binocular microscope (Leica model MZ16) at magnifications up to 115Â.Male genitalia dissection as well as morphological features and genitalia followed the procedures by Tan andRobillard (2021, 2022).Male tegminal veins and cells followed the terminology by Desutter-Grandcolas (2003) and Robillard and Desutter-Grandcolas (2004b).Male genitalia were named according to Desutter (1987), modified from Desutter-Grandcolas (2003) and Robillard and Desutter-Grandcolas (2004b).Photographs of male genitalia were captured using a binocular microscope Leica MZ16 F I G U R E 1 Hypotheses for generic relationships within Eurepini from previous studies.

Knowledge database and identification key
The identification key of Eurepini genera has been generated using the online collaborative software Xper3 Xper3 also makes possible a more flexible, interactive key that can be used even if some characters are lacking or not observable.

Taxon sampling
In total, we sampled 69 specimens (Table S1) representing 61 described species in five genera.Most specimens are holotypes deposited at ANIC, that were collected in the late 1960s, and a small proportion were collected more recently.Most specimens were preserved in 70% ethanol, and the remaining ones were preserved dry and pinned with needles.All specimens were photographed and made accessible by ANIC.

DNA extraction, library preparation and sequencing
Molecular work was performed at ANIC and MNHN.DNA was generally extracted from median legs detached from specimens following Gilbert et al.'s (2007) preparation protocol, which keeps the specimen or part of the specimen used intact while extracting DNA and can, thus, leave the external morphology of the specimen undamaged.
Total genomic DNA was quantified using the Qubit dsDNA

Sequence assembly and annotation
The resulting reads were assembled using Geneious v. 9.0.2 (KM853388, 2582 bp) of Eneoptera guyanensis Chopard were selected.A reference was then mapped to the total filtered reads to extract the reads of interest.After removing the reference, we ran de novo assembly to assemble new contigs.Generating consensus sequences from the contigs, the correct sequences were selected as new references to map to the filtered reads to extend contigs using multiple iterations of the mapping procedure until a complete circular sequence was assembled.Alternatively, several contigs with sufficiently long overlaps were assembled into longer contigs by running de novo assembly.Nuclear genes were also assembled with the same procedure.The newly obtained mitogenomes and genes were used as new reference sequences for further de novo assembly.All of the newly obtained sequences have been submitted to GenBank (accession numbers are listed in Table S1).
Mitogenome annotation was initially carried out on the MITOS web server (Bernt et al., 2013) with the invertebrate mitochondrial genetic code.However, annotations for protein-coding genes (PCGs) with MITOS have been shown to be unreliable (Cameron, 2014).
Therefore, we manually annotated PCGs using gene alignments of the three aforementioned reference species.The range of two rRNA genes was extended to the boundaries of their adjacent tRNA genes.

Sequence alignment and dataset partitions
All of the mitochondrial genes (13 PCGs, 22 tRNA and two rRNA genes) and three nuclear genes (H3, 18S and 28S) were individually extracted using Geneious v. 9.0.2 and saved as fasta format files, and then separately aligned online using the MAFFT server v. 7.0 (Katoh et al., 2019) with default parameters.The ambiguously aligned regions were manually removed using MEGA v. 7 (Kumar et al., 2016); the indels within the datasets were left untouched.For preliminary analyses, individual gene (all 22 tRNA genes were concatenated and set as one dataset) trees were generated using IQ-TREE v. 1.6.12(Trifinopoulos et al., 2016) with default settings.Then, the 40 gene datasets were concatenated into one large dataset using SequenceMatrix (Vaidya et al., 2011).The resulting supermatrix was divided into 19 partitions, one partition for each gene of 14 PCGs and four rRNA genes and one partition for all of the 22 tRNA genes, to obtain a specimen-level tree using IQ-TREE with settings as abovementioned.
Before further analyses, the supermatrix was initially divided as follows: one partition per codon position for each PCG, one for each rRNA gene (12S, 16S, 18S and 28S) and one for all of the 22 tRNA genes.Then, the best partitioning scheme and substitution models for each partitioning scheme were searched with PartitionFinder v. 1.1.1(Lanfear et al., 2012), using the greedy search algorithm, and the Bayesian information criterion to compare the fit of the different models (Ripplinger & Sullivan, 2008), with branch lengths linked and all models searched.

Phylogenetic analyses
Molecular data for five cricket species used as outgroups were downloaded from GenBank and added to the datasets before Phylogenetic analyses were carried out using maximum likelihood (ML) and Bayesian inference (BI) methods.The resulting subset partitions in the best partitioning scheme were used as input to submit to the IQ-TREE server v. 1.6.12,and models of nucleotide substitution were searched automatically across all available models, including the free rate model (+R) that relaxes the assumption of gammadistributed rates (Soubrier et al., 2012), which is recommended for analysis of large datasets.ML analysis was performed on the IQ-TREE server v. 1.6.12.Nodal support was assessed with 1000 replicates of ultrafast bootstrap (UFB) (Hoang et al., 2018;Minh et al., 2013), as well as 1000 replicates of Shimodaira-Hasegawa approximate likelihood ratio test (SH-aLRT) (Guindon et al., 2010).Nodes with support values of both SH-aLRT ≥80% and UFB ≥95% were considered well supported, nodes with one of SH-aLRT <80% or UFB <95% were considered weakly supported, and nodes with both SH-aLRT <80% and UFB <95% were considered unsupported.
Because some models (TVM, TrN and TrNef) are not implemented in MrBayes, they were replaced by the GTR model before BI.BI analysis was conducted using MrBayes v. 3.2.7 (Ronquist et al., 2012) with two independent runs and each with four Markov chain Monte Carlo chains, running for 20 million generations, and sampling every 1000 generations.The first 25% of sampled trees were discarded as burn-in, and the 50% majority-rule consensus trees were used to estimate posterior probabilities (PPs).Convergence of the runs and effective sample size (ESS) were checked using Tracer v. 1.7.2 (Rambaut et al., 2018) to ensure sufficient sampling (ESS ≥ 200) for all parameters.PP was estimated from the remaining trees.Nodes with a support value of PP ≥0.95 were considered strongly supported (Erixon et al., 2003), 0.94-0.90moderately supported, 0.89-0.85weakly supported and <0.85 unsupported.Consensus trees were initially visualised using FigTree v. 1.4.4 (Rambaut, 2019) and then edited using Inkscape v. 1.2 (https://inkscape.org/release/inkscape-1.2/).
As additional analyses, we first assessed the effect of excluding the third codon positions on tree topology and branch support in the 13 mitochondrial PCGs.We found that the resulting ML tree topology is roughly congruent with that of the ML tree inferred with full data, with the exception of several unresolved short branches that radiated recently in the clade Eurepella (Figures S1 and 3).More unsupported nodes were also found, especially for internal nodes located within the clade Myara (redefined as Piestodactylus, see Results section).In addition, we performed Xia's test (Xia et al., 2003;Xia & Lemey, 2009) using DAMBE v.7 (Xia, 2018) to individually test for substitution saturation of the third codon positions in the 13 PCGs and found that these positions had not experienced full substitution saturation, which suggested that they were still useful for phylogenetic inference.Given the negative impact of excluding the third codon positions and the lack of saturation in most of the third positions, we included the third codon positions in the datasets for further phylogenetic analyses.

