The phylogeny of fossil whip spiders
© The Author(s). 2017
Received: 14 September 2016
Accepted: 9 March 2017
Published: 21 April 2017
Arachnids are a highly successful group of land-dwelling arthropods. They are major contributors to modern terrestrial ecosystems, and have a deep evolutionary history. Whip spiders (Arachnida, Amblypygi), are one of the smaller arachnid orders with ca. 190 living species. Here we restudy one of the oldest fossil representatives of the group, Graeophonus anglicus Pocock, 1911 from the Late Carboniferous (Duckmantian, ca. 315 Ma) British Middle Coal Measures of the West Midlands, UK. Using X-ray microtomography, our principal aim was to resolve details of the limbs and mouthparts which would allow us to test whether this fossil belongs in the extant, relict family Paracharontidae; represented today by a single, blind species Paracharon caecus Hansen, 1921.
Tomography reveals several novel and significant character states for G. anglicus; most notably in the chelicerae, pedipalps and walking legs. These allowed it to be scored into a phylogenetic analysis together with the recently described Paracharonopsis cambayensis Engel & Grimaldi, 2014 from the Eocene (ca. 52 Ma) Cambay amber, and Kronocharon prendinii Engel & Grimaldi, 2014 from Cretaceous (ca. 99 Ma) Burmese amber. We recovered relationships of the form ((Graeophonus (Paracharonopsis + Paracharon)) + (Charinus (Stygophrynus (Kronocharon (Charon (Musicodamon + Paraphrynus)))))). This tree largely reflects Peter Weygoldt’s 1996 classification with its basic split into Paleoamblypygi and Euamblypygi lineages; we were able to score several of his characters for the first time in fossils. Our analysis draws into question the monophyly of the family Charontidae.
Our data suggest that Graeophonus is a crown group amblypygid, and falls within a monophyletic Paleoamblypgi clade, but outside the family Paracharontidae (= Paracharonopsis + Paracharon). Our results also suggest a new placement for the Burmese amber genus Kronocharon, a node further down from its original position. Overall, we offer a broad phylogenetic framework for both the fossil and Recent whip spiders against which future discoveries can be tested.
Fossil whip spiders
Whip spiders are rare as fossils. The oldest potential evidence for this group comprises some Middle Devonian (Givetian: ca. 390 Ma) cuticle fragments named Ecchosis pulchribothrium Selden & Shear, 1991 which have a trichobothrium on the patella —a character otherwise only seen in living Amblypygi today. Unequivocal Palaeozoic members of the clade include five species from the Late Carboniferous Coal Measures (ca. 315–305 Ma) of Europe and North America. Scudder  described an isolated opisthosoma from Cape Breton in Nova Scotia as Libellula carbonaria Scudder, 1876 (his Figure 2, referred to as Fig. 1 in the text) and interpreted it as a dragonfly larva. Based on the discovery of a better preserved specimen from Mazon Creek in Illinois (; see also ). Scudder later reinterpreted this find as a whip spider and created a new genus Graeophonus Scudder, 1890 to accommodate it as Graeophonus carbonarius. In a widely overlooked move, Pocock  did not feel that Scudder’s two North American specimens were conspecific and renamed the younger one Graeophonus scudderi Pocock, 1911. In the same monograph focussing on the British Middle Coal Measures Pocock  described Graeophonus anglicus Pocock, 1911 which—as the best preserved species—is the focus of the present study. The species is known from several well-preserved specimens . Dunlop  described a Carboniferous whip spider from the Writhlington Geological Nature Reserve known only from its ventral surface. A full overview of historical work on Palaeozoic taxa is provided by Dunlop et al. .
The next amblypygids found in the fossil record are Cretaceous in age and come from the ca. 115 Ma Crato Formation of Brazil—Dunlop, & Martill  described a species known from a ventral prosoma and limbs, including distinctive amblypygid pedipalps, and Dunlop & Barov  augmented this with further details of the sternal region and details of the walking limbs. These confirmed its referral to the modern family Phrynidae. Engel & Grimaldi  documented a further Cretaceous species, Kronocharon prendinii Engel & Grimaldi, 2014 from the ca. 99 Ma Burmese amber from Myanmar, with high levels of morphological detail preserved. In the same publication these authors described a whip spider from the Eocene (ca. 52 Ma) Cambay amber of India—Paracharonopsis cambayensis Engel & Grimaldi, 2014, which has a similarly high fidelity of preservation. The youngest records are from Miocene (ca. 16 Ma) Dominican Republic amber [19, 20] and the probably contemporary Chiapas amber of Mexico [21, 22]. Petrunkevitch’s amber species was recently shown to be a nomen dubium , while at least the Dominican amber material is barely distinguishable from a modern species of Phrynus Lamarck, 1801 (Phrynidae) found in the Caribbean.
Crown groups, stem groups and microtomography
When studying fossils a key question is when the oldest crown-group representatives of a given clade first appear (i.e. species descended from the most recent common ancestor of all extant members of that group). These data provide calibration points for molecular clock studies estimating dates of cladogenesis. Approaches vary, but time-calibrated phylogenies have traditionally been created using node-dating priors for relaxed clocks , and recent developments have shown that fossils can be coded directly into morphological matrices based on extant taxa, and used as calibration points—so-called total-evidence or tip-dating . Coal Measures arachnids are particularly interesting for these purposes given both their great antiquity—more than 300 million years—and often surprisingly good preservation, particularly in siderite nodules. For example, recent molecular clock work on harvestman (Opiliones) has been successful in demonstrating that Carboniferous fossils which unequivocally resolve as members of two modern suborders (Eupnoi and Dyspnoi) lived alongside harvestmen with extinct body plans placed in a new suborder, Tetrophthalmi [26, 27].
