The enduring controversy about the appearance of animals in the evolutionary record takes a fresh twist with an analysis of molecular fossils that places the rise of the sponge lineage before 635 million years ago.
Charles Darwin was famously sceptical about the sudden appearance of fully formed animals (metazoans) in the Early Cambrian fossil record, beginning some 542 million years ago. To a degree, he has been vindicated by the discovery of animal and animal-like fossils extending throughout the preceding Ediacaran Period, which followed the end of the second of the great Cryogenian ice ages around 635 million years ago (Fig. 1). But there the trail runs out. So is this really where metazoan life began? Or is it merely the point at which a capricious fossil record disappears?
In the absence of shells or bones, the fossil record of animals can fade away to localized snapshots, such as the remarkable diversity of early animal-like fossils in the Doushantuo biota of southern China1. However, estimates of evolutionary first appearance require a fundamentally more reliable type of data2. This is where Gordon Love and colleagues3 (page 718 of this issue) check in with their analysis of fossil biomarkers — geologically robust and taxonomically distinctive hydrocarbon molecules, derived primarily from the lipid membranes of once-living organisms. And when it comes to tracking primitive animals, the key biomarker is a 30-carbon steroid called 24-isopropylcholestane (24-ipc). The only known sources of this compound are species of the Demospongiae, one of the three main classes of extant sponges (phylum Porifera).
Love et al.3 focused on an unusually complete sequence of sedimentary rock in Oman. They not only document an abundance of 24-ipc throughout the Ediacaran, but also trace it into underlying Cryogenian strata — compelling evidence that the organisms producing this signal were present before the end of the 635-million-year-old glacial event (Fig. 1). Crucially, the authors show that the biomarkers could not have migrated from younger rocks. They achieved this by catalytically cracking the immobile organic matrix in the sediments, releasing 24-ipc biomarkers in similar high abundance to those in the associated soluble extracts.
This rigorous screening procedure was not used in a previous report4 of 24-ipc from much older rocks, extending back to about 1,600 million years ago. Love et al.3 attribute the conspicuously lower concentrations of the compound in that report to an alternative, non-sponge source, although this interpretation diminishes the value of 24-ipc as a taxonomically diagnostic biomarker. However, there is a strong possibility that the low concentrations of 30-carbon steroids in pre-Cryogenian rocks represent secondary contamination5. If so, the new data3 provide an even crisper signal for a Cryogenian first appearance of 24-ipc and its biological source.
So, what exactly were the organisms that produced these biomarkers? The most obvious answer, and the one that the authors3 plump for, is that demosponges had evolved and become ecologically prominent by at least the late Cryogenian. But this conclusion overlooks the evolutionary nature of biological taxa and the incremental assembly of defining characteristics along (now-extinct) 'stem lineages' (Fig. 2). It is only with a full complement of such characteristics — in the last common ancestor of the extant 'crown group' — that modern taxonomic boundaries apply6. It is certainly possible, perhaps even likely, that the biomarkers from Oman reflect the existence of true, multicellular sponges with a water-canal system. But this conclusion depends on the evolutionary relationships between extant sponges (represented principally by the demosponges, hexactinellids and calcareans) and their adjacent sister groups (the single-celled choanoflagellates, and the eumetazoans; this latter group includes all metazoans apart from sponges).
A defining characteristic of crown-group demosponges, hexactinellids and calcareans is widely understood to be the development of mineralized skeletal structures, or spicules. The absence of convincing spicules in the Ediacaran or Cryogenian fossil record7 implies that the modern poriferan classes were not fully defined until the Cambrian — and even then, seemingly bizarre combinations of spicule characteristics in Middle Cambrian fossils8 suggest a delayed arrival of poriferan crown groups. Assuming that pre-Cambrian 24-ipc biomarkers originated from a sponge stem-group certainly does not rule out their derivation from a true sponge, and some evolutionary scenarios for the distribution of 24-ipc yield this as a unique solution (Fig. 2a,b). Other interpretations, however — including that from the most recent and comprehensively sampled analysis of hexactinellid relationships9 — allow the biomarker biosynthesis to extend back into stem-group forms that were not sponges, and potentially not even multicellular (Fig. 2c,d). Such a possibility has important implications for the ecological interpretation of 24-ipc and the way it is applied to molecular clocks.
Despite the ambiguities, Love and colleagues' positive identification of 24-ipc in the late Cryogenian marks a considerable advance in resolving early animal evolution — particularly in light of the latest and most comprehensive molecular analysis of metazoan relationships, which no longer identifies sponges as the most primitive living animals10 (Fig. 2d). The next steps are to find out how far back the signal can be traced in time, and how to interpret negative results. There are currently fewer than half a dozen reports of convincing biomarker occurrences of Cryogenian age, and the conspicuously low abundances of 24-ipc in post-Cambrian sediments stands at odds with the proliferation of presumed demosponge reefs in succeeding periods of Earth history.
Further sampling of the Cryogenian is clearly in order, but so too is the search for independent proxies of early animal life. Like the first predatory animals in the Ediacaran, which seem to have induced a fundamental shift in both organismal morphology and evolutionary dynamics11, stem-group sponges may have left an indirect ecological fingerprint. It is possible, for example, that the novel feeding habits of sponges — based on the circulation of sea water through a sophisticated water-canal system — may have impinged sufficiently on the marine carbon cycle to register in the biogeochemical record12. Combined with new biomarker data and molecular phylogenomics, the identification of such signals promises to pinpoint the first appearance of our earliest animal ancestors.
Xiao, S. & Laflamme, M. Trends Ecol. Evol. 24, 31–40 (2009).
Butterfield, N. J. Integr. Comp. Biol. 43, 166–177 (2003).
Love, G. D. et al. Nature 457, 718–721 (2009).
McCaffrey, M. A. et al. Geochim. Cosmochim. Acta 58, 529–532 (1994).
Brocks, J. J., Grosjean, E. & Logan, G. A. Geochim. Cosmochim. Acta 72, 871–888 (2008).
Budd, G. Nature 412, 487 (2001).
Xiao, S., Hu, J., Yuan, X., Parsley, R. L. & Cao, R. Palaeogeogr. Palaeoclimatol. Palaeoecol. 220, 89–117 (2005).
Botting, J. P. & Butterfield, N. J. Proc. Natl Acad. Sci. USA 102, 1554–1559 (2005).
Dohrmann, M., Janussen, D., Reitner, J., Collins, A. G. & Wörheide, G. Syst. Biol. 57, 388–405 (2008).
Dunn, C. W. et al. Nature 452, 745–749 (2008).
Peterson, K. J. & Butterfield, N. J. Proc. Natl Acad. Sci. USA 102, 9547–9552 (2005).
Sperling, E. A., Pisani, D. & Peterson, K. J. in The Rise and Fall of the Ediacaran Biota (eds Vickers-Rich, P. & Komarower, P.) 355–368 (Geol. Soc. Lond., 2007).
Carr, M., Leadbeater, B. S. C., Hassan, R., Nelson, M. & Baldauf, S. L. Proc. Natl Acad. Sci. USA 105, 16641–16646 (2008).
About this article
Nature Ecology & Evolution (2018)