cambrian lobopodians: a review of recent progress in our understanding of their morphology and...

Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2013) xxx–xxx

PALAEO-06532; No of Pages 12

Contents lists available at SciVerse ScienceDirect

Palaeogeography, Palaeoclimatology, Palaeoecology

j ourna l homepage: www.e lsev ie r .com/ locate /pa laeo

Cambrian lobopodians: A review of recent progress in our understanding of theirmorphology and evolution

Jianni Liu a,⁎, Jason A. Dunlop b,1

a Early Life Institute, The Key Laboratory of Continental Dynamics, Department of Geology, Northwest University, Xi'an 710069, Chinab Museum für Naturkunde, Leibniz Institute for Research on Evolution and Biodiversity at the Humboldt University Berlin, D-10115 Berlin, Germany

⁎ Corresponding author. Tel./fax: +86 29 88303706.E-mail addresses: [email protected] (J. Liu), Jason.dun

1 Tel.: +49 30 2093 8516; fax: +49 30 2093 8528.

0031-0182/$ – see front matter © 2013 Elsevier B.V. Allhttp://dx.doi.org/10.1016/j.palaeo.2013.06.008

Please cite this article as: Liu, J., Dunlop, J.A.,evolution, Palaeogeography, Palaeoclimatolo

a b s t r a c t
a r t i c l e i n f o
Article history:Received 13 February 2013Received in revised form 26 April 2013Accepted 6 June 2013Available online xxxx

Keywords:CambrianLobopodiansEvolution

Lobopodians are an important group of organisms which appeared during the Cambrian Explosion. The under-lying morphology is invariably a worm-like body bearing multiple pairs of legs. Yet in detail these animals pre-serve a range of morphologies and have attracted much paleontological attention; particularly since thisassemblage probably includes the ancestors of living velvet worms (Onychophora), water bear (Tardigrada)and arthropods (Arthropoda). In recent years, knowledge of Cambrian lobopodians has increased dramaticallybased on numerous new records. However, there have been few comprehensive reviews of these animalssince Ramsköld & Chen's study in 1998. In the present paper, new insights into Cambrian lobopodians arepresented. The legs of Aysheaia pedunculata have a strong attachment with the body, like those of lobopodiansin the Chengjiang Fauna. Hallucigenia fortis has a pair of eyes, two pairs of tentacles are observed in the ‘neck’ re-gion while a bivalved head shield is unequivocally lacking. Some new characters for, and the orientation of,Hallucigenia sparsa are discussed. Longitudinal wrinkles on the body of Xenusion auerswalde are regarded hereas putative muscles. Cardiodictyon sinicum bears doublure structures at the anterior margin of head and a pairof eye spots; the shape of dorsal plates is also reinterpreted. Onychodictyon has a pair of anterior appendages,but no sclerotized head shield. The affinities of Miraluolishania haikouensis are clarified and the proposal thatM. haikouensis is a junior synonym of Luolishania longicruris is refuted. The large lobopodians, Kerygmachela,Jianshanopodia and Megadictyon – all with frontal appendages, gill-like limbs and tree-like or lamellate-likebranches – may be swimming predators.

© 2013 Elsevier B.V. All rights reserved.

1. Introduction

Since the last comprehensive studies of Cambrian lobopodians(e.g. Ramsköld and Chen, 1998; Budd, 1999; Bergström & Hou, 2001),muchnewmorphological information on the Chinese Chengjiangmate-rial in particular has come to light; some of it published by Liu et al.(2004, 2006, 2007, 2008, 2011a,b). These data – togetherwith advancesin our understanding of other Cambrian lobopodians – are summarizedhere on a genus by genus basis, including a reinterpretation of severalfeatures. Since relationships between the various taxa remain in astate of flux, we present the lobopodians in a largely chronological se-quence here based on when the taxa were established.

2. Material and methods

Lobopodian specimenswere examined either from the first author'spersonal collection, or from museum material; in particular from theSmithsonian Institution, Washington (USA), the Geological Museum,

[email protected] (J.A. Dunlop).

rights reserved.

Cambrian lobopodians: A revgy, Palaeoecology (2013), htt

Copenhagen (Denmark), and the Museum für Naturkunde Berlin(Germany). These investigations were supplemented by a review ofthe literature and individual specimens and/or publications are docu-mented in detail below. The fossils were photographed with Olympuscamera through the ocular of an Olympus SZX12 stereomicroscope.Drawings weremadewith a camera lucida on an Olympus SZX9 stereo-microscope. Figures were prepared with Adobe Photoshop CS3 andCoral Draw 11.

3. Results and discussion

3.1. Aysheaia

3.1.1. Aysheaia pedunculataAysheaia pedunculata Walcott, 1911 was originally described as a

fossil annelid worm from the Middle Cambrian Burgess Shale of Canada.It was subsequently re-interpreted as a velvetworm (Onychophora) – ora close relative thereof – by authors such as Brues (1923), Walton(1927), Hutchinson (1930), Walcott (1931), Whittington (1978) andRobison (1985). It was also compared with tardigrades (Delle Cave andSimonetta, 1975). A. pedunculata is a worm-like animal, ranging from 1to 6 cm in length and about 5 mm in width, bearing ten pairs of clawed,

iew of recent progress in our understanding of their morphology andp://dx.doi.org/10.1016/j.palaeo.2013.06.008

http://dx.doi.org/10.1016/j.palaeo.2013.06.008

mailto:[email protected]

mailto:[email protected]


http://www.sciencedirect.com/science/journal/00310182


2 J. Liu, J.A. Dunlop / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2013) xxx–xxx

spiny limbs on the ventro-lateral side of its body. Based on observationsof specimens of A. pedunculata (specimens 365608 and 83942A) in theSmithsonian Institution, we found one pair of sub-rounded structuresper body segment on the dorsal–lateral side of A. pedunculata (Fig. 1A–D), which suggests to us that the legs had a strong attachment to thebody, similar to those of lobopodians found in the Chinese ChengjiangFauna such as Paucipodia and Miraluolishania (see Sections 3.8 & 3.11).The whole morphology of A. pedunculata is somehow similar toOnychodictyon ferox (see Section 3.6) in that both of them bear tubercleson the trunk and limbs, the lobe-like limbs are short and stubby and thefrontal appendages of A. pedunculata are similar to feathery, antenniformappendages of ‘O. ferox’ (Fig. 1A–D, Ou et al., 2012, figs. 2c–e, see belowfor a discussion of Onychodictyon).