Both these harvestman studies were aided by the application of X-ray microtomography: a technique that allows digital visualisation of the void left by the organism within the nodule. It is a methodology that has proved especially useful in revealing distal details of appendages, as well as ornamentation such as spines or tubercles, which are often buried deep in the host nodule, and are difficult to resolve using traditional methods of light microscopy [28–32].
Three specimens of Graeophonus anglicus, were scanned from the Natural History Museum, London (NHMUK) collections: NHM In31234, NHM In31248, and NHM In31257. All originate from the Coseley Lagerstätte, a productive site from which specimens were collected in the late 1800s and early 1900s. The fossils—including these specimens—are typically preserved as three-dimensional voids within siderite concretions, sometimes with partial kaolinite infill. The deposit is Late Carboniferous (Bashkirian/Moscovian: Duckmantian) in age, or Westphalian B in traditional terminologies. The Duckmantian straddles the Bashkirian/Moscovian boundary, which Pointon et al.  place at around 315 Ma. Fossils were compared to Recent whip spiders in the collections of the Museum für Naturkunde Berlin. Comparative photographs of a syntype of the rare living species Paracharon caecus Hansen, 1921 (Fig. 2) were kindly provided by Jan Pedersen and Nikolaj Scharff from the Statens Naturhistoriske Museum, Københavns Universitet.
Cladistic Taxa and Characters Added
In order to assess the phylogeny of the fossil whip spiders we first re-ran the matrix of Weygoldt , but were unable to reproduce these results, recovering instead a polytomy with no reliable placement for fossil taxa. This is likely to result from differences in the search strategies and consensus tree calculations between Hennig86 and TNT, the latter used in the current study. Instead we chose to code fossil whip spiders into the matrix of Garwood et al , modified after Garwood & Dunlop , and based on the matrix of Pepato et al. , which in turn built on that of Giribet et al. . In order to obtain internal resolution within the amblypygids, and to fully encompass their morphological disparity, we added numerous characters. The are fully described in the supplementary character statements (Additional file 2), and we summarise relevant character additions here. Character 38 reflects the number of cheliceral teeth within the amblypygids, which is either four or three depending on taxon (after Weygoldt ; character 1), and character 39 whether the distal-most tooth is bicuspate (; character 2). Character 53 reflects the presence of a serrula; modified setae found on the cheliceral free finger in amblypygids and Schizomida. Character 54 further codes for whether the cheliceral serrula is more rounded—as in the whip spiders and the two Protoschizomidae genera of the schizomids—or toothlike as seen in the schizomid family Hubbardiidae (after Cokendolpher & Reddell ).
A key difference seen in the pedipalps of Paracharon and Graeophonus, in contrast with other whip spiders, is a more vertical plane of motion (Figs. 2 and 3), associated with a presumably plesiomorphic pediform state for these limbs. The modification in plane of motion is probably associated with prey capture, and has been added as character 62. Much of the internal resolution within Amblypygi is based on the arrangement and nature of the spines of the pedipalp, associated with the limbs’ raptorial nature. Character 64 reflects whether the pedipalps have a row of femoral spines (; character 7), character 65 on whether the trochanter has a distinctive ventral apophysis, and character 66 on whether this is present as a spine (; character 4). Character 69 has been modified to reflect whether palpal tarsus is subdivided or fused. This character is modified after Weygoldt (; character 14); we highlight that our pedipalp tarsus is equivalent to Weygoldt’s distitarsus (the two parts of which are fused within the Neoamblypygi clade; see below), and distal part of this we term the apotele. Character 71 reflects the absence or presence of a dorsal row of patellar spines on the pedipalp which, in some groups, form a catching basket, and in others (e.g. Phrynichidae) is more of a prehensile structure (a distinction coded in character 75). This row often consists of three principal spines, with the distal-most largest, which is coded in character 72 (; character 8). Whether these sequentially decrease in length towards the base (i.e. proximally) is coded as absent or present in character 73 (; character 10), and the subsequent character, 74, records if the most proximal spine of the three principal ones is highly reduced (; character 11). Character 76 codes for the presence of spines on the palpal tarsus.
We have additionally added a limb character to reflect the whip-like first limbs of whip spiders; character 83 is coded based on the number of subdivisions of the tibia of the first leg. This segment was chosen as it is more likely to be preserved in fossils, being more proximal than the (also subdivded) tarsus, and also on the basis of Weygoldt (; character 18). We recognise five states: unmodified, 16 segments, up to 23, 25 segments, and more than 25 segments. Whilst this is largely only informative for the living species, it can also be coded for the amber fossils Paracharonopsis from the description of Engel & Grimaldi . Whilst a rough figure for a Kronocharon species figured by Wunderlich  is obtainable, we have coded this as unknown for this species due to difficulty in differentiating the podomeres. We highlight that as part of ongoing development of the matrix we have added a number of further characters. These are listed in the character statements supplementary file (Additional file 2).
The previously published arachnid phylogeny of Garwood et al  included three whip spider genera; Charinus, Musicodamon, and Paraphrynus, as well as a generic coding for Amblypygi as a whole. In order to include a representative of every extant family—and to test the position of the extinct Kronocharon in more detail—we have added the genera Charon, Paracharon and Stygophrynus. The former were coded from Weygoldt ([5, 34]; character 18). We note that the new photos of the type specimen of Paracharon, suggest lateral eyespots are present (Fig. 2a). Hence we have coded lateral eyes as present herein, but the number of lenses and the nature of the rhabdomes as unknown. Stygophrynus  was coded on the suggestion of a reviewer, as it differs in some important aspects of the pedipalp from Charon even though both genera are currently placed in the same family. We also added the three known and reasonably complete fossil genera: Graeophonus, coded from the results herein and redescription of Dunlop et al. , and Paracharonopsis and Kronocharon from Engel & Grimaldi . We omitted the three species described by Petrunkevitch  as we consider the species descriptions unreliable, and the species described by Schawaller [19, 20] which is essentially modern.
