Onychodictyon ferox, from the Lower Cambrian Chengjiang Lagerstätte, is a worm-like armored lobopod with 10 pairs of dorsal plates and a pair of probable sensory appendages (“antennae”). Newly discovered complete specimens show that Onychodictyon is represented by two species: O. ferox with a “head” bearing a pair of dorsal “antennae” and a trunk with 11 pairs of limbs, whereas O. gracilis has a blunt anterior end without any appendages and a trunk with 12 pairs of limbs. Because of the close resemblance of the “antennae” of O. ferox and those of the lobopod Miraluolishania, they are considered to be homologous structures. The “antennae” of Cambrian lobopodians are proposed to be homologous with the arthropod antennulae.
The first described lobopodian from the Cambrian was Aysheaia pedunculata, reported by Walcott (1911) from the Burgess Shale of Canada. Conway Morris (1977) identified another lobopodian, Hallucigenia sparsa, from the same beds. Their morphology and affinities have received broad attention (Whittington 1978; Dzik and Krumbiegel 1989; Hou and Chen 1989; Hou et al. 1991; Ramsköld and Hou 1991; Chen et al. 1995; Budd 1997, 1999; Budd and Peel 1998; Ramsköld and Chen 1998). The Lower Cambrian Chengjiang fauna in Yunnan, south China, has yielded nine further taxa (Cardiodictyon catenulum, Microdictyon sinicum, Hallucigenia fortis, Onychodictyon ferox, Paucipodia inermis, Luolishania longicruris, Miraluolishania haikouensis, Jianshanopodia decora, and Megadictyon haikouensis', Hou and Chen 1989; Hou et al. 1991; Ramsköld and Hou 1991; Chen et al. 1995; Ramsköld and Chen 1998; Luo et al. 1999; Liu et al. 2004, 2006, 2007), adding significantly to the knowledge of the group. Here we reappraise the lobopodian Onychodictyon, an armored animal with rigid plates and spines. O. ferox was first described by Ramsköld and Hou in 1991, after which there has been some debate about its morphology and affinities (Ramsköld 1992; Ramsköld and Chen 1998; Chen 2004; Hou et al. 2004). The focus of this ongoing debate is whether or not O. ferox had antennae. In this paper, we offer some new data on the morphology of this animal and report a new species, or a case of sexual dimorphism, of Onychodictyon: O. gracilis sp. nov., discovered at Haikou village, Yunnan, from the Qiongzhusi (Chiungchussu) Formation, Yu'anshan Member (Eoredlichia Zone), Lower Cambrian.
Material and methods
The material described here was recovered from a grayish-green and grayish-yellow mudstone of the Yu'anshan Member (Eoredlichia Zone), Helinpu (Chiungchussu) Formation, Lower Cambrian at Haikou, Kunming, Yunnan. Details of the locality and stratigraphy were given in Zhang et al. (2001). More than 70 specimens have been collected from these localities by the working team of the Early Life Institute, Northwest University.
In general, the morphological terms used in the description are those of Ramsköld and Chen (1998) whenever appropriate. The specimens were photographed with an Olympus OM-4 Ti camera. The direction of illumination was varied where necessary to illustrate different features.
Phylum Lobopodia Snodgrass, 1938
Class Xenusia Dzik and Krumbiegel, 1989
Order Xenusiida Dzik and Krumbiegel, 1989
Family Onychodictyidae Hou and Bergström, 1995
Genus Onychodictyon Hou, Ramsköld, and Bergström, 1991
Type species: Onychodictyon ferox Hou, Ramsköld, and Bergström, 1991 (from Maotianshan, Chengjiang, Lower Cambrian).
Emended diagnosis.—A robust lobopod, subcylindrical in outline, consisting of a “head” and an elongate trunk. The trunk with 10 segments, each segment possessing a pair of dorsal round plates and lobe-like appendages on the ventrolateral surface.
Remarks.—The genus Onychodictyon was first described by Hou et al. (1991) from Chengjiang. Our materials provide new information, indicating different species or sexual dimorphs that differ with respect to the presence or absence of “antennae”. Accordingly, the generic diagnosis is revised herein.
Onychodictyon ferox Hou, Ramsköld, and Bergström, 1991
Material.—All specimens came from a grayish-green and grayish-yellow mudstone from the Jianshan section in Haikou village, Haikou County, and the Erjie section in Erjie village, Anning County, Kunming, Yunnan. Helinpu (Chiungchussu) Formation, Yu'anshan Member (Eoredlichia Zone), Lower Cambrian.
Diagnosis.—Robustly built lobopod, the body comprises a rounded “head” with “antennae”, the trunk with 10 segments and 11 pairs of limbs, long papillae on body and limbs, 10 pairs of plates on the dorsal of the body.
