Phylogenomics of darkling beetles (Coleoptera: Tenebrionidae) from the Atacama Desert

Insect collection

Tenebrionid beetles from the Chilean Atacama Desert (30 genera, 14 tribes) were collected by hand between 2017 and 2021 (Table 1; collecting permits CONAF No 005/2017, 105/2020, 016/2021). The collected specimens were either transferred directly into 96% ethanol for DNA and RNA analyses or transported alive for RNA extraction from fresh material; RNA extraction was then carried out in the Cologne laboratory. Furthermore, we collected samples of 51 tenebrionid genera (33 additional tribes) from Central Chile (collecting permits CONAF No 005/2017), Germany, Italy, Spain, Portugal (collecting permit No 757-758/2021/CAPT), Namibia (collecting permit NCRST RPIV01042034) and Peru (collecting permits SERFOR Nr D000019-2022) to improve taxon sampling for phylogenetic analyses. In addition, published peptide precursor sequences of Tribolium castaneum (Herbst, 1797) (Triboliini Gistel, 1848), Zophobas atratus (Fabricius, 1775) (Tenebrionini Latreille, 1802) (Marciniak, Pacholska-Bogalska & Ragionieri, 2022) and Tenebrio molitor Linnaeus, 1758 (Tenebrionini) (Li et al., 2008; Veenstra, 2019; Marciniak, Pacholska-Bogalska & Ragionieri, 2022) were added to our dataset, while peptide precursor sequences of Neomida bicornis (Fabricius, 1777) (Diaperinae: Diaperini Latreille, 1802) were obtained by Blast searches in the NCBI database (https://www.ncbi.nlm.nih.gov/Traces/wgs?val=GDMA01). RNA was additionally extracted from seven taxa of Tenebrionoidea Latreille, 1802 (families Ciidae Leach, 1819, Meloidae Gyllenhaal, 1810, Mycetophagidae Leach, 1815, Pyrochroidae Latreille, 1807, Salpingidae Leach, 1815, Zopheridae Solier, 1834 and one Cleroidea (Melyridae Leach, 1815) (Table 1), which were included in the phylogenetic analyses. Taxonomic determination was carried out by Álvaro Zúñiga-Reinoso and Reinhard Predel.

RNA extraction, cDNA library preparation and sequencing

Total RNA was extracted from samples stored in absolute ethanol or from individuals kept alive until tissue dissection. To avoid excessive RNA degradation in specimens stored in ethanol, head and pronotum of the beetles were separated from the rest of the body before transferring them into ethanol. In larger species, the body was additionally opened longitudinally with sterilized scissors. Without any treatment prior to storage in ethanol, the RNA was usually highly degraded, suggesting limited penetration of ethanol across the cuticle. Grinding of whole insects was avoided in order to enable the intestine to be removed later. Insects alive until tissue dissection were kept at 4 °C for 10 min before preparation. In most individuals (both ethanol and fresh material), after removal of the appendages (legs, elytra, antennae), the body was opened dorsally with sterilized scissors, the intestine was removed and the central nervous system (CNS) was carefully dissected. In small species, representing the genera Ammobius Guérin-Méneville, 1844, Achanius Erichson, 1847, Colydium Fabricius, 1792, Cordibates Kulzer, 1956, Corticeus Piller & Mitterpacher, 1783, Dichillus Jacquelin du Val, 1861, Discopleurus Lacordaire, 1859, Eledona Latreille, 1796, Melanimon Steven, 1829, Oochrotus Lucas, 1852, Synchita Hellwig, 1792, and Thinobatis Eschscholtz, 1831, the CNS was not dissected. For all other samples, total RNA was extracted from CNS and remaining tissues separately using one mL of TRIzol (Thermo Fisher Scientific, Darmstadt, Germany) following the manufacturers recommendations. Total RNA from each sample was quantified using Qubit RNA Assay Kit (Thermo Fisher Scientific) and subsequently subjected to quality control and RNA integrity number (RIN) as implemented in the Agilent 2100 Bioanalyzer system (Agilent Technologies, Waldbronn, Germany). Finally, RNA from CNS and remaining tissue from each sample were pooled together in equimolar concentrations for library preparations. This approach improved the detection of peptide precursor sequences, whose genes are mainly expressed in the CNS. Sequencing libraries (double-indexed) were prepared using 1 µg of total RNA with the Illumina^® TruSeq^® stranded RNA sample preparation kit (Cat.20020594; Illumina, San Diego, CA, U.S.A.). If the total RNA concentration was insufficient for standard library preparation, at least 2 ng of extract was pre-amplified using the Ovation RNA-Seq System V2 (NuGen, San Carlos, CA, USA). The library preparation of pre-amplified samples was performed according to the Nextera XT DNA sample preparation protocol (part no. 15031942 Rev. C). Subsequent sample preparation and sequencing was carried out at the Cologne Center for Genomics on an Illumina HiSeq 4000 and Illumina NovaSeq 6000 systems as described in Ragionieri & Predel (2020) with 75 bp or 100 bp paired end reads.