We analysed this matrix using TNT v.1.1. ( ; made available with the sponsorship of the Willi Hennig Society), using a traditional search and unordered multistate characters. The data matrix is available as Additional file 3 in a TNT-file ready format. Our searches comprised tree bisection-reconnection [TBR] with 1000 replicates, saving 100 trees per cycle. These were carried out under equal and implied weights. For all analyses, TNT was used to create a strict consensus tree which was exported as an SVG into Inkscape. For our implied weights analyses, we present the strict consensus of an arachnid-wide analysis at a k value of 3, and for the amblypygids, an additional strict consensus of all tree topologies recovered from 88 k values ranging from 0.001 to 122.0. We further explored the matrix by running the analyses with differing taxa and characters excluded to explore their impact. Resampling of all analyses was conducted within TNT. For our equal weight analysis, we provide jackknife (; 33% removal probability, 10,000 replicates), bootstrap ( ; 10,000 replicates) and Bremer support . For the implied weights analysis, k =3, supports are included through symmetric resampling ( ; change probability 33%, 10,000 replicates). Where applicable, nodal support values are shown as absolute frequencies.
Additional file 4: A video showing the tomographic reconstruction of Graeophonus anglicus Pocock, 1911 presented herein; inferred/reconstructed anatomy partially transparent. (AVI 87.2 mb)
The chelicerae of Graeophonus anglicus were also largely equivocal in previous studies, but the CT data reveals them as two-segmented, clasp-knife structures similar to those of living species (Fig. 3e). In modern whip spiders the fang opposes a series of internal teeth (Fig. 1), whose number and structure can be phylogenetically informative (; character 13). The CT scan was able to resolve the presence of four internal teeth in G. anglicus, although the most proximal one may not be preserved at its full length. Also of significance is the fact that, as in Paracharon, the uppermost or distal tooth has only one cusp. In all other living whip spiders the upper tooth has two cusps. The coxo-sternal region is well-preserved and for the first time the CT scan picks up the small coxae of leg I (Fig. 3a). The coxae of legs II-IV are very clearly preserved and coxa II in particular is quite tuberculate (Fig. 3b).
Pedipalp morphology is a key character for resolving whip spider relationships. Previous studies  recognised two dorsal spines on the femur of the pedipalp in Graeophonus anglicus. This is again similar to the condition in Paracharon caecus (Fig. 2c) and was used as evidence for referring the Carboniferous fossil to Paracharontidae. The CT scans allowed us to investigate pedipalp morphology in more detail (Fig. 3c, d, f, g). We could confirm the general observation that the pedipalps in G. anglicus primarily articulate up and down in a vertical plane (Fig. 3b). This is more like the condition in P. caecus (Fig. 2e); in more derived whip spiders the pedipalps primarily articulate from side to side.
In detail, we could resolve that the trochanter of the pedipalp has a ventral apophysis (Fig. 3d,f)—a character for all Amblypygi—and that in Graeophonus anglicus it is more like a flange and not reduced to a spine (; character 4). We could confirm that, like Paracharon caecus (Fig. 2c), the femur of the pedipalp in G. anglicus lacks a prominent row of dorsal spines (Weygoldt , character 7) and has only two short spines here (Fig. 3g). Weygoldt , like many authors, referred to the next pedipalp article in whip spiders as the tibia, however we follow Shultz  and recognise a more conventional series of articles for an arachnid pedipalp: namely a femur, patella, tibia, tarsus and apotele. In this scheme, we can argue that G. anglicus has two patellar spines (Fig. 3f). It also has, like P. caecus (Fig. 2c), only two prominent tibial spines (Fig. 3c). However, in the photograph of P. caecus (Fig. 2c) there is a small (proximal) tibial spine which was not shown in Weygoldt’s drawings. Three prominent tibial spines are seen in more derived living species (; character 8). The distal tip of the pedipalp in G. anglicus appears to be offset against the preceding element forming a pretarsus or apotele (Fig. 3g). In Weygoldt’s  character 14 terminology this would be referred to as a “divided distitarsus”. We prefer to interpret and score it as a tarsus which is (still) separate from its apotele. In more derived whip spiders (namely the Unidistitarsata; see below) the tarsus and apotele are fused.
As in living whip spiders, the first leg of Graeophonus anglicus appears to be long and slender. The CT scan can trace leg I down to the first few articles of what is clearly a subdivided tibia (Fig. 3a,b, Additional file 1). This implies that leg I was antenniform as in Recent taxa, but the exact number of tibial segments (; character 18) remains equivocal. Legs II-IV of G. anglicus are more complete, and III and IV can now be traced down to near their tips (Fig. 3a-b). Some legs show evidence for a pair of tarsal claws, but the presence or absence of a pad-like pulvillus between these claws in G. anglicus is equivocal. The femur is quite robust and, at least in legs II and III, it is preserved with the prolateral surface uppermost. This may well reflect its orientation in life since modern whip spiders have a similar leg configuration to help them to crawl into narrow spaces. The patella is small and bell shaped. The tibia of leg IV is divided into a basi- and distitibia as expected, and the distitibia itself appears to be undivided (; character 23). The opisthosoma is oval with a series of tergites (Figs. 3a-b and 4) which are shorter towards the anterior and posterior ends. Ventral sacs on the underside of the opisthosoma are equivocal. A reconstruction of the likely appearance in life is presented in Fig. 4.