3.2. Hallucigenia

3.2.1. Hallucigenia sparsaThis iconic species was described as Canadia sparsa Walcott, 1911

and, like Aysheaia pedunculata, was originally interpreted as an anne-lid worm. Another specimen was illustrated twenty years later(Walcott, 1931). Detailed study by Conway Morris (1977) demon-strated that this material did not belong to Canadia, or to the annelidsat all. Conway Morris' famous reconstruction showed a rather bizarreanimal, walking on spines and with dorsal tentacles interpreted as afeeding apparatus. He named the new genus Hallucigenia because ofits ‘dreamlike’ appearance and also in recognition of its uncertain

Fig. 1. Aysheaia pedunculata Walcott, 1911 and Hallucigenia sparsa Conway Morris, 1977,sub-rounded structures and a pair of frontal appendages. B. Camera lucida drawing of spesub-rounded structures. E. Specimen of H. sparsa 198658. F. Camera lucida of specimen offa: frontal appendages; sp.: spine; spl: the left spine; spr: the right spine; srs: sub-rounded

Please cite this article as: Liu, J., Dunlop, J.A., Cambrian lobopodians: A revevolution, Palaeogeography, Palaeoclimatology, Palaeoecology (2013), htt

affinities. The subsequent discovery of similar-looking lobopodians inthe exceptionally preserved Lower Cambrian Chengjiang Fauna ofChina (see below) suggested that the previous reconstruction was up-side down: the spines of H. sparsa were actually designed to protectthe dorsal surface of the animal. New data showed that the supposedtentacles represented just one row of a series of paired legs; theother row being buried in the matrix (cf. Ramsköld and Hou, 1991;Ramsköld, 1992). The anteroposterior orientation was also reversed,with the former swollen ‘head’ interpreted as possible decay fluidsseeping from the body (Ramsköld, 1992).

Based on observations of specimens of H. sparsa in the SmithsonianInstitute, we can largely accept the morphological description ofConway Morris, albeit with the orientation sensu Ramsköld. However,the location of first pair of legs in H. sparsa is still debatable. Ramsköld(1992) recognized three pairs of tentacles in the ‘neck’ region, with nocorresponding legs below the first pair of spines. Later Ramsköld andChen (1998) proposed that there are two pairs of tentacles in the‘neck’ region while the third pair of tentacles represents the first pairof legs and in turn corresponds to the first pair of spines. Obviously,the problem is which (and where) are the first pair of legs?

Morphologically, the three pairs of tentacles are muchmore slenderthan the legs and they lack annulations or distal claws. Thus the thirdpair of tentacles should probably not be treated as the first pair oflegs. However, topologically each pair of spines corresponds with apair of legs, and in this scenario the third pair of tentacles would indeedcorrespond with being the first pair of legs, since they are located

from Burgess Shale. A. Complete specimen of A. pedunculata 365608 with one pair ofcimen 365608. C, D. Enlargement of specimens 365608 and 83942A, arrows indicateH. sparsa 198658; showing the three pair of tentacles. Abbreviations are as follows:structures; te: tentacle.



3J. Liu, J.A. Dunlop / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2013) xxx–xxx

ventro-laterally beneath thefirst pair of spines. If the last pair of tentaclesare the first pair of limbs, then all the legs maintain the same distancefrom each other (i.e. about 0.3 cm in specimen 198658) (Fig. 1E, F).Therefore, it is hard to judge which set of appendages is the first pairgiven that morphology and topology offer contradictory interpretations.

3.2.2. Hallucigenia fortisHallucigenia was first recorded outside the Burgess Shale by Chen

(1991) who found another species belonging to this genus in theChengjiang fauna. The Chinese species was briefly described from an in-complete specimen as H. fortis Hou and Bergström, 1995. Subsequently,Ramsköld and Chen (1998) added new data based on fifteen specimens,and noted several errors of interpretation by Hou and Bergström (1995).Our present collection of twenty-three well-preserved specimens ofH. fortis reveals some important novel data.

The well-defined head of H. fortis is preserved in fifteen of ourtwenty-three specimens. The head is sub-rounded in shape but,depending on preservation, different specimens show different appear-ance, e.g. the head of specimen ELI-JS0013 is smooth (Fig. 2A), while thehead of specimen ELI-JS0007 is rough and sclerotized (Fig. 2B). LikeCollins (2002), we recognize a pair of black spots (0.4 mm long and0.2 mmwide) in the anterior third of the head (Fig. 2A, C) quite similar

Fig. 2. Hallucigenia fortis Hou and Bergström, 1995, from Chengjiang Lagerstätte. A, B. ComELI-JS0013, showing the eye, tentacle and terminal mouth. D. Enlargement of posterior rnote the first pair of tentacles overlies the head of H. fortis and gives the impression that tseven spines directing in different ways. Abbreviations are as follows: ca: circular area; lspine; te: tentacle; tel: the left tentacle; ter: the right tentacle.


to the eyes of the lobopodian Miraluolishania haikouensis (Liu et al.,2004; Xiao, 2004; Liu et al., 2006, 2007) in both their location andtheir preserved black color. We thus interpret the black spots inH. fortisas evidence for a pair of eyes in this species too.

As noted byRamsköld and Chen (1998), a gentle neckwith twopairsof tentacles connects the sub-rounded head to the trunk (Fig. 2B). Thesetwo tentacle pairs – see above for the discussion of whether tentacles inHalucigenia represent legs – are exceedingly long and slender, and lackeither annulations or claws. Indeed, these two pairs of tentacles inH. fortis are anatomically very similar to the dorsal antennae ofM. haikouensis (Liu et al., 2004, figs. 2(a)–(d)), although we shouldcaution that the position of these structures is different in the differentanimals. Sometimes, the first pair of tentacles overlies the head ofH. for-tis (Fig. 2A–C, E) and gives the impression of bivalved sclerites coveringthe head (Hou and Bergström, 1995). However, this seems to be an ar-tifact of preservation, thus we follow the view of Ramsköld and Chen(1998) and regard this observation of a bivalved structure as erroneous.

The sub-rounded trunk bears seven pairs of dorso-lateral plates andspines, plus a short posterior projection (Fig. 2A, F). The first and lastplate pairs and spines are slightly smaller and shorter than the otherintervening plate pairs. As noted by Ramsköld and Chen (1998), thedorsal spines do not all point in the same direction: the four anterior

plete specimen ELI-JS0013 and ELI-JS0007. C. Enlargement of head region of specimenegion of specimen ELI-JS0013, showing a small projection. E. Specimen ELI-ECH001,here are bivalved sclerites covering the head. F. Specimen ELI-JSH011A with the rightb: limb; lbl: the left limb; lbr: the right limb; pl: plate; pp: posterior projection; sp:




pairs usually point forwards, while the three posterior pairs normallypoint backwards (Fig. 1F). Eight pairs of limbs are observed on thetrunk of H. fortis, the last limb pair located ventro-laterally of the shortposterior projection (Fig. 2A, D); the other limbs attach via a circulararea beneath each plate (Fig. 2B). No claws have been observed on thelimbs, although this may be an artifact of preservation.

3.2.3. Hallucigenia hongmeiaRecently, another new Hallucigenia species – H. hongmeia Steiner

et al., 2012 – was discovered in the Guanshan Biota (Series 2, Stage4, Cambrian; ca. 515–510 Ma) of China. All specimens are unfortunatelyincomplete, with the head region and the anteriormost appendages infront of plate pair 1 missing. Compared with H. fortis, the dorsal spinesof H. hongmeia show an anterior–posterior differentiation with small,cone-shape sclerites in the middle part and the largest spinose scleritesat the posterior end (Steiner et al., 2012, figs. 4, 5, 6A–J). Claws ofH. hongmeia are well-preserved (Steiner et al., 2012, figs. 6D–E, 11).

