Biologia

Anatomy and relationships of the Early Cambrian worm Myoscolex
Blackwell Publishing Ltd.

JERZY DZIK

Accepted: 20 March 2003

Dzik, J. (2004). Anatomy and relationships of the Early Cambrian worm Myoscolex. — Zoologica Scripta, 33, 57 – 69. Numerous fossil specimens of Myoscolex ateles Glaessner, 1979 from the late Early Cambrian Emu Bay shale of Kangaroo Island, South Australia withphosphatized organic matter-rich tissues show its muscular body wall penetrated by rows of rod-like structures — possible chaetae. The body wall was composed of an external layer with transverse (circular) fibres. This layer was thickest in lateral parts of the body and very thin dorsally. In the ventro-lateral quarter of the body circumference, a belt of longitudinal fibres extended along the body.Longitudinal fibres also occurred in the dorsal region of the body. Along the venter extended a narrow longitudinal belt of probably oblique cords, crossing themselves perpendicularly. In having a virtually smooth, laterally flattened body, Myoscolex closely resembles the slightly geologically younger Pikaia from the Burgess shale of British Columbia, generally believed to be one of the oldestchordates. Being the oldest probable annelid, at least superficially similar to the opheliid polychaetes, Myoscolex may appear not too distant from the ancestor of the phylum. The lateral body flattening of Myoscolex was apparently an adaptation to swimming by undulation of the body in transverse plane, similar to today’s errant polychaetes but without using chaetae or appendages in propulsion. Jerzy Dzik,Instytut Palaeobiologii PAN, Twarda 51/55, 00-818 Warszawa, Poland. E-mail: dzik@twarda.pan.pl

Introduction
An ancient geological age of an organism does not guarantee that its anatomy is primitive. However, this increases the probability that a fossil would show a morphological proximity to the ancestor of its clade. This is an unavoidable implication from the basic assumption of anyphylogenetic inference: that there is a correspondence between time which passed since the divergence of lineages and their morphological difference. From this point of view the oldest known anatomically preserved worms of once proposed polychaete or echiurid affinities are of special importance, even if their systematic affiliations largely remain controversial. Myoscolex ateles Glaessner, 1979 from thelate Early Cambrian Emu Bay shale of Kangaroo Island, South Australia is the most informative of them, owing to numerous specimens with phosphatized soft tissues (Glaessner 1979; Briggs & Nedin 1997; Nedin 1997). The oldest fossils attributable to the Annelida are also important in the context of the recent progress in molecular phylogenetics. Some of the ideas forwarded by molecular biology pose achallenge to palaeontology as it is expected that the fossil record can potentially be used to test various phylogenetic trees based on molecular evidence. At least the

fossil data should not be incompatible with those offered by the analysis of the genome. In the case of the high rank taxonomic groupings of animal phyla proposed on the basis of molecular evidence (e.g. De Rosa et al. 1999),the Deuterostomia and Ecdysozoa are well supported by the palaeontological evidence (the latter was in fact first inferred from fossils; Dzik & Krumbiegel 1989). To the contrary, the superphylum Lophotrochozoa to which the Annelida belong, remains controversial. Despite some physiological affinities, for instance the use of haemerythrin as an oxygen transport protein (van Holde 1998) or intermediatefilament protein sequences (Erber et al. 1998), as well as support from embryology (Bartolomaeus 2001), the lophotrochozoan phyla are dramatically different from each other in their anatomies. There were attempts to connect phylogenetically brachiopods, molluscs, and polychaetes via the extinct Cambrian halkieriids (Conway Morris & Peel 1995), but this requires a series of transitional stages...