Shauna wins enough votes from the audience in both rounds win a Princess Key. Ash and friends find and heal an injured Spoink in the Lumiose Badlands. The Spoink takes them to the area's oasis, where Team Rocket and a Grumpig bully have taken over the food and water. The Grumpig chases the group into a cave where Team Rocket traps Ash and friends inside a prison cell. Ash escapes, and while he defends Goomy from Grumpig, Goomy evolves into Sliggoo. Sliggoo's defeats Grumpig with its Dragon Breath attack.
Grumpig returns the stolen food to the rest of the oasis' Spoink and promises not to steal again. Ash and friends sneak inside. Luxio evolves into Luxray. A Parting of the Ways! Jessie and Wobbuffet are separated from James and Meowth after being sent flying by Pikachu's Thunderbolt. A doctor named White saves Jessie from drowning in a river. Jessie falls in love with White and decides to quit Team Rocket to live with him. Ash and Serena temporarily team up with Team Rocket to defeat the hunter's Rhyperior.
A Splendid Performance Battle!! She challenges Serena to a double battle to show how Serena, Pancham and Fennekin can work as a team. The battle between the two performers is interrupted when Aria is called to an appointment elsewhere. During the battle, Fennekin evolves into Braixen. Alain and Steven must find a way to stop an apocalyptic clash between Groudon, Kyogre and Rayquaza before these titans turn the Hoenn Region into a ruin.
Ash reunites with Tierno and decides to prepare for his Lumiose Gym match by challenging Tierno to a double battle. When Pumpkaboo starts a forest fire while trying to catch Pikachu, Sliggoo stops the fire with its Rain Dance. Citroid VS Black Citroid!! Shitoroido tai Burakku Shitoroido!! Ash and the others arrive in Lumiose City and reunite with Clemont and discover that a robot builder called Belmondo had invented a Dark Clembot robot, which he had been using to commit crimes all over the city to frame Clemont's Clembot.
Clemont discovers Clembot's blueprints were stolen in a hack of the Prism Tower computer. He traces the hack to Belmondo. The Clembots are forced to rest after Team Rocket fails to kidnap them. Clemont battles Belmondo himself and wins after Belmondo fails to treat his Magnezone with respect. Pikachu beats Clemont's Bunnelby by hitting the ground with Iron Tail while Bunnelby digs underground, then finishes Bunnelby by hitting it with another Iron Tail attack. Ash then uses Hawlucha's speed to beat Heliolisk during the split-second Heliolisk leaves its body open to attack. Clemont rejoins Ash, Serena and Bonnie on their journey.
However, when Team Rocket crashes the experiment and gets away with the Mega Stone and Garchomp, Ash attaches a tracking device to them for him and the others to follow Team Rocket's escaping mecha. Team Rocket tries using a mind control ray on Garchomp before attempting to make it Mega Evolve. Goodra protects its friends from being attacked by Florges.
When it seems that Goodra has won, Jessie's Pumpkaboo and James's Inkay create a distraction, allowing Florges to retreat. Numelgon, Go Over the Rainbow!! Numerugon Niji no Kanata ni!!
Team Rocket deceives Florges into an alliance to help the sickly Floette it is taking care of. They double-cross Florges when they drain all the water in the spring Florges had taken over to heal Floette. Goodra uses its Rain Dance and Florges uses its Grassy Terrain to restore the dried up wetlands to life.
Believing that Goodra belongs in the wetlands, Ash decides to leave it behind. Bonnie and Dedenne find themselves lost in the forest with Meowth, with Bonnie and Meowth getting stuck in a vine together. Bonnie gets knocked unconscious by Foongus's Spore attack. Meowth gets Dedenne to find a Chesto Berry to wake Bonnie up. The three of them get washed down a river, which leads to Bonnie and Meowth being freed.
Ash and his friends find themselves in a mansion called the Scary House to shelter from the rain. Its owner Lon offers them some food. A Gastly, Haunter and Gengar then scare Ash and his friends as a playful way of showing hospitality. Ash, Pikachu, Serena and Bonnie fall down some trapdoors into the basement.
File The Quest For Truth
After rescuing them, Clemont finds a year-old journal which talks about the secret room the trapdoors lead to which was created for Lon to hide in when he was a boy. Ash and the others are shocked to discover a year-old photograph of Lon and that Lon is no longer alive. Afterwards, everyone wakes up in the forest, with Ash, Serena and Clemont unsure if they were dreaming. Ash and his friends finally arrive in Laverre City and immediately head over to the Gym only to find it closed due to its seasonal Fashion Show.