3.3. Xenusion

3.3.1. Xenusion auerswaldaeOriginally described from a glacial erratic in Germany, X. auerswaldae

Pompeckj, 1927 is a longish, metameric animal with paired legs; see alsoJaeger andMartinsson (1967). In Berlin in 2011, we had the opportunityto examine the original type specimen of X. auerswaldae, which is about90 mm long and 12 mmwide, excluding the appendages. The proboscisis not preserved but is, however, seen in another specimen described byDzik andKrumbiegel (1989), and allows us to recognize the anterior endof the original holotype. Dzik and Krumbiegel (1989) proposed that bothspecimens represent exuvia. We cannot agree with the statement, sincethe holotype of X. auerswaldae retains a degree of three-dimensionalityin its preservation which seems unlikely for an exuvium in a clasolite

Fig. 3. Xenusion auerswaldae Pompeckj, 1927, from the Kalmarsund sandstone. A. type spesub-rounded base for a spine, the wrinkles on the trunk and fine longitudinal striations on thlw: longitudinal wrinkles; tr: transverse rings.


matrix. Furthermore, the limbs of X. auerswaldae are preserved ventro-laterally in the anterior region and laterally in the posterior region. Wefeel this is more consistent with the life condition of X. auerswaldae,and further implies to us that this is a body fossil, and not an exuvium.

The trunk of X. auerswaldae bears fourteen segments. Each segmenthas a pair of humps, and there is a sub-rounded base for a spine in themiddle of each hump (Fig. 3A, B). The intervals between adjacenthumps possess numerous transverse wrinkles. It is worth mentioningthat there are numerous longitudinal wrinkles between the humpsand the margin of trunk (Fig. 3A, B). These were regarded by Jaegerand Martinsson (1967) as structures effacing the continuation of theintersegmental transversal depression across the dorsal zone. Wesuggest that these wrinkles could be the impressions of longitudinalmuscles. The limbs of X. auerswaldae are coniform in shape; each limbbearing several transverse rings and fine longitudinal striations(Fig. 3A, B). Left limb 7 reveals a row of spines pointing backward(Fig. 3A, B).

3.4. Microdictyon

3.4.1. Microdictyon sinicumThe genus Microdictyon was established by Bengtson et al. (1986)

based on isolated, phosphatic plates found at a number of localitiesworld-wide. Subsequent finds revealed that these structures are partof a lobopodian (Ramsköld and Hou, 1991) which bears nine pairs ofplates on the dorso-lateral surface and ten pairs of legs ventrally. Theplates of the Chengjiang species,M. sinicum Chen et al., 1989, were de-scribed in detail by Chen et al. (1995a) and Zhang and Aldridge (2007),but their orientation remains controversial. Ramsköld (1992) claimedthat the extended, limbless part represents the anterior end; a viewendorsed by Chen et al. (1995a) and Ramsköld and Chen (1998). Bycontrast, Hou and Bergström (1995) proposed a reversed orientation.

cimen of X. auerswaldae (coated). B. Camera lucida of the type specimen, showing thee limb. Abbreviations are as follows: bs: base for spine; fls: fine longitudinal striations;




Based on the known specimens and the arguments presented in theliterature, we favor the orientation sensu Ramsköld (1992). Further-more, there is usually a cuticular ring observed at – in this scenario –

Fig. 4. Microdictyon sinicum Chen et al., 1989 and Cardiodictyon catenulumHou et al., 1991, fromment of anterior part of ELI-JSMS0011A, showing the cuticular ring at the anterior part. C. ComELI-JS0002 with 23 plates. E. Specimen of C. catenulum ELI-JS0011A with a pair of eyes. F & G.V-shaped plates, saddle-shaped structure and plates of the spines. H & I. Dorso-ventrally preseris actually an artifact of preservation. J. Enlargement of specimen ELI-JS0100, showing the shieldof the spines; ssp: shield-shape plates; sss: saddle-shaped structure; te: tentacle; vsp: V-shaped


the anterior end of M. sinicum (Fig. 4A, B), which is similar to theperibuccal lamella of water bears (Tardigrada) (cf. Brusca and Brusca,1990).

the Chengjiang Lagerstätte. A. Complete specimen ofM. sinicum ELI-JSMS0011A. B. Enlarge-plete specimen of C. catenulum ELI-JS0035 with 25 plates. D. Nearly complete specimenLaterally preserved specimens of C. catenulum ELI-JS0054A and ELI-JS0112A showing theved specimens of C. catenulum ELI-JS0032 and ELI-JS0100 show shield-shape plates, which-shape plates. Abbreviations are as follows: cr: cuticular ring; lb: limb; pl: plate; ps: platesstructure.




3.5. Cardiodictyon

3.5.1. Cardiodictyon catenulumThe species Cardiodictyon catenulum Hou et al., 1991 is another

lobopodian from Chengjiang. A second specimen with an almostcompletely exposed head was illustrated by Ramsköld (1992). Thisspecimen, plus a further one, was studied by Hou and Bergström(1995). The latter authors identified a pair of sclerites covering mostof the head, leaving only the anterior tip free. After careful observa-tion of our 120 specimens, we found no evidence for such scleritesin the head of C. catenulum (Fig. 4C–E). Two pairs of tentacles occuron the ventro-lateral side of the head (Fig. 4D). As in H. fortis andM. haikouensis, a pair of eyes occur on the head of C. catenulum(Fig. 4E). The number of plate pairs in the three well-preserved, com-plete specimens of C. catenulum (ELI-JSC0035) is twenty-five. Noexamples of twenty-three or twenty-four plates were observed inour complete specimens (cf. Ramsköld and Chen, 1998). Therefore,we found no evidence for variation in the number of segments inC. catenulum and suppose that this species had a fixed number ofsegments, as seems to be the case in other Cambrian lobopodians.

Ramsköld and Chen (1998) indicated that the shape of the plateschanged during ontogeny, from V-shaped structures in small adults toventrally pointed, shield-shaped structures in large adults. Our materialrevealed both small and large adults bearing V-shaped plates; or to bemore accurate, the shape of the plates resembles the seeds of a sunflower.In the laterally preserved specimens, we can clearly see that the platesof each pair meet on the dorsal midline and – probably because of com-pression during preservation – the bases of the ‘seeds’ come together toform a saddle-shaped structure (Fig. 4F,G). There is a small spine at thetip of the ‘seed’, usually the spines are not well-preserved, but the smallsub-rounded plates of the spines are clear (Fig. 4F,G). In dorso-ventrallypreserved specimens, both ‘seeds’ are compressed to form a rhombus-shaped structure, and if the rhombus was squeezed strongly it presum-ably acquired the previously observed a shield-shape: an artifact ofpreservation (Fig. 4H–J).

3.6. Onychodictyon

3.6.1. Onychodictyon feroxA particularly striking armored lobopodian from Chengjiang is

Onychodictyon ferox Hou et al., 1991. The fossils are characterized byten pairs of dorsal plates, while a particular focus of debate has beenwhether there is a pair of sensory appendages (antennae) on the head.In the earlier descriptions Ramsköld & Hou (1991) and Hou andBergström (1995) provisionally identified antennae as present, butthese structures were later considered to be absent by Ramsköld andChen (1998) and Hou et al. (2004a). In 2008, complete specimens fromChengjiang showed thatOnychodictyon is in fact represented by two spe-cies (Liu et al., 2008). Onychodictyon ferox has a ‘head’ bearing a pair ofdorsal ‘antennae’ and a trunk with eleven pairs of limbs. However, Ouet al. (2012) provided an emended diagnosis of Onychodictyon ferox inwhich they think it bears twelve pairs of limbs. By contrast, we supposehere that their specimens should in fact be Onychodictyon gracilis; notOnychodictyon ferox (see below). With regards to a sclerotized headshield in Onychodictyon ferox, we could not confirm this from the mate-rial available. Our complete and extended specimen lacks sclerotizedstructures (Liu et al., 2008, figs. 1A1–A5, 2A1, A3). One of Ramsköld andChen's (1998) specimens with a putative sclerite (Ramsköld and Chen,1998, fig. 3.5C, 3.6B) may be a preservation artifact, given that the spec-imen is bent in this area such that the condensed annuli could give a falseimpression of a head shield.

