[Palaeontology, Vol. 51, Part 4, 2008, pp. 865–879] SKELETAL MICROSTRUCTURE INDICATES CHANCELLORIIDS AND HALKIERIIDS ARE CLOSELY RELATED by SUSANNAH M. PORTER Department of Earth Science, University of California at Santa Barbara, Santa Barbara, CA 93106, USA; e-mail: [email protected] Typescript received 7 November 2005; accepted in revised form 7 September 2007 Abstract: Chancelloriids are problematic, sac-like animals and siphogonuchitid sclerites are homologous. The difference whose sclerites are common in Cambrian fossil assemblages. in chancelloriid and halkieriid body plans can be resolved in They look like sponges, but were united with the slug-like two ways. Chancelloriids either represent a derived condition halkieriids in the group Coeloscleritophora Bengtson and exhibiting complete loss of bilaterian characters or they repre- Missarzhevsky, 1981 based on a unique mode of sclerite con- sent the ancestral condition from which the bilaterally sym- struction. Because their body plans are so different, this pro- metric halkieriids, and the Bilateria as a whole, derived. The posal has never been well accepted, but detailed study of their latter interpretation, proposed by Bengtson (2005), implies sclerite microstructure presented here provides additional that coeloscleritophoran sclerites (‘coelosclerites’) are a plesi- support for this grouping. Both taxa possess walls composed omorphy of the Bilateria, lost or transformed in descendent of a thin, probably organic, sheet overlying a single layer of lineages. Given that mineralized coelosclerites appear in the aragonite fibres orientated parallel to the long axis of the fossil record no earlier than c. 542 Ma, this in turn implies sclerite. In all halkieriids and in the chancelloriid genus Archi- either that the Ediacaran record of bilaterians has been misin- asterella Sdzuy, 1969, bundles of these fibres form inclined terpreted or that coelosclerite preservability increased at the projections on the upper surface of the sclerite giving it a beginning of the Cambrian Period. The former is difficult to scaly appearance. On the lower surface of the sclerite, the pro- reconcile with Ediacaran trace and body fossil evidence, but jections are absent. This microstructure appears to be unique the latter may be possible, reflecting either independent min- to chancelloriids, halkieriids, and their relatives, siphogonu- eralization of organic-walled sclerites in chancelloriids and chitids and sachitids. (The sclerites of another putative hal- halkieriids or the opening of a taphonomic window that kieriid relative, Wiwaxia Walcott, 1911, are unmineralized, favours coelosclerite preservation. making direct comparisons impossible.) Thus, similarity both at the level of sclerite construction and the level of sclerite Key words: chancelloriids, halkieriids, biomineralization, microstructure suggests that chancelloriid, halkieriid, sachitid, microstructure, sclerites, Coeloscleritophora. Chancelloriids were Cambrian animals that pos- Butterfield and Nicholas (1996) rejected a coeloscleritoph- sessed a radially symmetrical, sac-like body covered with oran affinity and instead supported the original interpre- an array of star-shaped sclerites. Their phylogenetic affini- tation of chancelloriids as sponges. Mehl (1996) also ties are controversial; first interpreted as sponges (Walcott rejected a coelosceritophoran affinity, and speculated that 1920; Sdzuy 1969), they were later united with the halki- chancelloriids may be related to ascidians. Conway Morris eriids, wiwaxiids, sachitids, and siphogonuchitids into the and Chapman (1996) suggested that the chancelloriid and Coeloscleritophora Bengtson and Missarzhevsky, 1981, a halkieriid body plans are so different that it is likely their group characterized by a multisclerite exoskeleton and sclerites are convergent. In a later paper, they recognized mineralized sclerites displaying a prominent internal only halkieriids, siphogonuchitids, and wiwaxiids as a cavity, a restricted basal foramen, and no evidence of natural group (Conway Morris and Chapman 1997). accretionary growth (Bengtson and Missarzhevsky 1981). Most recent systematic treatments are agnostic about the Later, Bengtson et al. (1990) and Bengtson (2005) added higher-level relationships of chancelloriids (e.g. Mehl some microstructural observations in support of the 1998; Ferna´ndez Remolar 2001; Janussen et al. 2002; hypothesis. Randell et al. 2005); only Bengtson and Hou (2001) The affinities between chancelloriids and other coelo- and Bengtson (2005) have argued recently in favour scleritophorans have never been well accepted, however. of a coeloscleritophoran affinity. Criticism of a ª The Palaeontological Association doi: 10.1111/j.1475-4983.2008.00792.x 865 866 