Back to Table of Contents

43893-AC8
Carbonate-Hosted Ediacaran Fossils from Siberia

N. J. Butterfield, University of Cambridge

Due to mid-grant personnel change, the breadth of this project has been expanded to include the exceptional preservation of Cambrian sponges.  It nevertheless remains linked to the original theme through its focus on the role of carbonate diagenesis in capturing aspects of early animal evolution.  The first year’s work on carbonate-hosted Ediacaran fossils from Siberia has just been published in Geology, where Grazhdankin et al. (2008) provide a detailed account of the Khatyspyt biota and how its taphonomic expression can be used to illuminate the constitution of these enigmatic organisms.   

In the past year post-doc Alex Page and I have been investigating the preservation and phylogenetic affiliations of Early-Middle Cambrian sponge fossils using detailed petrographic and geochemical analyses.  The study has focused primarily on Protospongia, Halichondrites and Eiffelia in the Burgess Shale of western Canada, but we have also examined type materials from the Menevian Group, Wales, and closely comparable material in Quebec, NW Canada and Ireland.  Although previous work readily classified all of these taxa within extant poriferan groups, it is becoming increasingly clear that many exhibit extinct character combinations.  Fossils exhibiting mixtures of primitive and derived characters almost certainly represent the “stem-groups” of modern demosponges, hexactinellids, silicisponges and/or calcareans (e.g., Botting and Butterfield 2005).  As such, they hold the clue to the earliest steps in animal evolution.

We are particularly interested in the mineralogy of early sponge spicules, though this tends to be obliterated by early diagenetic pyrite replacement.  In the case of Protospongia, the simple stauract/hexact nature of its spicules has been used to infer its affiliation to extant siliceous hexactinellids.  Although the type material of Protospongia and Halichondrites has been entirely pyritized, we note that this is not the case for specimens in the Burgess Shale.  Moreover, unlike almost all other Burgess fossils (including sponges), these two taxa are preserved without any sign of vertical compaction. This 3-D habit is only otherwise seen in Burgess Shale fossils with low-Mg calcite skeletons such as trilobites.

We analyzed polished sections of Burgess Shale Protospongia and Halichondrites petrographically and with cathode luminescence (CL), backscatter SEM (BSEM), energy dispersive X-rays (EDX), and electron microprobe.  Remarkably, these fossil spicules contain no silica and are instead composed of monocrystalline calcite with a rind of aluminosilicate minerals (chlorite and muscovite).  The aluminosilicates are unquestionably secondary – related to the greenschist metamorphism experienced by the whole of the Burgess Shale – but there is good reason to believe that the calcite is primary:  co-occurring trilobite cuticle has also been replaced by aluminosilicates, but with residual inclusions of (presumably primary) calcite (Butterfield et al. 2007).  As such, we interpret Protospongia and Halichondrites spicules as originally calcitic.

Insofar as Protospongia would be classified as a hexactinellid on the basis of its spicule geometry, but as a calcarean by virtue of its mineralogy, it represents an extinct character combination:  it is correctly positioned on the stem-group of an extant poriferan clade.  On the basis of quite different characters, this was also found to be the case with the Burgess Shale fossil Eiffelia (Botting and Butterfield 2005).  Together with extant sponge groups, these fossils allow a more general analysis of character acquisition in early sponge evolution.  In a cladistic analysis using PAUP and a consensus topology for the major clades of extant sponges (see ‘Nugget’) we found Eiffelia to be a stem-group silicisponge (hexactinellids + demosponges) while Protospongia was most closely related to extant calcareans.  This is a radical departure from their conventional taxonomic placement, and has important implications for understanding the inter-relationships of extant sponge groups (Page et al., submitted).

By recognizing differences in fossil expression, and the underlying taphonomic mechanisms, we have been able to identify the original calcareous mineralogy of Cambrian sponges that have conventionally been interpreted as siliceous.  Carbonate mineralogy and taphonomy also figures large in Ediacaran palaeobiology (e.g., Grazhdankin et al. 2008), and this research will contribute importantly to its further understanding, not least in providing a search image for the last common ancestor of animals.

References
Botting & Butterfield 2005. Reconstructing early sponge relationships using the Burgess Shale fossil Eiffelia globosa, Walcott. PNAS, 102, 1554-1559.

Butterfield, N. J., Balthasar, U., and Wilson, L. A. 2007. Fossil diagenesis in the Burgess Shale. Palaeontology, 50, 537-543.

Grazhdankin, D. V., Balthasar, U., Nagovitsin, K. E., and Kochnev, B. B. 2008. Carbonate-hosted Avalon-type fossils in arctic Siberia. Geology, 36, 803–806.

Page et al., submitted. Character acquisition in early sponge evolution inferred from Cambrian fossils. Proceedings of the Royal Society, B.