3.6.2. Onychodictyon gracilisThe second species, Onychodictyon gracilis Liu et al., 2008 has, by

contrast, a blunt anterior end without any appendages, and a trunkwith twelve pairs of limbs. The first pair of limbs are set at a greater


distance from the second pair of limbs than the second pair of limbsfrom the third pair of limbs (Liu et al., 2008, fig. 2A1, A3). The speci-mens of Ou et al. (2012) also bear twelve pairs of limbs and thearrangement of the legs here is essentially the same. Furthermore,the putative proboscis of Ou et al. (2012, fig. 2f–h) is very similar tothe small head of O. gracilis (Liu et al., 2008, fig. 2A1–A5). With regardsto the branched antenniform appendages, the key feature of Ou et al.'sspecimen, we think their specimens are especially well preserved, butthat our specimens did not preserve these branched structures. Inter-estingly, these feathery, antenniform appendages are very similar tothe first pair of ‘great appendages’ in Aysheaia (see above).

3.7. Kerygmachela

3.7.1. Kerygmachela kierkegaardiThe Sirrius Passet, part of the Buen Formation in northern Greenland,

yielded the remarkable Kerygmachela kierkegaardi Budd, 1993. This largelobopodianwas described as havingwell-developed frontal appendages,a pair of terminal tail spines, a trunk bearing nodes and a series offlap-like lateral lobes above the lobopodous appendages; see also Budd(1999). Examination of some specimens in the Geological Museum ofCopenhagen, indicated that the frontal appendages, tail spines, trunkwith nodes and lateral lobes are all very clear, however the lobopodousappendages are relatively poor preserved (Fig. 5A–C). Based on its ‘gilledstructure’, Budd (1993) suggested that K. kierkegaardi could form part ofa clade with animals like Opabinia and Anomalocaris – which also havelateral lobes – all three together forming the sister group of the arthro-pods. Budd (1999) developed this hypothesis further in a more detailedstudy, also taking into account some intervening opinions and develop-ments. Kerygmachela kierkegaardi was included among the lobopodiansbased on its gross morphology, but Budd (1999) began to questionlobopodian monophyly; recognizing an ‘AOPK’ group (i.e. Anomalocaris,Opabinia, Pambdelurion, Kerygmachela), but suggesting that this mayrepresent a grade of animals (rather than a clade)which became increas-inglymore arthropod-like going from Kerygmachela toAnomalocaris, andeventually to Arthropoda. A further hypothesis of some significancederiving from the Sirrius Passet material (Budd, 1999) is the idea thatbiramous limbs could have evolved through fusion of the lobopod limbto the corresponding lateral (gill) lobe.

3.8. Paucipoda

3.8.1. Paucipodia inermisCompared to other armored lobopodians, Paucipodia inermis (Chen

et al., 1995b) is a relatively featureless animal. It was originally describedas having six pairs of legs. Later, Chen et al. (2002) found some speci-mens with nine pair of legs and subsequently regarded it as anotherspecies Paucipodia haikouensis Luo et al., 1999 (see below). However,Hou et al. (2004b) thought it was the complete specimen of Paucipodiainermis. Here, based on careful observation of the material, we acceptthe idea of Chen et al. (2002). Paucipodia inermis is of small size, about2 cm long; cylindrical in shape, and is composed of a slim head regionand trunk. Length is about two times of the width in the head region.The trunk region is more robust than the head region, with six pairs oflegs ventrolaterally splaying on both sides. The last pair of legs is locatedat the end of the trunk, compared with the other five pairs of limbs, andthis pair of limbs is much shorter (Chen et al., 2002, pp. 240, fig. 372).

3.8.2. Paucipodia haikouensisFrom the other species in this genus, P. haikouensis Luo et al., 1999,

three complete specimens and eight incomplete specimens have beendiscovered; the average length of complete specimens being about10 cm (Fig. 6A). However, one incomplete specimen (ELI-JSP008)with only six pairs of limbs reaches 11 cm (Fig. 6B), which indicatesthat P. haikouensis could achieve larger sizes. The whole shape ofP. haikouensis is cylindrical and the anterior part is curved and sharply



Fig. 5. Kerygmachela kierkegaardi Budd, 1993, from Sirrius Passet, Greenland. A. MGUH 22.083 (holotype, part) (uncoated). B. Enlargement of specimen MGUH 22.083 (uncoated),showing the frontal appendages. C. MGUH 25.047 (uncoated), showing the trunk with nodes. Abbreviations are as follows: fa: frontal appendages; la: lobopodous appendages;ll: lateral lobes; ts: tail spines.


cone-shaped; its length about three times its width. The mouth is ter-minal and no pharyngeal tooth or brain (cf. Hou et al., 2004b) can beobserved. The trunk shows fine annulations, 2–3 per mm, but lacks pa-pillae and/or paired sclerites. However, the attachment area of the legsis somehow sclerotized (Fig. 6C), which resembles the condition seen inother armored lobopodians. Nine pairs of legs protrude from theventro-lateral side of the body (Fig. 6A). Two claws occur at the end ofthe legs (Fig. 6D). However, in specimen ELI-JSP001, a claw-like struc-ture is observed on the cavity of the leg, about two-thirds of the waytowards the end (Fig. 6E). Comparing this claw-like structure with theclaws, they are similar, black, curved and hard. This leads us to questionthe function of these structures and to speculate whether there wereclaws for locomotion or other claws – perhaps connected to the bodycavity – which had another role, such as dispensing venom for preycapture?

3.9. Megadictyon

3.9.1. Megadictyon haikouensisThe large, robust lobopodianM. haikouensis Luo et al., 1999 was first

described from an incomplete specimen. At most, only seven segmentsof the trunk region are preserved; each bearing a pair of strong, stout


appendageswith appendicules. Unfortunately, the anterior and posteri-or ends of the original material are poorly preserved. In 2006, the firstauthor and a colleague at the Early Life Institute found a nearly completespecimen revealing some new characters. The specimen is 24 cm long,including a pair of strong frontal appendages. The mouthparts ofM. haikouensis are ‘Peytoia’-like, in other words they are very similarto those of the ‘AOPK’ group (see above). Note that the gut contents ofspecimen ELI-M001 of M. haikouensis are filled with a mineral mold(Liu et al., 2007, figs. 1A, B, 2K and 3A).