PALAEONTOLOGY, VOLUME 51 chancelloriid-coeloscleritophoran relationship has been Swedish Museum of Natural History, Stockholm; SM, Sedgwick based primarily on the dissimilarity between chancelloriid Museum, University of Cambridge; GPIN, Nanjing Institute of and wiwaxiid sclerites (a criticism that is undermined by Geology and Paleontology, Nanjing; and GSC, Geological Survey the questionable status of the wiwaxiids themselves; see of Canada, Ottawa. Butterfield 1990), but rejection of the concept reflects a more general unease about the fundamental difference in chancelloriid and halkieriid body plans (e.g. Dzik 1994; MICROSTRUCTURE OF Conway Morris and Chapman 1996). Halkieriids clearly COELOSCLERITOPHORAN SCLERITES possessed bilateral symmetry and exhibited additional characters, including accretionary shells and a slug-like Halkieriid sclerites foot, that suggest a relationship with members of crown- group Bilateria, in particular, members of the Lophotro- Porter (2004b) studied well-preserved sclerites belonging chozoa (Conway Morris and Peel 1995). They thus to halkieriids from the Middle Cambrian Georgina presumably possessed other bilaterian or eumetazoan char- Basin, Australia, and provided a reconstruction of scler- acters such as a gut, internal organs, and a coelom. Chan- ite microstructure. Comparison with other halkieriid celloriids, in contrast, lack bilateral symmetry, as well as a species described in the literature indicates that there is gut, internal organs, and a coelom (Bengtson 2005). a microstructural theme that characterizes the entire Given such distinct differences in body plans, it is pos- group. It consists of the following four characters (see sible that the similarity in halkieriid and chancelloriid Text-fig. 1): sclerite construction noted by Bengtson and Missarzhev- 1. A thin outer layer, possibly organic in original compo- sky (1981) is the result of convergent evolution. But even sition. Almost all halkieriid casts exhibit this outer layer, when skeletal elements are similar at one level, conver- preserved through secondary phosphatization. It can be gence can often be detected by examining other structural identified by its smooth or granular, wrinkled appearance. levels (Hickman 2004). Here, I build on the observations It is often partly degraded, particularly where it covers of Bengtson et al. (1990), Porter (2004b), and Bengtson tubercles or plates (Pl. 1, figs 4, 6, 15, 17, 20); fibrous ele- (2005) as well as new observations of the chancelloriid, ments of the inner layer are occasionally visible beneath Archiasterella hirundo Bengtson, in Bengtson et al. 1990, (Pl. 1, fig. 4). The outer layer is distinct from any second- presented here, to show that the sclerites of chancelloriids, ary diagenetic coating, which is usually thicker, consists halkieriids, sachitids, and siphogonuchitids not only have of radially orientated crystals, and obscures detail on the a unique mode of sclerite construction but also share a sclerite surface (e.g. Pl. 1, figs 7–8). A primary origin is unique sclerite microstructure. also suggested by its contact with internal moulds (Porter 2004b). An originally organic composition, now replaced by calcium phosphate, is suggested by two observations. MATERIAL AND METHODS First, it sometimes appears wrinkled, indicating a degree of pliability (Pl. 1, figs 11–12). Second, its texture, con- Sclerites of Archiasterella hirundo were picked from acid sisting of spheroidal, randomly orientated calcium phos- maceration residues obtained during an earlier study of phate crystals, appears similar to that observed in Early Cambrian fossils from South Australia (Bengtson secondarily phosphatized fossils known to have been et al. 1990). The fossils come from samples UNEL1763a originally organic (Porter 2004b). Whether or not the and UNEL1848 from the Parara Limestone, Curramulka, layer may have been lightly mineralized is unclear. For a Yorke Peninsula (localities described in Bengtson et al. more detailed discussion of primary versus secondary 1990). Sclerites are preserved as phosphatic casts and diagenetic coating and evidence for original composition, moulds. Specimens were coated with gold and palladium see Porter (2004b). and viewed using a JEOL JSM-6300V scanning electron 2. An inner layer composed of aragonite fibres orientated microscope at 15 kV. Specimens of Australohalkieria parallel to the long axis of the sclerite and either parallel superstes Porter, 2004b and Wiwaxia Walcott, 1911 were to the sclerite surface or inclined toward its distal tip. processed in an earlier study of Middle Cambrian phos- Phosphatic casts of longitudinally orientated fibres may phatized assemblages from the Georgina Basin, Australia be observed where the outer organic layer is degraded