Cladistic results and discussion
Traditionally, whip spiders were broadly divided into so-called pulvillate and apulvillate taxa (e.g. ) based on the presence or absence of this small fleshy pad called the pulvillus at the ends of legs II-IV. Weygoldt  tested this hypothesis cladistically and recovered (Paracharontidae (Charinidae (Charontidae (Phrynichidae + Phrynidae)))), rendering the ‘pulvillate’ taxa paraphyletic. As an alternative way of expressing this, he recognised two suborders: Paleoamblypygi (for Paracharontidae) and Euamblypygi for the other four families. The euamblypygids were further subdivided into the infraorders Charinidae (for the family of the same name) and Neoamblypygi. Finally, neoamblypygids were divided into the superfamilies Charontoidea (for Charontidae) and Phrynoidea (for Phrynichidae and Phrynidae). Note that the phrynoids are the most derived group in this scheme, and equivalent to the ‘apullvillates’ in Quintero’s earlier scheme. Weygoldt’s model was largely adopted in subsequent classifications, such as the catalogue of Harvey , albeit with the modification of the infraorder name Charinidae to Charinina in Prendini . Within the neoamblypygids, the recent study of amber fossils by Engel & Grimaldi  introduced another clade name, Unidistitarsata, which encompassed their extinct genus Kronocharon Engel & Grimaldi, 2014 + Phrynoidea (but see below).
Looking at the character distribution on this tree in detail, Amblypygi is defined in our analysis by six apomorphies, although not all of them can be seen in all the fossils. These are: (1) the presence of serrula on the chelicerae, (2) a ventral apophysis on the pedipalp trochanter, (3) presence of a palpal cleaning organ, (4) a dorsal row of patellar spines on the pedipalp, (5) presence of a pulvillus on legs II-IV, albeit with a reversal in the most derived genera, and (6) the presence of ventral sacs. Note that serrula, a pulvillus and ventral sacs are not unique for Amblypygi and can also occur in some other arachnid orders.
The analysis supports a monophyletic suborder Paleoamblypygi sensu Weygoldt , consisting of Graeophonus, Paracharon and Paracharonopsis (Fig. 5). A single putative apomorphy supports Paleoamblypygi in our analysis, namely the anterior projection at the front of the prosomal dorsal shield. This is not seen in the Euamblypygi (see below), in which the anterior margin is straight or only slightly rounded. Were this carapace projection to be interpreted as a plesiomorphic character, we must entertain the possibility that the paleoamblypygids are a grade rather than a clade, but this was not the most parsimonious result in the present analysis. Outgroup comparison does suggest that other paleoamblypygid features like the vertically articulating pedipalps are probably plesiomorphic (as it is for other Pedipalpi).
In contrast to Dunlop et al. , we now formally exclude Graeophonus from Paracharontidae (see Systematic Palaeontology) and treat it as a stem-paleoamblypygid. Our justification for this is that Graeophonus anglicus has a much simpler pattern of pedipalp spines compared to other whip spiders. Characters relating to this spination explicitly group Paracharon closer to Paracharonopsis (see below). This result is similar to Tetlie & Dunlop’s  conclusions about the Coal Measures whip scorpions. The earliest fossil whip scorpions lack projecting apophyses on the pedipalp. These define the (Cretaceous-Recent) whip scorpion family Thelyphonidae as a crown group which can be defined by more raptorial—and presumably more efficient—pedipalp used for prey capture. Likewise, the palp of Graeophonus suggests that whip spiders also seem to show a trend in which the pedipalps become increasingly better adapted for restraining their prey (Fig. 6).
With respect to the other Coal Measures whip spiders, the original figures of Sorellophrynus carbonarius (Petrunkevitch, 1913) from Mazon Creek suggest that this genus also has the paleoamblypygid character of a carapace projection. The status of Thelyphrynus elongatus Petrunkevitch, 1913, also from Mazon Creek, is harder to assess. The carapace is figured as more rounded anteriorly, but the original figures imply that the carapace and pedipalp are not well-preserved. In lieu of a formal redescription we provisionally place this genus as a paleoamblypygid too.
The family Paracharontidae is restricted here to Paracharon and Paracharonopsis (Fig. 6). It is defined in our dataset by the presence of several small spines on the pedipalp tarsus (Fig. 2c), a character which is—convergently in our analysis—present in Charinus too. This character is reversed in the more derived whip spiders in which the tarsus and apotele also fuse into a single (and spineless) tip. More generally, we suggest that the pedipalps of the amber and living palaeoamblypyid genera are more heavily spined than the pedipalps of Graeophonus anglicus. For example the patella bears three spines in Paracharontidae, but only two in G. anglicus (Figs. 3 and 6).
Weygoldt’s  suborder Euamblypygi—i.e. (Charinus (Stygophrynus ((Kronocharon (Charon (Musicodamon + Paraphrynus))))) among our terminal taxa—is defined in our dataset by (1) pedipalps which articulate primarily in a horizontal plane and (2) a row of dorsal spines on the pedipalp femur, rather than just a couple of isolated spines as in Paleoamblypygi; see also Fig. 6. The biology of living Paracharon caecus is unknown but we suspect that both this species, and the fossil paleoamblypygids, grab down on prey items immediately in front of them. Almost all living whip spiders are euamblypygids. We presume that the horizontal, side-to-side, action of their pedipalps conferred a considerable evolutionary advantage, enabling broad sweeping movements to capture prey which is further away. The elongation of the pedipalps in derived euamblypygids supports this general hypothesis.