3.10. Luolishania

3.10.1. Luolishania longicrurisA rare lobopodian described from a single specimen by Hou and

Chen (1989) (Fig. 7A–C), this species was initially characterized as aworm-like animalwithfifteen pairs of lobe-like limbs. The body consistsof a head and a trunkwith fifteen segments, each segment bearing threetubercles (Fig. 7F). Later Ramsköld and Chen (1998) found another lesswell-preserved specimen and indicated that the numbers of legs is six-teen; one more than originally stated. Here, we also accept that thereare sixteen pairs of legs in the holotype (Fig. 7A, B). Ramsköld andChen (1998) also observed a large, rounded structure in the anterior



Fig. 6. Paucipodia haikouensis Luo et al., 1999, from the Chengjiang Lagerstätte. A. Complete specimen ELI-JSP035 with nine pair of legs and terminal mouth. B. Incomplete specimenELI-JSP008 with fine annulations on the trunk. C. Specimen ELI-JSP016B shows that the attachment area of legs is somehow sclerotized. D. The leg of specimen ELI-JSP030 bears twoclaws. E. Specimen ELI-JSP001 with a claw-like structure on the coelom of the leg. Abbreviations are as follows: an: annulation; dls: disc-like structure.


part of thehead in theholotype (Fig. 7D, E). Anteriorlywithin this area isa transversely arranged pair of tubercles (Fig. 7D, E) (see also Ramsköldand Chen, 1998, figs. 3.4A, 3.5B). For this structure, Ramsköld and Chen(1998) proposed two alternative interpretations: (1) a preservationalartifact given that it only occurs in the holotype, or (2) that it could rep-resent legs, eyes or antennae. Here, we prefer the artifact interpretationsince the head of holotype is strongly sclerotized and has somehowbro-ken; thus we cannot see any sign of legs, eyes or antennae. See below(Section 3.11) for a discussion of Ma et al.'s (2009) redescription ofthis species and comparison with another Chengjiang lobopodian.


3.11. Miraluolishania

3.11.1. Miraluolishania haikouensisMiraluolishania haikouensis Liu & Shu, 2004 was originally described

from 53 specimens (Liu et al., 2004), with another twelve found subse-quently. The body is divided into awell-defined head area and trunk re-gion. The head possesses a pair of eyes and a pair of antennae, themouth was situated antero-ventrally, while the trunk region comprisesfourteen segments, each bearing a pair of ventro-lateral limbs. Thetrunk of M. haikouensis shows partial differentiation and can be



Fig. 7. Luolishania longicruris Hou and Chen, 1989, from the Chengjiang Lagerstätte, holotype NIGPAS 108741. A. Part of NIGPAS 108741a. B. Camera lucida of NIGPAS 108741a,showing 16 pairs of limbs and three tubercles on each segment. C. Counterpart of NIGPAS 108741b. D. Anterior portion of part of NIGPAS 108741a, note the rounded structureand pair of tubercles in the head? E. Enlargement of head of NIGPAS 108741a, showing the strongly sclerotized and somehow broken head region. F. Enlargement of part ofNIGPAS 108741a, showing three tubercles. Abbreviations are as follows: pp: posterior projection; rs: rounded structure; sh: sclerotized head. tu: tubercles.


subdivided (see below). Since its original description as a cover articlein the Chinese Science Bulletin, a number of workers have acceptedthe original interpretation (e.g. Xiao, 2004; Mayer, 2006; Maas et al.,2007; Schoenemann et al., 2009; Whittle et al., 2009). By contrast, Maet al. (2006, 2009) doubted its status as a distinct species, consideringM. haikouensis to be a junior synonym of L. longicruris (see Section3.10.1).

Here, we discuss alternative interpretations of these fossils (Table 1)and reiterate the differences between, what for us, are two separatespecies (Table 2). In detail, we only see two sclerites on each segment


of M. haikouensis, regardless of whether the specimen is straight ortwisted (Liu et al., 2004, figs. 1(a) (b) (i) (j); Liu et al., 2007, fig. 1(h)).A mid sclerite identified by Ma et al. (2009, figs. 3A, C, J) is here treatedas the imprint of a dorsal spine which has fallen down to form apreservational artifact. None of the specimens figured byMa et al. clearlyshow an antenniform outgrowth protruding from the very anterior partof the head. Our dorso-laterally preserved specimen ELI-M0021 (see alsoLiu et al., 2004, figs. 2(a)–(d)) indicates antennae protruding from thedorsal–posterior part of the head; about a third of the distance fromthe back. Furthermore, in the ventro-dorsally preserved specimen



Table 2Summary of the differences recognized here between M. haikouensis and L. longicruris(cf. Table 1) which we believe justifies maintaining them in separate genera.

M. haikouensis L. longicrus

1. Eyes Present Absent2. Antenniform outgrowth Present Absent3. Setae on limbs Present Absent4. Tubercles per segment 2 35. Tubercles with spines Present Absent6. Number of limb pairs 14 167. Locality Haikou Chengjiang


ELI-M0010 (Liu et al., 2004, figs. 2(e)–(f)), part of an antenna is observedon the left side of the head, and is quite slim when compared withthe complete antennae of specimen ELI-M0021 (Liu et al., 2004, figs.2(a)–(d)). Thus we deduce that this part of the antenna onlyrepresents the tip. If the antenna really did originate from the front ofthe head, this part should be almost complete, and not just a slendertip. This supports the original observation that the antennae originatenearer the back of the head.

Based on two purplish red, smooth lines preserved at the anterioredge of the head of one specimen, Ma et al. (2009) inferred a headshield covering the head, with an opening at the anterior end; sup-posedly similar to that of C. catenulum (Hou et al., 1991) and H. fortis(Hou and Bergström, 1995). We do not accept a head shield neither inL. longicruris (Fig. 7A, B, D, E) nor M. haikouensis (Liu et al., 2004, figs.1(a), (b)). These two lines probably indicate some sort of doublure.Furthermore, neither C. catenulum (Hou et al., 1991) nor H. fortis(Hou and Bergström, 1995) actually have a head shield (see above).Ma et al. (2009) also suggested that the mouth is situated anteriorlyrather than anteroventrally (as per Liu et al., 2004). Based on Maet al. (2009, figs. 3F–H and 6D), we consider that the alimentarycanal does not terminate at its anterior end – still a small distancefrom the margin – and that there is a round ‘soft’ structure protrudingfrom the mouth (Ma et al., 2009, figs. 3H–J), which we regard as some-thing equivalent to the hypopharynx of an insect (Liu et al., 2004, fig.2(i) (j)). Ma et al. suggested this could be a proboscis-like extensionof the gut. Yet if this were the case, tiny spines or hooks might beexpected. By contrast, this structure is smooth and does not seem toexpand like, for example, the proboscis of Priapulida.

Ma et al. (2009) supposed that all the lobopods of M. haikouensisbore setae, with at least three rows along the lobopod rather than thetwo rows described by Liu et al. (2004). As clarified previously (Liuet al., 2008), this appears to be another artifact, and some traces of thesetae sensuMaet al. (2009,figs. 2A, 9B) are actually indications of annu-lations. This phenomenon also occurs in our specimens; e.g. it can beseen by enlarging the photographs of Liu et al. (2004, fig. 2(g), (h)).Regarding the number of claws per limb, the extensive preparationwork around the four radiating claws of specimen YKLP 11277 (Maet al., 2009,fig. 10A, B) lead us to doubt that these four brown structuresreally are terminal claws. Instead, they could be setae of the anterior andposterior margin of the limbs. We prefer the original interpretation ofone sclerotized claw at the limb tip.