Sequencing results and general characters of mitogenome organization in Eurepini
Our sampling comprised 61 of 64 described Eurepini species, including 55 holotypes (Table S1).Only three described species were unavailable to be sampled for this study: Eurepella narranda Otte & Alexander (the holotype, housed at the South Australia Museum, was not accessible), Eurepa bifasciata (Chopard) and Myara wintrena Otte & Alexander.
Of the 69 sequenced specimens, 50 produced complete mitogenomes and 11 generated partially complete mitogenomes, in which part of the AT-rich region or the region between ND3 and ND5 was not recovered; and several long segments were generated for each of the remaining four specimens.Sequences of three nuclear genes were also partially recovered from nearly all of these specimens.We were unable to retrieve any data from four holotype specimens.
Therefore, we excluded these four species from subsequent phylogenetic analyses.Taken together, we obtained molecular data for 57 of the 64 described species.All mitogenome and gene sequences, with GenBank accession numbers, are listed in Table S1.
The lengths of complete mitogenomes ranged from 15,838 bp (Eurepa quabara Otte & Alexander) to 17,213 bp (Eurepella ballina Otte & Alexander) (Table S1).For better readability, we use Salmanites obscurifrons Chopard as a representative to present the general features of mitogenome organization in Eurepini (Figure 2 and Table S2).As with other insects, mitogenomes of Eurepini consisted of a conserved set of 37 genes (13 PCGs, 22 tRNA genes and two rRNA genes) and a control region, with the exception of Eurepella wanga Otte & Alexander, in which trnE was not found.Of these genes, 20 were encoded on the majority strand (J strand), and the remaining 17 on the minority strand (N strand).The start and stop codons were consistent with those reported for invertebrates in general (Cameron, 2014) and Grylloidea in particular (Ma & Li, 2018).
The gene arrangement and orientation were the same as the three