The next clade recovered in our analysis is (Stygophrynus (Kronocharon (Charon (Musicodamon + Paraphrynus)))). The principal apomorphy which supports this group is the tendency for the distal ventral patellar spines on the pedipalp to form a ‘catching basket’ (subsequently modified into the phrynichid ‘hand’). It is also supported by a longer number of tibial elements in the first pair of legs and the loss of the coxal gland opening on the segment bearing the third pair of legs. Traditionally Stygophrynus and Charon were placed together as the family Charontidae. Our results challenge the monophyly of this family (see also below) especially given the fact that Stygophrynus retains a separate tarsus and apotele in the pedipalp while Charon has the more derived character of a fusion of these elements. In our analysis the two extant genera are also split apart from each other by the extinct Kronocharon which also has a fused pedipalp tip (Fig. 6; although this does move up-tree in some implied weights analyses). To accommodate this result we could assign Kronocharon to Charontidae too, but this would still leave the family paraphyletic and containing genera having quite different pedipalp architectures. Since we were not comprehensive in covering extant genera we prefer for now to place Stygophrynus as the sister group to the Unidistitarsata (see below) and restrict Charontidae to the genus Charon as Charontidae sensu stricto. We concede that it may prove necessary to recognise a separate family for Stygophrynus, but we are reluctant to add a monogeneric family to the nomenclature at this stage.
This clade is defined in our dataset by the key character of the fusion of the tarsus and apotele into a single element. In our analysis the Burmese amber genus Kronocharon resolved as sister group to the Neoamblypygi (but see also implied weights analysis). This position for Kronocharon differs from the original interpretation of Engel & Grimaldi , who proposed that the fossil genus was the sister group of the Phrynoidea (i.e. Musicodamon + Paraphrynus), forming the Unidistitarsata based on the diagnostic apomorphy of an undivided pedipalp tip (Fig. 6). However, we contend that—contra Engel & Grimaldi —Charon also has, like phrynoids, an undivided distitarsus, as shown in Figure 15 of Weygoldt , and reflected in the scoring of his character 14 (see also  for example). As noted above Stygophrynus, by contrast, retains a pedipalp tip which is divided. The most derived whip spiders also have the greatest number of elements in the leg I tibia. It may be possible to score the number of tibial elements in leg I for Kronocharon based on Wunderlich , who reported 65 elements beyond the patella. However, since the exact number of tibial as opposed to tarsal elements is not given here we have chosen to treat this character in the amber genus as equivocal.
The neoamblypgids, (Charon (Musicodamon + Paraphrynus)) from our selected terminal taxa, are defined in this analysis by the apomorphy of a spine-like ventral apophysis on the trochanter of the pedipalp. By contrast in Kronocharon this apophysis is described as being a large and carina-like , rather than explicitly a spine. As noted above, in our scheme Charontidae is no longer monophyletic, and we recognise Charontidae sensu stricto (consisting of Charon only) as the sister-group to Phrynoidea.
Finally the two most derived genera in our dataset (Musicodamon and Paraphrynus) belong to the Phrynoidea (or Phrynida), which is defined here by (1) the reduction of four cheliceral teeth to three and (2) the loss of the pulvillus. Indeed, as noted above, an older name for the same clade in the literature is Apullvillata.
Suborder PALEOAMBLYPYGI Weygoldt, 1996
Genus †Graeophonus Scudder, 1890
†Graeophonus anglicus Pocock, 1911
Description as Dunlop et al. : here we primarily focus on novel morphological features revealed through microtomography (Fig. 3). Total length 11.5 mm. Prosomal dorsal shield reniform, but with wide anterior projection; shield length 4.6 mm, maximum width 5.3 mm. Dorsal shield with median longitudinal depression (the fovea) and several depressions radiating out from this structure. Pair of median eyes on the anterior projection, borne on a tear-drop shaped tubercle; small lateral eye tubercles in an anterolateral position can also be resolved; number of individual lenses equivocal.
Chelicerae small, of the ‘clasp-knife type, projecting forwards beyond the anterior margin prosomal shield and composed of two articles (Fig. 3e): a basal element (or paturon; min. 8 mm long) opposed by a gently curving and tapering distal fang (1 mm in length). Basal element bears four teeth, the most distal of which has only one cusp. Pedipalps robust and bearing numerous spines (Fig. 3c,d,f,g); total length ~7 mm. Trochanter with flange-like ventral apophysis. Femur broadens distally and bears only two small dorsal spines. Patella broad, slightly procurved on the mesal margin and here bearing two spines, the proximal one slightly shorter (0.3 mm) than the distal one (0.6 mm). The same article has a further distal spine on the outer surface (0.4 mm). Tibia with two mesal spines, proximal shorter (0.3 mm) than distal (0.6 mm), and a prominent outer distal spine. Tarsus separate from terminal apotele; both without spines.
Legs gracile, leg I antenniform. Leg I coxa small, triangular. Leg I trochanter more than twice as wide as long. Leg 1 femur slender, narrowing slightly distally. Leg 1 patella small, bell shaped. Leg 1 tibia incomplete but was evidently subdivided; more distal elements equivocal but overall habitus implies an antenniform appendage. Legs II-IV more robust. Coxae subtriangular; trochanters bell-shaped. Femora somewhat flattened, apparently with thin margins and becoming slightly narrower distally. Femur IV with hook-like ventral projection at its distal margin. Patellae short and bell shaped. Tibia slender and divided into a basi- and distitibia; these elements are not further subdivided. Basitarsus undivided, but tarsus divided into three tarsomeres, with terminal claws visible in leg III. Pulvillus equivocal. Tritosternum or additional expected sternal elements between the leg coxae equivocal.