Ma et al. (2009) interpreted the trunk as being divided into twoparts between T5 and T6. We prefer a trunk subdivided into threedivisions: the fore-trunk (up to segment 6, whereby segment lengthincreases from anterior to posterior), the mid-trunk (segments 7–9,all segments nearly the same length), and rear-trunk (behind segment9, whereby segment length decreases from anterior to posterior). Thisscheme can easily be recognized in our specimens (Liu et al., 2004,figs. 2(a)–(d)) and even in Ma et al. (2009, fig. 1A).

The major point of conflict (Tables 1–2) is the number of limbs. Liuet al. (2004) regarded M. haikouensis as having fourteen pairs of limbs,whileMa et al. (2009) interpreted L. longicruris (as its synonym) as hav-ing fourteen to sixteen pairs. In other words, Liu et al. (2004, 2008)

Table 1Alternative interpretations of specimens attributed to M. haikouensis Liu et al., 2004.

Liu et al. (2004, 2008) Ma et al. (2009)

1. Number of sclerites per segment 2 32. Origin of antenniform outgrowth Rear of head Front of head3. Head shield present No Yes4. Location of mouth Anteroventral Anterior5. Rows of setae along the lobopod 2 >36. Number of claws per limb 1 47. Division of the trunk 3 parts 2 parts8. Number of limb pairs 14 14–169. Paleoenvironment Haikou Haikou


regarded their fossils as belonging to a distinct genus based on the dif-ferent number of appendages. Ma et al. (2009) rejected the significanceof this as a generic character, and suggested that differences in the num-ber of limb pairs could simply be amanifestation of sexual dimorphism.We find this argument difficult to accept. First, the number of limbs inthe holotype of L. longicruris is unequivocally sixteen pairs (Fig. 7A, B),as indicted by Hou and Bergström (1995) and Ramsköld and Chen(1998). Second, the situation for L. longicruris and M. haikouensis is notnecessarily the same as in extant onychophorans. Although males andfemales of individual species of living velvet worms can indeed havedifferent numbers of limbs (Tait, 2001), their other somatic charac-ters are typically largely the same. For L. longicruris and M. haikouensisthere remain a suite of somatic differences (Table 2), in addition tothe limb number. Thus we maintain the earlier position in which M.haikouensis bears fourteen pairs of limbs – and belongs to a distinctgenus and species – while L. longicruris has sixteen.

Another point to note is that all ofMa et al.'s specimenswere collectedfrom Haikou; in the same area which yielded the specimens of Liu et al.and about 40 km from the type locality of L. longicruris (Maotianshan,Chengjiang). As indicated by Chen (2004), the Maotianshan deposit ofthe Chengjiang area was formed in a marine slope-shelf transitional set-ting of comparatively deeper seawater with poorer light transmission.By contrast, the Haikou deposit occurs in a shallower setting with higherseawater transparency. Thus Chen named them separately as theChengjiang biota and the Haikou biota respectively. Morphologically ofthe ‘Luolishania’ described by Ma et al. differs from the holotype ofLuolishania longicruris. We believe that the ‘Luolishania’ of the Haikouarea does belong to a distinct genus,Miraluolishania, whereas Luolishaniais typical for – and perhaps restricted to – the Maotianshan area ofChengjiang County.

3.12. Jianshanopodia

3.12.1. Jianshanopodia decoraA large animal discovered in the Chengjiang fauna, with a body

length with frontal appendages probably reaching 26 cm, J. decora Liuet al., 2006 preserves three body regions: a head, trunk and tail. Thehead consists of a pair of stout, strongly wrinkled appendages about40 mm long flanking the anterior region, with five cuneiform platesattached along the inner margin of the appendages (Liu et al., 2006,figs. 1B1, B2, 2B1). A small bradoriid can be observed located on themost anterior frontal plate, which could imply that J. decorawas a pred-ator. This lobopodian species bears a short, dilatable mouth tube thatleads into a funnel-like pharynx surrounded by the bases of the largefrontal appendages. A pharynx of this nature could perhaps have beenused for swallowing and masticating the captured prey. The worm-like trunk region seems to be sub-cylindrical in cross section. It is com-posed of at least nine segments, probably up to twelve segments basedon the entire morphology of J. decora. The most striking feature of thisspecies is its complex appendages, which bear two rows of tubercleson their dorsal side, with tree-like or lamellate branches protrudingfrom the tubercles. In terms of function, we speculate that they mayhave increased the surface area, as a type of gill to help J. decora carryout gas exchange. Perhaps this anatomy could also have helped the




animal to swim; at least temporarily. In support of this, the tail fan couldhave acted to add balance when J. decora was swimming.

3.13. Antennacanthopodia

3.13.1. Antennacanthopodia gracilisA particularly interesting recent find is the onychophoran-like

A. gracilis Ou et al., 2011. Only two specimens have been found, onefrom Erjie section, Kunyang, the other from the Jianshan section(~10 km northeast of Erjie), Haikou. Both specimens show a body isdivided into three parts: a head, trunk and caudal region. The headbears at least two appendage pairs, i.e., frontal antennae and an abuttingpair of second antennae, as well as paired lateral ocelli situated at thebase of the frontal antennae (Ou et al., 2011, figs. 1.1–1.5). The trunkis composed of nine segments and a terminal projection, without anyobvious trunk annuli or sclerotized plates. Each segment bears a pairof unjointed, stubby, robust legs (i.e. lobopods) ventrolaterally splayingon both sides of the body. Thorn-like, sharply pointed spines are ar-ranged on the surface of each lobe-like limb. At the end of each lobopoda walking pad can be observed (Ou et al., 2011, figs. 1.6–1.9), which issimilar to that of modern Onychophora (Brusca and Brusca, 1990).The caudal region of this new species is a little confusing, since thereis a pair of soft, cirriform structures protruding from the terminalprojection (Ou et al., 2011, figs. 1.1, 1.2, 1.10). Compared to otherlobopodians, this paired structure is morphologically most similar tothe caudal furcae in, e.g., Kerygmachela (see above). In term of its func-tion, these furcae could be accessory tactile/chemosensory receptors, orperhaps a prehensile device to aid in locomotion, gas exchange, or a de-fensive structure to confuse predators. In general A. gracilis is the most‘modern’ looking lobopodian from the Cambrian discovered so far –

the next fossil closely resembling a living velvetwormbeingHelenodorainopinata Thompson & Jones, 1980 from the Late Carboniferous CoalMeasures of the USA (see also Haug et al., 2012).

3.14. Diania

3.14.1. Diania cactiformisThe armored Chengjiang lobopodian Diania cactiformis Liu et al.,

2011a,b is about 6 cm long with a body divided into a proboscis andtrunk (Fig. 8). The proboscis of Diania is slightly expanded anteriorly,without any spines or tubercles but with some wrinkles. The trunk ofD. cactiformis is composed of nine segments, bearing rows of transverseannulations, each with numerous tubercles. Significantly, each segmentpossesses a pair of robust appendages with primary articulation andspines (Liu et al., 2011a, figure 2i–l). Compared with the hard, and possi-bly jointed appendages, the body of Diania is worm-like and relativelysoft. D. cactiformis may offer evidence that arthropodization (sclerotiza-tion of the limbs) preceded arthrodization (sclerotization of the body).In the original description, the presence of articulated limbs was used toargue that this fossil may be close to the origins of the arthropods (Liuet al., 2011a,fig. 4, Liu et al., 2011b,fig. 3), although the exact phylogenetic

Fig. 8. 3D reconstruction of Diania, f


position of this controversial species has since been questioned (cf. Legget al., 2011; Liu et al., 2011b; Mounce and Wills, 2011).