Transcriptome assembly, evaluation of cross-contaminations and statistics

Raw data (FASTQ files format) were filtered by removing adapter sequences and low quality using Trimmomatic 0.38 (Bolger, Lohse & Usadel, 2014). The resulting filtered RAW reads were submitted to NCBI (BioProject: PRJNA884860 Sequence Read Archives (SRA): SRR22314233, SRR22314232, SRR22314230, SRR22314229, SRR22314228, SRR22314227, SRR22314226, SRR22314224, SRR22314223, SRR22314225, SRR22314212, SRR22314210, SRR22314208, SRR22314209, SRR22314222, SRR22314221, SRR22314220, SRR22314218, SRR22314217, SRR22314216, SRR22314214, SRR22314213, SRR22314207, SRR22314215, SRR22314206, SRR22314211, SRR22314219, SRR22314193, SRR22314192, SRR22314191, SRR22314199, SRR22314205, SRR22314204, SRR22314203, SRR22314202, SRR22314201, SRR22314200, SRR22314198, SRR22314197, SRR22314196, SRR22314194, SRR22314195, SRR22314190, SRR22314188, SRR22314187, SRR22314186, SRR22314184, SRR22314183, SRR22314182, SRR22314181, SRR22314180, SRR22314189, SRR22314178, SRR22314177, SRR22314185, SRR22314179, SRR22314172, SRR22314170, SRR22314168, SRR22314175, SRR22314174, SRR22314173, SRR22314167, SRR22314166, SRR22314176, SRR22314164, SRR22314163, SRR22314162, SRR22314161, SRR22314160, SRR22314169, SRR22314159, SRR22314171, SRR22314165, SRR22314158, SRR22314157, SRR22314156, SRR22314154, SRR22314153, SRR22314152, SRR22314151, SRR22314150, SRR22314148, SRR22314147, SRR22314146, SRR22314145, SRR22314155, SRR22314149, SRR22314144, SRR22314143, SRR22314142, SRR22314231). Filtered reads were de novo assembled using Trinity v2.2.0 (Grabherr et al., 2011; Haas et al., 2013) with the read normalization option. All transcriptome assemblies were checked for potential cross-contaminations due to multiplex sequencing of several libraries using CroCo v1.1 (Simion et al., 2018) which removes potential sources of contamination using both transcriptome assemblies and the corresponding paired raw data (Table S1). This strategy uses sequence similarities and abundances to detect potential cross-contaminations. For closely related species that are analysed together, this can lead to an overestimation of cross-contamination (Simion et al., 2018). CroCo was run with the following settings: fold-threshold 2, minimum-coverage 0.1, overexp FLOAT 300, minimum percent identity between two transcripts to suspect across contamination 98%, minimum length of an alignment between two transcripts to suspect a cross contamination 180. Finally, we checked for and eliminated additional contamination of vector and linker/adapter using UniVec database (http://www.ncbi.nlm.nih.gov/tools/vecscreen/univec/). The transcriptomes assembled in this study lost on average about 2% of their sequence information due to the cross-contamination check. The quality and the completeness according to conserved single-copy ortholog content of transcriptome assemblies were evaluated using the Perl script (TrinityStats.pl) included in Trinity and BUSCO v3 based on an Endopterygota obd9 dataset (Seppey, Manni & Zdobnov, 2019), respectively. The filtered transcriptome assemblies were submitted to NCBI Transcriptome Shotgun Assembly database (Table 1) and used for the large scale data set phylogenetic reconstruction.