Family PARACHARONTIDAE Weygoldt, 1996
Suborder EUAMBLYPYGI Weygoldt, 1996
Family CHARINIDAE Quintero, 1986
Clade UNIDISTITARSATA Engel & Grimaldi, 2014
Infraorder NEOAMBLYPGI Weygoldt, 1996
Superfamily CHARONTOIDEA Simon, 1892
Family CHARONTIDAE Simon, 1892 sensu stricto
Superfamily PHRYNOIDEA Blanchard, 1852
Family PHRYNIDAE Blanchard, 1852
Family PHRYNICHIDAE Simon, 1892
Damon C.L. Koch, 1850; Euphrynichus Weygoldt 1995; Musicodamon Fage, 1839; Phrynichodamon, Weygoldt 1996; Phrynichus Karsch, 1879; Trichodamon Mello-Leitao, 1935; Xerophrynus Weygoldt, 1996.
Tomographic investigation of the Carboniferous amblypygid Graeophonus anglicus reveals the least modified pedipalps of any whip spider, living or extinct. These appendages still primarily articulate up and down, in common with typical arachnid walking legs. They have relatively few spines to facilitate prey capture: namely two small dorsal spines on the femur, two larger spines on the patella and three (two mesal, one lateral) on the tibia. G. anglicus is placed in the Carboniferous–Recent suborder Paleoamblypygi (Fig. 7), which can be defined based on the projecting anterior part of the carapace (Fig. 6). However, the fossil species can be excluded from the Eocene–Recent group Paracharontidae which has more pedipalp spines; specifically three patella spines and additional small tarsal spines. Most whip spiders belong to the Cretaceous–Recent suborder Euamblypygi (Fig. 7), in which the pedipalps primarily articulate from side to side, i.e. with a horizontal plane of motion. The Cretaceous phrynid from the Crato Formation of Brazil implies that all three euamblypyid families should have been present in the Cretaceous (Fig. 7). The relationships we recover (Fig. 5) largely support the 1996 phylogeny of Weygoldt, although the recently described Burmese amber genus Kronocharon is placed a node deeper in the tree compared to its original interpretation. This analysis demonstrates that whip spider phylogeny is effectively reflected in the evolution and modification of the group’s pedipalps (Fig. 6). Trends which can be recognised include (1) fusion of the tarsus and apotele to define the Unidistitarsata, (2) appearance of a spine-like ventral apophysis on the pedipalp trochanter, and a tendency to concentrate the patella spines distally into a ‘catching basket’ which define the Neoamblypygi, and (3) loss of the pulvillus on the legs and reduction in the number of cheliceral teeth defining the Phrynoidea. Several of the most derived phrynoid genera also have extremely long pedipalps, in which the distally highly concentrated patellar spines form the so-called phrynichid ‘hand’. These results provide a framework for testing the position of future fossil discoveries.
We thank Jan Pedersen and Nikolaj Scharff (Statens Naturhistoriske Museum, Københavns Universitet), for providing photographs of Paracharon, and Claire Mellish (Natural History Museum, London) for providing access to fossils in her care. RJG is a member of the Interdisciplinary Centre for Ancient Life (UMRI). We are grateful for the comments of two anonymous reviewers, which improved the manuscript.
RJG’s visit to Berlin was supported by a European Union SYNTHESYS grant. We would like to acknowledge the assistance provided by the Manchester X-ray Imaging Facility, which was funded in part by the EPSRC (grants EP/F007906/1, EP/F001452/1 and EP/I02249X/1). BJK’s work was supported by a WIU Undergraduate Research Grant and the WIU Department of Geology.
Availability of data and materials
The dataset supporting the conclusions of this article are available in a Zenodo repository (DOI: 10.5281/zenodo.437900).
The datasets supporting the conclusions of this article are included within the article (and its additional files).
RJG and JAD wrote most of the text and descriptions and carried out the cladistic analysis. Tomographic images and the reconstruction were produced by RJG, BJK and TAH. All authors contributed intellectually to the paper. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
All authors have provided consent to publish.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Garwood RJ, Dunlop J. Three-dimensional reconstruction and the phylogeny of extinct chelicerate orders. PeerJ. 2014;2:e641.View ArticlePubMedPubMed CentralGoogle Scholar
- Shultz JW. A phylogenetic analysis of the arachnid orders based on morphological characters. Zool J Linn Soc. 2007;150:221–65.View ArticleGoogle Scholar
- Sharma PP, Kaluziak ST, Pérez-Porro AR, González VL, Hormiga G, Wheeler WC, et al. Phylogenomic interrogation of Arachnida reveals systemic conflicts in phylogenetic signal. Mol Biol Evol. 2014;31:2963–84.View ArticlePubMedGoogle Scholar
- Shultz JW. Muscular anatomy of a whipspider, Phrynus longipes (Pocock) (Arachnida: Amblypygi), and its evolutionary significance. Zool J Linn Soc. 1999;126:81–116.View ArticleGoogle Scholar
- Weygoldt P. Whip Spiders. Stenstrup: Apollo Books; 2000.Google Scholar
- Harvey MS. Catalogue of the smaller arachnid orders of the world. 1st ed. Collingwood: CSIRO Publishing; 2003.Google Scholar
- Harvey MS. Whip Spiders of the World, version 1.0 [Internet]. West. Aust. Mus. Perth. 2014. Available from: http://museum.wa.gov.au/catalogues-beta/whip-spiders. Accessed 5 Sept 2016.