They instead meet Sawyer, who had a Gym battle with Clemont recently and gets to challenge the Gym Leader Valerie during the show. The Beautiful Fairy Trap!! Ash challenges Valerie to a two-on-two Gym battle for the Fairy Badge. Sawyer accidentally loses his notebook, which is found by Team Rocket who think it belongs to Steven Stone, the Hoenn Region Champion and that the information inside it will help them with their plans. Before parting ways, Ash and Sawyer decide to have a three on three battle where the former helps give the novice trainer gain useful experience.
Ash wins the battle, during which Sawyer's Treecko evolves into a Grovyle. When Hawlucha hands the egg to Ash, it hatches into a Noibat. Although Noibat is unable to fly very well, Team Rocket plots to take it and make it evolve into a more useful Noivern.
Join Kobo & start eReading today
Hawlucha tries teaching Noibat how to glide on the wind currents when Team Rocket steals both of them. Hawlucha and Noibat break free and hide in a dark cave. Aided by Noibat's echolocation, Hawlucha knocks off Team Rocket's noise-canceling headphones. Noibat frightens Team Rocket out of the cave with its Supersonic attack. After Ash defeats Team Rocket, he invites Noibat to come with him and catches it. Team Rocket, expecting Noibat to evolve during the race, enters the race with Pumpkaboo, Meowth piloting a Pelipper-shaped machine , and Inkay to steal Noibat when it evolves.
Their plan fails during the second leg of the race when Meowth's machine collides with Jessie and James. Ash and Noibat narrowly lose to Orson, the previous Sky Relay winner, in the anchor leg. How to Pikachu The Movie! Ash and his friends arrive at a village, where they meet a group of Pikachu and their owners, Frank and Jean. The story is set when Hoopa's companions, Baraz and Meray, were still children. Everyone there loved Hoopa's rings, which could make anything and everything appear!
One day, the three of them were helping an old woman make donuts. Hoopa's eyes started glittering at the sight of these round treats it had never seen before! When Hoopa's rings get used to make donuts, could things get silly? Pikachu gets in a fight with Meowth over the suitcase at the factory, which Pikachu wins. A Captivating Fiery Performance!! Serena wins her first Princess Key after the audience votes Pancham and Braixen's performance in the freestyle round their favorite.
They meet a timid Rotom once owned by the front desk clerk Weston. Rotom uses a television to show Ash and the others that ten years ago at the founding of the hotel, Weston had lost ownership of the hotel to Mantle by forfeiting a match after Rotom was scared off. Rotom sends Ash and his friends ten years into the past.
Mantle apologizes for his wrongdoing. Ash and the others return to the present, and discover Mantle was given a job at the hotel by Weston ten years ago. Jessie's Pumpkaboo then evolves into Gourgeist. Prince Pumpkaboo rejects Gourgeist, having preferred her when she was a Pumpkaboo. Jessie and Count Pumpka both agree to have Gourgeist traded back to Jessie. Ash and his friends convince Count Pumpka that Team Rocket had tricked him. Ash and his friends ride on some Mamoswine to climb a snowy mountain which they have to pass through to get to Anistar City.
The Mamoswine lead them to a cave where an Abomasnow is taking care of a sick Snover. Ash and his friends go looking for plants that grow on the mountain to make medicine to help Snover. Hajimete no o Tsukai!! Clemont starts repairs on the generator, while Ash fixes the roof. Silurian millipedes exhibit the earliest direct evidence for air breathing in the form of spiracles Wilson and Anderson This geological antiquity has brought Myriapoda to the forefront in considerations of the timing of terrestrialization in animals and plants Kenrick et al.
Such questions of timing naturally intersect with phylogenetics in the realm of molecular dating. Recent timetrees for myriapods in the context of arthropods as a whole estimate the origin of Myriapoda by the early Cambrian and the divergence of its 4 classes in the late Cambrian Rota-Stabelli et al. The open phylogenetic questions about how the main groups of myriapods are related limit how we interpret the timing of the diversification of a major component of the soil arthropod biota.
We thus present a large injection of Illumina transcriptome data for myriapods, including previously unsampled millipede orders and a substantially expanded taxonomic coverage for centipedes. Furthermore, we investigate missing data in relation to potential confounding factors in phylogenomic reconstruction e.