Phylogenetic analyses
Preliminary ML analyses show no major conflict between 10 PCGs, 12S, 16S and 28S.Although the consensus tree differs in topology from ATP6, COX3, ND4L and 22 concatenated tRNA genes, and strikingly from that of H3 and 18S, both trees are poorly resolved (Figure S2).
The supermatrix included 70 taxa, of which 65 were Eurepini, comprising 19,798 aligned sites, of which 7669 were parsimony informative.
The best partitioning scheme constituted 17 subset partitions (see Table S3 for details).Both ML and BI analyses generated topologically identical trees (Figures 3 and S4, also see Figure S3 for the tree inferred from 19 gene partitions in the preliminary analyses); most of the nodes are strongly supported on the ML and BI trees, whereas only the extent of support of a minority of nodes is lower on the ML tree than BI tree.
Diagnosis (emended from Otte & Alexander, 1983).Size is small to average for the tribe.All members of this genus have a relatively stocky shape and are characterized by the following characters: the head is large compared with the rest of the body; the fastigium (rostrum) is wide, ca.1.8-2.8times as wide as the scape width; the transfrontal furrow is well marked and located below the median ocellus.
The pronotum is short and wide.The hind wings are short in both sexes and do not exceed FWs.Male FWs are usually reaching the abdomen mid-length but are not widened laterally; the harp is large, with 3-4 oblique veins, two of them being connected to the diagonal vein; the mirror is not rounded, its length being 1.5 times its width.
The male genitalia possess cup-like pseudepiphallic parameres (ectoparameres) located apically, below the median process ending the pseudepiphallic sclerite.to that of Salmanites, explaining that these two genera have been confused in the past.They share similar small sizes (but Napieria are relatively smaller), and wide fastigium, characters are also shared with Miripella gen.nov.Napieria includes the smallest species in the tribe.It differs from all other genera by the head shape with rounded fastigium, the presence of unique dorsal abdominal glands on male tergites IV-V, and by harp veins including, with only one oblique vein connected to the diagonal vein (two in all other genera, except in Miripella gen.nov., which has only one oblique vein).From Salmanites, Napieria also differs by the following characters: fastigium is not angular at apex, with a nearly complete continuity between the frons and the fastigium (no angle permit to distinguish the two parts in lateral view); the pronotum is rectangular (trapezoidal in Salmanites); male FWs (except for apterous species) are shorter and never reach the abdomen apex; mirror is longer than wide (wider than long in Salmanites).
Male genitalia have very short rami (longer in Salmanites) and are characterized by the presence of strong black setae on pseudepiphallic slerite in most species.From Miripella gen.nov., Napieria differs by its face shape; its head is wider than high, whereas it is as wide as high in Napieria.Napieria also differs from Miripella gen.nov.by its longer male FWs (not reaching the abdomen mid-length in Miripella gen.Alexander, 1983).They were both transferred to Eurepini by Robillard and Su (2018), who tentatively included them in Eurepa and Salmanites, respectively.The phylogenetic position of Miripella miripara as the sister species of all the other genera, except Salmanites, combined with its original morphological characters (see below), supports the hypothesis that the species belongs to a new genus characterized by strong and short hind legs, face wider than high, and very short wings in both sexes, without a clearly differentiated mirror in males.The second species, formerly known as Lebinthus bifasciatus, also shares these characteristics while differing from M. miripara in terms of male genitalic characters (pseudepiphallic parameres are large, recalling those of Eurepa).While the species is not included in our phylogenetic study, we tentatively place it in the new genus Miripella gen.nov.
Etymology.The genus name is derived from a combination of the names 'miripara' (type species' epithet) and -pella, from Eurepella, itself derived from Eurepa, the type genus of the tribe.
Common name.Frog-faced ground crickets.
Diagnosis.Size is small for the tribe, shape is stocky.The new genus is characterized by the following characters: the fastigium is very strong and wider than long, slightly divergent apically; the face is wider than high; the transfrontal furrow is variably marked and located just below the median ocellus; the face is slightly rounded in lateral view.The pronotum is rectangular.Hind legs are strong and short.FWs are very short in both sexes and are straight along the abdomen; hind wings are absent.Male FWs are characterized by a large harp, with 1-2 oblique veins, one being connected to the diagonal vein; the mirror is not differentiated; the apex of the dorsal field is plicated transversally, which delimits a small apical flap expanding the median fold dorsally.The male metanotum has no glandular structures.Male genitalia are characterized by short triangular apical lophi; pseudepiphallic parameres are relatively large; the rami are wide and long, located laterally; the ectophallic apodemes are short and wide apically.Female FWs are variable in length among species; ovipositor is longer than the FIII.
Differential diagnosis.The members of Miripella gen.nov.resemble Salmanites, Arilpa and Napieria by their small size, short wings and wide fastigium, and to Eurepa by the relatively large pseudepiphallic parameres.Miripella gen.nov. is unique in the tribe in its shorter, more muscular hind legs, its fastigium widened apically (straight or convergent in all other genera), its shorter wings in both sexes (not reaching abdomen mid-length in males, short but variable in females), male metanotal glands are absent, male FWs' harp have one or two oblique veins (three or more in other genera), and mirror is not differentiated (well differentiated in other genera).The head is wider than high, whereas it is as wide as high in Napieria and a few species of Piestodactylus (P.unicolor), and higher than wide in other genera.Female ovipositor is longer than the FIII (shorter in Salmanites and Arilpa), like in Napieria and other genera.
Vertex wide and relatively flat, directly prolonged by very wide fastigium, more than three times as wide as scape and slightly widened apically.Eyes of average size for the tribe, in dorsal view eyes combined width represents ca.30% of head width.Head wider than high in facial view.Transfrontal furrow weak, located just below median ocellus (Figure 5g); face slightly rounded in lateral view (Figure S11C).
Scapes small.Pronotum dorsal disk rectangular to slightly trapezoidal (Figure 4g).TI with one outer tympanum typical of tribe, thin and elongate.TI and TII each with one anterior and two posterior spurs, as other genera of tribe.FIII muscular and short, without a linear region before knee (Figures 4g and S11).TIII short, its dorsal side with four curved subapical spurs on inner margin, and four straight spurs on outer margin, with rather strong spines above and between spurs, progressively thicker towards posterior end of tibia; apex of TIII with typical spurs of tribe and subfamily.TaIII-1 with two dorso-apical Ectophallic apodemes short and wide apically, their bases linked by a thick arc; ectophallic fold bilobate apically, slightly sclerotized laterally.
Female.FWs variable in length, either shorter or as long as in male, either well separated (M.miripara) or slightly overlapping (Miripella bifasciata); dorsal field and lateral field with strong longitudinal veins.Ovipositor longer than FIII, its apex acute, typical of tribe.
Species list (species name mentioned in bold is not included in the phylogeny).Taxonomic remark.The two species belonging to this genus were previously included in the paraphyletic genus Myara (now Piesdodactylus).
Etymology.The genus is named after the fictional desert planet featured in the Dune series of novels by Frank Herbert, in reference to the desert regions where the members of this genus are distributed.
Common name.Long-nosed desert crickets.
Diagnosis.Size is average to large for the tribe, shape is slender.
Arrakis is characterized by its bulging face, due to fastigium being longer than wide, narrower than two times the scape width, and prolonged by a bulbous extension of the clypeus (resembling a long nose in lateral view).The median ocellus is located on the dorsal part of the fastigium, at mid-length.The transfrontal furrow is well marked and located dorsally near the median ocellus.FWs are reaching beyond 3 /4 of the abdomen length in both sexes and are almost straight along the body.The male FWs are characterized by a narrow harp, with three oblique veins, two of them being connected to the diagonal vein; the mirror is well differentiated, not rounded anteriorly and slightly longer than wide.Hind wings are slightly exceeding the FWs in both sexes.
The female ovipositor is very long, more than four times as long as the FIII.Female.FW dorsal field and lateral field with strong longitudinal veins (Figure S12D).Ovipositor very long (Figure 4h,i), more than four times as long as FIII.its apex acute, typical of tribe.
Type locality.Australia, Queensland, Cape York Peninsula, Iron Range.
Remark.The female holotype of S. iknurra was located by one of us after the description of Macrobinthus kutini.There can be no doubt, based on general head morphology and tympana, that it is the same species as the one described as M. kutini (see Robillard & Su, 2018 for complete description).The type locality is also similar between the two names (Queensland, Iron Range).The morphology shows that the species does belong to tribe Lebinthini and not to Eurepini.Consequently, M. kutini must be considered as a junior synonym of Macrobinthus iknurra comb.nov.

Knowledge database and identification key to Eurepini genera
The interactive version of the key is available at https://eurepinigenera.identificationkey.org.
Relatively small for the tribe: 5.

DISCUSSION
Importance of museum specimens and efficiency of genome skimming NHCs are an essential resource for the scientific community for studying biodiversity (Graham et al., 2004;Suarez & Tsutsui, 2004;Yeates et al., 2016).These collections are important for the study of systematics, evolution, ecology and a variety of other disciplines (Bakker et al., 2016;Bradley et al., 2014).Sampling from NHCs often presents a colossal time saving relative to fieldwork.Although there is often due reticence in reliance on sampling of NHCs for genetic resources, in geographically vast and biodiversity-rich regions like Australia, they often represent an efficient way to sample geographically disparate or rare taxa for molecular phylogenetic studies.Furthermore, NHCs are sometimes the only potential source of genetic data.
Advances in sequencing technologies greatly improve the capacity and efficiency of obtaining genetic information from historical specimens and can circumvent the obstacles that traditional sequencing methods are faced with due to the difficulty of primer-based amplification of highly degraded DNA.Therefore, immense numbers of specimens from NHCs can now be integrated into molecular studies, whereas they were considered less usable a decade ago.
In this context, we took advantage of these advances and applied the genome skimming method to sequence relatively old Eurepini specimens including nearly all types of specimens.As genome skimming has been successfully used to recover considerable genetic information from a variety of historical specimens of various ages (e.g., Bakker et al., 2016;Besnard et al., 2014;de Abreu et al., 2020;Guschanski et al., 2013;Jin, Zwick, Slipinski, De Keyzer, & Pang, 2020;Jin, Zwick, Slipinski, Marris, et al., 2020;Salazar & Nattier, 2020;Zedane et al., 2016;Zeng et al., 2018), we managed to recover substantial amounts of genetic data, including complete or partially complete mitogenomes and nuclear rDNA for most of the specimens that were collected more than 50 years ago, demonstrating the high efficiency of genome skimming on such old materials.We were unable to recover genetic data for only four relatively old holotype specimens, which might be due to high degradation of DNA lead-