- Prendini L. Amblypygi Thorell, 1883. Anim Biodivers Outl High-Level Classif Surv Taxon Richness Zootaxa. 2011;3148:154.Google Scholar
- Selden PA, Shear WA, Bonamo PMP. A spider and other arachnids from the Devonian of New York, and reinterpretations of Devonian Araneae. Palaeontology. 1991;34:241–81.Google Scholar
- Scudder SH. New and interesting insects from the Carboniferous of Cape Breton. Can Nat Q J Sci. 1878;8:88–90.Google Scholar
- Scudder SH. Illustrations of the Carboniferous Arachnida of North America of the Orders Anthracomarti and Pedipalpi. Mem Boston Soc Nat Hist. 1890;4:443–56.Google Scholar
- Petrunkevitch AI. A monograph of the terrestrial Palaeozoic Arachnida of North America. Trans Conn Acad Arts Sci. 1913;18:1–137.Google Scholar
- Pocock RI. A monograph of the terrestrial Carboniferous Arachnida of Great Britain. Monogr Palaeontogr Soc. 1911;64:1–84.Google Scholar
- Dunlop JA, Zhou GRS, Braddy SJ. The affinities of the Carboniferous whip spider Graeophonus anglicus Pocock, 1911 (Arachnida: Amblypygi). Earth Environ Sci Trans R Soc Edinb. 2007;98:165–78.Google Scholar
- Dunlop JA. An Upper Carboniferous amblypygid from the Writhlington Geological Nature Reserve. Proc Geol Assoc. 1994;105:245–50.View ArticleGoogle Scholar
- Dunlop JA, Martill DM. The first whipspider (Arachnida: Amblypygi) and three new whipscorpions (Arachnida: Thelyphonida) from the Lower Cretaceous Crato Formation of Brazil. Trans R Soc Edinb Earth Sci. 2002;92:325–34.View ArticleGoogle Scholar
- Dunlop JA, Barov V. A new fossil whip spider (Arachnida: Amblypygi) from the Crato Formation of Brazil. Rev Ibérica Arachnol. 2005;12:53–62.Google Scholar
- Engel MS, Grimaldi DA. Whipspiders (Arachnida: Amblypygi) in amber from the Early Eocene and mid-Cretaceous, including maternal care. Novit Paleoentomologicae. 2014;9:1–17.Google Scholar
- Schawaller W. Erstnachweis der Ordnung Geißelspinne in Dominikanischem Bernstein:(Stuttgarter Bernsteinsammlung: Arachnida, Amblypygi). Stuttg Beitr Zur Naturkunde Ser B Geol Paläontol. 1979;50:48–66.Google Scholar
- Schawaller W. Neue Befunde an Geißelspinne in Dominikanischem Bernstein (Stuttgarter Bernsteinsammlung: Arachnida, Amblypygi). Stuttg Beitr Zur Naturkunde Ser B Geol Paläontol. 1982;86:1–12.Google Scholar
- Petrunkevitch A. Chiapas amber spiders. II. Univ Calif Publ Entomol. 1971;63:1–44.Google Scholar
- Poinar Jr GO, Brown AE. A new whip spider (Arachnida: Amblypygi), Phrynus mexicana, is described from Mexican amber. In: Wunderlich J, editor. Beitr. Zur Araneol. 2004. p. 1881–5.Google Scholar
- Dunlop JA, Mrugalla B. Redescription of the Chiapas amber whip spider Electrophrynus mirus (Amblypygi). J Arachnol. 2015;43:220–3.View ArticleGoogle Scholar
- Warnock RCM, Yang Z, Donoghue PCJ. Exploring uncertainty in the calibration of the molecular clock. Biol Lett. 2012;8:156–9.View ArticlePubMedGoogle Scholar
- Ronquist F, Klopfstein S, Vilhelmsen L, Schulmeister S, Murray DL, Rasnitsyn AP. A total-evidence approach to dating with fossils, applied to the early radiation of the Hymenoptera. Syst Biol. 2012;61:973–99.View ArticlePubMedPubMed CentralGoogle Scholar
- Garwood RJ, Dunlop JA, Giribet G, Sutton MD. Anatomically modern Carboniferous harvestmen demonstrate early cladogenesis and stasis in Opiliones. Nat Commun. 2011;2:444.View ArticlePubMedGoogle Scholar
- Garwood RJ, Sharma PP, Dunlop JA, Giribet G. A Paleozoic stem group to mite harvestmen revealed through integration of phylogenetics and development. Curr Biol. 2014;24:1017–23.View ArticlePubMedGoogle Scholar
- Garwood RJ, Dunlop JA, Sutton MD. High-fidelity X-ray micro-tomography reconstruction of siderite-hosted Carboniferous arachnids. Biol Lett. 2009;5:841–4.View ArticlePubMedPubMed CentralGoogle Scholar
- Garwood RJ, Dunlop JA. Morphology and systematics of Anthracomartidae (Arachnida: Trigonotarbida). Palaeontology. 2011;54:145–61.View ArticleGoogle Scholar
- Garwood RJ, Sutton MD. X-ray micro-tomography of Carboniferous stem-Dictyoptera: new insights into early insects. Biol Lett. 2010;6:699–702.View ArticlePubMedPubMed CentralGoogle Scholar
- Garwood RJ, Sutton MD. The enigmatic arthropod Camptophyllia. Palaeontol Electron. 2012;15:12p.Google Scholar
- Spencer ART, Hilton J, Sutton MD. Combined methodologies for three-dimensional reconstruction of fossil plants preserved in siderite nodules: Stephanospermum braidwoodensis nov. sp. (Medullosales) from the Mazon Creek lagerstätte. Rev Palaeobot Palynol. 2012;188:1–17.View ArticleGoogle Scholar