Concurrently, we present a morphological character set for the same set of species as sampled transcriptomically and code a set of key fossil species for their preserved morphological characters in order to place them phylogenetically on the myriapod tree These data sets with a precise placement of the fossils permit divergence times to be better explored for Myriapoda. Twenty-five species representing 3 of the 4 major groups of myriapods Diplopoda, Chilopoda, and Symphyla were collected and newly sequenced for this study.
Our sampling was designed to maximize representation at the ordinal level in millipedes and at the family level in centipedes. Information on sampling localities and accession numbers in the Sequence Read Archive database for each transcriptome can be found in Table 1 and in the Dryad package for this article. The following taxa were used as outgroups: In addition, a seventh chelicerate outgroup, Anoplodactylus insignis Pycnogonida , was newly sequenced for this study.
All data sets included in this study were sequenced with the Illumina platform. In addition, we retrieved all available sequence data from the pauropod Euripauropus spinosus Regier et al. S1 available on Dryad at http: For further details about collection site, BioProject and BioSample accession numbers and links to the MCZ database please see the Dryad package associated to this manuscript.
Sequenced results were quality filtered accordingly to a threshold average quality Phred score of 30 and adaptors trimmed using Trimgalore v 0. Each sample was sequentially aligned to the index allowing up to 2 mismatches via Bowtie 1. Strand-specific de novo assemblies were done individually for each specimen in Trinity Haas et al. The path reinforcement distance is the minimum read overlap required for path extension in the De Bruijn graph in the Trinity assembly.
Higher values reduce the probability of constructing chimeric contigs. Resulting assemblies were processed in TransDecoder Haas et al. Peptide sequences with all final candidate ORFs were retained as multifasta files. In order to explore the trade-off between number of genes and matrix completeness, 3 supermatrices were constructed by varying gene occupancy thresholds: For this purpose, we analyzed the signal-to-noise distribution of the genes of supermatrix III using the method described in Townsend et al.
This method estimates the state space and the evolutionary rates of characters to approximate the probability of phylogenetic signal versus noise due to convergence or parallelism. At every site, based on the rate of character evolution and the character state space, phylogenetic signal is characterized by the probability of observing a parsimony informative synapomorphic site pattern at the leaves of the taxa, whereas phylogenetic noise is characterized by the probability distribution function over time for homoplastic site patterns that mimic the correct pattern and mislead analyses.
Supermatrix IV included the 40 genes with the highest values of phylogenetic informativeness see Supplementary Fig. S2, available on Dryad. Likewise, to construct supermatrix V, we included only the genes in the 2 upper quartiles in their ranking of phylogenetic informativeness 62 orthologs. To account for the effect of evolutionary rate, we ordered the orthogroups in supermatrix I based on their increasing rate and selected the most conserved genes supermatrix VI; Similar criteria have been applied to construct matrices for exploratory purposes in other phylogenomic analyses e.
Only supermatrix III showed signs of non-heterogeneous gene composition measured as gene putative function , with 40 genes corresponding to ribosomal proteins. Two multigene submatrices were then constructed with identical taxon sampling, relatively comparable lengths and missing data percentages but different gene content analogous to the analyses of Nosenko et al. These steps were not necessary in the remaining matrices as they were derived from supermatrices I and III.
The aligned, masked orthogroups were then concatenated using Phyutility 2. In order to further test for the effect of heterotachy and heterogeneous substitution rates, we also analyzed some of the matrices in PhyML v. In this analysis, the starting tree was set to the optimal parsimony tree and the FreeRate model Soubrier et al. Convergence of chains was assessed both at the level of the bipartition frequencies with the command bpcomp and the summary variables displayed in the trace files with the command tracecomp.
MH370: ‘No one will ever know the truth’
For practical reasons and due to the similar results obtained for the different phylogenetic analysis see Results and Discussion , not all the analyses were implemented in all the supermatrices but at least 1 ML and 1 Bayesian inference analysis per supermatrix were explored Fig. Summary of analyses of myriapod relationships. Checked matrices in each node represent high nodal support for the different analyses in supermatrices I—VIII see Material and Methods for further information. Each matrix is represented by a different color, following the legend of the figure.