Systematics of Eurepini
In this study, we inferred the first well-resolved molecular phylogeny of the Australian endemic crickets corresponding to the tribe Eurepini.
Using genome skimming, we were able to recover substantial molecular data from a comprehensive sampling of Eurepini specimens, including a large proportion of relatively old-type specimens.(Figure 7a).The genus Arilpa also has the same species list as in the study by Otte and Alexander (1983).Male genitalia for these crickets also show distinctive features corresponding to their cup-shaped pseudepiphallic parameres (Figure 7c).(1983), the genus Salmanites is here redefined, including the two species transferred from Eurepella, which form a coherent species list in terms of morphological features (male FW venation and genitalia), as previously hypothesised by Rentz and Su (2019).
With the exception of the type species Salmanites obscurifrons, all the other species previously classified in the genus Salmanites are here transferred to the genus Napieria, which was defined by Baehr (1989) based on the type species N. muta.This name previously considered as a synonym of Salmanites by Robillard and Desutter-Grandcolas (2008) is revalidated in this study.Although the species N. muta itself appears as an exception in terms of morphological features (e.g., males are wingless; tympanum and abdominal glands are absent), the priority rule makes it the type species of the genus.In addition, the close phylogenetic affinity of N. muta with other Napieria (formerly Salmanites) species is strongly supported by our results.
nov., includes two species that were tentatively placed in two different genera, and were initially described under the genus Lebinthus in the tribe Lebinthini.Our results clearly show that M. miripara form an independent lineage.The revision of the morphological study by Robillard and Su (2018) also indicates that the species M. bifascitata shares several features with M. miripara (Figures 4g and S11).In particular, these crickets share small and stocky body dimensions, and the face is wider than high (Figure 5g), justifying their common name of 'frog-faced ground crickets'.Monophyly of this genus will be tested by including molecular data of the species M. bifasciata in further phylogenetic studies.
Our results show that the generic relationships are significantly incongruent with those of previous studies (Figure 1), mainly due to very limited taxon sampling and data for Eurepini species in these studies, leading to unstable topologies and unresolved generic relationships.In our study, phylogenetic inferences with a large dataset consisting of mitogenomes and three nuclear genes covering nearly all described species in the tribe generated highly resolved and congruent topologies.Our results confirm the hypothesis that the genus Napieria (formerly Salmanites) diverged earlier than most other clades, as indicated in previous studies (Tan et al., 2021;Vicente et al., 2017), whereas Eurepella lately split from the other clades (Robillard & Desutter-Grandcolas, 2004b;Robillard & Desutter-Grandcolas, 2006;Vicente et al., 2017).However, hypotheses for generic relationships for other clades are not supported by our study.Overall, the generic and species relationships are resolved with high support in our study, which supports the power of coupling comprehensive taxon sampling with a large number of molecular data in resolving the systematics of a diverse insect group.

CONCLUSIONS
NHCs represent an essential biobank of specimens for studying biodiversity and a source of genetic information for molecular studies.Our extensive sampling of museum specimens for the cricket tribe Eurepini, combined with genome skimming approach, has allowed us to recover a large amount of molecular data for almost all specimens attempted.These data allowed us to infer a well-resolved phylogeny, which sheds light on the evolutionary history of Eurepini and will allow us to put forward related evolutionary hypotheses for this tribe.
Our study furthermore provides the first comprehensive phylogenetic framework of Eurepini and lays the foundation for future evolutionary studies.
Photoshop CS6 Extended to adjust levels, contrast, exposure and sharpness, and to add scale bars (1 mm).

with
Photonics, China).Image editing was accomplished using CombineZP ver.1.0 and Adobe Photoshop CC2014.To highlight the structural components of genitalia, water solution containing a drop of JBL Punktol was used.To fix orientations and stabilization of genitalia for photography, a clear and viscous Hand Sanitizer was used followingSu (2016).
High-sensitivity assay kit (Life Technologies, Paisley, UK) and the Fluorescence Microplate Reader in a 1.0 μL of sample.DNA was then indexed, and libraries were constructed using the NEBNext Ultra II DNA Library Prep Kit (New England BioLabs, Ipswich, MA, USA) following the protocol ofMeyer and Kircher (2010) with minor modifications.After library preparation, total DNA was quantified again as in the previous step, and libraries were quantified on the 2100 Bioanalyzer using the DNA 1000 series II chip (Agilent Technologies, Santa Clara, CA, USA; High Sensitivity DNA Assay).Pooled libraries were subsequently sequenced as paired-end reads (150 bp) on a NovaSeq 6000 sequencing platform (Illumina, CA, USA).
eneopterine cricket mitogenomes that were used as reference sequences.One gene rearrangement was found in all of the Eurepini species, corresponding to a local inversion of the ancestral trnN-trnS1-trnE to trnE-trnS1-trnN.All of these complete mitogenomes showed a high bias to A and T nucleotide content, which is typical of most insects, ranging from 70.4% in Napieria muta Baehr to 75.2% in Eurepa wirkutta Otte & Alexander.