4. Conclusions

Lobopodians continue to fascinate as organisms which may wellhave given rise to both velvet worms and arthropods. Yet we shouldnot forget that in the early Palaeozoic they seem to have been a diverseand significant group in their own right. Recent studies – particularlythose based on Cambrian fossils from China – continue to recognizeboth new taxa, and new character sets, with important implications forour understanding of the origins and early evolution of the arthropods.At the same time interpreting lobopodians is often challenging. As dem-onstrated in the present review, different authors have reconstructed(or even orientated) the same animal in sometimes quite differentways, and in the absence of a direct modern equivalent (the velvetworm being a derived, terrestrial form) it is sometimes hard to teaseout genuine morphological features from artifacts of preservation. Thisproblem is not trivial, since correctly interpreting the presence and/orsegmental position of features like, for example, eyes, antennae, a pro-boscis, or the differentiation of the trunk and its appendages can impactin a major way on constructing arguments for homology with similarfeatures in arthropods. In the same vein, using lobopodians as anoutgroup to polarize arthropod character states requires a decision tobe made about which lobopodian(s) to use, and a clear understandingof the chosen taxa's morphology. The sometimes equivocal nature ofthe anatomical data may also explain why it has proven challenging toconstruct a robust cladogram for internal relationships between the gen-era outlined in the present review. Are the lobopodians a clade, or morelikely a paraphyletic grade, of ‘worms with legs’with multiple divergentlineages? New discoveries should help resolve their phylogeny andaffinities.

Acknowledgments

This work was financially supported by 973 project of the Ministryof Science and Technology of China (Grant 2013CB837100), NationalNatural Science Foundation of China (Grants 41222014 and D020101),financial supports from Ministry of Education of China, the Programmeof Introducing Talents of Discipline to Universities the 111 project(P201102007), Shaanxi Bureau of Science and Technology (FJ11366),the MOST Special Fund from the State Key Laboratory of ContinentalDynamics, Northwest University and the support for young talentsof Northwest University. Many thanks for Dr Steiner of Free UniversityBerlin for discussion, Dr Weber for taking the photo of Kerygmachelafrom the Geological Museum, Copenhagen (Denmark), Dr Neumann fortaking photos of Xenusion in Museum für Naturkunde, Berlin (Germany),and the Ph.D. student Sun, Chenyang for taking photos of Luolishanialongicruris in Nanjing Institute of Geology and Palaeontology, ChineseAcademy of Sciences, Nanjing (China).

rom the Chengjiang Lagerstätte.




References

Bengtson, S., Matthews, S.C., Missarzhevsky, V.V., 1986. The Cambrian netlike fossilMicrodictyon. In: Hoffman, A., Nitecki, M.H. (Eds.), Problematic Fossil Taxa, vol. 5.Oxford U. P., N.Y., & Clarendon, Oxford, pp. 97–115.

Bergström, J., Hou, X.G., 2001. Cambrian Onychophora or Xenusians. ZoologischerAnzeiger 240, 237–245.

Brues, C.T., 1923. The geographical distribution of the Onychphora. The AmericanNaturalist 57, 210–217.

Brusca, R.C., Brusca, G.J., 1990. Invertebrates. Sinauer, Sunderland Mass.Budd, G.E., 1993. A Cambrian gilled lobopod from Greenland. Nature 364, 709–711.Budd, G.E., 1999. The morphology and phylogenetic significance of Kerygmachela

kierkegaardi Budd. Transactions of the Royal Society of Edinburgh: Earth Sciences89, 249–290.

Chen, J.Y., 1991. Report. In: Walliser, O.H. (Ed.), IGCP 216, Bio-Events: Supplement tothe Annual Report, 1991. Springer-Verlag Berlin Heidelberg, Göttingen, p. 5.

Chen, J.Y., 2004. The Dawn of Animal World. Jiangsu Science and Technology Press,Nanjing (In Chinese).

Chen, J.Y., Hou, X.G., Lu, H.Z., 1989. Early Cambrian netted scale-bearing worm-like seaanimal. Acta Palaeontologica Sinica 28, 1–16.

Chen, J.Y., Zhou, G.Q., Ramsköld, L., 1995a. The Cambrian lobopodianMicrodictyon sinicum.Bulletin of the National Museum of Natural Science 5, 1–93 (Taichung, Taiwan).

Chen, J.Y., Zhou, G.Q., Ramsköld, L., 1995b. A new Early Cambrian onychophoran-likeanimal, Paucipodia gen. nov., from the Chengjiang fauna, China. Transactions ofthe Royal Society Edinburgh: Earth Sciences 85, 275–282.

Chen, L.Z., Luo, H.L., Hu, S.X., Yin, J.Y., Jiang, Z.W., Wu, Z.L., Li, F., Chen, A.L., 2002. EarlyCambrian Chengjiang Fauna in Eastern Yunnan, China. Yunnan Science and TechnologyPress, Kunming (In Chinese, with English summary).

Collins, D.H., 2002. The Paleontological Association 46th annual meeting. Book ofAbstracts.University of Cambridge (85 pp.).

Conway Morris, S., 1977. A new metazoan from the Burgess Shale of British Columbia.Palaeontology 20, 623–640.

Delle Cave, L., Simonetta, A.M., 1975. Notes on the morphology and taxonomic positionof Aysheaia (Onychophora?) and of Skania (undetermined phylum). MonitoreZoologica Italiano, n. s. 9, 67–81.

Dzik, J., Krumbiegel, G., 1989. The oldest ‘onychophoran’ Xenusion: a link connectingphyla? Lethaia 22, 169–181.

Haug, J.T., Mayer, G., Haug, C., Briggs, D.E.G., 2012. A Carboniferous non-onychophoranlobopodian reveals long-term survival of a Cambrian morphotype. Current Biology22, 1673–1675.

Hou, X.G., Bergström, J., 1995. Cambrian lobopodians—ancestors of extant onychophorans?Zoological Journal of the Linnaean Society 114, 3–19.

Hou, X.G., Chen, J.Y., 1989. Early Cambrian arthropod–annelid intermediate sea animal,Luolishania gen.nov. from Chengjiang, Yunnan. Acta Palaeontologica Sinica 28,207–213 (In Chinese, with English summary).

Hou, X.G., Ramsköld, L., Bergström, J., 1991. Composition and preservation ofthe Chengjiang fauna—a lower Cambrian soft-bodied biota. Zoologica Scripta 20,395–411.

Hou, X.G., Aldridge, R.J., Bergström, J., Siveter, D.J., Siveter, D.J., Feng, X.H., 2004a. TheCambrian Fossils of Chengjiang, China: the Flowering of the Early Animal Life.Blackwell Publishing Company, Oxford 83–93.

Hou, X.G., Ma, X.Y., Zhao, J., Bergström, J., 2004b. The lobopodian Pancipodia inermisfrom the Lower Cambrian Chengjiang fauna, Lower China. Lethaia 37, 235–244.