Orthology assessment and alignment of neuropeptide precursors

Available amino acid sequences of neuropeptide precursors of Tr. castaneum and Te. molitor (Li et al., 2008; Veenstra, 2019; Marciniak, Pacholska-Bogalska & Ragionieri, 2022) were used as initial queries to search for orthologous sequences in the transcriptome assemblies. The assembled transcripts were analysed with the tblastn algorithms provided by NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi) or BioEdit version 7.0.5.3 (Hall, 1999). In case of missing data, precursor sequences of closely related taxa were used as alternative query sequences. Candidate nucleotide precursor gene sequences were translated into amino acid sequences using the ExPASy Translate tool (Artimo et al., 2012; http://web.expasy.org/translate/) with the standard genetic code. Orthologous neuropeptide precursor sequences were aligned using the MAFFT-L-INS-i algorithm (Katoh & Standley, 2013) (dvtditr (amino acid) Version 7.299b alg=A, model=BLOSUM62, 1.53, −0.00, −0.00, noshift, amax =0.0); terminal sequences which were only found in few species were manually trimmed. The results were then manually checked for misaligned sequences using, e.g., N-termini of signal peptides and conserved amino-acid residues (cleavage signals, Cys as target for disulfide bridges) as anchor points. Individual amino acid alignments of each group of orthologous neuropeptide precursors were concatenated with catsequences 1.3 (https://zenodo.org/record/4409153#.YmJYT35Byot). The average evolutionary divergence for each neuropeptide precursor was calculated as in Bläser & Predel (2020). Briefly, overall mean distances (± standard error after 500 bootstrap generations) were computed with MEGA X (Kumar et al., 2018) implementing the Poisson correction model (Zuckerkandl & Pauling, 1965). Amino acid compositions and parsimony informative sites of the combined alignment were calculated using MEGA X.

Compilation of an orthologous gene dataset of Tenebrionidae

A Coleoptera orthologous reference gene set was compiled using OrthoDB v10. This approach provides reliable markers for phylogenomics (Misof et al., 2014; McKenna et al., 2019). Single copy genes shared across species of Coleoptera (Taxonomy ID: 7041) were selected for analysis. Orthograph (Petersen et al., 2017) was used to generate a profile hidden Markov model from the amino acid sequences of transcripts of each reference gene on the filtered transcriptome assemblies. Initially, we obtained 2,689 orthogroups (OGs) shared among Coleoptera, which were subsequently aligned using the MAFFT-L-INS-I algorithm (Katoh & Standley, 2013). Alignment ambiguities or spurious sequences in each OG were identified and removed using trimAL 1.2 (Capella-Gutierrez, Silla-Martinez & Gabaldon, 2009) with residue overlap threshold (-resoverlap 0.75) and sequence overlap threshold (-seqoverlap 90). With that approach, 947 out of 2,689 OGs were removed from the initial data set. Finally, all OGs were concatenated in a single partitioned super-alignment using catsequences.

Genome sequencing, assembly and identification of myosuppressin genes

Whole genome extraction was carried out using thoracic muscles of a single individual of Nycterinus abdominalis Eschscholtz, 1829 collected in Talcahuano, Chile. High molecular weight genomic DNA was purified using MagAttract^® HMW DNA Kit (Ref. 67563, QIAGEN GmbH, Hilden, Germany). DNA concentration was determined using Qubit 2.0 Fluorometer (Thermo Fisher Scientific). Fragment size was verified using DNA integrity number as implemented in the Agilent 2100 Bioanalyzer system. Genomic DNA library was prepared using the Illumina TruSeq Nano DNA High Throughput Library Prep Kit (Illumina, Cat. No 20015965) with modifications of the protocol (TruSeq DNA Nano Reference Guide, Document # 1000000040135 v00, October 2017). Only one cycle of polymerase chain reaction (PCR) was conducted to complete adapter structures in order to avoid PCR bias. Library validation and quantification were carried out as implemented in Agilent TapeStation, and subsequently the library was pooled and quantified using the Peqlab KAPA Library Quantification Kit (Roche Sequencing Solutions, Inc., USA; KK4835-07960204001) on an Applied Biosystems 7900HT Sequence Detection System and finally sequenced on an Illumina NovaSeq 6000 sequencer with 150 bp paired end reads. Raw data (FASTQ files format) were filtered by removing adapter sequences and low quality reads using Trimmomatic 0.38 (Bolger, Lohse & Usadel, 2014). Filtered raw data were assembled using the programs SOAPdenovo2 (Luo et al., 2012) using different k-mer values. The myosuppressin precursor was identified as described above (2.4). Genomic nucleotide sequences containing introns were subsequently aligned manually in BioEdit version 7.0.5.3 (Hall, 1999).