(
SH-aLRT = 100%, UFBoot = 100%; PP = 1).Our analyses recovered seven well-supported monophyletic groups and a paraphyletic group, of which five were large clades.The crown node of Eurepini involved Mitogenome organization of Salmanites obscurifrons.The 13 protein-coding genes (PCGs), 22 transfer RNA (tRNA) genes, two ribosomal RNA (rRNA) genes and AT-rich region are coloured in green, blue, red and orange, respectively.The direction of transcription is indicated by an arrow.Genes encoded on the majority (minority) strain are denoted with clockwise (counterclockwise) arrows.the divergence of a small clade with three species (Eurepella budyara Otte & Alexander, Eurepella tumbiumbia Otte & Alexander and Salmanites obscurifrons) from all the remaining Eurepini.Here, we redefine this small clade as the genus Salmanites sensu stricto.The second deepest node involved the divergence of Salmanites miripara from the rest of the Eurepini (SH-aLRT = 96%, UFBoot = 98%; PP = 1), which we define here as the new genus Miripella gen.nov.At the third deepest node, a large clade comprising most of the Salmanites species split from the other Eurepini (SH-aLRT = 99.9%,UFBoot = 100%; PP = 1), which we redefine here as the genus Napieria Baehr.At a shallower node, the genus Eurepa was recovered as sister to the remaining Eurepini species (SH-aLRT = 100%, UFBoot = 100%; PP = 1).A small clade containing two species previously classified in Myara (M.mabanuria Otte & Alexander and M. yurgama Otte & Alexander) diverged from the remainder of Eurepini (SH-aLRT = 97.8%,UFBoot = 99%; PP = 1), and is defined here as the new genus Arrakis gen.nov.At the next shallowest node, the genus Arilpa, a relatively small clade, was sister to a large clade comprised of most of the Eurepella and Myara species (SH-aLRT = 100%, UFBoot = 100%; PP = 1).The clade corresponding to 'Myara' sensu stricto, that has to be named Piestodactylus Saussure based on the rule of priority (see taxonomic section below) was marginally well supported as the sister clade to the rest of Eurepini in the ML analysis (SH-aLRT = 93%, UFBoot = 94%), whereas this relationship was strongly supported in the BI analysis (PP = 1).Finally, the remainder of Eurepini was composed of two main lineages: Piestodactylus yabmanna (Otte & Alexander) emerged as a divergent lineage independent from the rest of Piestodactylus, and was found to be sister to Eurepella, which represents the most diversified group in Eurepini (SH-aLRT = 100%, UFBoot = 100%; PP = 1).P. yabmanna together with the other seven Piestodactylus species formed a paraphyletic group.The sister relationship of P. yabmanna to the genus Eurepella renders Piestodactylus a paraphyletic genus; however, there is no strong morphological evidence to support this relationship.FI G U R E 3 Molecular phylogeny of Eurepini.The trees inferred from concatenated datasets (13 PCGs, 12S, 16S, 22 tRNAs, H3, 18S and 28S) using Maximum likelihood (ML) and Bayesian inference are integrated as one tree.ML bootstrap support and BI posterior probability values are indicated at nodes (SH-aLRT/UFBoot/PP).Type species of each genus are highlighted in bold, and new genera are indicated with an asterisk (*).Dotted branches in outgroups have been shortened to maximise readability.Photo: male of Eurepa marginipennis (copyright David Rentz, with permission) TAXONOMIC PART Family Gryllidae Subfamily Eneopterinae Saussure, 1874 Tribe Eurepini Robillard, 2004 Diagnosis: Members of the tribe Eurepini are characterized by the following set of characters: the fronto-clypeal (epistomal) suture of the face is interrupted between the tentorial insertions, completed with a variable transfrontal furrow delimiting an extension of the clypeus on the frons variable in size, and sometimes forming a bulging expansion on the face.The forelegs have only one tympanum, oval, on the outer surface; its membrane is flat; the inner tympanum is absent.The TaIII-1 has one spine on the outer side.The male metanotum shows variable dorsal glands in most genera; the postnotum is narrow and partly fused with the first abdominal tergite.The hind wings are short or absent, usually not exceeding the FWs in both F I G U R E 4 Body in dorsal and lateral views of eight genera in Eurepini: (a) Eurepa yumbena, (b) Eurepella mjobergi, (c) Arilpa allara, (d) Piestodactylus warrantina, (e) Salmanites obscurifrons, (f) Napieria noccundris, (g) Miripella miripara gen.nov., (h) Arrakis mabanuria gen.nov.(female), (i) Arrakis yurgama gen.nov.(male).Scale bar: 1 cm.sexes.The male FW venation includes a strong CuP vein, reaching beyond the file vein.The male subgenital plate lacks lateral swellings on dorsal margin.The male genitalia have ramal plates and articulated rami; the lateral arms of the endophallic sclerite are directly prolonged by sclerotization of the ectophallic fold.The female ovipositor has an apex with ventral valves completely hidden by the dorsal valves.