The abbreviation of the analyses in each matrix is as follows: Filled squares indicate nodal support values higher than 0. White squares indicate lower nodal support; visually, every matrix filled with color indicates that all analyses support that node. Nodal support for the clades represented with letters A—U, and the alternative topologies to the 3 conflicting nodes named 1, 2, and 3 are shown. The relative composition frequency variability RCFV measures the absolute deviation from the mean for each amino acid for each taxon It is calculated as the summation of the difference between the frequency of each amino acid in a specific taxon and the mean frequency of that amino acid over all taxa, and divided by the total number of taxa Zhong et al.
The higher the RCFV value, the more the amino acid composition of an individual sequence differs from the overall trend in a data set. Gene trees were decomposed into quartets with SuperQ v. All Python custom scripts can be downloaded from https: As our data set includes several ancient lineages in the order of several hundreds of Ma with a low diversity of extant species, we further evaluated the potential effect of long branch attraction LBA in 3 lineages that show less extant diversity than their extant closest relatives: Craterostigmomorpha, Scutigeromorpha, and Polyxenida.
Previous morphological and molecular analyses have recovered Scutigeromorpha as the sister group to all other centipede orders e. The order comprises 3 families and a total of ca.
A fine WordPress.com site
More striking is the case of Craterostigmomorpha, with only 2 extant species see Results and Discussion. These 2 orders originated more than Ma Murienne et al. In millipedes, the order Polyxenida includes only 89 species, and it is universally recognized as sister group to the rest of Diplopoda.
As the earliest diverging lineages of both centipedes and millipedes involve long branches, we also explored the effect of LBA in a fourth lineage with long branches and poor taxon representation, Symphyla, which was recovered as sister group to Polyxenida in some of our analyses see Results and Discussion. This method is based on the removal of taxa in a pair suspected to be affected by LBA, one at a time.
If after re-running the analysis either of the taxa appears at different branch points in the absence of the other, LBA is postulated. As rogue taxa can frequently have a negative impact on topology or in a bootstrap analysis see a review in Goloboff and Szumik , we also explored the presence of putative wild card taxa in our data set with RogueNaRok Aberer et al. Individual gene trees were run in RAxML 8. Best-scoring ML trees were inferred for each gene under the selected model from replicates of parsimony starting trees.
One hundred bootstrap replicates for each gene were also inferred. ASTRAL estimates the species tree that has the maximum number of shared induced quartet trees with the given set of gene trees. Morphological characters for the set of species for which transcriptomes were available or newly generated were coded, principally drawing on existing data sets. These include characters bearing on myriapod phylogeny Rota-Stabelli et al. The objective of analyzing the morphological data set was to establish the systematic position of fossils used for calibration.
As such, node calibration is based on hypotheses that are consistent with the precise taxonomic sampling used for molecular analyses e. Sources for morphological information used for coding fossil terminals are detailed in Supplementary Material. These data were analyzed using equally weighted parsimony and implied weighting Goloboff in TNT Goloboff et al. In order to infer divergence time in the myriapod phylogeny, we included 7 Palaeozoic and Mesozoic fossils in our data set: In addition, we included 2 outgroup fossils, a crustacean Rehbachiella kinnekullensis , and a scorpion Proscorpius osborni.
Fossils used for calibration and minimum age in millions of years employed for dating the respective crown groups. Temporal data for the fossils listed above provide constraints on divergence dates. The resolution of Crussolum as stem-group Scutigeromorpha in our morphological cladogram Fig. S3 available on Dryad constrains the split of Scutigeromorpha and Pleurostigmophora, i. To allow for uncertainty over the phylogenetic position of Devonobius and Craterostigmus , the former was conservatively used only to constrain crown-group Pleurostigmophora minimum of Crown-group Epimorpha is constrained to a minimum of That species is resolved as total-group Scolopendromorpha based on our morphological analysis Supplementary Fig.
S3 available on Dryad. Kachinophilus pereirai , originally assigned to Geophilidae Bonato et al. As such, it constrains crown-group Adesmata, i. Crown-group Chilognatha is constrained by the occurrence of Cowiedesmus eroticopodus , resolved in our morphological analysis as total-group Helminthomorpha S3 available on Dryad , they do not unambiguously contribute to calibrate additional nodes.
Rehbachiella kinnekullensis has been interpreted as stem-group Anostraca, i. Our morphological data agree with either of these in resolving it as more closely related to Daphnia than to Calanus. This provides a minimum date for crown-group Altocrustacea sensu Regier et al. Because of uncertainty with regards to the interrelationships between major pancrustacean groups including Branchiopoda, Copepoda and Insecta, from which the 3 exemplars used here were sampled this date is conservatively applied only to crown-group Pancrustacea.