Description.—One complete adult specimen of Onychodictyon ferox is about 50–70 mm long (Fig. 1A1–A3). It possesses a small head region, which is semi-round in shape. In specimen ELIJSO-013, the “head” is a little convex dorsally, with a semi-elliptical profile and a maximum sagittal length of 2 mm, approximately 50% of its width (Fig. 1A1–A5). A pair of flagellum-like structures is about 4 mm long and 0.5 mm wide, protruding from the dorso-lateral side of the “head”; compared with the trunk limbs, the flagellum-like structures are thin and gentle with no annuli or nodes. Therefore, they are most likely sensory appendages(“antennae”) (Fig. 1A1–A5). Similar appendages also occur in dorsolaterally preserved specimen ELIMF-015B (Fig. 3D). No eyes and head shield are preserved.
The worm-like trunk region is sub-rounded in cross-section. It is composed of 10 segments, with a number of strong and clear annuli (Figs. 1A1–A5 and 3C1, C2). The anterior portion of the trunk in front of the 3rd segment comprises 2–3 annuli, the 4th, 5th, 6th, and 7th segments of the trunk possesses 5–6 further annuli, the posterior portion of the trunk behind the 7th segment bears 3 annuli (Fig. 1A1, A2). Each annulus has a vertical row of “tubercles”; there are usually 10 tubercles in one row observed around the circumference (Fig. 1A1, A2). Long, finely annulated appendicules extend from these “tubercles”. In each annulus, the appendicules are symmetrically placed relative to the dorsal midline. As in the trunk limbs, fine canals can be observed in these protrusions (Fig. 1B1, B2).
Ten pairs of plates are located dorsally in turn in the 10 segments. Each plate is sub-rounded, shows three-dimensional structures (convexity, ridges, rims), with a central broad-based spine directed obliquely upward and out (Figs. 1A1–A3, B1 and 2B).
There are 11 pairs of limbs, the anterior 10 attach at a circular area beneath each plate, the last pair located at the end of the trunk (Fig. 1B1). The limbs are cylindrical, protruding ventro-laterally from the trunk, all possessing thin, transverse annuli, about 30 in number from the base to the tip (Fig. 1A1, A2, B2). Each limb carries along its length at least two rows of tubercles, similar to those of the trunk (Fig. 1A1, A2, B2). The number of tubercles in each row is at least 10, probably up to 20, and small appendicules can be observed protruding from these tubercles, as in specimen ELIJSO-002A (Fig. 2C), Compared to those of the trunk, the appendicules of the limbs are shorter and thinner. A pair of claws occurs at the end of each limb (Fig. 3E), the direction of each claw varies depending on the preservation of the specimens, except for the claws of limb 11, where the direction is usually bent anteriorly (Fig. 1B1). The posture further underlines the adaptation of the posterior limb pair to an anchoring function, as suggested by Chen et al. (1995) and Ramsköld and Chen (1998).
The gut is straight and simple (Fig. 1A1–A3), usually preserved in a flattened black band, which lies near the ventral margin. No anus has been found.
Onychodictyon gracilis sp. nov.
Figs. 2A1–A7, 3A1–C2.
Holotype: ELIJSO-0003 (Fig. 2A1–A7); from the Jianshan section, Haikou county, Lower Cambrian.
Derivation of the name: Gracilis refers to its delicate appearance.
Diagnosis.—Relatively small to up to 50 mm, the body comprises a rounded “head” without “antennae”, the trunk with 10 segments and 12 pairs of limbs, the posterior of the trunk bearing a short projection. Compared to Onychodictyon ferox, O. gracilis is smaller, more fragile-looking with more pairs of limbs.
Description.—Three complete specimens (ELIJSO-0003, ELIJSO-0005, and ELIEJO-0017) are known. All three were laterally compressed. The body can be divided into a welldefined head area, a trunk with 10 segments and a small posterior projection. The maximum length (sagittal) of the three specimens is 50 mm (ELIJSO-0003) (Fig. 2A1), 35 mm (ELIEJO-0017) (Fig. 3B1) and 12mm (ELIJSO-0005) (Fig. 3A1), transverse width amounts to 5 mm (ELIJSO-0003) (Fig. 2A1), 3.5 mm (ELIEJO-0017) (Fig. 3B1) and 1.5 mm (ELIJSO-0005) (Fig. 3A1).
The head area is clearly differentiated from the trunk, the ridge on the “head” is very smooth on the antero-ventral side, and a little sclerotized on the dorsal side of specimens ELIJSO-0003 (Fig. 2A1, A3–A5) and ELIEJO-0017 (Fig. 3B1–B3). However, there is no sign of “antennae” or any other structures on the “head” (Figs. 2A1, A3–A5, 3A1–A3, B1–B3). According to the black trace interpreted as the gut, the mouth is located on the antero-ventral side of the “head” (Figs. 2A1, A3–A5,3A1–A3, B1–B3).