F
I G U R E 5 Head in anterior view of (a) Eurepa yumbena, (b) Eurepella mjobergi, (c) Arilpa allara, (d) Piestodactylus warratina, (e) Salmanites obscurifrons, (f) Napieria noccundris, (g) Miripella miripara gen.nov., (h) Arrakis mabanuria gen.nov.Scale bar: 1 mm.Diagnosis (emended from Otte & Alexander, 1983).Size average for the tribe.The members of the genus are mostly characterized by the male genitalia with large, laterally flattened pseudepiphallic parameres (=ectoparameres in Otte & Alexander, 1983).The fastigium (=rostrum in Otte & Alexander, 1983) is longer than wide, narrower than two times the scape width.The transfrontal furrow is strongly marked and relatively low on face.FIII is short and thin.FWs are straight along the body, reaching beyond 3 /4 of abdomen length in males, variable in size in females (females in Woortooa species groupwith very short FWs).The hind wings are short but slightly exceeding the FWs in both sexes.The male harp is large, with 3-4 oblique veins, two of them connected to the diagonal vein; the mirror is welldifferentiated, usually slightly longer than wide.The female ovipositor is long, from 1.5 to 3.5 times as long as the FIII.Differential diagnosis.The genus differs from all other genera by male genitalia with large, laterally flattened pseudepiphallic parameres.
Taxonomic remarks.Baehr (1989) defined the monospecific genus NapieriaBaehr, 1989, with the type species N. muta.This genus was later synonymised under Salmanites byRobillard and Desutter- Grandcolas (2008) based on the fact that the type species was similar to the species previously included in this genus.Our phylogenetic results lead to a new definition of the genus Salmanites (corresponding to the type species Salmanites obscurifrons and two species transferred from Eurepella), which excludes the remaining species formerly known as Salmanites sp., which form a clade that has to be defined as a separate genus.Given the fact that this clade includes the species S. muta, the type species of Napieria, the latter genus name is here revalidated and used as the genus name of this clade including all diminutive Eurepini species.We revalidate here the species Paraenepterus handschini Chopard, 1931, previously considered byOtte and Alexander (1983)  as nomen dubium under the genus Salmanites.Although the species was described from a female type only, the original description and the type localities are clear enough to recognize the species as a member of the genus Napieria, to be redescribed in a future study based on newly collected materials.New diagnosis.Size is very small for the tribe.The members of this genus are characterized by the following characters: head is as wide as high in facial view; the transfrontal furrow is weak and located just below the median ocellus; face is slightly rounded in lateral view.Fastigium is very wide, more than three times as wide as the scape width, without a clear delimitation between the dorsal part of the fastigium and the frons, forming a regular curve.The pronotum is rectangular.Hind wings are short in both sexes.Male FWs are short, barely reaching 3 /4 of the abdomen length, and are variably rounded laterally; the harp is large, with three or more oblique veins, only one of them being connected to the diagonal vein; the mirror is little-differentiated and longer than wide.The genus is characterized by the presence of dorsal glands near the abdomen mid-length on tergites IV-V.Male genitalia are small and characterized by short triangular pseudepiphallic lophi curved dorsally, and by small pseudepiphallic parameres located apically; rami are very short; in most species, dorsal surface of pseudepiphallic slerite presents strong black setae.Female FWs are absent or very short, their length being less than half of the pronotum length.Female ovipositors are variable but are usually longer than the FIII.Differential diagnosis.The members of Napieria are most similar nov.).The female ovipositor is longer than FIII (shorter in Salmanites and Arilpa), like in most other genera.Female FWs are very short, as in Miripella gen.nov.
spines and a row of spines on outer dorsal edge; with a row of lateral outer spine.FWs short in both sexes, not reaching abdomen midlength; hind wings absent.Abdomen: Tergites without longitudinal bands.Cerci of average length.Male.Metanotal glands absent.FWs venation (Figure 6g): 1A vein (file) forming a wide curve (>120 ).CuP strong, extended posterior to file level.CuA weak and straight.Harp occupying most of dorsal field, flat; with one or two oblique veins; main one straight and connected to diagonal vein.Mirror (d1) undifferentiated.Cell c1 large along diagonal vein.Apical field small, with a transversal fold delimiting a small apical flap expanding median fold dorsally.Lateral field with veins M and R strong and straight, Sc straight without bifurcating veins, and three more ventral veins.Male genitalia (Figures 7g and S11E-G): pseudepiphallic sclerite forming a triangle, with short triangular apical lophi; pseudepiphallic parameres relatively large; rami wide and long, located laterally.
and S12B,F).TIII short, its dorsal side with four curved subapical spurs on inner margin, and four straight spurs on outer margin, with rather strong spines above and between spurs, progressively thicker toward posterior end of tibia; apex of TIII with typical spurs of tribe and subfamily.TaIII-1 with two dorso-apical spines and a row of spines on outer dorsal edge; without lateral outer spines.Legs short, in particular hind legs, barely reaching beyond abdomen apex; FIII narrow, little muscular.FWs relatively long in both sexes (Figure S12D,H), reaching beyond 3 /4 of abdomen length, narrow along body; hind wings present in both sexes, slightly longer than FWs.Abdomen: Tergites dark, without longitudinal bands.Cerci of average length.Male.Metanotal glands present.FW cells and veins translucent brown (Figure 6h), cells finely wrinkled longitudinally (mostly visible in harp and mirror).FW venation: 1A vein (file) forming a wide curve (>120 ).CuP strong, extended posterior to file level.CuA straight.Harp longer than wide, relatively narrow, flat, with three oblique veins; posterior ones sub-straight and connected to diagonal vein, anterior one connecting 1A and CuA.Cell c1 narrow, well separated from c2.Chord veins delimiting a narrow, semi-circular cell.Mirror (d1) well differentiated, slightly longer than wide, its anterior part forming an acute angle, its posterior part rounded.Apical field as long as mirror, including 3-4 cell alignments.Lateral field with veins M and R strong and straight, Sc straight with bifurcating veins along its whole length.Male genitalia (Figures 7h and S12i-k): pseudepiphallic sclerite forming an elongate triangle with short triangular apical lophi; anterior edge deeply indented, its lateral parts folded dorsally; pseudepiphallic parameres relatively large; rami long, slightly convergent.Ectophallic apodemes forming wide plates.Ectophallic fold slightly sclerotized laterally.Endophallic sclerite small.
lus for nomenclatural reasons (see below), and members of the genus roughly correspond to the classification of Myara by Otte andAlexander (1983), with the exception of two species, which form an independent clade defined as the new genus Arrakis.Even though these nomenclatural changes, Piestodactylus remains a paraphyletic group due to the sister relationship of P. yabmana to the genus Eurepella, whereas P. yabmana shares more common morphological characters with other Piestodactylus species than with Eurepella.This might be due to Piestodactylus being weakly defined in terms of morphological features: in contrast to other genera in the tribe, it is mostly characterized by 'average' characters, and this may thus lead to several lineages, suggesting paraphyly or even polyphyly of the genus.This hypothesis will be tested in further studies with more species close to Piestodactylus that were not sampled in this study, including a number of undescribed species.As mentioned above, two species previously classified in the genus Myara form a small distant clade from Piestodactylus, which is defined here as the new genus Arrakis gen.nov., with clear morphological coherence (Figures4h,i and S12).They are distributed in desert areas and characterized by a very long fastigium leading to a bulging face (Figure5h), justifying their common name of 'long-nosed desert crickets'.

Figure S4 .
Figure S4.Trees inferred from concatenated datasets using the substitution models and corresponding subset partitions in the best partitioning schemes (Table S3) searched by PartitionFinder v. 1.1.1 with ML and BI methods.Values of Bootstrap support and posterior probability are indicated at nodes (SH-aLRT/UFBoot).