Proscorpius osborni is resolved by our morphological data data set as closest relative of the extant scorpion exemplar Centruroides , setting a constraint on the split of Scorpiones from Tetrapulmonata. The relevant crown-group based on our taxon sampling is Arachnopulmonata, dated to at least Nodes that were calibrated with fossils are indicated with a star placed at the age of the fossil. The split between Onychophora and Arthropoda was dated between million years the minimum age for Arthropoda used by Lee et al. The latter was used as the root of Panarthropoda Lee et al.
Divergence dates were estimated using the Bayesian relaxed molecular clock approach as implemented in PhyloBayes v. An auto-correlated relaxed clock model was applied as it has been shown to provide a significantly better fit than uncorrelated models on phylogenomic data sets Lepage et al.
The calibration constraints specified above were used with soft bounds Yang and Rannala under a birth-death prior in PhyloBayes, this strategy having been found to provide the best compromise for dating estimates Inoue et al. Two independent MCMC chains were run for — 7, cycles, sampling posterior rates and dates every 10 cycles. Posterior estimates of divergence dates were then computed from the remaining samples of each chain. We unsuccessfully tested 2 different software tools for total evidence dating analysis with supermatrix III gene-matrix: The chains did not reach convergence, despite multiple trials with different parameters.
All analyses conducted, regardless of methodology or data set, recover monophyly of Myriapoda, Diplopoda, Chilopoda, and Symphyla, but the interrelationships among the myriapod classes are less stable Fig. With regards to the position of Symphyla, 2 alternatives are found: Symphyla either unites with Diplopoda as predicted by the Progoneata hypothesis morphological evidence discussed by Dohle ; Edgecombe , or Chilopoda and Diplopoda unite as a clade to the exclusion of Symphyla.
The latter result was also found in many of the analyses by Rehm et al. In addition, analyses of supermatrices IX and X including ribosomal and non-ribosomal genes from supermatrix III, respectively suggest that the grouping of Symphyla and Diplopoda is driven by the ribosomal proteins: S5a, available on Dryad , which is obviously artifactual.
On the contrary, the ML hypothesis of supermatrix X i. S5b, c, d available on Dryad. A second factor influencing the positioning of Symphyla is rooting. S6 available on Dryad. S7 available on Dryad. Rogue taxa did not appear to influence this result, none being detected by RogueNaRok. A chilopod—diplopod clade has not been anticipated morphologically, though certain morphological characters fit such a grouping.
For example, chilopods and diplopods share a series of imbricated comb lamellae on the mandibles that are lacking in symphylans and pauropods Edgecombe and Giribet A second conflicting area concerns the relationships among the 5 orders of centipedes. Half of the analyses retrieved the widely accepted division of Chilopoda into Notostigmophora Scutigeromorpha and Pleurostigmophora the other 4 orders.
However, in the other half, Scutigeromorpha and Craterostigmomorpha were recovered as a clade. The latter result has never been advocated morphologically. Both clades are almost invariantly at the base of the tree, as recovered in all the SAW analyses Supplementary Fig. S4 available on Dryad.
However, their interrelationships remain unresolved. A possible cause of this phylogenetic conundrum could be the low diversity of Craterostigmomorpha, which comprises only 2 extant species, and the long branch to the origin of the clade. S6E available on Dryad. By exploring different data matrices with different occupancy, we were able to identify potential discordance between some of our matrices, which led to some of the additional analyses removing distant outgroups and identifying the type of proteins that went into these matrices.
These results are difficult to disentangle from the problem of missing data, inversely correlated to gene occupancy. Assessing the effect of missing data on phylogenetic inference has received substantial attention from phylogeneticists over many years e. Properties of the data sets themselves, such as different rates of evolution and compositional heterogeneity, can have a strong influence on the accuracy of phylogenetic inference, irrespective of the amount or the pattern of missing data, whereas for many phylogenomic data sets large amounts of missing data have not been a major problem. The combined effect of these factors complicates the problem, and no study has dissected out in depth the effect of each parameter independently.