The trunk region comprises 10 segments, each bearing a number of strong annuli (Figs. 2A1–A3, A6, 3B1, B3i C1, C2), which range between 1 and 6 in number (Fig. 2A1, A3). The number of annuli is in direct proportion to the longitudinal length of the segments. There are only 1–2 in the first segment, increasing to 4 in the 5th segment. The 6th, 7th, 8th segments have 4–6 annuli. The remainder, except for the small projection, all have annuli, the number decreasing from 5 to 2 (Fig. 2A1, A3). As in Onychodictyon ferox, each annulus has a vertical row of “tubercles”, and long, finely annulated appendicules extend from these (Figs. 2A1, A3, A6, 3A1, A2, B1, B3, C1, C2 Similar to those of O. ferox, the plates are rounded, dorso-ventrally elongate, convex outward, with a ridge near the edge, with the structure of plates showing the same pattern (convexity, ridges, rims) (Figs. 2A1, A3, 3A1–A3). A pointed spine protrudes from each plate (Fig. 3A1–A3, B1, B3, C2), according to preservation, in some cases only the basal plate of the spines is observable (Fig. 2A1–A3).
Twelve pairs of limbs are observed in Onychodictyon gracilis, the first pair located anterior to the first plate pair, and, compared to the other limbs of the trunk, this pair of limbs is fat and strong (Figs. 2A1, A3–A5, 3B1–B3). As Ramsköld and Chen noted, this pair of anterior limbs is set at a greater distance from Limb1 than Limb1 is from Limb2 (Fig. 2A1, A3). The limb-pairs 2 to 10 are located ventro-laterally of the plate-pairs 1 to 9, respectively (Figs. 2A1, A3, 3B1, B3); limb-pairs 11 and 12 are based anteriorly and posteriorly below plate-pair 10 (Fig. 2A1, A3). As in the limbs of O. ferox, there are about 30 rows of thin, transverse annuli from the base to the tip (Fig. 2A1, A3, A6). Tubercles are present in two rows along the lengths of the limbs, similar to those of the trunk (Fig. 2A1, A3, A6), the number of series being about 10 to 20, and small appendicules are observed protruding from these tubercles (Fig. 2A1, A3, A6). A pair of claws occurs at the tip of Limb 8, and these two claws point in the opposite directions (Fig. 2A1, A3, A6).
A short posterior projection is well preserved in specimens ELLJSO-0003 (Fig. 2A1, A3) and ELIJSO-0005 (Fig. 3A1, A2, A4), and is similar to that in Microdictyon. The projection of Onychodictyon gracilis protrudes postero-ventrally from the margin, slightly ventral to the posterior end of the body, and is about 2.5 mm long by 2 mm wide in specimen ELIJSO-0003 (Figs. 2A1, A3). Probably because of the state of preservation, the ridge of the projection in ELIJSO-0003 shows some signs of small spines (Fig. 2A1, A3), but that of the projection in ELIJSO-0005 is smooth (Fig. 3A1, A4).
The alimentary canal is expressed as a trace with a dark band about 1mm wide. It is simple and straight and extends through much of the body, mostly in a subventral position (Figs. 2A1, A3, 3A1, A2, B1, B3). No anus is apparent.
The similarites of both Onychodictyon ferox and O. gracilis are clear, implying a close relationship for the two species of Onychodictyon: (1) both are subcylindrical in outline; (2) there are similar dorsal plates and appendicules; (3) both bear lobe-like limbs with tubercles and appendicules; and (4) two claws are present in these two species. However, there are also significant differences between them: (1) O. ferox possesses a pair of “antennae” on the “head”, while O. gracilis does not; (2) 11 pairs of limbs occur on the trunk of O. ferox, with 12 pairs of limbs only observed in O. gracilis with the first pair of limbs located anterior to the first plate pair; (3) the last limbs of O. ferox are located at the end of the trunk with the claws usually bent and directed anteriorly, while in O. gracilis the direction of the claws of last limb bent according to the posture of the trunk; and (4) O. gracilis bear a small projection on the trunk, while O. ferox does not. Accordingly, we interpret Onychodictyon ferox and O. gracilis as two different species. However, because different sexes within the same species probably had different morphologies, there is the alternative possibility that the two species represent sexual dimorphs of Onychodictyon.