Phylogenetic analysis of neuropeptide precursors and a large scale orthologous gene dataset

FASTA files of aligned peptide precursor sequences were converted into PHYLIP and NEXUS formats using AliView 1.18-beta7 (Larsson, 2014). After defining the N-terminus of each neuropeptide precursor as starting partition, best-fit partitioning schemes and substitution models for subsequent phylogenetic analyses were predicted with ModelFinder (Chernomor, von Haeseler & Minh, 2016; Kalyaanamoorthy et al., 2017; Minh et al., 2021) implemented in IQ-TREE release 2.1.4b (Minh et al., 2020). Models and concatenated alignments for all analyses of both data sets are listed in Data S1 and S2. All phylogenetic analyses have been rooted using the Cleroidea Melyris sp. Bayesian inference (BI) analyses were run with MrBayes, with four runs, using eight chains and a sample frequency of 1,000 until convergence was achieved (PSFR value between 1.00–1.02) with a 10,000,000 generations (Ronquist et al., 2012). Maximum likelihood (ML) analyses were carried out using IQ-TREE 2.1.4b. ML analyses of both data sets were ran with the nearest-neighbour interchange search to consider all possible nearest-neighbour interchanges (-allnni) and branch support was evaluated with 1,000 ultra-fast bootstrap (UFBoot) (Hoang et al., 2017) and the Shimodaira–Hasegawa-like approximate likelihood ratio test (SH-aLRT) (Guindon et al., 2010). Trees were visualized using FigTree 1.4.2 (http://tree.bio.ed.ac.uk/) and designed in Inkscape 1.0 (https://inkscape.org/).

(A) Analysis of neuropeptide precursors

The primary matrix comprises 6,457 amino acids from 35 neuropeptide and neuropeptide-like precursors; information on sequence length and sequence coverage is provided in Table S2. The average evolutionary divergence over all sequences of the precursor dataset is 0.25 (± 0.03) and differs considerably between the different precursors (Table S2). The best fitting models according to ModelFinder are listed for each partition in Data S1, which also contains the concatenated alignment.