The different matrices constructed and the multitude of analyses conducted to address the potential effects of missing data, compositional heterogeneity, heterotachy, differential evolutionary rate, levels of phylogenetic informativeness, model misspecification, concatenation, gene-tree incongruence, gene composition and rooting issues yield largely congruent results see Fig. S2, S3, S8, S9, available on Dryad. The main 3 conflicting nodes marked as 1, 2, and 3 in Fig. The hypotheses most strongly supported by the current data set, after considering all these analyses are summarized in Figure 4.
Schematic summary of the interrelationships within Myriapoda based on a diversity of phylogenomic analyses. Lineages presently lacking transcriptomes shown in light color their placement is based on published data sets from morphology or targeted sequencing. Dashed lines indicate conflict between analyses. Recent studies on the effects of missing data in phylogenetic reconstruction have focused on data sets with hundreds or thousands of orthologous genes e.
Missing data may reduce detection of multiple substitutions, exacerbating systematic errors such as LBA, as incomplete species are less efficient in breaking long branches Roure et al. Perhaps for this reason other studies have shown that adding incomplete taxa is not deleterious per se as long as enough informative character states are present for each species, but analyzing too few complete characters could reduce accuracy because of a lack of phylogenetic signal Wiens , ; Philippe et al.
In this study, we show that high matrix occupancy, in some specific cases, can lead to anomalous inferences: Reasons for this may lay in the high expression level of these genes, which therefore are differentially selected when using transcriptomes for phylogenomic analyses even though they are not necessarily the genes with the highest phylogenetic informativeness. Also, evidence indicates that most or all ribosomal proteins are encoded by two or more highly similar gene family members at least in plants: Our finding that matrices with low levels of missing data show high levels of intragene conflict and yield unusual tree topologies that conflict with trees based on other data sets may have profound implications for the experimental design of phylogenomic data sets.
The conflicting nodes between centipede orders are unlikely to result from incomplete taxon sampling, as our data set includes all extant families of centipedes with the exception of 2 geophilomorph families Zelanophilidae and Gonibregmatidae and 1 monotypic family of Scolopendromorpha Mimopidae. In this context, biological explanations for the incongruence, such as incomplete lineage sorting consistent with a scenario of rapid radiation, or lineage-specific high extinction, may need to be considered. In fact, Craterostigmomorpha, the taxon responsible for some of the major instability, comprises only 2 species in Tasmania and New Zealand, constituting an old lineage with depauperate extant diversity and conservative morphology, some of the conditions often used to refer to living fossils Werth and Shear The 4-gene analysis including the sequences available for a pauropod recovered Pauropoda as sister group to Symphyla with a posterior probability of 0.
S1 available on Dryad. Within Diplopoda, most analyses but see Supplementary Fig. S5 and S6 available on Dryad. Chilognatha is united by a calcified cuticle and the male actively transferring sperm to the female genital opening, among other apomorphic characters Enghoff ; Blanke and Wesener The fundamental division within Chilognatha likewise corresponds to traditional morphology-based systematics, including a sister group relationship between Pentazonia generally pill millipedes and Helminthomorpha long-bodied millipedes.
Relationships within Pentazonia depart from the standard morphological hypothesis. The latter unites Glomerida and Sphaerotheriida in a clade named Oniscomorpha. S5a, b available on Dryad. The only previous molecular studies to include all 3 orders of Pentazonia are the 3-gene analysis of Regier et al. Monophyly of Oniscomorpha has been defended based on the presence of a collum much smaller than the following tergites, the second tergite being much larger than following tergites, and on the lack of coxal pouches on legs Blanke and Wesener The status of these characters as synapomorphies is disputed by our trees.
In Helminthomorpha, the traditional clades Colobognatha and Eugnatha are each monophyletic, as found in prior analyses of largely the same set of millipede species Brewer and Bond , as well as by sampling different exemplars for 3 nuclear coding genes Miyazawa et al. Colobognatha and Eugnatha were originally proposed in the classical era of myriapod systematics Attems ; Verhoeff , but were discarded in the influential millipede classification by Hoffman , who considered that the characteristic reduced mouthparts of Colobognatha were prone to convergence.
They have, however, consistently been endorsed in morphological cladistic analyses Enghoff ; Sierwald et al. As likewise found by Brewer and Bond , Eugnatha includes 2 groups that have a long history of usage in morphology-based systematics. One of these unites Chordeumatida Cleidogona and Callipodida Abacion , which are classified together as Coelochaeta. Another clade consists of the ring-forming millipedes, Juliformia.