Figure S12 .
Figure S12.Body in dorsal (A, E) and lateral (B, F), head in anterior (C, G), FW in dorsal (D, H) and male genitalia in dorsal (I), ventral (J) and lateral (K) views of Arrakis mabanuria.Scale bar: 1 mm.
Distribution.Australia (widespread except for cold areas).
Species listMarginipennis species group: Eurepa eeboolaga Otte & Alexander, 1983 Eurepa marginipennis (White, 1841) Eurepa tanderra Otte & Alexander, 1983 Eurepa yumbena Otte & Alexander, 1983 Nurndina species group: Eurepa nurndina Otte & Alexander, 1983 Eurepella Otte & Alexander, 1983: 279; Desutter-Grandcolas, 1990: 239 (Eneopteridae: Eneopterinae); Otte, 1994: 66; Rentz, 1996: 140 (key); Robillard & Desutter-Grandcolas, 2004a: 275 (morphological phylogeny), 2004b: 578, 2006: 644 (molecular phylogeny); 2008: 66 (Eurepini); 2011: 637 (bioacoustics); Robillard et al., 2007: 1266 (bioacoustics); Nattier et al., 2011: 2201 (molecular phylogeny); Chintauan-Marquier et al., 2016: 71 (molecular phylogeny); Vicente et al., 2017: 2203 (historical biogeography); Rentz & Su, 2019: 207; Cigliano et al., 2023 (Orthoptera Species File online).Type species.Eurepella quarriana Otte & Alexander, 1983, by original designation.Taxonomic remark.The species group Moojera defined by Otte and Alexander (1983) is validated by the phylogenetic results and remains unchanged.Based on the phylogenetic results the species group Quarriana is unchanged, except for two species forming a separate monophyletic group corresponding to the new species group Lewara, including the species E. lewara Otte & Alexander, 1983, and E. nakkara Otte & Alexander, 1983.The members of the Budyara species group (two species) are transferred here to Salmanites (see below).diagonalvein; the mirror is well differentiated, generally wider than long.Female FWs are of average size, longer than pronotum, but usually do not reach beyond the abdomen mid-length.The hind wings are short in both sexes and never exceed the FWs.The ovipositor is short for the tribe, between 0.8 and 1.2 times as long as FIII.Differential diagnosis.Eurepella is similar to Napieria and Salmanites by male FWs (when present) widened around the abdomen, unlike other genera where FWs are as wide or narrower than abdomen.From Eurepa, it also differs by male genitalia without large, laterally flattened pseudepiphallic parameres.Fastigium width average, wider than in Eurepa, Piestodactylus and Arrakis gen.nov., but thinner than in Salmanites, Napieria and Arilpa.Common name.Little long-tailed crickets.Eurepella arowacka Otte & Alexander, 1983 Eurepella ballina Otte & Alexander, 1983 Eurepella kulkawirra Otte & Alexander, 1983 Eurepella mataranka Otte & Alexander, 1983 Eurepella meda Otte & Alexander, 1983 Eurepella mjobergi (Chopard, 1925) Eurepella quarriana Otte & Alexander, 1983 Eurepella tjairaia Otte & Alexander, 1983 Eurepella torowatta Otte & Alexander, 1983 Eurepella wanga Otte & Alexander, 1983 Eurepella waninga Otte & Alexander, 1983 Lewara species group: Eurepella lewara Otte & Alexander, 1983 Eurepella nakkara Otte & Alexander, 1983 Nomen dubium Eurepella curvatifrons (Chopard, 1951)-Otte & Alexander (1983) Eurepella subaptera (Chopard, 1925)-Otte & Alexander (1983) Type species.Arilpa wirrilla Otte & Alexander, 1983, by original designation.Taxonomic remark.The species list remains as defined by Otte The female FWs are shorter than the pronotum length.The female ovipositor is short and slightly curved, about as long as the FIII length.Here, we resurrected Piestodactylus as a valid genus, of which Myara is a junior synonym.Compared with Myara as defined by Otte and Alexander (1983), the species list is similar except for two species excluded to form the new genus Arrakis gen.nov.(Myara mabanuria Salmanites until now.They mostly share similar a small size (but Salmanites tend to be relatively larger) and a wide fastigium, characters also shared with Miripella gen.nov.Salmanites and Miripella gen.nov.differ from Napieria by the following characters: the fastigium forms an angular rostrum in lateral view in Salmanites and Miripella gen.nov., whereas there is a nearly complete continuum between the frons and the fastigium in Napieria (no angle to distinguish the two parts in lateral view).From Miripella gen.nov., Salmanites differs by head shape in facial view: its head is wider than high in Differential diagnosis.From Eurepa, Piestodactylus and Arrakis gen.nov., Arilpa differs by its shorter male FWs, as in Salmanites and Eurepella; the FWs are longer than in Miripella gen.nov.From Eurepella, Arilpa differs by male FWs that are not widened laterally and by its shorter female ovipositor.Arilpa resembles Salmanites and Miripella gen.nov.by its short to average size and its stocky shape; it is, however, larger than the diminutive Napieria.The female ovipositor is short, as in Salmanites and Miripella gen.nov., while generally longer than FIII in other genera.Common name.Moon crickets.Distribution.Western Australia.Eurepa Walker, 1869, then in Myara Otte & Alexander, 1983.Myara Otte & Alexander, 1983 syn.nov.As remarked by Braun (in Cigliano et al., 2023), Chopard (1951) synonymised Piestodactylus brevipennis, type species of Piestodactylus Saussure, 1878, under Myara sordida.Since the genus Myara includes the type species of the genus Piestodactylus, the principle of priority implies that Piestodactylus is the valid genus name (name resurrected), of which Myara is a junior synonym.Type species.PIatydactylus brevipennis Brunner von Wattenwyl, 1898, by original designation.The type species is currently a junior synonym of Salmania sordida Walker, 1869 (=Piestodactylus sordida).Taxonomic remarks.Piestodactylus was previously synonymized under Eurepa by Kirby (1906), under which it remained until now, even if the type species was synonymised with M. sordida (Walker, 1869).New diagnosis.The size is average to large for the tribe, with species being average and stocky in shape.Piestodactylus is characterized by its fastigium (rostrum) longer than wide, rectangular and narrower than two times the scape width.The transfrontal furrow is strongly marked and is usually low on face.The FWs are reaching beyond 3 /4 of the abdomen length in both sexes and are almost straight along the body; in male FWs, the harp is large, with 3-4 oblique veins, two of them being connected to the diagonal vein; the mirror is well differentiated and rounded, and is nearly as long as wide.In most species, the hind wings are slightly exceeding the FWs in both sexes.The female ovipositor is long, its length ranging from 0.7 to 3.1 times as long as FIII.Differential diagnosis.The members of Piestodactylus