The phylogenetic tree of the concatenated neuropeptide precursor dataset (Fig. 2) recovered Tenebrionidae as monophyletic. Tenebrionidae are separated into one clade containing all the Pimeliinae analysed and a second clade containing the taxa of Alleculinae, Blaptinae, Diaperinae, Lagriinae Latreille, 1825, Stenochinae Kirby, 1837, and Tenebrioninae. All Atacama genera of Pimeliinae belong to a clade with worldwide distribution which is recovered as sister to Akis Herbst, 1799 (Akidini Billberg, 1820) and Pimelia Fabricius, 1775 (Pimeliini Latreille, 1802) from the Mediterranean region. The Atacama Pimeliinae are separated into a lineage containing South American genera of Elenophorini Solier, 1837, Nycteliini Solier, 1834, Physogasterini (Lacordaire, 1859), Praociini Eschscholtz, 1829, and Stenosini Schaum, 1859 and a second lineage including all remaining Pimeliinae from the Atacama Desert. Both clades also include taxa from other regions. Internal branches of the first clade generally show high support. This clade is first divided into a group containing Stenosini and a group containing the Elenophorini, Nycteliini, Physogasterini, and Praociini from the Atacama Desert. Within the Stenosini with Chilean species of Discopleurus and Hexagonochilus Solier, 1851 nests Mediterranean Leptoderis Billberg, 1820 (Elenophorini) as sister to Mediterranean Dichillus (Stenosini). The only described member of the Elenophorini from the Atacama region, Psammetichus Latreille, 1829, appears on a branch with the southern African Eurychora Thunberg, 1789 (Adelostomini Solier, 1834) and Mediterranean Sepidium Fabricius, 1775 (Sepidiini Eschscholz, 1829). The remaining taxa of this large clade branch into Mediterranean Alphasida Escalera, 1905 (Asidini Fleming, 1821) and the South American Nycteliini, Physogasterini, and Praociini with representatives from the Atacama Desert. The Nycteliini, which appear as sister to Praociini and Physogasterini, are represented by the genera Auladera Solier, 1836, Callyntra Solier, 1836, Nyctelia Laterille, 1825, and Psectrascelis Solier, 1836, and the sister taxa thereof, Gyriosomus Guérin-Méneville, 1834 + Pilobalia. While the Physogasterini Philorea Erichson, 1834 + (Physogaster Lacordaire, 1830 + Entomochilus Solier, 1844) occur as monophyletic in our analysis, the Praociini are polyphyletic, with Gyrasida Koch, 1962 as sister to (Praocis Eschscholtz, 1829 + Falsopraocis Kulzer, 1958) + Physogasterini. Antofagapraocis occurs as sister to the latter group. The topology of the second lineage with Pimeliinae from the Atacama Desert shows Caenocrypticoides Kaszab, 1969 (Caenocrypticini, Koch 1958) separated from the rest. The remaining taxa split into a heterogeneous group comprising southern African Zophosini Solier 1834 and Adesmiini Lacordaire, 1859, Mediterranean Erodiini Billberg, 1820, and Tentyriini Eschscholtz 1831 and a branch with the Atacama genera of Edrotini Lacordaire, 1859, Epitragini Blanchard, 1845, Evaniosomini Lacordaire, 1859, Thinobatini Lacordaire, 1859, and Trilobocarini Lacordaire, 1859. Within the latter branch the topology shows with maximum branch support the evaniosomin Melaphorus Guérin-Ménéville, 1834 + Evaniosomus Guérin-Ménéville, 1834 and Aryenis Bates, 1868 as sister to Trilobocara Solier, 1851 (Trilobocarini) and these four taxa appear as sister to the rest of this clade. Within these remaining taxa the epitragin Geoborus Blanchard, 1842 + Nyctopetus Guérin-Ménéville, 1831 (Central Chile) and Salax Guérin-Méneville, 1834 (Trilobocarini) are sister to Achanius, Arthroconus Solier, 1851, Aspidolobus Redtenbacher, 1868, Cordibates, Eremoecus Lacordaire, 1859, Hylithus Guérin-Méneville, 1834, and Thinobatis. While [Hylithus (Edrotini) + Thinobatis (Thinobatini)] + Cordibates (Thinobatini) form a well-supported monophyletic group, the sister group relationships of Achanius (Evaniosomini), Arthroconus (Edrotini), Aspidolobus (Epitragini), and Eremoecus (Trilobocarini) are not fully resolved.

The sister group of Pimeliinae includes in our analyses the subfamilies Lagriinae, Stenochinae, Blaptinae, Alleculinae, Tenebrioninae, and Diaperinae; the latter two being non-monophyletic. The three analyzed taxa of Lagriinae (incl. Adeliini Kirby, 1828, Cossyphini Latreille, 1802, Lagriini Latreille 1825; without representatives in the Atacama Desert) form the sister group to the remaining species of this clade, which in turn is separated into Tenebrio + [Bolitophagus Illiger, 1798 + Eledona Latreille, 1797] from Europe (both Bolitophagini Kirby 1837) and the rest. The latter group contains Tribolium + European Melanimon (Melanimini Seidlitz, 1894) as sister to the remaining taxa. These remaining taxa are further divided into Blaptinae (incl. Blaptini Leach, 1815, Dendarini Mulsant & Rey, 1854, Opatrini Brullé, 1832, Pedinini Eschscholtz, 1829, Platynotini Mulsant & Rey, 1853) with Blapstinus Dejean, 1821 (Opatrini) from the Atacama Desert and a second clade which consists of Alleculinae, Diaperinae, Stenochinae, and several Tenebrioninae. The first branch of that diverse clade separates European Nalassus Mulsant, 1854 (Tenebrioninae: Helopini Latreille 1802) from the rest, which is further separated into Stenochinae (without representatives in the Atacama Desert) and a clade consisting of Alleculinae, Diaperinae, and Tenebrioninae. Within the latter, some members of the polyphyletic Diaperinae (Crypticini Brullé, 1832 and Hypophlaeini Billberg, 1820 from Europe) form together with A. diaperinus (Tenebrioninae) the sister to the rest. The latter clade splits into monophyletic Alleculinae (incl. Alleculini Laporte, 1840, Cteniopodini Solier, 1835) with an undescribed species from the periphery of the Atacama Desert (Alleculinae gen. nov.) and a subclade containing further Diaperinae and Tenebrioninae. The Diaperinae of this subclade, including Phaleria (Phaleriini Blanchard, 1845) from the beaches of the Atacama Desert and Holarctic Diaperini, are sister to the Mediterranean Scaurus Fabricius, 1775 and a clade consisting of Scotobiini Solier, 1838 /Amphidorini LeConte, 1862 from the Atacama Desert and the Neotropical Z. atratus (Tenebrionini). Within the latter group the genus Nycterinus Eschscholtz, 1829 (incertae sedis) is sister to Scotobiini (Ammophorus Guérin-Ménéville, 1830 + [Scotobius Germar, 1824 + Diastoleus Solier, 1838]) and Z. atratus.