Hutchinson, G.E., 1930. Restudy of some Burgess Shale fossils. Proceeding of the UnitedStates National Museum 78 (11), 1–24.

Jaeger, H., Martinsson, A., 1967. Remarks on the problematic fossil Xenusionauerswalde. Geologiska Foreningens i Stockholm Forhandlingar 88, 435–452.

Legg, D.A., Ma, X.Y., Wolfe, J.M., Ortega-Hernández, J., Edgecombe, G.D., Sutton, M.D.,2011. Lobopodian phylogeny reanalysed. Nature 476, E2.

Liu, J.N., Shu, D.G., Han, J., Zhang, Z.F., 2004. A rare lobopod with well-preserved eyesfrom Chengjiang Lagerstätte and its implications for origin of arthropods. ChineseScience Bulletin 49, 1063–1071.

Liu, J.N., Shu, D.G., Han, J., Zhang, Z.F., Zhang, X.L., 2006. A large xenusiid lobopod withcomplex appendages from the Chengjiang Lagerstätte (Lower Cambrian, China).Acta Palaeontologica Polonica 51, 215–222.


Liu, J.N., Shu, D.G., Han, J., Zhang, Z.F., Zhang, X.L., 2007. Morpho-anatomy of the lobopodMegadictyon cf. haikouensis from the Early Cambrian Chengjiang Lagersträtte, SouthChina. Acta Zoologica 88, 209–288.

Liu, J.N., Shu, D.G., Han, J., Zhang, Z.F., 2008. The comparative study of Cambrianlobopods Miraluolishania and Luolishania. Chinese Science Bulletin 53 (1), 87–93.

Liu, J.N., Steiner, M., Dunlop, J.A., Keup, H., Shu, D.G., Ou, Q., Han, J., Zhang, Z.F., Zhang,X.L., 2011a. An armoured Cambrian lobopodian from China with arthropod-likeappendages. Nature 470, 526–530.

Liu, J.N., Steiner, M., Dunlop, J.A., Keup, H., Shu, D.G., Ou, Q., Han, J., Zhang, Z.F., Zhang,X.L., 2011b. Reply-Replying to. C. P. Mounce & M. Wills. Nature 476. http://dx.doi.org/10.1038/nature10266.Legg, D.A., et al. Nature 476, http://dx.doi.org/10.1038/nature10267. Nature 476, E3-E4

Luo, H.L., Hu, S.X., Chen, L.Z., Tao, Y.H., 1999. Early Cambrian Chengjiang Fauna fromKunming Region, China. Yunnan Science and Technology Press, Kunming (In Chinese,with English summary).

Ma, X.Y., Bergström, J., Hou, X.G., 2006. A systematic study of lobopodians from theChengjiang biota. Abstracts of the Second International PalaeontologicalCongress.University of Science and Technology of China Press, Hefei 307–308.

Ma, X.Y., Bergström, J., Hou, X.G., 2009. Morphology of Luolishania longicruris (LowerCambrian, Chengjiang Lagerstätte, SW China) and the phylogenetic relationshipswithin lobopodians. Arthropod Structure & Development 38, 271–291.

Maas, A., Mayer, G., Kristensen, R.M., Walsozek, D., 2007. A Cambrian micro-lobopodianand the evolution of arthropod locomotion and reproduction. Chinese Science Bulletin52, 3385–3392.

Mayer, G., 2006. Structure and development of onychophoran eyes: what is the ancestralvisual organ in arthropods? Arthropod Structure & Development 35, 231–245.

Mounce, R.C.P.,Wills,M.A., 2011. Phylogenetic position ofDiania challenged. Nature 476, E1.Ou, Q., Liu, J.N., Shu, D.G., Han, J., Zhang, Z.F.,Wan, X.Q., Lei, Q.P., 2011. A rare onychophoran-

like lobopodian from the Lower Cambrian Chengjiang Lagerstätte, Southwestern China,and its phylogenetic implications. Journal of Paleontology 85 (3), 587–594.

Ou, Q., Shu, D.G., Mayer, G., 2012. Cambrian lobopodians and extant onychophoransprovide new insights into early cephalization in Panarthropoda. Nature Communication3 (1261), 1–6.

Pompeckj, J.F., 1927. Ein neues Zeugnis uralten Lebens. Paläontologische Zeitschrift 9,287–313.

Ramsköld, L., 1992. Homologies in Cambrian Onychophora. Lethaia 25, 443–460.Ramsköld, L., Chen, J.Y., 1998. Cambrian Lobopodians: morphology and phylogeny. In:

Edgecombe, G.D. (Ed.), Arthropods Fossils and Phylogeny. Columbia UniversityPress, New York, pp. 107–150.

Ramsköld, L., Hou, X.G., 1991. New early Cambrian animal and onychophoran affinitiesof enigmatic metazoans. Nature 251, 225–228.

Robison, R.A., 1985. Affinities of Aysheaia (Onychophoran): with description of a newCambrian species. Journal of Paleontology 59, 226–235.

Schoenemann, B., Liu, J.N., Shu, D.G., Han, J., Zhang, Z.F., 2009. A miniscule optimisedvisual system in the Lower Cambrian. Lethaia 42, 265–273.

Steiner, M., Hu, S.X., Liu, J.N., Keupp, H., 2012. A new species of Hallucigenia from theCambrian Stage 4Wulongqing Formation of Yunnan (South China) and the structureof sclerites in lobopodians. Bulletin of Geosciences 87 (1), 107–124.

Tait, N.N., 2001. The Onychophora and Tardigrada. In: Anderson, D.T. (Ed.), Inverte-brate Zoology, 2nd edition. Oxford University Press, Oxford, pp. 206–224.

Thompson, I., Jones, D.S., 1980. A possible onychophoran from the Middle PennsylvanianMazon Creek beds of northern Illinois. Journal of Palaeontology 54, 588–596.

Walcott, C.D., 1911. Middle Cambrian annelids. Cambrian geology and paleontology, II.Smithsonian Miscellaneous Collections 57, 109–144.

Walcott, C.D., 1931. Addenda to descriptions of Burgess Shale fossils (with explanatorynotes by Charles E. Resser). Smithsonian Miscellaneous Collections 85, 1–46.

Walton, L.B., 1927. The polychaete ancestry of the insects. I. The external structure. TheAmerican Naturalist 61, 226–250.

Whittington, H.B., 1978. The lobopodian animal Aysheaia pedunculata Walcott, MiddleCambrian, Burgess Shale, British Columbia. Philosophical Transactions of the RoyalSociety of London B 284, 165–197.

Whittle, R.J., Gabbott, S.E., Aldridge, R.J., Theron, J., 2009. An Ordovician lobopodianfrom the Soom Shale Lagerstätte, South Africa. Palaeontology 52, 561–567.

Xiao, S.H., 2004. An arthropod sphinx. Chinese Science Bulletin 49, 983–984.Zhang, X.G., Aldridge, R.J., 2007. Development and diversification of trunk plates of the

Lower Cambrian lobopodians. Palaeontology 50, 401–415.


http://refhub.elsevier.com/S0031-0182(13)00285-X/rf0240





































































http://dx.doi.org/10.1038/nature10266
























































cambrian lobopodians: a review of recent progress in our understanding of their morphology and...

Documents