are very similar to Arrakis gen.nov., including the hind wings slightly exceeding the FWs in most species.From Eurepa, they differ by their slightly larger size and stockier shape, and by the male genitalia without large, laterally flattened pseudepiphallic parameres (ectoparameres in Otte & Alexander, 1983).From Arrakis gen.nov., Piestodactylus differs by its face generally flat (very bulging in Arrakis gen.nov.), with median ocellus anterior to scapes, at apex of fastigium (on dorsal part of fastigium in Arrakis gen.nov.),FIIImuscular(thinner in Arrakis gen.nov.), and shorter ovipositor (except in P. unicolor).From Salmanites, Napieria, Miripella gen.nov.andArilpa,Piestodactylusdiffers by its larger size, longer FWs and by fastigium longer than wide.Piestodactylus erola (Otte & Alexander, 1983) comb.nov.transferredfromMyara.Piestodactylus merimbula (Otte & Alexander, 1983) comb.nov.transferredfromMyaraPiestodactylusmuttaburra(Otte& Alexander, 1983) comb.nov.transferredfromMyaraPiestodactyluspakaria(Otte& Alexander, 1983) comb.nov.transferredfromMyaraPiestodactylussordida(Chopard,1951) comb.nov.-transferredfromMyaraPiestodactylusunicolor(Chopard,1951) comb.nov.-transferredfromMyaraPiestodactyluswarratinna(Otte& Alexander, 1983) comb.nov.-2011:637(bioacoustics);Robillardetal., 2007: 1266 (bioacoustics); Nattier et al., 2011: 2201 (molecular phylogeny); Vicente et al., 2017: 2203 (historical biogeography); Robillard & Su, 2018: 244; Rentz & Su, 2019: 217; Cigliano et al., 2023 (Orthoptera Species File online).Type species.Salmanites obscurifrons Chopard, 1951, by original monotypy.Taxonomic remarks.Otte and Alexander (1983) defined Salmanites as including all the small-sized species of Eurepini.Based on the phylogenetic results, we show here that the type species S. obscurifrons does not belong to the same clade as most of these species, but that it forms a clade with two species formerly classified as Eurepella, with which S. obscurifrons shares many characters of male genitalia.Consequently, except for the type species, all the species previously classified under Salmanites are here transferred to the genus Napieria Baehr, which is resurrected (see below).Salmanites sensu stricto is here redefined and redescribed.At the same time, two species from the genus Eurepella appear to belong to the same clade as S. obscurifrons and are transferred here to Salmanites (comb.nov.).Finally, the species Salmanites iknurra Otte & Alexander, 1983 shown to belong to the tribe Lebinthini and is transferred to the genus Macrobinthus Robillard & Dong, 2016 (see below).morethan three times as wide as the scape width, and forming an angular rostrum in lateral view.Pronotum is trapezoidal.Hind wings are short in both sexes and never exceed the FWs.The male FWs reach the abdomen apex; they are rounded laterally and slightly widened around abdomen; the harp is large, with 3-4 oblique veins, two of them being connected to the diagonal vein; the mirror is well-differentiated; it is not rounded and nearly as long as wide.Male genitalia have short triangular pseudepiphallic lophi and small, bilobate pseudepiphallic parameres located apically.Female FWs are very short, their length being less than half of the pronotum length.Female ovipositor is short and curved, its length being shorter than the FIII length.Differential diagnosis.The members of Salmanites are most similar to that of Napieria, explaining that these two genera have been considered as gen.nov., whereas it is as wide as high in Salmanites.In Salmanites, the male FWs are longer, and the mirror is nearly as long as wide, whereas the mirror in Napiera is always longer than wide (except for apterous species), and the mirror is not differentiated in Miripella gen.nov.The male genitalia are larger in Napieria; Salmanites never show strong setae on the dorsal part of the pseudepiphallus as in most Napieria species; rami are longer than the rest of the pseudepiphallic sclerite in Salmanites, whereas the rami are usually very short in Napieria, and long but wide in Miripella gen.nov.Female ovipositor is short, as in Arilpa, unlike in most other genera.Female FWs are very short, as in Napieria.Common name.Dark ground crickets.Distribution.Northern Territory and Queensland.
Differential diagnosis.The members of Arrakis gen.nov.are similar to Piestodactylus and Eurepa by their hind wings slightly exceeding FWs and by body size being average to large; they are similar to Eurepa by the FIII relatively thin and short.Arrakis gen.nov.differs from these genera by its bulging face due to the fastigium/clypeus extension, the median ocellus located on the dorsal part of fastigium (at apex of fastigium in other genera), the transfrontal furrow located dorsally near median ocellus (lower on face in Eurepa and Piestodactylus), and by the very long female ovipositor.From Salmanites, Napieria, Miripella gen.nov.and Arilpa, Arrakis gen.nov.differs by its larger size, its fastigium longer than wide, longer FWs, and hind wings being present and longer than the FWs.
This species was described as Piestodactylus siamensis based on one female specimen from Thailand.The characters present in the original description are not informative, but the type locality makes it very likely the species does not belong to the tribe Eurepini.The type is lost, making the previous hypothesis impossible to test.We consequently consider this name as nomen dubium.
For genus Eurepella, with the exception of two species previously classified in the Budyara species group being transferred to the genus Salmanites, as redefined in this study, the remaining species are all kept within the current definitions of Eurepella.The two species transferred to Salmanites were classified into a particular species group byOtte and Alexander (1983)who recognised their particularities.More recently, Rentz and Su (2019) suggested that the two species did not belong to Eurepella based on morphological observations and proposed to move them out of this genus.The species group Moojera defined by Otte and Alexander (1983) remains unchanged, and the species group Quarriana is also roughly unchanged, with the exception of two species, E. lewara and E. nakkara, forming a separate monophyletic group, which is split from Quarriana species group and defined as a new group Lewara.In conclusion, the genus Eurepella remains largely stable.
Since the type species of Salmanites is Salmanites obscurifrons, and our results show that this species forms a clade with two species previously classified in the Budyara species group of Eurepelle instead of the other species attributed to this genus by Otte and Alexander