Onychodictyon were first described, based on two specimens, as a caterpillar-like animal with “papillae” surrounding the mouth and a pair of “antennae” on the “head” (Ramsköld and Hou 1991; Hou et al. 1991). Later, these “papillae” were regarded as jaws (Ramsköld 1992; Hou and Bergström 1995) and the pair of “antennae” was thought to be a misinterpretation (Bergström and Hou 2001; Hou et al. 2004). Then, Ramsköld and Chen (1998) and Chen et al. (2004) observed 10 new specimens and interpreted the “jaw” structure as a head shield with a pair of appendages anterior to the pair beneath the first plate pair. Based on this information, we can see that the focus of the debate is: (1) whether Onychodictyon possessed “antennae” or not; (2) is there a jaw, a head shield, or anything else; (3) is there a pair of anterior appendages? After examination of 70 specimens, we consider mat there indeed exists a pair of “antennae” on the “head” of Onychodictyon ferox (Fig. 1A1–A5), which is quite similar to that of the coexisting lobopod Miraluolishania haikouensis in both its location and morphology (Liu et al. 2004; Xiao 2004; Brigitte et al. in press), both of them located on the dorsal of the “head” and flagellum-like in shape. Furthermore, the “antennae” of Miraluolishania are regarded as quite similar to those of arthropods, such as Fuxianhuia (Hou and Bergström 1997). Thus, this pair of “antennae” is not only an homologous structure in both O. ferox and Miraluolishania, but also suggests that such “antennae” might be a primitive link between lobopods and arthropods. Arthropod researchers generally agree that the antennules, or first antennae, of crustaceans are homologous to the antennae of other arthropod groups. Moreover, the antennae of all arthropods seem to be associated with the unfused anterior-most monosegments (Edgecombe 1998; Schram and Hof 1998). Besides, it is virtually certain that antennae of all arthropods were originally uniramous (Hou and Bergström 1997; Dzik 2003). New data indicate that onychophoran antennae are not homologous with those of arthropods; they are innervated by protocerebral and deutocerebral nerves respectively (Scholtz and Edgecombe 2005, 2006). Evaluation of which part of the cerebrum innervated the antennae of O. ferox and Miraluolishania is beyond the scope of this paper, but considering other distinct differences between Onychodictyon and onychophorans, such as no jaw or oral papillae in Onychodictyon, we are inclined to believe that the antennae of lobopods are closely implicated in the origin of antennae in arthropods. Concerning the “jaw” or “head shield” we think that both phenomena are artifacts of preservation. Although the dorsal cephalic ridge of O. gracilis is a little sclerotized, the head area of both O. ferox and O. gracilis shows no sign of jaw-like structures or a head shield (Figs. 1A1–A5, 2A1–A5, 3A1–A3, B1–B3), and this is a typical character of most lobopods, such as Cardiodictyon and Hallucigenia (Ramsköld and Chen 1998). A pair of anterior appendages occurs on the trunk of O. gracilis, and it appears that these appendages, which are based laterally on the trunk, and also anterior appendages in the same are present in Aysheaia (Ramsköld and Chen 1998).
We conclude that, thus far, the common features of O. ferox and O. gracilis strongly indicate that they belong to the same genus. However, the significant differences between them show that they are likely to be different species or possibly sexual dimorphs. Together with Miraluolishania, the “antennae” of these species are crucial to any exploration of the origin of antennae in arthropods.
This work was supported by the National Natural Science Foundation of China (grants 40332016 and 04062003), the Program for Changjiang Scholar and Innovative Research Team in the Universities, and Ministry of Sciences and Technology of China (PCSIRT, grant 2000077700), “973” Program (grants 2006CB806401), Ph.D. Innovative Program of Northwest University (grant 07YYB02). The authors are indebted to Simon Conway Morris (Department of Earth Science, University of Cambridge, Cambridge, UK) for helpful discussion, Susan Turner (Queensland Museum, Australia) for improving the English, Zhengqian Luo for improving the line drawings, Paul Taylor (Natural History Museum, London, UK) for linguistically adjusted our text, and Yanchun Yao (both Northwest University, Xi'an, China) for field work help. Special thanks are given to Huilin Luo and Shixue Hu (both Yunnan Institute of Geological Science, China) for advice.
Synonymies.—Based on the distinct differences between Onychodictyon ferox and O. gracilis, here we give remarks to help the reader to distinguish them. The material in Ramsköld and Hou (1991), Hou et al. (1991), and Hou and Bergström (1995) is the same specimen NIGPAS 115271 and is O. ferox. Ramsköld and Chen (1998) show another three specimens, of which ELRC 33011 possesses 11 pairs of limbs and no projection at the end of the trunk, here considered to be O. ferox. The other two specimens, ELRC 33022 and ELRC 33009, are partly shown in their figure; it is hard to decide to which species they belong. The specimen RCCBYU10250 in Hou et al. (2004) is incomplete and, as we have not studied this fossil personally, we cannot make a decision about it. The specimens in Ramsköld (1992) and Bergström and Hou (2001) are all reproduced from Ramsköld and Hou (1991) and should be O. ferox.