Overall, in the neuropeptide tree few branches show low support (Fig. S1). These branches include the position of Achanius to Arthroconus (SH-aLRT = 3.4, UFBoot = 43), Salax as sister to Nyctopetus + Geoborus (SH-aLRT = 14.7, UFBoot = 65), Praocis + Falsopraocis (SH-aLRT = 51.4, UFBoot = 89), Auladera as sister to Callyntra + Psectrascelis (SH-aLRT = 38, UFBoot = 79), Sepidium + Psammetichus (SH-aLRT = 3, UFBoot = 47), Diaperis Geoffroy, 1762 + Neomida Latreille, 1829 (SH-aLRT = 6.9, UFBoot = 86), Nestorinus Guerrero, Vidal & Zúñiga-Reinoso, 2022 + Heliofugus Guérin-Méneville 1831 (SH-aLRT = 28.9, UFBoot = 82), Isomira Mulsant, 1856 as sister to Omophlus Dejean, 1834 + Heliotaurus Mulsant, 1856, and Diaperini/Phaleriini as sister to a clade with Scotobiini/Nycterinus Solier, 1835 /Zophobas Dejean, 1834 + Scaurini Billberg, 1820 (SH-aLRT = 62.9, UFBoot = 91).

All analysed taxa of Alleculinae, Blaptinae, Diaperinae, and Stenochinae, as well as those taxa of Tenebrioninae that nest within the sister clade of Blaptinae (this clade is marked with an arrow in Fig. 2), have a distinct synapomorphy in common, namely an insertion of eight amino acids in the myosuppressin precursor (Fig. 3A; see Data S3 for full sequences). This insertion does not result from differential transcription, but it is indeed manifested at the gene level. This could be verified by genome sequencing of a N. abdominalis specimen and a subsequent comparison of the myosuppressin gene structures (exons) of N. abdominalis and Tr. castaneum (Fig. 3B).

(B) Analysis of a large scale dataset of orthologous genes

The partitioned and concatenated alignment is composed of 1742 OGs with an overall length of 788,676 amino acid sites (Data S2). The best fitting models according to ModelFinder are listed for each partition in Data S2, which also contains the concatenated alignment. The topology of the resulting tree (Fig. 4) is largely congruent with that of the neuropeptide precursor data set. Differences are mainly observed for several of those branches with low support in the neuropeptide precursor tree (see Fig. S1): Salax as sister to a clade comprising Achanius, Arthroconus, Aspidolobus, Cordibates, Eremoecus, Hylithus, and Thinobatis; Praocis as sister to Falsopraocis + Physogasterini; Auladera + Nyctelia as sister to Callyntra + Psectrascelis; Sepidium as sister to Psammetichus + Eurychora; Alleculinae as sister to Scotobiini/Nycterinus/Zophobas + Scaurini; Nestorinus as sister to Heliofugus + Cuphotes Champion, 1887; and Isomira + Prionychus as sister to Omophlus + Heliotaurus. In addition, Discopleurus is sister to a main branch of Pimeliinae (Fig. 4), including, among other tribes, also the Stenosini; and Nycterinus changed its position and was recovered as sister to Diastoleus + Scotobius. In the large scale data set of orthologous genes, the branches with low support (Fig. S2) include that with Zophobas as sister to Nycterinus, Scotobius and Diastoleus (SH-aLRT = 8.2/UFBoot = 61). In both data sets, Corticeus has the same position, but the corresponding branches are very long.