- Petrunkevitch AI. Arachnida. In: Moore RC, editor. Treatise Invertebr. Paleontol. Part P Arthropoda 2. Lawrence: University of Kansas Press; 1955. p. 42–162.Google Scholar
- Weygoldt P. Evolutionary morphology of whip spiders: towards a phylogenetic system (Chelicerata: Arachnida: Amblypygi). J Zool Syst Evol Res. 1996;34:185–202.View ArticleGoogle Scholar
- Pointon MA, Chew DM, Ovtcharova M, Sevastopulo GD, Crowley QG. New high-precision U–Pb dates from western European Carboniferous tuffs; implications for time scale calibration, the periodicity of late Carboniferous cycles and stratigraphical correlation. J Geol Soc. 2012;169:713–21.View ArticleGoogle Scholar
- Garwood RJ, Ross A, Sotty D, Chabard D, Charbonnier S, Sutton MD, et al. Tomographic reconstruction of neopterous Carboniferous insect nymphs. PLoS One. 2012;7:e45779.View ArticlePubMedPubMed CentralGoogle Scholar
- Garwood RJ, Dunlop JA. The walking dead: Blender as a tool for paleontologists with a case study on extinct arachnids. J Paleontol. 2014;88:735–46.View ArticleGoogle Scholar
- Sutton MD, Garwood RJ, Siveter DJ, Siveter DJ. Spiers and VAXML: A software toolkit for tomographic visualisation, and a format for virtual specimen interchange. Palaeontol Electron. 2012;15:14p.Google Scholar
- Garwood RJ, Edgecombe GD, Charbonnier S, Chabard D, Sotty D, Giribet G. Carboniferous Onychophora from Montceau-les-Mines, France, and onychophoran terrestrialization. Invertebr Biol. 2016;135:179–90.View ArticlePubMedPubMed CentralGoogle Scholar
- Pepato AR, Rocha CEF, Dunlop JA. Phylogenetic position of the acariform mites: sensitivity to homology assessment under total evidence. BMC Evol Biol. 2010;10:1–23.View ArticleGoogle Scholar
- Giribet G, Edgecombe GD, Wheeler WC, Babbitt C. Phylogeny and systematic position of Opiliones: A combined analysis of chelicerate relationships using morphological and molecular data. Cladistics. 2002;18:5–70.PubMedGoogle Scholar
- Cokendolpher JC, Reddell JR. Revision of the Protoschizomidae (Arachnida: Schizomida) with notes on the phylogeny of the order. Tex Meml Mus Speleol Monogr. 1992;3:31–74.Google Scholar
- Wunderlich J. New and rare fossil Arachnida in Cretaceous Burmese amber (Amblypygi, Ricinulei amd Uropygi: Thelyphonida). Beitr Zur Araneol. 2015;9:409–36.Google Scholar
- Rahmadi C, Harvey MS. A first epigean species of Stygophrynus Kraepelin (Amblypygi: Charontidae) from Java and adjacent islands, Indonesia with notes on S. dammermani Roewer, 1928. Raffles Bull Zool. 2008;56:281–8.Google Scholar
- Goloboff PA, Farris JS, Nixon KC. TNT, a free program for phylogenetic analysis. Cladistics. 2008;24:774–86.View ArticleGoogle Scholar
- Farris JS, Albert VA, Kallersjo M, Lipscomb D, Kluge AG. Parsimony jackknifing outperforms neighbor-joining. Cladistics. 1996;12:99–124.View ArticleGoogle Scholar
- Felsenstein J. Confidence limits on phylogenies: An approach using the bootstrap. Evolution. 1985;39:783–91.View ArticleGoogle Scholar
- Bremer K. Branch support and tree stability. Cladistics. 1994;10:295–304.View ArticleGoogle Scholar
- Goloboff PA, Farris JS, Källersjö M, Oxelman B, Ramirez MJ, Szumik CA. Improvements to resampling measures of group support. Cladistics. 2003;19:324–32.View ArticleGoogle Scholar
- Petrunkevitch AI. A Study of Palaeozoic Arachnida. Trans Conn Acad Arts Sci. 1949;37:69–315.Google Scholar
- Quintero D. Revision de la clasificacion de amblypygidos pulvinados: creacion de subordenes, una nueva familia y un nuevo genero con tres nuevas especies (Arachnida: Amblypygi). In: Ebehard WG, Lubin YD, Robinson BD, editors. Proc. Ninth Int. Congr. Arachnol. Panama 1983 [Internet]. Smithsonian Institution Press; 1986 [cited 2016 Sep 5]. p. 203–212Google Scholar
- Puttick MN, O’Reilly JE, Tanner AR, Fleming JF, Clark J, Holloway L, et al. Uncertain-tree: discriminating among competing approaches to the phylogenetic analysis of phenotype data. Proc R Soc B. 2017;284:20162290.View ArticlePubMedPubMed CentralGoogle Scholar
- Congreve CR, Lamsdell JC. Implied weighting and its utility in palaeontological datasets: a study using modelled phylogenetic matrices. Palaeontology. 2016;59:447–62.View ArticleGoogle Scholar
- Tetlie OE, Dunlop JA. Geralinura carbonaria (Arachnida; Uropygi) from Mazon Creek, Illinois, USA, and the origin of subchelate pedipalps in whip scorpions. J Paleontol. 2008;82:299–312.View ArticleGoogle Scholar
- Harvey MS, West PLJ. New species of Charon (Amblypygi, Charontidae) from Northern Australia and Christmas Island. J Arachnol. 1998;26:273–84.Google Scholar