Our analyses variably resolve the 3-taxon problem of the juliform orders with either Julida most analyses or Spirobolida as sister group of Spirostreptida. Two contentious issues in millipede systematics involve the relationships of Polydesmida and of Stemmiulida. The latter group, named for the shared presence of preanal spinnerets, traditionally includes Stemmiulida, Callipodida, and Chordeumatida Enghoff ; Sierwald et al. Our results, like those of Brewer and Bond , do not retrieve monophyly of Nematophora in its traditional guise: Optimization on the topology of Brewer and Bond as well as those obtained here suggests that spinnerets are an autapomorphy of Eugnatha as a whole and were lost in Juliformia.
All chilopod orders are well supported and relationships within each of the large orders correspond to previous phylogenetic hypotheses, and for most to traditional taxonomy. The 3 families of Scutigeromorpha resolve with Pselliodidae Sphendononema as sister group to Scutigerinidae Scutigerina and Scutigeridae Scutigera. The same topology has consistently been inferred using targeted sequencing of a few loci Edgecombe and Giribet , ; Butler et al.
Lithobiomorpha comprises 2 clades that correspond to its 2 families, Henicopidae and Lithobiidae. Scolopendromorpha consists of a blind clade and an ocellate clade. The number of events of eye loss in Scolopendromorpha had been a subject of some debate Vahtera et al. All analyses herein resolve Scolopocryptopidae as paraphyletic, Scolopocryptops being more closely related to Theatops Plutoniumidae than to Newportia. This result implies that the shared presence of 23 leg-bearing trunk segments in Scolopocryptopinae Scolopocryptops and Newportiinae Newportia is either acquired convergently from a segmented ancestor or is homologous but was reversed to 21 segments in Plutoniumidae.
Monophyletic Scolopendridae consists of Otostigminae Alipes and Rhysida and Scolopendrinae Akymnopellis and Scolopendropsis , precisely mirroring classical and current taxonomy. Geophilomorpha also resolves along traditional lines into Placodesmata Mecistocephalidae: Mecistocephalus and Tygarrup and Adesmata, a clade composed of all other geophilomorphs. Resolution within Adesmata conflicts with the most recent analysis of geophilomorph relationships Bonato et al.
Our 3 Geophilidae representatives of Geophilidae and 2 recently synonymized families Dignathodontidae and Linotaeniidae constitute a well-supported and stable clade. The dated phylogeny for Supermatrix III reflects a diversification of Myriapoda in the early and middle Cambrian, supporting a Cambrian diversification event, as suggested recently for this lineage Rota-Stabelli et al.
This dating continues to substantially predate body fossil or even trace fossil evidence for Myriapoda, the oldest trackways that can be ascribed to myriapods with reasonable confidence being Late Ordovician in age Johnson et al. Diversification of Diplopoda is inferred to occur in the late Cambrian to mid Silurian, and the diversification of Chilopoda in the Early Ordovician to Middle Devonian Fig.
The diversification times of the millipede and centipede orders included in our analyses are thus generally congruent with the timing inferred in previous studies Meusemann et al. On the other hand, an estimated Cambrian diversification of myriapods implies a large gap in the fossil record.
Fossil candidates for stem-group Myriapoda in marine, freshwater or terrestrial sediments of Cambrian age remain unknown or unidentified Edgecombe ; Shear and Edgecombe Nonetheless, the occurrence of crown-group Pancrustacea as early as Cambrian Stage 3 Edgecombe and Legg predicts a ghost lineage for Myriapoda to at least that time, ca. In this study, the inferred diversification dates ranged from the Early Ordovician to the Middle Devonian for Pleurostigmophora, from the Middle Ordovician to the Early Carboniferous for Amalpighiata, and from the Late Devonian to the mid Carboniferous for Epimorpha.
The controversy over the systematic position of Craterostigmus i. Despite a divergence from other living chilopod orders by the Late Silurian, the 2 extant species of Craterostigmus have a mean date for their divergence from each other in the Cretaceous, and these species are almost indistinguishable morphologically Edgecombe and Giribet Likewise, Scutigeromorpha, with which Craterostigmus groups in various analyses, has a stem dating to the Silurian, but the deepest split between its 3 extant families has a mean date in the Triassic.
Although the group is also conservative morphologically, we have maximized phylogenetic diversity here for a clade that has been well resolved phylogenetically and which increased its diversification rate around million years ago Giribet and Edgecombe